The Chinese Economy After WTO Accession (The Chinese Economy Series)

THE CHINESE ECONOMY AFTER WTO ACCESSION The Chinese Economy Series This series examines the immense importance of Chi...

Author: Shuming Bao | Shuanglin Lin | Changwen Zhao

67 downloads 2157 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

THE CHINESE ECONOMY AFTER WTO ACCESSION

The Chinese Economy Series This series examines the immense importance of China within the global economy. Books in the series view the Chinese economy in many ways, such as: a transition economy; a bridge between the developing and developed nations; a vital member of the WTO; and even as a potential rival to the US. Providing readers with high quality monographs and edited volumes by authors from East and West, this series is a truly global forum on one of the world’s key economies. Series Editors Aimin Chen, Indiana State University, USA Shunfeng Song, University of Nevada, USA Recent Titles in the Series China’s Agricultural Development Challenges and Prospects Edited by Xiao-yuan Dong, Shunfeng Song and Xiaobo Zhang ISBN 0 7546 4696 3 China’s Rural Economy after WTO Problems and Strategies Edited by Shunfeng Song and Aimin Chen ISBN 0 7546 4695 5 Chinese Youth in Transition Edited by Jieying Xi, Yunxiao Sun and Jing Jian Xiao ISBN 0 7546 4369 7 Grains in China Foodgrain, Feedgrain and World Trade Edited by Zhang-Yue Zhou and Wei-Ming Tian ISBN 0 7546 4280 1 Critical Issues in China’s Growth and Development Edited by Yum K. Kwan and Eden S.H. Yu ISBN 0 7546 4270 4 The Efﬁciency of China’s Stock Market Shiguang Ma ISBN 0 7546 4241 0

The Chinese Economy after WTO Accession

Edited by SHUMING BAO The University of Michigan at Ann Arbor, USA SHUANGLIN LIN The University of Nebraska at Omaha, USA CHANGWEN ZHAO Sichuan University, China

© Shuming Bao, Shuanglin Lin and Changwen Zhao 2006 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the publisher. Shuming Bao, Shuanglin Lin and Changwen Zhao have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identiﬁed as the editors of this work. Published by Ashgate Publishing Limited Gower House Croft Road Aldershot Hampshire GU11 3HR England

Ashgate Publishing Company Suite 420 101 Cherry Street Burlington, VT 05401-4405 USA

Ashgate website: http://www.ashgate.com British Library Cataloguing in Publication Data The Chinese economy after WTO accession. - (The Chinese economy series) 1. World Trade Organization - China 2.Globalization Economic aspects - China 3.China - Economic conditions 2000- 4.China - Foreign economic relations 5.China Commercial policy 6.China - Economic policy - 2000I.Bao, Shuming II. Lin, Shuanglin III. Zhao, Changwen 337.5'1 Library of Congress Cataloging-in-Publication Data The Chinese economy after WTO accession / edited by Shuming Bao, Shuanglin Lin, and Changwen Zhao. p. cm. -- (The Chinese economy series) Includes index. ISBN 0-7546-4482-0 1. China--Economic conditions--2000- 2. China--Economic policy--2000- 3. Privatization--China. 4. Globalization--Economic aspects--China. 5. World Trade Organization--China. 6. China--Foreign economic relations. I. Bao, Shuming. II. Lin, Shuanglin. III. Zhao, Changwen. IV. Series. HC427.95.C486 2006 330.951--dc22 2005035223 ISBN-13: 978-0-7546-4482-8 ISBN-10: 0 7546 4482 0 Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire.

Contents List of Figures List of Tables List of Contributors Introduction Shuming Bao, Shuanglin Lin and Changwen Zhao PART I

Globalization and Privatization

Chapter 1

A New Round of Economic Growth: Causes, Mechanisms and Characteristics Shijin Liu

vii ix xii 1

9

Chapter 2

China One Year After Its WTO Entry Aimin Chen

17

Chapter 3

WTO and Private Enterprises: A Case Study of China Feiyue Shunfeng Song and Hong Cheng

33

Chapter 4

Local Government and Private Sector Development Yifan Zhang

45

Chapter 5

Globalization and Privatization: Evidence from China Jian Su

63

PART II

Fiscal Policy Reform and Financial Development

Chapter 6

China’s Capital Tax Reforms in an Open Economy Shuanglin Lin

89

Chapter 7

On the Intertemporal Sustainability of Fiscal Debt Ying Wu

111

Chapter 8

Sequencing Domestic Financial Reform: Country Experiences and China’s Roadmap Yanqing Yang

125

vi

The Chinese Economy after WTO Accession

Chapter 9

Public Venture Capital and Its Private Strategies in China Changwen Zhao, Shuming Bao and Chunfa Chen

Chapter 10

Financial Development and Urban-Rural Income Disparity in China Qi Zhang, Mingxing Liu, Yiu-Por Chen and Ran Tao

155

181

PART III

International Trade and Industrial Development

Chapter 11

China in the Antidumping War Against China Jason Z. Yin

209

Chapter 12

Price Competition in the Chinese Pharmaceutical Market Y. Richard Wang

225

Chapter 13

The Performance of Commodity Futures Markets: A Comparison of China and US Wheat Futures Wen Du and H. Holly Wang

237

Home Market Effect and Its Impact on Production and Trade: An Empirical Study of China and the US Fan Zhang and Zuohong Pan

257

Chapter 14

PART IV

Economic Performance and Labor Market

Chapter 15

Comparing Selected Aspects of Economic Performance: China versus India Peter E. Koveos and Yimin Zhang

273

Low Wage and Low Labor Standards in China: A Substitute Explanation of ‘The Race-to-the-Bottom’ Sheng Li

291

Chapter 16

Chapter 17

Migration and Regional Development in China Shuming Bao, Anqing Shi and Jack W. Hou

Chapter 18

Taiwan Cross-Strait Economic Relations in the Era of Globalization Yongjun Chen

Index

307

335

351

List of Figures Figure 1.1 Figure 1.2 Figure 1.3

Changes in the Diffusion Index of the Growth of Chinese Industry 10 Monthly Growth Indices of Major Industrial Output (2001–2002) 11 Monthly Outputs of Some Major Industries (1999–2002) 12

Figure 6.1

Foreign Capital Actually Utilized by China

Figure 7.1

The Long-run Equilibrium Relationship and Short-run Adjustment (Unrestricted Co-integration) The Long-run Equilibrium Relationship and Short-run Adjustment (Restricted Co-integration)

Figure 7.2

Figure 10.1 Figure 10.2 Figure 10.3

Figure 12.1 Figure 12.2

97

119 121

Urban-Rural Income Inequality Across Provinces (1978–1998) 183 Urban-Rural Income Inequality and Per Capita Income (Full Sample, 1978–1999, Average Values) 183 Economic Development and Financial Intermediation Development 188 Distribution of Global (N=229) and Local (N=983) Products by Main Therapeutic Class (1-digit ATC code) Distribution of Global and Local Product: Observations by Quarter (1999–2002)

230 230

Figure 13.1 Figure 13.2 Figure 13.3

Annual Wheat Futures Trading in CZCE (1993–2002) 238 CZCE and CBOT Wheat Futures Price for September Contract 244 Squared First Difference of CZCE and CBOT Wheat Futures Prices 245

Figure 15.1 Figure 15.2 Figure 15.3 Figure 15.4 Figure 15.5 Figure 15.6 Figure 15.7 Figure 15.8 Figure 15.9 Figure 15.10

Per Capita GDP in US$ Per Capita GDP Annual Growth Rate PCPDI of China and India Annual Growth Rate of PCPDI PCLSI of China and India Annual Growth Rate of PCLSI Comparison between China and India Relative Per Capita Production Index Relative Per Capita Living Standard/Consumption Index Time Lag of PCPDI and PCLSI Index, India and China

275 276 278 278 278 279 279 281 281 281

viii

The Chinese Economy after WTO Accession

Figure 16.1 Figure 16.A-1 Figure 16.A-2 Figure 16.A-3 Figure 16.A-4 Figure 16.A-5

Structure of China’s Segmented Labor Market Scales of China’s Segmented Labor Markets (unit: million) Wages of Different Sectors (unit: yuan) Labor Market Scale in Industry (unit: million) Labor Market Scale in Construction (unit: million) Labor Market Scale in Wholesale, Retail Trade and Catering Services (unit: million) Figure 16.A-6 Comparison of Urban Low-end Labor Market and Rural Market (unit: yuan)

305

Figure 17.1 Figure 17.2

318 319

The Percentage of Net Migration, by Province (1995–2000) The Percentage of Profession Migration by Regions

296 303 303 304 304 305

List of Tables Table 1.1

Table 2.1 Table 2.2

2002 Diffusion Indexes for 18 Major Industrial Branches and Their Contributions to Industrial Growth

11

Table 2.3

Foreign Direct Investment and Foreign Trade in China Selected World’s Highest Productions of Manufacturing Goods by China (2002) Assessment of Sectoral Impact of WTO Entry after One Year

23 29

Table 3.1 Table 3.2 Table 3.3 Table 3.4

China’s Export of Sewing Machines (1992–2000) Major Indicators of Feiyue (1998–2002) Overseas Subsidiary Companies of Feiyue International Expos Feiyue Participated In (2001–2002)

35 36 40 41

Table 4.1 Table 4.2 Table 4.3

Private Sector Development in China (1992–2002) Percentage of Private Sector in Total Labor Force by Region Private Sector Development and Regional Characteristics (1997–2001) Summary Statistics of the Variables Correlation Patterns PRIVATE Regression I PRIVATE Regression II

46 47

Table 4.4 Table 4.5 Table 4.6 Table 4.7 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6

Number of LMIEs and SOEs (1995–2000) The Channels of Globalization and Data Availability Descriptive Statistics for Variables Used in the Regressions Results with One-year-lag Data Results with Four-year-average Data The Economic Signiﬁcance of the Effects of Globalization on Probability of Privatization Table 5.7 Regression Result with Dummy Variables of Globalization Table 5.8 The Impact of Privatization on Globalization Table 5.9 A Hypothetical Firm Table 5.A-1 Concordance of Ownership Classiﬁcations (1992–2001) Table 5.A-2 Industrial Code Table 6.1

Types of Revenue as Percentage of Total Central Government Revenue

19

55 56 56 57 58 66 68 72 73 76 77 78 79 81 84 86

94

x

Table 6.2

The Chinese Economy after WTO Accession

Table 6.3

Tax Revenues from Foreign-Investment Enterprises and Foreign Investment in 1998 Sources of Foreign Investment to China (US$ million)

Table 7.1 Table 7.2 Table 7.3

Interest Rates and GDP Growth Rates in China (1990–2000) Augmented Dickey-Fuller Test for Unit Roots* Johanson Cointegration Test

115 117 118

Table 8.1 Table 8.2 Table 8.3 Table 8.4 Table 8.5 Table 8.6 Table 8.7

NPL Disposition and Asset Recovery (December 2001, RMB bn) China’s Central Budgetary Revenue/Expenditure (% of GDP) Contingent Liabilities of China’s Government China’s Liabilities versus Assets China’s AMCs: Policy-based NPL Transfers (1999–2000) Percent of Reported CAR in SOCBs (1998–2004) Initial Condition of Financial Reform

129 130 131 132 133 134 145

Table 9.1

Venture Capital Pools and Disbursements in China by Year (in $US millions) Venture Capital Investments in China by Sector (1998–2002, in $US millions*) Volume of IFSTF in China Ways to Use Innovation Funds for Small Technology-based Firms Provincial and Municipal Venture Capital Funds in China (in $US millions)

Table 9.2 Table 9.3 Table 9.4 Table 9.5

Table 10.1 Table 10.2 Table 10.3 Table 10.4 Table 10.5 Table 10.6 Table 10.7 Table 10.8 Table 11.1 Table 11.2 Table 11.3 Table 12.1 Table 12.2

Financial Development Comparison Across Countries (%) How Do the Four State-Owned Commercial Banks Matter? Banking Market Structure in China (%) Descriptive Statistics (Average Values, 1978–1998) Financial Intermediation and URID: Empirical Result I Financial Intermediation and URID: Empirical Result II Financial Intermediation and URID: Empirical Result III Financial Intermediation and URID: Empirical Result IV Financial Intermediation and URID: Empirical Result V

96 98

161 162 165 166 168 185 186 192 193 195 197 199 201

Antidumping Cases in which China is Involved (1/1/1995– 12/31/2003) 210 Antidumping Cases for Major Reporting Parties and Affected Parties (1/1/1995–12/31/2003) 211 AD Measures Taken against China, by Sector (1/1/95–12/31/03) 214 Characteristics of Global and Local Product Observations Generalized Linear Model Results on the Determinants of Global and Local Product Price

231 232

Migration and Regional Development in China

Table 12.3 Table 12.4

Characteristics of 1996 and after Global Product Observations Generalized Linear Model Results on the Determinants of Global Product Price for 1996 and after Global Products

xi

232 233

Table 13.1 Table 13.2 Table 13.3

Estimates of Selected Univariate ARCH/GARCH Models 247 Estimates of Selected Multivariate ARCH/GARCH Models 249 Within and Cross Market Effects of Bivariate BEKK–ARCH(1) 250

Table 14.1 Table 14.2 Table 14.3 Table 14.A-1

Sector Level Production on Factor Endowments Sector Level Production on Economic Geography Disaggregated Estimation on Industry Level Sector Classiﬁcation

264 265 266 269

Table 15.1 Indices Selected for PCPDI and PCLSI Classiﬁed by Group Table 15.2 Growth Competitiveness and Business Competitiveness Table 15.A-1 The GDP Growth Rate Contributed by Investment, Consumption and Net Export

277 283

Table 16.1

Wage Regression Results

300

Table 17.1

Migration Flows between City, Town and Rural Areas (1995–2000) The Intra-province and Inter-province Migration Flows Changes in the Inter-province Immigration Flows: By Destination Region Changes in the Inter-province Emigration Flows: By Originating Region Net Migration by Regions (in 10,000 persons) Inter-province Net Migration Flows between Regions (1995–2000) Inter-province Migrants as Percentage of Total Migrants Distribution of Migrants by Occupation and Region Migration (both inter- & intra-province) Motivation by Gender (1985–90) Migration (both inter- & intra-province) Motivation by Gender (1995–2000) The Migration Preference Index by Province The 2000 Population and Migration by Region OLS Estimates of the Modiﬁed Narayana Model

Table 17.2 Table 17.3 Table 17.4 Table 17.5 Table 17.6 Table 17.7 Table 17.8 Table 17.9a Table 17.9b Table 17.10 Table 17.11 Table 17.12

288

312 313 314 314 315 316 317 320 321 321 323 325 327

List of Contributors Shuming Bao University of Michigan Bao received his Ph.D. in applied economics from Clemson University in 1996. He was a research scientist at MathSoft from 1996 to 1997, and is currently the senior research coordinator for China initiatives of the International Institute and the senior research associate of the China Data Center at the University of Michigan in Ann Arbor. His primary research interests are in GIS, regional economics, spatial statistics and econometrics. Aimin Chen Sichuan University Chen is currently Vice President of Sichuan University and professor of economics at Indiana State University. Chunfa Chen Sichuan University Chen is currently a Ph.D. candidate in Finance in the Business School, Sichuan University. He earned his B.A. and M.A. from Southwest China Normal University and got his MBA from the University of Illinois at Chicago. His research focuses on venture capital and mutual funds. He is now a professor of English at Sichuan University. Yiu-Por Chen Brown University Chen is a research fellow at Population Studies and Training Center, Brown University. He received his Ph.D. from Columbia University. Dr. Chen’s areas of interest include urbanization, labor migration, the political economy of development, and institutional analysis. Yongjun Chen Xiamen University Chen received his Ph.D. in Economics from Xiamen University in 1992. He is currently a Professor and the Deputy Dean of the Graduate School of Xiamen University. He was the Director of Social Science Research Administration Division of Xiamen University in 1998-2001, a visiting scholar at Stanford University in 2000, and a visiting scholar at John Hopkins University in 2004-2005. Hong Cheng Columbia University Cheng is a MPA student at the School of International and Public Affairs of Columbia University. He received his M.S. in Economics from the University of Nevada, Reno. He was Executive President of Zhejiang Provisional Student Federation during his senior year in college in China.

List of Contributors

xiii

Wen Du Washington State University Du received her Ph.D. from the Department of Agricultural and Resource Economics at Washington State University. She was a Sylvia Lane Research Fellow at the Department of Agricultural and Consumer Economics, University of Illinois at Urbana-Champaign in the summer of 2002. Jack W. Hou California State University, Long Beach Hou earned his B.A. from National Taiwan University and his Ph.D. from Yale University. His research focuses on the Chinese economy (and economic history), labor economics, and international trade/ﬁnance. He has published many articles and written several book chapters on these subjects. He has been a visiting professor at Purdue University, Soochow University, and most recently at UCLA. He is currently a member of the Board of Directors for the Chinese Economists Society, and on the Executive Council of the Western Social Science Association, as well as an ofﬁcer in the Western Economic Association. Peter E. Koveos Syracuse University Koveos received his Ph.D. in economics from Pennsylvania State University in 1977. He is currently a professor of ﬁnance at the School of Management of Syracuse University. He serves as the Chair of the Finance Department of Syracuse University (1986-88, 1991-97 and 2002-present), Senior Director for International Programs of Syracuse University (2001-present), Associate Dean for Master’s Programs of Syracuse University (1997-2001), Director of Kiebach Center for International Business Studies of Syracuse University (1990-present) and Interim Dean of the School of Management of Syracuse University (1988-90). His primary research interests are in Finance, International Finance and Global Economy. Sheng Li University of Utah Li is a Ph.D. candidate at the University of Utah. She earned her B.A. from University of Electronic Science and Technology of China, and her M.A. from Central University of Finance and Economics. Her research concentrates on theory and policy, the construction industry, and the Chinese economy. She has published in journals, including Finance Studies, Investment Research, and China Soft Science. She received a graduate honor roll scholarship from the University of Utah in 2002. Shuanglin Lin University of Nebraska at Omaha Lin is Lindley Professor of economics at the University of Nebraska at Omaha, Research Associate of the East Asian Institute of National University of Singapore, and Guest Professor of Peking University. He has published numerous journal articles and book chapters on public ﬁnance, economic growth, and Chinese economy. He was President of the Chinese Economists Society (2002/2003).

xiv

The Chinese Economy after WTO Accession

Mingxing Liu Peking University Liu is an assistant professor at the School of Government, Peking University. He received his Ph.D. from Peking University. Dr. Liu’s research interest includes rural governance, and the political economy. Shijin Liu The Development Research Center of the State Council Liu received his Ph.D. from the Chinese Academy of Social Science and is a Senior Research Fellow in The Development Research Center of the State Council (DRC) since 1994. He was the Deputy Director-General of the Institute of Market Economy and the Research Department of Macro-economy. He currently holds the positions of Director-General of the Research Department of Industrial Economy, DirectorGeneral of the General Ofﬁce and Secretary-General of the Academic Committee of DRC. Zuohong Pan Western Connecticut State University Pan received his Ph.D. in economics from Wayne State University in 1995. He is currently an associate professor at the Department of Social Sciences of Western Connecticut State University. He teaches macroeconomics, microeconomics, intermediate macro/micro, money and banking, mathematical economics, economic development, economic applications on the Internet, contemporary domestic economic issues, and applied econometrics. He has published a number of papers in Urban Studies, Computational Economics, Journal of Computational Intelligence in Finance, and other journals. Anqing Shi The World Bank Shi is a Researcher in the Development Research Group of The World Bank. Prior to this he was a visiting scholar at the Institute of Asian Studies at Columbia University, New York. He received UN Population Fellowship and was elected as a member of International Union of Scientiﬁc Study of Population. One of his research areas is demographic dynamics and sustainable development. He has published papers in this area. Shunfeng Song University of Nevada, Reno Song received his Ph.D. in economics from the University of California at Irvine in 1992. He is currently the Chair of the Department of Economics at the University of Nevada, Reno. Dr. Song has had his work published in the Journal of Political Economy, Journal of Urban Economics, Urban Studies, Land Economics, Contemporary Economic Policy, China Economic Review, Current Anthropology, Review of Regional Studies, Socio-economic Planning Sciences and Journal of Quantitative Anthropology. He was the recipient of the Best Research of the Year Award in 1995 from the Alpha Chapter of BETA GAMMA SIGMA, which is the highest honor society for collegiate schools of business in the United States. He was a director/vice-president of the Chinese Economists Society in 1996–1998. Dr. Song is included in Who’s Who in the World, 17th edition, 2000.

List of Contributors

xv

Jian Su Peking University Su received his Ph.D. in international economics and ﬁnance from Brandeis University in 2004. He is currently an associate professor and Deputy Chair of the Economics Department of Peking University. His primary teaching and research interests are in macroeconomics, development/transition and international trade. Ran Tao China Center for Agricultural Research, Chinese Academy of Sciences Tao is a research fellow at China Center for Agricultural Policies, Chinese Academy of Sciences. He received his Ph.D. from Chicago University. His research interest includes rural economic development and growth, rural governance and institutional analysis. H. Holly Wang Washington State University Wang received her Ph.D. in agricultural economics from Michigan State University in 1996. She is currently an associate professor at the Department of Agricultural and Resource Economics of Washington State University. Her research interests are in risk management in agriculture and agricultural economics in China, and spatial statistics applications. Y. Richard Wang AstraZeneca Pharmaceuticals LP Wang received his Ph.D. in health economics from the Wharton School of University of Pennsylvania in 1999. He is currently a Director of Health Services Research and Policy Analysis at the Public Policy Department of AstraZeneca Pharmaceuticals. He has been an adjunct senior fellow at Leonard Davis Institute of Health Economics of the University of Pennsylvania from 1999 to present. Ying Wu Salisbury University Wu received her Ph.D. in economics from the University of Oregon in 1992. She is currently an associate professor of economics at the Franklin P. Perdue School of Business of Salisbury University. Her specialized ﬁelds are macroeconomics and monetary economics, open-economy macroeconomic issues, and international economics. Yangqing Yang Johns Hopkins University-SAIS Yang received her Ph.D. in economics from Fudan University in 1998. She was a senior reporter of Jiefang Daily from 1998-2002, a visiting scholar at the Johns Hopkins University in 2002-2003 and is currently a leading member at the Preparatory Ofﬁce of National Economic Review. Jason Z. Yin Seton Hall University Yin is a professor at the Stillman School of Business of Seton Hall University. He served as the president of the Chinese Economists Society in 2002. Dr. Yin has published intensively in areas of strategic and technology management, WTO and international trade and business environment in China.

xvi

The Chinese Economy after WTO Accession

Fan Zhang Peking University Zhang received his Ph.D. in economics from Wayne State University. He is currently an associate professor of economics at Peking University. His research ﬁelds include industrial organization, international economics and urban economics. Qi Zhang Institute of World Economy & Politics, Chinese Academy of Social Sciences Zhang is an assistant research fellow at Institute of World Economy & Politics, Chinese Academy of Social Sciences. He received his Ph.D. in economics from Peking University. Dr. Zhang’s areas of interest include rural economic development and growth, the political economy of development, and international comparison of development. Yifan Zhang University of Pittsburgh Zhang is a Ph.D. candidate of at the Department of Economics of the University of Pittsburgh. He has received a number of academic awards, including China Council Pre-dissertation Grants, University of Pittsburgh in 2002, Chancellor’s Fellowship, University of Pittsburgh in 1998, and Guanghua Excellence Award, Renmin University of China in 1994, 1995 and 1996. He was a visiting research fellow, Center for China Studies, Tsinghua University, 1999-2000, and a research assistant, University of Pittsburgh for 2002-2004. Yimin Zhang Shanghai University of Science & Technology Zhang is a professor at Shanghai University of Science & Technology and a senior research fellow at the School of Management of Syracuse University. Changwen Zhao Sichuan University Zhao received his Ph.D. in economics from Southwest University of Finance and Economics in 1995. He is the chair professor of Finance at the Business School of Sichuan University, president of the Board of Directors of the High-tech Group of Sichuan University, and vice president of Sichuan University. He was head of the West Development Academy of Sichuan University (2000-2002), deputy dean of the Business School of Sichuan University (2000-2002), and a Fulbright visiting scholar at the International Institute of the University of Michigan in 2002-2003.

Introduction Shuming Bao, Shuanglin Lin and Changwen Zhao

More than a century ago China’s door was opened to the world. In 1948 China became one of the twenty-three original signatories of the General Agreement on Tariffs and Trade (GATT). In 1949 the People’s Republic of China was established and the Nationalist government retreated to Taiwan. China became un-uniﬁed. The government in Taiwan later announced that China would leave the GATT system. The Beijing government never explicitly recognized this withdrawal decision, but had no contact with the GATT. For many years after 1949, China emphasized selfsufﬁciency and the Chinese economy was essentially closed. In 1978 China started economic reforms and voluntarily opened its door to the entire world. Since then, the Chinese economy has been growing at a rapid rate and international trade and foreign investment has increased steadily. In 1986, China informed the GATT of its wish to resume its status as a GATT member and negotiations soon began. In 1995 the GATT was replaced by the World Trade Organization (WTO). After 15 years of extensive negotiations, China entered the WTO on 11 December 2001. The accession of China to WTO brought many opportunities as well as new challenges to China and the world. After accession, China would have a lower cost of accessing foreign markets, a more equitable trading mechanism based on WTO rules and multilateral dispute resolution processes, more foreign investment and advanced foreign technology, and a legitimated role in setting and enforcing international trading rules. However, according to the agreement, China must undertake a series of important commitments to open its markets and to offer a more predictable environment for trade and foreign investment. The responsibilities include: providing non-discriminatory treatment to all WTO members, eliminating dual pricing practices as well as differences in treatment accorded to goods produced for sale in China in comparison to those produced for export, preventing price controls to be used for the purpose of protecting domestic industries or services, revising its existing domestic laws and enacting new legislation fully in compliance with the WTO Agreement, eliminating all export subsidies on agricultural products, implementing the Trade-related Aspects of Intellectual Property Rights Agreement, eliminating trade barriers and expanding market access to goods from foreign countries, and lowering average tariff level. Many believed that the WTO entry would beneﬁt China in the long run, but would hurt the Chinese economy in the short run, particularly in the agriculture and service areas. The results of China’s WTO entry are surprising. China’s agricultural industry held up after the WTO entry, despite predictions that it would be severely battered by increased competition. In the ﬁrst year of WTO entry, the value of exported

2

The Chinese Economy after WTO Accession

agricultural goods increased by 11.5 percent, while imports fell by 0.4 percent and the trade surplus rose by 52.3 percent. The Chinese economy grew at a rate of 7.8 percent in 2002, and 9.3 percent in 2003. Apparently, the WTO entry appears to have beneﬁted China even in the short run. However, it would be a mistake to underestimate the challenges of globalization. China must phase in most of its WTO commitments within the ﬁrst ten years after WTO entry. In agriculture, China must limit its subsidies for agricultural production to 8.5 percent of the value of farm output. In the telecommunication industry, China must permit foreign companies to establish joint venture enterprises, with foreign investment share being no more than 35 percent within one year of accession, no more than 49 percent within three years, and no geographic restrictions within ﬁve years. In the banking industry, foreign ﬁnancial institutions are permitted to provide local currency services to Chinese enterprises within two years of accession, and to provide services to all Chinese clients within ﬁve years. In the insurance industry, foreign non-life insurers have been permitted to establish as a branch or as a joint venture with 51 percent foreign ownership, to establish as a wholly-owned subsidiary within two years; and foreign life insurers should be permitted 50 percent foreign ownership in a joint venture with the partner of their choice. For large scale commercial risks, reinsurance and international marine, aviation and transport insurance and reinsurance, joint ventures with foreign equity of no more than 50 percent are permitted upon accession, increase to 51 percent within three years, and reach 100 percent within ﬁve years. China faces serious problems, including low labor productivity in agriculture, growing unemployment, inefﬁcient SOEs (State Owned Enterprises), an inefﬁcient banking system, growing central and local government debt, reluctant consumers and weak domestic demand, inequality, weak social insurance system, and a weak legal system. China is also under increasing pressure to revaluate its currency. At a crucial moment when many problems related to China’s WTO entry began to surface, the Chinese Economists Society (CES) organized an international conference on ‘Chinese Economy after WTO: Opportunities and Challenges of Globalization’ at the University of Michigan in Ann Arbor, Michigan on 2-3 August 2003 More than two hundred CES members and other scholars from Europe, Asia, and North America attended the conference, and over sixty papers were presented. These papers evaluate the impact of China’s WTO entry on the Chinese as well as on the world economy, analyze the challenges and opportunities faced by China and other countries, discuss China’s current economic problems, and provide urgently needed policy recommendations to policy makers. This volume consists of eighteen selected papers from the conference. All of these papers were reviewed anonymously by scholars in the ﬁeld. The authors then revised their papers following the comments made by the reviewers. The volume provides a remarkable background of information and ideas about Chinese economy after WTO and analyses of many important issues concerning China’s foreign trade, ﬁscal and ﬁnancial reforms, privatization, migration, and regional development. The volume is

Introduction

3

an indispensable source for scholars and students interested in Chinese economic studies and many chapters should also be of interest to a wide range of readers. An Overview This book is organized in four parts. The ﬁrst part addresses globalization and privatization. The second part deals with ﬁscal policy reform and ﬁnancial development. The third part focuses on international trade and industrial development. The fourth part is devoted the labor market and economic integration. The ﬁrst chapter in Part I by Shijin Liu gives an observation and analysis on a ‘turning point’ of Chinese economy after WTO accession. After China entered the WTO, the Chinese economy has experienced an accelerated growth. This may be a surprise to some scholars who believed that China might suffer in the short run and raised some concerns about the sustainability of the growth of the Chinese economy. From an insider’s viewpoint, Shijin Liu describes the major characteristics of the new round of economic growth under globalization, analyzes the driving forces and mechanisms for the high growth, and discusses the new pattern of economic development in China. The second chapter analyzes the impact of the WTO entry on China after its ofﬁcial accession based on empirical observations. By comparing the industrial structure and expected impact prior to the entry, Aimin Chen ﬁnds that while China’s agriculture sector has dodged the expected squeeze in the ﬁrst year, the manufacturing sector has gained from the entry and the service sector calmly allowed unprecedented breaking into many of its industries. The entry has led to institutional and systemic changes that have made China a more open society and an environment more conducive to the development of a private market economy. In the third chapter, Shunfeng Song and Hong Cheng present a case study of a private company that manufactures sewing machines in Zhejiang province. China produces more than half of the world’s total sewing machines with an export-production ratio of 0.82 in 2002. With WTO, China will see a large increase in the demand for its sewing machines in the international market as textile exports increase due to lower tariffs and elimination of export quotas. Meanwhile, China will also face more competition as more foreign brands ﬂow into the Chinese market. China’s entry into WTO poses challenges and opportunities to Chinese enterprises in the sewing machine industry. Based on his ﬁeld survey results, Shunfeng provides insights on how private enterprises, export-oriented companies, and traditional producers react to changes resulting from China’s accession into WTO. The next chapter, by Yifang Zhang, discusses the impacts of local government reform on heterogeneous development of private sectors in China. Yifang examines the relationship between local government and private sector in China based on the existing theories on economic transition and recent observations. The results suggest that government intervention, law enforcement and government policy all have strong and enduring effects on private sector development while the evidence on the inﬂuence of ﬁscal autonomy is relatively weak. In chapter 5, Jian Su examines the interaction between

4

The Chinese Economy after WTO Accession

globalization and privatization with evidence from China. Jian Su identiﬁes three channels of globalization: the outﬂow of goods, the inﬂow of foreign capital (FDI), and the inﬂow of knowledge. By building some regression models with pooled datasets, Jian Su ﬁnds that globalization and privatization are mutually reinforcing. With the WTO entry commitment, China is under increasing pressure to continue its ﬁnancial system reform. Part II discusses ﬁscal sector reform in China. In chapter 6, Shuanglin Lin analyzes the effects of China’s upcoming capital tax reform of switching from a dual tax system to a uniﬁed system. Shuanglin found that a decrease in the tax rate on domestic capital has no effect on the domestic interest rate, capital-labor ratio, or output-labor ratio, and leads to an increase in domestic capital, a decrease in foreign capital, and an increase in the trade surplus. An increase in the tax rate on foreign capital increases the domestic interest rate and decreases the capital-labor ratio, the output-labor ratio, domestic capital, as well as foreign capital and the trade surplus. In chapter 7, Ying Wu analyzes the ﬁscal sustainability of China’s government debt with the 1979–2001 budgetary data. Applying the present value budget constraint and performing a co-integration analysis for public expenditures and tax revenues, Ying ﬁnds that the government expenditures and tax revenues in China are co-integrated and the co-integration vector is signiﬁcantly close to the theoretical value for ﬁscal sustainability. The tax revenue plays a key role in sustaining ﬁscal debt in the sense of the present value budget constraint. The analysis also shows that high GDP growth relative to the interest rate favors the sustainability of China’s government debt over the long run in spite of recent growing budget deﬁcits and deﬂation. Although the importance of China’s ﬁnancial reform has been widely realized, there have been many discussions about the sequence of its ﬁnancial reform for adaptation of WTO regulation. Before China’s ﬁnancial system is integrated into the global ﬁnancial market, it has to recapitalize banks and liberate them from the huge burden of nonperforming loans (NPL). In chapter 8, Yanqing Yang suggests a radical approach and illustrates several avenues to solve China’s NPL problem. A set of preconditions needs to be met before full-ﬂedged capital mobility. In coordination with interest rate liberalization, capital account liberalization is likely to initiate around 2008–2010. In a comparison of public venture capital and private venture capital, Changwen Zhao, Shuming Bao and Chunfa Chen discuss the different roles and impacts of private and public venture capitals in ﬁnancing the high-tech businesses in chapter 9. The results show that high-tech development in China suffers from a lack of venture capital, both public and private, and that public venture capital can ﬁll the gap between the demands for and supplies of venture capital and help improve the investment climate for private venture capital, which may be much more important than the ﬁnancial transfer from government. In chapter 10, Qi Zhang, Mingxing Liu, Yiu-Por Chen, and Ran Tao analyze how ﬁnancial intermediation development affects urbanrural income disparity in China. Zhang et al. demonstrate that rural ﬁnancial systems were inefﬁcient in allocating funds for supporting rural economic development during the early period of the market reform (1978–88), which was primarily affected by ﬁscal policy, and the ﬁnancial intermediation signiﬁcantly manifested its effect

Introduction

5

during the second period of the reform (1989–98). By introducing endogeneity, the robust results were estimated to show that the urban-rural income disparity does not depend upon the sector structure in the province, which is consistent with traditional urban-bias hypothesis. Part III discusses the challenges for some industrial sectors in China. In chapter 11, Jason Z. Yin addresses the antidumping war against China. He conducted a survey of the WTO antidumping data, analyzes the structure and characteristics of all antidumping cases against Chinese enterprises, and offers some strategic and policy recommendations for the Chinese government and its enterprises to better prepare for the discriminatory provisions in China’s WTO accession protocol and for the escalating antidumping trade disputes in protection for its economic and trade interests. In chapter 12, Y. Richard Wang studies the international competition in the pharmaceutical market between global products and local products in China. Wang ﬁnds that the price of global products is less responsive to local generic competition than the price of local products, and his results suggest a highly competitive local generic market while the evidence on therapeutic competition is not conclusive. The agriculture future market is new in China, which can have an important impact on the future development of the agriculture sector in China. In chapter 13, Wen Du and H. Holly Wang compare the wheat market in the US and China in an analysis of the performance of commodity futures markets in China. The China Zhengzhou Commodity Exchange (CZCE) has been in stable development since its establishment and was expected to be integrated in the world market after China joined WTO. Du and Wang compare the price behavior of CZCE with that of the Chicago Board of Trade (CBOT) in the US by using univariate and multivariate time series methods in their price models. Their results show both markets can be modeled by an ARCH or a GARCH model and the models have a better ﬁt when conditional error variance is t distributed. The interrelations between the two markets are signiﬁcant but asymmetric. CBOT holds a dominant position in the interactions while CZCE is more like a follower. In chapter 14, Fan Zhang and Zuohong Pan analyze the border effect of WTO on China’s trade. China’s entry into WTO will signiﬁcantly reduce its trade barriers to the outside world. Using a theoretical gravity equation developed by Anderson and Van Wincoop, Zhang and Pan estimate the trade barriers between non-WTO member countries and WTO member countries. Their results suggest that trade barriers reduced trade by an average of 35 percent before a country like China enters into WTO, and entry into WTO will remove this barrier in a few years and increase trade by about 54 percent. Part IV is devoted to the issues on international economy and labor market in China. In chapter 15, Peter E. Koveos and Yimin Zhang compare several selected aspects of economic performance between China and India, which are experiencing similar transitions in their development. China and India have made signiﬁcant changes to their economies, as they are making the transition to a market-based economic framework. Meanwhile, China and India are attaining considerable stature within the global economy. In a comparative analysis, Peter and Yimin construct two performance indices by proposing a per capita production index for measuring a

6

The Chinese Economy after WTO Accession

country’s output and a per capita living standard index for measuring a country’s consumption activities. They ﬁnd that China outperforms India by using both indices and the country ranking provided by the World Economic Forum. In chapter 16, Sheng Li investigates the determinants of low wage and labor standard in China by analyzing the segmented labor markets and immigration from rural to urban. The special market structure shows by itself that the low wage and labor standard are endogenous and naturally determined by the agricultural sector. Wages may continue to drop once no entry barrier exists between rural and urban areas. The low labor cost in China is decided by its enormous labor supply instead of any external measure. In the analysis of the 2000 population Census data, Shuming Bao, Anqing Shi and Jack W. Hou conducted a comprehensive study on migration in China, which is presented in chapter 17. They analyze the changing patterns of migration in the last 30 years, and compare changes in spatial tendency of migration by different regions, such as changes in migration distance, places of origination and destination, and regional variations. By applying Roberto Bachi’s Migration Preference Index for inter-provincial migration, They investigate the changes in the migration patterns in western regions and other regions, built a migration model based on some selected factors and discuss the policy implications for regional development of China. The WTO could have a signiﬁcant impact on the relationship between mainland China and Taiwan. The ﬁnal chapter, by Yongjun Chen, discusses the major factors that may affect economic relations between the two sides of the Taiwan Strait in the new era of economic globalization. The chapter examines the trend of development of cross-strait economic relations, the structural changes in the two sides of the Taiwan Strait, and gives suggestions on how to enhance economic cooperation and development between them. Acknowledgments We thank all the participants of the Chinese Economists Society conference, ‘Chinese Economy after WTO: Opportunities and Challenges of Globalization,’ for their contributions and involvement. We appreciate the help from those who have offered their research assistance and editorial assistance in preparing the book for publication. We are grateful to Sichuan University Business School, the University of Michigan International Institute, the Ofﬁce of Vice President for Research, the Center for Chinese Studies, China Data Center, Business School, the Department of Economics, and the Williams Davidson Institute for their ﬁnancial support. Finally, we would like to express our special thanks to Terre Fisher, Angela Kuhlmann, Jackie Lynch, Miao Wang, Tingting Song, Yan Guan and Li Zheng for their editorial assistance.

PART I Globalization and Privatization

This page intentionally left blank

Chapter 1

A New Round of Economic Growth: Causes, Mechanisms and Characteristics Shijin Liu

The Chinese economy has picked up speed once again since 2002, the year after China acceded to WTO. In the ﬁrst half of 2003, the growth rate exceeded 8 percent despite the adverse impact of the SARS rampage. After several years of low growth, especially after China’s accession to WTO, a ‘turning point’ occurred in the economic growth. This situation has caught some people by surprise, and raised doubts about the sustainability of this relatively fast growth rate. It is, therefore, necessary to clearly explain the causes, mechanisms and salient features of the current new round of quick economic growth. New Pattern for Industrial Growth: The Rise of High-Growth Industries According to the results of the ‘Follow-Up Study of the Development of the Chinese Industry’ conducted by the Department of Industrial Economics of the State Council Research Centre, China’s industrial growth has assumed some new salient features beginning from 2002. Firstly, a striking gap has occurred in the growth of different industries. A number of new high-growth industries have emerged and are going strong. Figure 1.1 sheds some light on the trend of the diffusion index of the growth of the Chinese industry. Figure 1.1 shows that in the period from 1998 to the ﬁrst half of 2001, the industrial growth’s diffusion index was ﬂuctuating basically between 110 and 90, and growth was by and large balanced and stable, with none of the industrial branches standing out in performance. With the average level of the industrial growth’s diffusion index recovering steadily since the year-end of 2001, the gap between high-growth industries and low-growth industries has been widening. Table 1.1 gives some idea about changes in the average boom level and income level of 18 major industrial branches as well as in their contributions to industrial growth as a whole. As Table 1.1 indicates, 16 of the 18 industrial branches did better than the previous year while the automobile, coal-mining, food-processing and machinebuilding industries posted the best records. The boom levels of these four industries went up by more than 10 percentage points as compared with 2001. They accounted for 27.6 percent of the total industrial sales revenue but their rate of contribution

10

Figure 1.1

The Chinese Economy after WTO Accession

Changes in the Diffusion Index of the Growth of Chinese Industry

to the total industrial sales revenue went as high as 34.6 percent, indicating that they were the most important factors in boosting the industrial growth of 2002. The boom level for the pharmaceutical, building materials, power, textile and apparels, iron and steel, chemical and electronics industries was good, while gas and tap water supplies, papermaking, tobacco, household electric appliance and drinks and beverages industries were at an average boom level and remained more or less the same as the previous year. Secondly, the industrial growth has shown a striking relevance effect and group characteristics. A relevance effect with a well-deﬁned logic relationship has been observed among the high-growth industries. On this basis, a number of major highgrowth industrial clusters have emerged. One such cluster is headed by the automobile industry and includes the synthetic materials industry, the tire manufacturing industry, the iron and steel industry (devoted mainly to the production of thin steel plates and steel strips for the making of automobiles), and a branch of the machine-building industry devoted to the building of machine tools in general and digital-controlled ones in particular. Figure 1.2 is an indication of the relevance of the automobile industry to the above-mentioned industrial branches. Moreover, there is also a high-growth industrial cluster headed by the real estate industry, a high-growth industrial cluster headed by the machine-building industry, and a high-growth industrial cluster for the manufacturing of consumer goods that do not include housing and automobiles. The macroeconomic growth rate will go up with the strength of these highgrowth industrial clusters. This gives rise to another question: How did this pattern of economic growth occur and can this economic growth be sustained? To answer this question it is necessary to analyze and interpret the causes and mechanisms for this new round of economic growth.

A New Round of Economic Growth: Causes, Mechanisms and Characteristics

Table 1.1

2002 Diffusion Indexes for 18 Major Industrial Branches and Their Contributions to Industrial Growth

Industrial Sector

Average diffusion index in 2002

Difference from 2001

Portion in industrial sales revenue

Rate of contribution to industrial growth %

Automobile industry

137.858

36.163

0.057

11.26

Coal-mining industry

118.547

16.514

0.019

2.98

Food-processing industry

116.613

16.167

0.063

6.51

Machine-building industry

110.049

11.161

0.137

13.87

Pharmaceutical industry

109.274

6.755

0.023

2.21

Building materials industry

108.413

9.857

0.042

3.62

Power industry

108.020

7.147

0.090

9.75

Textile & apparels industry

106.194

7.705

0.088

7.93

Chemical industry

105.947

4.596

0.114

10.51

Iron & steel industry

105.339

4.440

0.065

6.22

Electronics industry

104.263

2.725

0.091

13.00

Gas and tap water supplies

103.839

4.569

0.007

0.74

102.945

2.732

0.027

2.53

102.755

3.230

0.020

1.78

102.461

2.540

0.039

3.92

Papermaking and printing industry Tobacco industry Household electric appliance industry Drinks and beverage industry

101.732

0.394

0.019

0.85

Nonferrous metal industry

97.979

-0.967

0.025

1.72

Oil and petrochemical industry

88.429

-10.147

0.074

0.62

Figure 1.2

11

Monthly Growth Indices of Major Industrial Output (2001–2002)

12

Figure 1.3

The Chinese Economy after WTO Accession

Monthly Outputs of Some Major Industries (1999–2002)

How Did the New Round of High Growth Occur? High-speed Economic Growth Depends on New Fast-growing Industries A salient feature of high-speed economic growth lies in the fact that this type of growth is not an outcome of balanced growth of various industrial branches. In each stage of such growth, a particular group of high-growth industrial clusters, or leading industrial clusters, occurs. The industries in these clusters grow much faster than the other industrial branches, thereby boosting the growth of the national economy as a whole. This phenomenon has been borne out more than once by the international experience. The robust growth of the Chinese economy since the adoption of the policy of reform and opening up to the outside world has also been fueled by these fastgrowing industries. In the past two decades and more, the Chinese economy has experienced two rounds of fast growth catalyzed exactly by fast-growing industries. The ﬁrst round, which took place during the early and middle 1980s, drew its impetus from the light and textile industries. The second round of fast growth, revved up by the remarks made by Deng Xiaoping during a south China tour in the early 1990s, was guided by a series of fast-growing industries, including the infrastructure and the basic industries (the construction of highways and harbours, and the power and iron and steel industries), the manufacturing of new-generation household electrical products (color television sets, refrigerators, washing machines and air-conditioners), and the real estate business. The slowdown of the Chinese national economy after 1997 may well be interpreted this way: the wind was taken out of the sails of those high-growth industries that rose in the early 1990s, while no new high-growth industries emerged to take over their position. In other words, the void thus created was not ﬁlled in a timely fashion.

A New Round of Economic Growth: Causes, Mechanisms and Characteristics

13

Major Causes for the Emergence of New High-Growth Industrial Clusters: Collapse of the Bottleneck in the Economic Cycle The void created by the lack of high-growth industries and the economic slowdown indicated that major problems had appeared in certain links in the economic cycle. Because of the interaction of numerous links in the economic cycle, people may have many reasons to account for the economic slowdown. However, it is necessary and possible to pinpoint the links and problems that were responsible for it. Observations of the process of the economic cycle in recent years indicate that the upgrading of the consumption mix of urban dwellers was mainly to blame for the economic slowdown. The difﬁculties confronting the upgrading of the consumption mix of urban dwellers stemmed neither from a potential lack of demand nor from a lack of supply of production elements, but rather, they stemmed from obstacles in the Chinese system and government policies, obstacles that manifest themselves in housing and transportation. The national economy has entered another round of relatively fast growth since the beginning of last year. The most important of all the factors behind this change is that the bottleneck in the effort to upgrade the urban dweller’s consumption structure is being broken through, and that major progress has been achieved in readjusting the relevant systems and policies. Free distribution of housing funded by the government has been abolished while a multiple-level housing market is burgeoning. China’s accession to WTO and the subsequent reduction of tariff on auto imports played only a marginal role in the major development of the Chinese automobile industry in 2002 – the major reason being that the government’s conscious or unconscious decontrol of access to this industry has boosted the number of automakers, facilitated and intensiﬁed competition, increased the variety of products, reduced the prices, thereby enabling consumers to readjust their anticipations and release their purchasing power, and putting automobile consumption and production on a ﬁne cycle of interaction. It is no exaggeration to say that the auto industry’s explosive 2002 growth was a typical case of industrial and economic growth precipitated by government decontrol of the access to the industry. If we bring this chain of events into focus, we can clearly see a logical process in which the bottleneck in upgrading the consumption structure was broken by restructuring the Chinese system and readjusting the government policies, and the upgraded consumption structure catalyzed the emergence of a group of highgrowth industries and ushered the national economy into a new round of fast growth. On this basis, we can come nearer to understanding the real cause behind this new round of fast economic growth. Major Characteristics of the New Round of Economic Growth The new round of economic growth is still in its initial stage. The following are our initial observations of some of its major characters and our judgments on it.

14

The Chinese Economy after WTO Accession

The Growth of the Housing, Automobile, Electronic Telecommunications and Other Consumer Goods Industries is Highly Sustainable because Their Goal is to Satisfy Consumer Demand The products of these industries not only meet the demand of a speciﬁc group of consumers; more importantly they are catered to the snowballing demand of the vast populace. In a given year, only a small high-income portion of the population can afford to buy houses and cars. However, because of the colossal size of the Chinese population, this portion of people is large enough in number to make substantial contributions to economic increments, and, what is more, their numbers are growing. Only by entering the stage of popular consumption can the housing and auto industries truly become mainstays for the growth of the national economy as a whole; only thus can their growth be really sustained. The characteristic of this new industrial cluster to meet the consumer demand of the vast populace can put economic growth on a solid basis and effectively prevent overall and continuous ‘bubbles’ from occurring. The Housing and Automobile Industries that are Leading this New Round of Economic Growth are in a Considerably Long Cycle of Fast Growth According to the experiences of the world’s major industrialized powers, the automobile industry maintains a period of fast growth for 20 to 30 years after it enters the stage of popular consumption. Urbanization is accelerating in this country. Studies indicate that improved housing conditions for urban dwellers and the entry of rural dwellers into cities can precipitate fast growth of the housing industry for at least two decades. Although industries have different cycles of growth, although the factors affecting economic growth are complicated and changeable, and although short-term or accidental factors may cause large or small economic ﬂuctuations, the rise of the housing and automobile industries will keep the Chinese economy in a considerably long period of relatively fast growth. An Extraordinary, Large Economic Scale Commensurate with the Capacity of the Chinese Market is Emerging China has a population of nearly 1.3 billion people, nearly 500 million of whom are urban dwellers. For a population this size to enter the middle stage of industrialization is something never experienced in human history. The nation’s urban population alone is already larger than the population of any industrialized nation. China is leading the world in the output of many traditional industrial and agricultural products. China is also among the world’s largest producers and sellers of some high-tech products, cell phones included. In conjunction with the new round of economic growth, some Chinese industrial products and services will, with their extraordinarily large market capacities, enter the front ranks of the world and become major economic phenomenon. For instance, in the next 10 or 20 years, it will be possible for China to

A New Round of Economic Growth: Causes, Mechanisms and Characteristics

15

become one of the countries in the world to possess and produce the largest numbers of automobiles. There are at least two strengths about this super-large market scale: ﬁrstly, large enterprises can emerge by relying on the domestic market alone; and secondly, several large enterprises that have measured up to the requirements for sizeable economic scales can coexist and form effective competition in the home market. In the New Round of Economic Growth, the Competitiveness of the Chinese Industry will be More Closely Associated with the Industrial Clusters The clustering of industries is nothing new in the process of industrialization. It has been growing robustly in China during the last few years, and particularly so in the developed southeastern coastal regions. This has enabled many industries to vastly cut their costs, thereby considerably raising their competitiveness. Take the Pearl River Delta for instance. In this area with a circumference of 100 kilometres, the purchasing prices for colour television sets, computers and some other products are 30 percent lower than in other regions; as a result, most of the leading enterprises in these industries have set up manufacturing bases in the Pearl River Delta. The new round of economic growth entails a stage in which industrial competitiveness is closely linked with the degree of industrial clustering. That is to say a relatively competitive industry will make its presence felt in a certain region where certain industries are thickly clustered. Interaction between Industry and Urban Industry in the New Round of Economic Growth will Catalyze the Formation and Expansion of a Number of Major Urban Belt Regions Urban development has been a highly controversial issue for many years. It is, however, becoming increasingly clear that the basic trend is not to emphasize the development of a certain type of city, but rather to form and expand several belts or circles each containing a number of cities of different sizes and types, to which the majority of the population and most of the resources will be gravitated. In an experience similar to that of the entire world, several city circles are emerging in the Chinese coastal regions, and so are a number of urban belts coming into being along the trunk transportation lines of the hinterland. The immense demand of the rising high-growth industries for the services industry in general, and the modern productive service industry in particular, will provide a major impetus for urban development. That is why quite a few typical cases have occurred in recent years, in which industries were born along with new cities. Among these cases are the growth of the urban functions of Hong Kong and Shenzhen alongside the development of the electronic telecommunications industry in Dongguan, Guangdong Province, and the development of the urban function of Shanghai in sync with the rise of the electronic telecommunications industry in Kunshan, Jiangsu Province. To seek the ‘industrial backing’ is increasingly becoming a common issue for various localities

16

The Chinese Economy after WTO Accession

in pursuing urban development. The fast growth of new high-growth industries will provide more opportunities for tackling this issue.

Chapter 2

China One Year After Its WTO Entry Aimin Chen

Introduction China formally became a WTO member on December 11, 2001. One year after its entry into the WTO, one is compelled to ask these questions: Has China become a more open society? Has China complied with the WTO rules? How have China’s sectors of agriculture, industry, as well as service, endured the entry impacts? Have there been systemic changes beyond economic growth and foreign trade? As the largest developing country of the world that has scored unprecedented economic growth in the past two decades, China’s WTO entry as well as its performance as a new member further integrated into the global economic system are of importance and concern to most. This study attempts to answer these questions. Addressing both aspects of the effects on China’s sectors of agriculture, industry, and services, and on systemic structure and institutional changes, the impact of entry one year after is analyzed in comparison with the industrial structure and expected impact prior to the entry. Collecting and organizing most recently released data by Chinese news media in the impact analysis, the author hopes to make a timely contribution to the literature on China’s WTO accession as ofﬁcial statistical volumes containing economic data after China’s WTO entry are yet to become available. The remainder of the chapter is organized as the follows. In Section II, we present an analytical summary of China’s pre-entry systemic and sectoral structure of the Chinese industry and expected impact of China’s WTO entry based on such structure. Section III, the focal section of this study, provides post-entry impact analysis structured into three parts. The ﬁrst part addresses the changes in openness of China to the world; the second, sectoral changes in the three sectors of agriculture, industry, and services; and the third, institutional and systemic changes to the Chinese society. In Section IV, the author presents a brief analysis of what is awaiting China after the ﬁrst year, which will suggest that the impact of WTO entry is yet to be seen, especially in agriculture and service sectors. The last section makes concluding remarks.

18

The Chinese Economy after WTO Accession

Pre-entry Structure and Anticipated Impacts To better understand what WTO entry has brought to China one year after its membership and to predict further impact in the years ahead, a good knowledge of the background under which China entered the WTO and the anticipated impact prior to the entry is helpful. We, therefore, in this section review, from both perspectives of systemic and scale characteristics, the structure of Chinese industry prior to the entry and the analysis of the anticipated impact, based primarily on the author’s previous research (Chen, 2002) on the impact of China’s WTO entry. To guide our analysis, a review of the classiﬁcation of China’s economic sectors is provided in Table 2.1. The Pre-entry Structure of the Chinese Industry Systemically, Chinese industry prior to WTO entry featured a signiﬁcantly weakened state sector. The state sector’s share of urban employment declined from 76.2 percent in 1980 to 31.9 percent in 2001 (Chen, 2002, and China Statistical Yearbook 2002, p. 120), while its share of the gross value of industrial output (GVIO) declined from 64.86 percent in 1985 to 28.1 percent in 1999 (Chen, 2002, and China Statistical Yearbook 2000, pp. 407–409). 1 Moreover, the state had declined signiﬁcantly in the industrial sector as a whole, but its dominance within the sector existed in the productions that were considered to be vital to national interests, such as in petroleum and natural gas extraction, coal mining and dressing, logging and transport of timber and bamboo, tobacco manufacturing, petroleum processing and coking, smelting and pressing of ferrous metals, production and supply of electrical power, gas, and tap water. Beyond the industrial sector, the state remained its dominance or monopoly position in the sectors of education, culture and arts, radio, ﬁlm, and television (97.07 percent of urban employment); in scientiﬁc research and polytechnic services (92.04 percent of urban employment); in services for farming, forestry, animal husbandry, and ﬁshery (89.06 percent); in health care, sports, and social welfare (87.2 percent); in productions and supply of electricity, gas, and water (85.9 percent); in mining and quarrying (83.26 percent); in real estate trade (70.58 percent); in banking and insurance (69.13 percent); and in transport, storage, post, and telecommunications (65.54 percent). The state sector had even expanded in absolute size in the public utilities sector, the banking and insurance sector, the education, culture and arts, radio, ﬁlm and television sector, the real estate sector, and the scientiﬁc research and polytechnic services sector (Chen, 2002). From the engineering and market concentrations perspective, China’s industrial sector prior to WTO entry characterized low concentrations and too small ﬁrms.

1 The measure of gross value of industrial output (GVIO) by state owned enterprises (SOEs) discontinued after 1999. Starting in 2000, the closest substitute measure is GVIO produced by state-owned and state-holding enterprises, which includes a large amount of nonstate elements and, therefore, is not comparable with the previous measure.

China One Year After Its WTO Entry

Table 2.1

19

Foreign Direct Investment and Foreign Trade in China

FDI % of FDI Export % of Export Import % of Import (bil. USD) Growth (bil. USD) Growth (bil. USD) Growth 1990 3.487 – 62.09 – 53.35 – 1991 4.366 25.20 71.84 15.70 63.79 19.60 1992 11.007 152.10 84.94 18.20 80.59 26.30 1993 27.515 150.00 91.74 8.00 103.96 29.00 1994 33.767 22.70 121.01 31.90 115.61 11.20 1995 37.521 11.10 148.78 22.90 132.08 14.20 1996 41.725 11.20 151.05 1.530 138.83 5.11 1997 45.257 8.46 182.79 21.00 142.37 2.55 1998 45.463 0.45 183.71 0.50 140.24 -1.45 1999 40.319 -11.30 194.93 6.11 165.70 18.15 2000 40.715 0.98 249.20 27.80 225.09 35.80 2001 46.878 15.14 266.15 6.80 243.61 8.22 2002 55.000 17.33 323.64 21.60 293.55 20.50 Note: The calculations of import and export ﬁgures of 2002 are based on the estimated growth rates of export of 21.6 percent and of import of 20.5 percent (CCTV, 12/12/02), the sum of which (617.19 billion) is slightly lower than the ﬁgure of 620.8 billion given by the deputy director of China State Statistical Bureau, Mr. Qiu Xiaohua (CCTV, 3/1/03). Sources: 1. Figures before 2002 are from China Statistical Yearbook 2002, pp. 612, 629 and the author’s calculations. 2002 FDI ﬁgure is from Financial and Economic News Brieﬁng, CCTV, 3/3/03 and the author’s calculations. The 2002 trade ﬁgures are from China Report, CCTV, 12/2/2002 and the author’s calculations. Year

China in 1997, for example, had 17,831 ﬁrms in plastic products, 58,662 in nonmetal mineral products, 28,283 in metal products, 27,837 in ordinary machinery, 18,332 in transport equipment, etc. The highest per ﬁrm sales revenues among these groups were 21.51 million yuan. The smallness of the ﬁrms was further evidenced by the auto industry in which 47 percent of ﬁrms in the auto industry in 1996 produced less than 1000 vehicles (Chen, 2002). Chinese industries also featured severe overcapacities. In the textile industry, for example, the state textile enterprises by October 1998 had idled 4.32 million spindles under a state ‘Aide’ project in order to eliminate overcapacities. Moreover, the author had developed a measure of R ratio, the ratio of total industrial output to the production capacity of key ﬁrms in the industry, as a criterion to see whether an industry has too many ﬁrms. When the ratio was greater than one, the industry allowed fringe ﬁrms to produce alongside of the key ﬁrms, but if the ratio was smaller than one, the key ﬁrms’ capacities alone exceeded the total industry’s output. It turned out that at the end of 1997, 28 industries out of the 33 listed had smaller than one R ratio, evidencing strongly the existence of overcapacities and too many ﬁrms (Chen, 2002). Three factors had been identiﬁed to have contributed to Chinese industry’s low concentration and the existence of too many ﬁrms. First, China lacked multi-plant

The Chinese Economy after WTO Accession

20

ﬁrms. Unlike in developed market economies where the number of ﬁrms (Nf) can be signiﬁcantly smaller than the number of factories or plants (Np) because many ﬁrms have multi-plants, Nf in China was roughly the same as Np because of its early market development. While largely statistical when counting the number of production units, such structure could result in real economic effects because of differences in coordination between plants within a ﬁrm and between ﬁrms. Second, local protectionism prevailed through which local governments protected inefﬁcient ﬁrms from bankruptcy by restricting trade of non-local ﬁrms in the local markets. Third, Chinese state enterprises had faced exit barriers because they must obtain their quota to exit their businesses for social stability and other considerations. Many small ﬁrms also suffered from being unable to ‘die’ (Chen, 2002). Accompanying the structural problems of the Chinese industry was rapid consolidations of Chinese ﬁrms through regrouping, bankruptcy, and growing more mature. As a result, the number of SOEs had declined from 118,000 in 1995 to 46,800 in 2001, while the size of SOE employment per ﬁrm has grown from 954 persons in 1995 to 1632 in 2001.2 Many industries, such as electronics, textile, toys, shipbuilding, have become increasingly competitive in the world market. Meanwhile, China’s agriculture sector, where production had been private family based, had been most closed to foreign competition. The sector also characterized low per capita land, low level of mechanization, and the existence of massive surplus labor. China’s service sector, on the other hand, had been heavily monopolized by the state except in retail sales as aforementioned. Expected Impact of the Accession The severity of the impact of WTO entry depends, in general, on the openness and the extent of monopolization. The more closed a sector is to foreign competition, the greater shock it will bear after the accession; the more monopolized an industry, the more dramatic the decline in the market shares of the monopolistic ﬁrms. Based on the pre-entry conditions, China’s three sectors were predicted to bear a diverse impact of its WTO entry. The agriculture sector would be impacted heavily as the sector has had a low degree of openness despite the low degree of monopolization. Resource reallocation, primarily rural labor, was expected to ﬂow toward other sectors. The manufacturing sector, on the other hand, was predicted to have unclear overall impact. Some sub-sectors were to beneﬁt from expanded foreign markets and would further extend their comparative advantage; some would contract in the effort to consolidate and become competitive; and yet others would shift resources to other industries. Thus, reallocation of resources in this sector would occur primarily among its own industries. China’s service sector was to experience a strong overall effect from the accession. Retail sales excepted, this sector characterized the most pronounced state 2

Calculated from China Statistical Yearbook 2002, pp. 406, and 121–22.

China One Year After Its WTO Entry

21

dominance prior to the accession, such as in banking, insurance, telecommunications, and wholesale distribution. WTO entry marked the sharing of the market for the ﬁrst time with any non-state enterprises in these industries. While state monopolies were ending in this sector, resources were also expected to ﬂow into the sector as the economy became more advanced and privatized. Systemically, WTO entry would change the ratio of SOE/(SOE+NSOE+FE) in the sectors, where NSOE symbolizes non-state enterprises and FE, foreign enterprises. In the agriculture sector, this ratio would decline, reﬂecting primarily the sharing of trading rights and distribution that had long been monopolized by the state. The same effect would take place in the service sector as a result of unprecedented breaking of entry barriers by FEs and NSOEs. But such an effect was unclear in the manufacturing sector where most exporting ﬁrms were state ﬁrms and were becoming still more competitive, while many small and private ﬁrms were unable to survive foreign competition especially when domestic conditions were not conducive to their development. The most striking systemic effect would be the creation of an entirely new growth environment for China’s private enterprises as state monopolies were to end in many services and public utilities where entry barriers had long lasted. Moreover, the negative impact on NSOEs would be transitory whereas the impact on SOEs would be permanent (Chen, 2002). An Analysis of the Impact What has happened to China one year after its long-sought entry into the WTO? Based on available data, we carry out our impact analysis ex post by attempting to answer the questions of whether China has become more open, what has happened to each of the sectors, and what institutional and/or systemic changes in a broader perspective have taken place following the entry. Our earlier analysis of the basic structure of the Chinese industry continues to serve as a useful background and the predicted effects ex ante provides a basis of comparison. A More Open Economy We use the extent of international involvement of the Chinese economy to measure its openness. Available data on FDI, exports and imports, tariff changes, breaking of entry barriers, as well as interaction of Chinese ﬁrms in the international markets afford us to take a glance at the issue. In 2002, China’s FDI had achieved a growth rate of 17.33 percent, the highest rate since 1995 (Table 2.1). With an accumulated value of 530 billion US dollars, China had surpassed the US to become the world’s number one country in attracting FDI. Its imports and exports had also grown briskly at the rates of 21.6 percent and 20.5 percent. Import tariffs have on average decreased by 35 percent and on some products by 45 percent (CCTV, 12/12/02).

22

The Chinese Economy after WTO Accession

China’s export expansion following the entry has also provoked many antidumping investigations against China. During January-June 2002, there had been 16 anti-dumping cases against Chinese ﬁrms in the textile industry alone. Meanwhile, Chinese ﬁrms in 2002 had responded to 100 percent of the dumping charges, a ratio much higher than before, suggesting that they have become more engaged in dealing with international affairs and disputes. Moreover, there had been 10 investigations against dumping in Chinese markets compared with six cases in 2001 (CCTV, 2/10/03). China claims repeatedly that it has complied with WTO rules as a new member. In 2002 it had made more than 300 brieﬁngs to the WTO on relevant new developments (Economic and Financial Report, CCTV, 2/10/03). In compliance with WTO rules, China has established many new laws and regulations regarding foreign investment, trade, and intellectual property rights, which we will discuss in a greater detail later. Moreover, China has opened many industries that were previously closed to foreign investors, allowing, for example, French capital to jointly run water supply and Singapore capital to jointly run public bus operation in Shanghai (China Report, CCTV, 1/16/03). More evidence of breaking of entry barriers will be presented when we present sectoral impact. Sectoral Effects Revisited The agriculture sectorChina’s agriculture sector has had a low level of mechanization and massive disguised unemployment and had been closed to foreign competition prior to its WTO entry, as aforementioned. It has thus been expected to be severely impacted by WTO entry, especially in the areas of bulk productions such as wheat, grain, and cotton. The impact had not been observed in 2002, however. Instead, China became a wheat exporter in 2002, with import declining by 12 percent to 600,000 tons and export growing by 51 percent to 690,000 tons. Poor harvests and price hikes by major exporting countries were responsible for the change (Financial News, CCTV, 2/22/03). The manufacturing sectorChina’s manufacturing sector has grown more competitive and become the engine of the country’s economic growth. The sector had contributed to 40 percent of China’s GDP growth and more than 50 percent of government revenues, absorbed about 50 percent of the nation’s total employment and about 50 percent of rural surplus labor. More than 80 percent of China’s export and more than 75 percent of foreign exchange come from the manufacturing industry (CCTV, 12/17/02). China has the world’s highest production of more than 100 products, and many of these products have more than majority shares of the world market (see Table 2.2). The automobile industry, however, was expected to take the major crash within the manufacturing sector, but the negative impact has been weaker than most had expected. In 2002, China imported 120,000 vehicles worth six billion US dollars. Of the 120,000, 70,000 were sedans and comprised 6 percent of total domestic market

China One Year After Its WTO Entry

Table 2.2

23

Selected World’s Highest Productions of Manufacturing Goods by China (2002)

Products Volume Color TV Sets 39.36 (million) Refrigerators 14.43 (million) Telephones 95.98 (million) Electronic Indicators 45.90 (million) Silk 73.3 (1000 tons) Penicillin NA Oxytetracycline NA Tractors NA Transport Containers NA Source: China Report, CCTV, 12/17/02.

Shares of World Market 29% 24% 50% 42% 70% 60% 65% 83% 83%

which during the year had expanded and prices declined rapidly. The year saw a 35 percent increase in auto consumption and more than 35 percent increase in auto production (CCTV, 12/12/02). The Chinese auto industry has been undergoing rapid consolidations among ﬁrms, resulting in concentrations of major production bases in Shanghai, Beijing, Changchun, Wuhan, and Chongqing. Production scales of major producers have been expanding rapidly. China in 2002 had produced a total of 3.251 million vehicles and become the ﬁfth largest auto producing country in the world after the US, Japan, Germany, and France (CCTV, 2/14/03), though joint venture ﬁrms produced much of the total volume. It seems that the unclear overall effect on this sector (Chen, 2002) has now become clear. China’s manufacturing sector has gained from the WTO entry through enhanced competitiveness and institutional changes that are conducive to more efﬁcient allocation of resources. The tertiary sectorThe tertiary sector has experienced profound changes after the WTO entry. The changes have been observed in the inﬂow of foreign investment in the sector and unprecedented breaking of entry barriers. We address these two aspects based on available data. The sector in the ﬁrst nine months of 2002 hosted new FDI of 9.5 billion US dollars, which was an increase of 24 percent over the same period a year previously. Of the 9.5 billion, technological and computer application related investment increased by 116 percent and wholesale and retail increased by 81 percent (CCTV, 12/12/02). The following developments further suggest breaking of entry barriers. In the insurance industry, China had issued 9 permits to foreign investors to open business. The total business volume of the industry had increased 35 percent and is still rapidly expanding. In banking, there had been 34 foreign banks by November 2002. The ﬁrst joint venture bank between the City Bank and Shanghai Pudong Development Bank

The Chinese Economy after WTO Accession

24

has been formed (ibid). More than 100 private banks of Chinese ownership have been on the market. In the retail industry, China had its ﬁrst foreign retail chain stores in 1995 from Germany. Currently, there are 16 of those (including Walmart). In January-June of 2002, the government had given permission to 8 foreign ﬁrms to open shops in China. In the plans are 8 new Walmart stores, 10 French Jialefu, and 8 German Maidelong.3 The entry of foreign retail chains had provoked mergers among domestic retail ﬁrms toward a greater concentration. Currently, domestic chain stores together contribute to 6 percent of total retail sales, compared with the US where Walmart alone accounts for 8 percent of total retail sales (ibid). FDI in telecommunications, according to WTO negotiations, may not exceed 49 percent and would be subject to geographical restrictions within one year of China’s WTO entry, and, within the second year, the share would go up to 50 percent and geographical restrictions lifted. In practice, China has allowed foreign capital to have controlling shares within one year and independent ownership within three years (ibid). Institutional/Systemic Effects The impact of China’s WTO entry goes well beyond the countable entries into Chinese industries by foreign ﬁrms. The systemic and institutional changes that accompany the inﬂow of foreign capital and the establishments of foreign services will bring to China more profound impact for years to come. We address these changes in the following four areas. Development of rule of lawFollowing the WTO entry, China is moving more rapidly toward establishing a legal infrastructure conducive to market development. Laws and regulations have been made, cleared, and terminated. In compliance with WTO rules, the National People’s Congress, China’s legislating apparatus, in 2002 amended laws regarding foreign investment, intellectual property rights, and foreign trade. They are Enterprise Law on China-Foreign Cooperative Management, Enterprise Law on China-Foreign Joint Investment, Foreign Enterprise Law, Patent Law, Trade Mark Law, Copy Right Law, and the Export and Import Commodity Inspection Law. Domestically, the Marriage Law, Cultural Relics Protection Law, Private Educational Establishments Law, and Contract Law, etc., have also been amended. Moreover, laws safeguarding citizens’ rights and supervising the government, such as the Civil Laws and Supervisory Law, have been undergoing debate and discussion since the 1990s, and progress has been made toward their ﬁnal establishment (China Report, CCTV, 3/5/03). The slow progress up to now, however, suggests their sensitive nature and the difﬁculties in moving toward political democracy. Meanwhile, the State Council has voided 151 outdated laws and regulations; clariﬁed more than 1100 regulations; and established many new ones. The new 3

The names of the French and German chains are according to Chinese pronunciation.

China One Year After Its WTO Entry

25

regulations established include PRC Anti-dumping Stipulation, PRC Antisubsidy Stipulation, the Stipulation on the Management of Foreign Investment in Telecommunications Industry, the Stipulation on the Management of Foreign Financial Institutions, and the Stipulation on the Management of Cooperative AudioVideo Distribution Firms (CCTV, 12/12/02). Changing role of the governmentThe government and its agencies are gradually changing their role from one of regulating and supervising to one of servicing. Here are some examples. Example 1: On January 5, 2003, the government issued for the ﬁrst time a stipulation on safe-guarding the constitutional rights of rural migrants and explicitly on changing its role from regulating and restricting migrants to servicing and helping to smooth market transaction (CCTV, 1/22/03). Example 2: Suzhou municipal government has established a ‘government supermarket’ to provide various consulting services to citizens on a walk-in basis (CCTV, 12/12/02). Example 3: The Communist Party of China (CPC) had amended its constitution at its 16th Party Congress as highlighted below. The CPC has redeﬁned its identity: ‘The Communist Party of China is the vanguard of the Chinese working class, the vanguard of the Chinese people and the Chinese nation . . .’ (CCTV, 11/14/02). The CPC has amended its rules for recruitment: ‘The CPC members are drawn from workers, farmers, members of the armed forces, intellectuals … outstanding elements from all social strata, committed to the party’s cause’ (ibid). This amendment has made it acceptable for private entrepreneurs to become CPC members. The CPC has modiﬁed the basic principles guiding the Party: ‘The CPC takes Marxism-Leninism, Mao Zedong Thought, Deng Xiaoping Theory, and Three Represents Thought as its guidance. The CPC should represent the advanced productive forces, the orientation of the advanced culture, and the fundamental interests of the overwhelming majority of the Chinese people’ (ibid). These changes suggest that China is not only further opening up its market, but also becoming a more approachable society politically. The economic changes have brought profound institutional changes that are leading to more democracy and better protected human rights. Continuing with state enterprise reformConsistent with the institutional developments, the systemic restructuring of the Chinese industry persists with further reform of state enterprises and mushrooming of private enterprises. While the reform of state enterprises continues with the basic principle of ‘holding on to the large and letting go of the small’ established at the 15th CPC Party Congress, with laying off workers and shutting down factories, and with debt-equity swap, the government has issued a stipulation ‘Utilizing Foreign Capital to Restructure Stateowned Enterprises,’ allowing foreign capital investment in, purchase of, or merger with Chinese SOEs. This move not only opens a new channel to reform China’s

26

The Chinese Economy after WTO Accession

SOEs, but also will attract more FDI through expanded cross-national mergers that account for 80 percent of FDI internationally but only 5 percent in China (CCTV, 12/12/02). Domestically, the government offered tax breaks to non-state enterprises for hiring former state workers (China News Digest, 10/04/02). It has been reported that in 2002 the number of SOEs had decreased, but overall performance improved (CCTV, 11/10/02). Private economy gaining ﬁrmer groundOn the front of private enterprise development, 2002 saw major improvements in their growth environment. Private enterprise employment had increased rapidly during the year. At the end of 2001, 12.68 million urban and 11.39 million rural, or a total of 24.07 million, people were employed by POEs (China Statistical Yearbook, 2002, p. 120). By the end of June 2002, employment by 2.21 million privately owned enterprises (POEs) had increased to about 30 million (CCTV, 10/17/02).4 The number of private enterprises has also mushroomed. The number of private enterprises in Beijing, for example, had reached 150,873 by 2002 (which ranked 6th among all provinces and municipalities) in addition to 312,932 proprietors,5 representing a growth rate over 2001 of 21.54 percent for the former and 20.77 percent for the latter (Financial and Economic Brieﬁng, CCTV, 3/3/03). Shanghai, which has the 4th largest number of private enterprises among all provinces and municipalities, averaged a daily increase of 3.5 newly established POEs (CCTV, 10/15/02). Private enterprises have especially made headway in foreign trade. Export value by Shanghai POEs had increased by more than 300 percent compared with the year before (CCTV, 10/15/02). Nationwide, more than 40,000 private ﬁrms have obtained export licenses. Growing at a faster pace than other types of enterprises, exports by private enterprises comprised 8.6 percent of the national total in 2002 (Financial and Economic Brieﬁng, CCTV, 2/26/03). Private investment had grown 65.6 percent from 735.5 billion yuan in 1998 to 1218.1 billion yuan in 2002. While China’s GDP grew at an average rate of 7.7 percent in the past ﬁve years, the private sector GDP grew by more than 10 percent and the public sector, 7 percent. As a result, the share of GDP by the private sector had grown from smaller than one-quarter in 1998 to greater than one-third in 2002. The deputy director of China’s State Statistical Bureau predicts the private sector will further grow at an annual rate of 15-20 percent in the next ﬁve years (Financial and Economic Brieﬁng, CCTV, 3/3/03).

4 If we account for nation-wide employment by non-state enterprises, including private enterprises, collectively owned enterprises, proprietors, foreign enterprises, and joint ownership enterprises, the share had reached 88.76 percent of all employment (including farming) (calculated from China Statistical Yearbook, 2002, p. 120). 5 Private enterprises in China ofﬁcially refer to enterprises that are owned by domestic private entrepreneurs with eight or more employees. This measure then leaves out proprietors, who are also private entrepreneurs, and rural family-based production units that are privately owned.

China One Year After Its WTO Entry

27

The rapid expansion of private enterprises thus far owes much to the improvement of their growth environment, i.e., the legitimatization of the role of private enterprises and breaking of entry barriers. At the 15th Party Congress, the CPC upgraded the role of private enterprises as being ‘supplementary to state enterprises’ to that of ‘equal importance.’ The 16th Party Congress further promoted the role by opening the door of the Party to private entrepreneurs and accepting them to become new members, as shown in the amendment of the CPC constitution. This is an unprecedented move that has signaled the green light to an unconstrained development of Chinese private enterprises. The release of ideological constraints has led to the gradual elimination of entry barriers to China’s industrial sectors. The government ofﬁcially denounced discrimination against private enterprises in their access to, for example, land use rights, bank loans, distribution channels, import and export rights, and to issuing corporate bonds, etc. China’s Premier, for example, emphasized at the Fifth Annual Session of the Ninth National People’s Congress (March 2002) that the banking system will change its ﬁnancing policies and pay special attention to, and support the ﬁnancial needs of, small and medium ﬁrms, namely, non-state enterprises. Private capital controlled banks are increasing rapidly, amounting to a total of about 100. They no longer have to exist in the form of Urban Credit Union or Urban Cooperative Bank. But, as of June 2002, China had not had an urban commercial bank of 100 percent private capital, and private banks dealt primarily with small businesses as a result of history, size, and ideological considerations.6 Before 2002, the auto industry was off limits to private enterprises, which were only allowed to produce parts. Private enterprises can now enter the auto industry without restrictions on what to produce, parts or the whole vehicle of any model. Currently, there is only one private auto maker, Zhejiang Jili (Economic and Financial News, CCTV, 12/18/02). Moreover, the year saw the ﬁrst private enterprise to ever obtain, through bidding, the right to operate a gold mine (CCTV, 10/17/02), suggesting potential breaking of entry barriers into other similar areas that have previously been off limits to private enterprises. The dynamic developments in both the state and the private sectors tend to suggest that the ratio of SOE/(SOE+NSOE+FE), an indicator of systemic structure of the Chinese industry, has declined not only in the service and agriculture sectors, but also in the manufacturing sector. Recall that prior to the entry, the direction of change in the ratio in the manufacturing sector was unclear because of rising competitiveness of state enterprises and uncertainty in the extent to which private enterprises would be allowed to develop (Chen, 2002). The unprecedented development of private enterprises and the consolidation of state enterprises have made the decline in the ratio more obvious. Ofﬁcial statistics that can lend more deﬁnitive support to this inference are still to become available, however.

6

Interview notes, Commercial Bank of Taizhou City, June 25, 2002.

28

The Chinese Economy after WTO Accession

Challenges It seems that China successfully coped with its ﬁrst year of WTO membership. While such success is indicative of the ability of China to further integrate itself into the world economy, China is by no means ‘home free.’ China is still facing many challenges even after its accession, both in terms of meeting more WTO requirements and coping with structural and systemic problems. The following analysis of the potential impact on China’s sectors builds on the investigations in the previous sections. We summarize the impact assessment in Table 2.3. In the agriculture sector, several problems have persisted. The sector is most closed, it has a low level of mechanization, and there is massive disguised unemployment. Meanwhile, the WTO accession requires that all tariff cuts be implemented by 2004 and all tariffs be bound (cannot be increased) and that full trading rights (the right to import and export) and distribution rights (wholesaling, retailing, maintenance and repair, transportation, etc.) be provided to foreign ﬁrms (White House Fact Sheets, 2000, Fact Sheets hereafter). China had dodged the impact in 2002, owing primarily to poorer world harvests and the resultant price increases. To successfully cope with the foreign competition on the horizon, China must create jobs for rural surplus labor, that is estimated to be at least 150 million, and the agriculture productions must be quickly restructured toward idiosyncratic Chinese products and toward higher technological elements. The agriculture sector was most vulnerable before China’s entry and remains most vulnerable one year after because of the predicaments and the shorter phase-in period compared with other sectors. Within-sector restructuring is still an urgent task for China’s industrial sector as well, despite its enhanced competitiveness and expanded exports that result primarily from goods such as electronic products, toys, and textiles in which China possesses comparative advantages. In this sector, China has to cut two thirds of tariffs by 2003, and the rest by 2005, with a limited number of exceptions. Moreover, as in the agriculture sector, China has to provide full trading rights and distribution to foreign ﬁrms in most industries (Fact Sheets, 2000). In industries such as the high tech and automobiles China has considerable catching up to do while time can run out quickly. In the high tech industry, for example, China has to reduce tariffs from 13.3 percent pre-accession level to zero for semiconductors, computers, computer equipment, telecommunications equipment and other information technology products. China has to eliminate most of the tariffs in this industry by 2003 and in all other industries by 2005 (Fact Sheets, 2000). In the automobile industry, China must lower tariffs from the current level of 50 percent to 25 percent by 2005; quotas will grow from 6.0 billion US dollars in the ﬁrst year to an additional 15 percent annually until eliminated by 2005. In addition, China has agreed to let companies, Chinese or US, distribute most products, including autos and auto parts, into any part of China three years after accession and be able to provide related services such as repair and maintenance services over a three year phase-in period (Fact Sheets, 2000). The auto industry prior to entry was a weak link and is remaining so at the current stage. Though China in 2002 became the ﬁfth

China One Year After Its WTO Entry

Table 2.3

29

Assessment of Sectoral Impact of WTO Entry after One Year

The Sectors Agriculture

Impact Assessment Strong overall impact predicted, but yet to be seen. Domestic sector at disadvantage in general. Sectorwide reallocation of labor to other sectors continues.

Industry and Construction

Tertiary The ﬁrst level: transportation, storage, postal and telecommunications, whole sale and retail trade

Unclear overall effect has become clear. China’s manufacturing sector has gained from the entry and has become more competitive and the major source of economic growth. Some industries are racing against time to grow out of infancy. Strong overall impact: Yes, and more yet to come. Unprecedented breaking of entry barriers in the ﬁrst and second levels of the sector, e.g. the wholesale and retail trade channels, telecommunications, banking, and insurance services. Ending of state monopolies.

The second level: banking, insurance, geological survey, water conservancy management, real estate; services for residents, agriculture; and forestry, animal husbandry, ﬁshery, subsidiary services for transportation and communications, comprehensive technical services, etc. The third level: education, culture and arts, broadcasting, movies, television, public health, sports, social welfare and scientiﬁc research, etc.

Many entry barriers to remain or uninterested by foreign competition in the third level of the sector. State monopolies to stay.

The fourth level: government agencies, political parties, social organizations, military and police service

The fourth level remains naturally intact.

largest auto producer in the world, much of the production had been done by joint venture ﬁrms. Can China’s high-tech and auto industries have enough time to grow mature and competitive? The answer remains to be seen.

30

The Chinese Economy after WTO Accession

More pervasive foreign competition is also on the horizon for China’s service industries. In 2002, only China’s key telecommunications services corridor in Beijing, Shanghai, and Guangzhou was open to foreign competition. By 2003, all telecommunications services became open, for example. In the insurance industry, ‘China promised to permit foreign property and casualty ﬁrms to insure large-scale risks nationwide immediately upon accession, and to eliminate all geographic limitations for future licenses within ﬁve years, allowing access to the key cities of priority U.S. interest within two to three years.’ China also promised to allow foreign ﬁrms the right to distribute video and sound recordings, and cinema ownership and operation, and to allow unrestricted access to the Chinese market for hotel operators with the ability to set up 100 percent ownership foreign hotels within 3 years after the WTO accession, while only majority ownership was allowed at accession (Fact Sheets, 2000). Facing the toughest challenge among all service industries is China’s banking sector, which is notoriously ridden with bad debts, heavy government intervention, extremely limited product variety and poor quality, unskilled stafﬁng, and other problems. Without protection, Chinese banks have little chance of surviving foreign competition. The ﬁrst year of the membership only entitled foreign banks to conduct local currency in four cities and with foreign clients. After ﬁve years, foreign banks would have full rights to handle both local and foreign currency business transaction; the rights to serve Chinese as well as foreign customers; and be allowed to liberalize investment. Thus what had happened in this sector during the ﬁrst year was merely symbolic. The difﬁculty of the banking sector goes beyond the infancy nature. Chinese banks have not been, in a real sense, commercial banks. They are deeply intertwined with the problems of China’s state-owned enterprises. If failing SOEs, who most likely owe large debts to the state banks, are allowed to go under, banks will suffer from considerable unrecoverable debts. If the failing SOEs cannot exit the industry, then the banks must continue to subsidize them via policy channels. The current debt-equity swap scheme functions only symbolically as a possible solution to the vicious circle for its limited coverage and intrinsic SOE problems of its own. Concluding Remarks The road to WTO membership for China was a long and arduous one, not only in terms of the difﬁculties and complexities in entry negotiations, but also of the anxiety experienced by China of facing coming ‘wolves.’ A year after China’s WTO entry, one is thus compelled to ask what has happened to China and how have China’s sectors endured the impact of foreign competition. It turns out that China has successfully coped with its ﬁrst year of WTO membership. Its exports in 2002 had grown 21.6 percent, and imports, 20.5 percent; it has become the largest host of foreign direct investment; its manufacturing sector has become increasingly more competitive and has expanded the shares of many productions in the world market; and its service sector has opened its door to foreign

China One Year After Its WTO Entry

31

investors and has been calmly restructuring. Most importantly, China, as a result of further opening up, has undergone systemic and institutional changes that are conducive to a more efﬁcient resource allocation and a more stable and democratic political system; the state sector in the economy has further declined and private enterprises gained unprecedented ground for development. All these have taken place while China’s agriculture sector, which has been expected to bear heavier crushing effect, has survived the ﬁrst year. Poorer world harvests, resultant price increases, as well as the rapid urbanization taking place in China explain part of the quietness in that sector. China is by no means home free, however. The WTO impact is yet to be seen for the years to come when China runs out of tariff and quota shelters. China’s sectors are racing against time. Moreover, China faces severe challenges to its sustainable growth and development, such as the bad-debt ridden state banks, the reform of state enterprises, the establishment of an effective social security system, a balanced regional development, urbanization of its rural sector, etc. Of these problems, the agriculture sector calls for most attention to its growing income disparity from the urban sector, the existence of massive surplus labor of at least 150 million, and the resultant urgent need to rapidly urbanize and create non-farm jobs. The agriculture sector thus remains most vulnerable followed by the service sector. What the next few years of the phase-in period bring to China will continue to call for attention by both policy makers and academicians. References Chen, Aimin (2002), ‘The Structure of Chinese Industry and the Impact from China’s WTO Entry,’ Comparative Economic Studies, Spring, pp. 72–98. ——— (2001), ‘Has China’s State Sector Really Turned the Corner?’ paper presented at a session of the ASSA annual meetings, January 4. ——— (1999), ‘Unemployment and Labor Market Development in China,’ presented at the International Symposium on 21st Century China and Challenge of Sustainable Development, September 3-5, Washington DC. ——— (1998), ‘Inertia in Reforming China’s State-owned Enterprises: The Case of Chongqing,’ World Development, Vol. 26, No. 3, pp. 479–95. China Central Television Station, various dates. China Statistical Yearbook, various issues. Cui, Minxuan, and Zhang, Cunping (1998), ‘The Comparison of Scale Economies of Chinese Enterprises,’ China Industrial Economy, No. 5, pp. 53–58. Frazier, W. Mark (1999), ‘Coming to Terms with the ‘WTO Effect’ on U.S.-China Trade and China’s Economic Growth,’ The National Bureau of Asian Research, September. Frazier, W. Mark and Hansen, M. Peter (1999), China’s Accession to the WTO: A Candid Appraisal from U.S. Industry, The National Bureau of Asian Research, September.

32

The Chinese Economy after WTO Accession

Hufbauer, Gary Clyde and Rosen, Daniel H. (2000), ‘American Access to China’s Market: The Congressional Vote on PNTR,’ International Economics Policy Briefs, April 20, Institute for International Economics. ‘Market Access and Protocol Commitments,’ (1999) Government Releases, April. Research Group, the Industrial Economy Research Institute, Social Science Academy of China (1997), ‘Chinese Industry’s Change from Quantity Expansion to Quality Enhancing,’ China’s Industrial Economy, June, pp. 5–14. White House Fact Sheets, (2000) February 17, www.uschina.org/public/wto/ factsheets/.

Chapter 3

WTO and Private Enterprises: A Case Study of China Feiyue Shunfeng Song and Hong Cheng1

Introduction In summer 2002, the ﬁrst author of this chapter joined a ﬁeld trip with a research group organized by the Chinese Economists Society (CES) and visited Taizhou, Zhejiang Province, China. In Taizhou, the group visited China Feiyue, a privately owned company that produces sewing machines as its main product. Delegates were impressed by the progress that Feiyue has achieved since its founding in 1986. To many Chinese, a sewing machine is a traditional household product. Yet, Feiyue introduces high-tech into the product and exports its product to over 100 countries. Over the past 17 years, Feiyue has gone through a typical path like many other successful private enterprises in China. In 1986, Qiu Jibao, president of China Feiyue, set up a family workshop with his savings and borrowed 300 yuan ($1=3.45 yuan at the 1986 exchange rate) from a local credit agency (xinyongshe). The workshop was under the title of ‘Jiaojiang Second Industrial Sewing Machine’ and registered as a collective-owned enterprise. It became Feiyue Sewing Machine Group Corporation in 1994 and Feiyue Co., Ltd in December 2000. In 2002, Feiyue produced an annual output of 2.1 billion yuan ($1=8.3 yuan at the current exchange rate) and an export of $101 million. It employed 1,750 workers. The history and growth of Feiyue evidences China’s economic reform and opening-up. It also evidences the improvement of the status of Chinese private enterprises. China’s sewing machine industry, although small and traditional compared with many other Chinese industries, produces more than half of the world’s total sewing machines. It is export-oriented, with an export-production ratio of 0.82 in 2002. With WTO, China will see a big increase in its textile export because of lower tariff and elimination of export quota. In turn, this will create a stronger demand for sewing machines. With further opening-up, China will also see more sewing machines coming to China and more foreign brands joining the direct competition with their domestic counterparts in the Chinese market. No doubt China’s entry into WTO 1 The authors thank the anonymous reviewer for constructive suggestions, and Shuming Bao for his editorial help. The authors are grateful to Qiu Jibao and Hong Xianzhou for supporting this research project and ﬁeld trip.

34

The Chinese Economy after WTO Accession

poses challenges and opportunities to Chinese enterprises in the sewing machine industry. This chapter uses Feiyue as a case study. Data mainly come from Feiyue, through interviews, ﬁeld visits, and correspondence. The authors hope that the research on Feiyue will provide insights on how private enterprises, export-oriented companies, and traditional producers react to changes brought by China’s accession into WTO. However, due to the lack of data, this research is more an informative study than an analytical one. The chapter is organized as follows. The next section gives an overview of the Chinese sewing machine industry, then we review Feiyue’s development in the past two decades. The penultimate section discusses strategies that Feiyue adopted to cope with challenges and seek development after China became a member of the WTO in 2001. The ﬁnal section provides conclusions. China’s Sewing Machine Industry China’s sewing machine industry is relatively traditional. Sewing machines were probably the ﬁrst modern machine that many Chinese saw before China’s openingup. Many people aged 40 and over remember wearing local tailor-made clothes. A sewing machine was an important piece of dowry for many countryside marriages. It was also a luxury possession for housewives in both cities and the countryside. Chinese regarded sewing machines as a symbol of the good life even in the 1980s. China’s sewing machine industry is also relatively small. In 2002, China had about 800 sewing machine manufacturers. They produced 13.75 million sets, including 4.35 million sets for industrial use. The industry contributed the country 18 billion yuan, only 0.176 percent of the national GDP. The top 51 enterprises of the industry employed 28,068 workers (Tian, 2003). However, the sewing machine industry exhibits several interesting characteristics. First, the industry produces more than half (three-quarters in 2002) of the world’s total sewing machines. In recent years, the world produces 18 million sets of sewing machine annually, with a value of $6-7 billion (China National Light Industry Web site). Other major producers include Japan, Germany, and Italy. Second, the industry is export-oriented. In 2002, it produced 13.75 million sets of sewing machines and exported 11.258 million to more than 150 countries. The exportproduction ratio is 0.82 (China Sewing Machinery Website). Table 3.1 lists the annual export volume and value for the period of 1992-2000. It shows increasing trends in both the export value and unit price of sewing machines. The unit price of exported sewing machines was $19.4 for domestic-use sewing machines and $99.6 for industrial-use sewing machines in 2002 (Tian, 2003). In the same year, China imported 795,000 sets of industrial-use sewing machines from Japan, Germany, and Italy, with an average price of $985 per set. The striking price gap between China-made and foreign-made sewing machines indicates a lower quality of Chinese brands. China imports sewing machines that are of higher quality and more technically advanced.

WTO and Private Enterprises: A Case Study of China Feiyue

Table 3.1 Year

35

China’s Export of Sewing Machines (1992–2000)

Export Volume Export Value (million sets) ($million) 1992 2.79 97.69 1993 3.10 119.06 1994 4.19 164.10 1995 4.52 208.12 1996 4.07 206.41 1997 4.09 223.55 1998 4.06 235.91 1999 4.42 237.06 2000 6.00 346.19 Source: China Statistical Yearbook, various years.

Average Price ($/set) 35.0 38.4 39.2 46.0 50.7 54.7 58.1 53.6 57.7

Third, the industry is highly concentrated. In 2002, the eight largest sewing machine producers in China produced 48 percent of the national total of industrial-use sewing machines. Their output accounted for 34 percent of the industry’s output and 45.4 percent of the industry’s export (China Sewing Machinery Website). Fourth, more foreign companies have become partners of Chinese sewing machine producers. The ﬁrst foreign company, Pegasus from Japan, started up a joint venture in the mid-1980s with Tianjin Tiangong Sewing Machine Company. Today, more than 60 joint ventures have been established, with major sewing machine producers from Japan, Germany, and Italy. Foreign companies move to China to enjoy scale and agglomeration economies in sewing machine production; they also enjoy the low cost of labor and resources. Chinese producers beneﬁt, too, mainly through technology transfer, management know-how, and international sales networks. Fifth, although the sewing machine industry is still labor intensive and resourceoriented, high technology is becoming more important. A survey of 51 major sewing machine enterprises shows that 19.8 percent of their work force are professional staff, technicians, and engineers. Of the 51 enterprises, 45 had R&D departments in 2002, with one at the national level and three at the provincial level. About half (25 out of 51) work with universities and research institutes to develop new products. More than 10 of these enterprises hired foreign technical experts. Yet, on average, only 1.6 percent of their total sales revenue goes to R&D, and many enterprises devote very limited resources to train their employees (Tian, 2003). Technology and R&D capacity are far behind those in advanced countries. Lastly, China’s accession into WTO brings opportunities and challenges to the sewing machine industry. On one hand, after China becomes a WTO member, the tariff on Chinese exported textile products will be signiﬁcantly reduced and the quota of Chinese exported textile products will be gradually eliminated. Both will promote China’s export. It is projected that China’s textile export will increase by 43 percent from $30 billion in 1999 to $43 billion by 2005 (China National Light Industry Website). To meet this increased demand, China will produce a lot more

The Chinese Economy after WTO Accession

36

sewing machines. On the other hand, more sewing machines will enter China, posing challenges on producers in the Chinese sewing machine industry. With WTO, China is lowering the tariff on imported sewing machines, reducing export tax reimbursement, and eliminating export subsidies. In addition, more foreign producers go to China either to start their own businesses or become partners with their Chinese counterparts. Thus, they compete with Chinese producers directly and indirectly for China’s domestic market. All these changes will weaken the export competitiveness of China’s sewing machines in the international market. The following sections will discuss Feiyue’s development and the new business strategies facing China’s accession into the WTO. Feiyue Group Feiyue Group is a privately-owned corporation located in Taizhou, on the east coast of Zhejiang Province. In 1986, Qiu Jibao, the President of Feiyue Group, set up a family workshop with his savings and borrowed 300 yuan from a local credit agency (xinyongshe). Over the past 17 years, Feiyue has grown into a leading sewing machine producer in China. Table 3.2 presents major indicators of Feiyue in the past ﬁve years. The appendix shows Feiyue’s milestones since 1986. In 2000, Zhu Rongji, former Premier of the People’s Republic of China, praised the company, saying: ‘Feiyue not only revitalized the industry but applied high and new technology, using high-tech to renovate traditional industry. You, Qiu Jibao, are a national treasure’ (Feiyue Path, 2001.1). Table 3.2

Year

Major Indicators of Feiyue (1998–2002)

Employment (person)

Output (Million yuan)

1998 1510 600 1999 1560 850 2000 1600 1510 2001 1670 1650 2002 1750 2100 Source: China Feiyue Group.

Export Volume (thousand sets) 250 300 350 420 700

Total Export Value Investment (Million ($million) yuan) 30.58 145 37.50 57 63.70 62 71.00 64 101.00 170

R&D investment (Million yuan) 7.4 12.0 35.0 45.0 55.0

The 1997 Asian ﬁnancial crisis threw thousands of companies out of business. Interestingly, however, it brought a turning point for Feiyue. Because of the crisis, many Asian countries’ currencies were devalued, especially relative to China’s currency. With this change and Chinese government’s policies to encourage the importation of high-tech equipment, Feiyue bought hundred of sets of technologically advanced processing centers from the US, Korea, and Taiwan. This move not only realized equipment renovation, but also saved investment of more than 70 million

WTO and Private Enterprises: A Case Study of China Feiyue

37

yuan and helped improve product quality and diversiﬁcation (Feiyue Path, 2001.1, 57–59). This advanced equipment has been used in the product design, product machining and quality control. They also helped Feiyue pave the way for its products to enter the EU, the US, and Japanese markets. Technological innovation and walking-out are two important strategies in Feiyue’s continual development. Accurate product orientation is the base of the enterprise’s rapid development. Feiyue’s insight of fashion trends such as undergarment out-wear, and knitting material development helped Feiyue win the sewing machine market. Feiyue is also the ﬁrst enterprise that began to produce the stretch sewing machine. In the past, the domestic sewing machine was one of the three most important pieces of family furniture. Nowadays, in most people’s minds, the domestic sewing machine is still black, heavy, pedal-moved, and single-functioned. Feiyue adopted new technology and new material to produce electrically-powered, multi-functional domestic sewing machines. Its domestic sewing machine was listed in the national ‘torch plan’ and has received quality certiﬁcates such as UL, CE, and GS. Recently Feiyue has developed the servo system that reached the world class and is more advanced than Panasonic and Mitsubishi to a certain range. Since 1999, ‘Feiyue’ has been named as ‘China Famed Brand.’ Today, Feiyue produces more than 31 series and 300 products of sewing machines. Doing international business has been Feiyue’s goal since the beginning. By the end of 2002, Feiyue had set up 18 subsidiary companies in 17 countries and employed 300 international dealers. It sells sewing machines to more than 100 countries and over 50 percent of them are for developed nations. Feiyue broke the unilateral status of sewing machine exportation from Japan to China. Table 3.2 shows the rapid growth of Feiyue’s export in a ﬁve-year period, from $30.58 million in 1998 to $101 million in 2002. People in the sewing machine industry call Feiyue’s exporting strategy ‘Feiyue mold.’ After China’s accession into WTO, Feiyue grows with China’s ambition to become a leading global market player. It considers WTO more an opportunity than a challenge. On March 26, 2002, Feiyue industrial park was grandly opened for construction. The park will cover an area of two square kilometers, with a total investment of 7.5 hundred million yuan. Once it is completed, the park will facilitate R&D, production, and marketing. The annual production capacity will reach 3.5 million sets of various sewing machines (Feiyue News, April 30, 2003). The industrial park will enable Feiyue to meet the increasing domestic and international demand for sewing machines and help Feiyue with the opportunities and challenges it faces after WTO. Strategies to Confront the WTO Not only is China the largest consumer market of sewing machines, it is also the largest sewing machine manufacturer in the world. With its complete production system and comparative lower labor cost, China is accelerating its process to become

38

The Chinese Economy after WTO Accession

the world sewing machinery center. Feiyue, as a leading producer of China’s sewing machine industry, shows how Chinese producers react to the changing environment after China’s accession into WTO. This section discusses two important strategies of Feiyue. One is to emphasize on R&D; the other is to walk out and do business internationally. Actively Engaging in R&D Feiyue considers technology as the driving force in its development. Within the country, Feiyue works with Tsinghua University, Zhejiang University, Shanghai Sewing Machine Institute, Beijing Mechanics and Electrics Institute to promote research and development work. On November 1, 2002, for example, Shanghai University won the bid with a ‘development platform for supporting intelligentized sewing machine innovation.’ The project is to design an intelligent sewing machine, which considers all factors to ensure harmony between them; concept design, structure design, control design, electronics, and the setting up of a repository (Feiyue Path, 2002.2, p. 79). Outside the country, Feiyue cooperates with sewing machine experts from Japan, Germany, and Italy. Currently, about 300 foreign experts are working at Feiyue’s headquarters and branches in R&D and design. In 1997, Feiyue founded a development center for sewing machine information and technology in Japan. In the same year, Feiyue founded its Development Center for Sewing Machine Information and Technology in Beijing in an attempt to keep pace with internationally-advanced levels. In October 2000, Feiyue set up Feiyue R&D Center in Taizhou, which has become the accelerator of the development of the enterprise. The center strengthens the cooperation with domestic and overseas universities and research institutes, and consummates its system of absorbing and training talent. It also provides the high level equipment for developing and testing, and helps Feiyue in its technological innovation and new product development. China’s sewing machine industry has one national-level R&D center and three provincial level R&D centers including Feiyue R&D center. The internal structure of Feiyue’s R&D center is the matrix, which consists of standard group, developing group, and information group. The developing group is further divided into 5 departments according to the type of the product. Today, the center employs 247 researchers. A survey of 51 Chinese sewing machine producers indicates that Feiyue makes a greater effort on R&D than all other enterprises. In 2002, Feiyue spent 16.5 million yuan on R&D and 38 million yuan on technological innovation and reconstruction, accounting for 17 percent and 19 percent of the total investment of the 51 enterprises, respectively (Tian, 2003). On average, the 51 enterprises devoted 1.6 percent of their sales revenues to R&D. This number is 5 percent for Feiyue. Between 20-25 new products are introduced into the market every year. In 2000, Feiyue gained nine patents at a national level. Technology innovation brought continual orders; the sales of products with high technology content accounted for

WTO and Private Enterprises: A Case Study of China Feiyue

39

70 percent of the total sales. In 2001, Feiyue received 14 national patents. New products contributed 75 percent to the annual production value. Feiyue’s R&D center was titled ‘provincial demonstration’ in Zhejiang Province. In 2002, Feiyue introduced more than 20 new products (Feiyue Path, various issues). At present, Feiyue produces more than 300 varieties in 31 series of sewing machines. Most of them meet international standards. It is worth mentioning that Feiyue’s technology innovation is directed by market. Market demand pulls technology innovation and pushes Feiyue to develop new products that the market urgently needs. To meet international market demand, for example, Feiyue emphasizes applying computer and electronic technology into sewing machines. It also constantly launches new products with comprehensive technology to move market expansions. Participating in International Competition One key to success for Feiyue is aiming at the international market and carrying out the strategy of ‘walking out.’ As early as in 1989, Feiyue people realized the importance of the international market. For this reason, the founder of Feiyue appeared in the China Export Commodities Fair (CECF) in Guangzhou (the Canton Fair). Back then, however, Feiyue was too small to get a booth at the fair. The founder bought one yellow page of the Hong Kong telephone book. He began calling Hong Kong’s sewing machine dealers and learned of a potential market in Latin America. He borrowed an interpreter from the provincial foreign affair ofﬁce and went to Latin America to promote sewing machines. By 1994, Feiyue’s export volume to Latin America had reached $10 million. At that time, an economic crisis was sweeping across Mexico and over other Latin American countries. Feiyue quickly adjusted its export policy and laid its eyes on Southeast Asian, Middle Eastern, and African countries. Within two years, Feiyue established business relations with more than 50 developing countries and its foreign export earnings rose to more than $20 million. As seen in Table 3.2, Feiyue exhibits a dramatic increase in its export, from $30.58 million in 1998 to $101 million in 2002. By the end of 2002, Feiyue had set up 18 subsidiary companies in 17 countries (Table 3.3) and 300 international dealers. It sold sewing machines to more than 100 countries and regions, and 50 percent of them were for the developed nations. In 2001, Feiyue was listed on China’s top 20 strongest enterprises in terms of export volume and foreign exchange earnings, among privately-owned export-oriented enterprises. The rank was honored by the Ministry of Agriculture and the Ministry of Foreign Trade and Economics Cooperation. Feiyue carries out its ‘walking-out’ strategy by establishing subsidiary companies in foreign countries, participating in numerous international fairs, and improving the quality of its sewing machines. Table 3.3 shows that Feiyue set up its ﬁrst overseas subsidiary company in Hong Kong in 1990. In 1997, it established two subsidiaries, one in Los Angeles and one in Miami to cover the entire US market, which has

40

Table 3.3

The Chinese Economy after WTO Accession

Overseas Subsidiary Companies of Feiyue

Location Hong Kong Los Angeles, USA Miami, USA Mexico Peru Thailand Turkey Italy India Germany Greece Bulgaria Russia Venezuela Indonesia Morocco Jakarta, Indonesia Brazil Source: Feiyue Path, various issues.

Year of Establishment 1990 1997 1997 1999 1999 1999 1999 1999 2000 2000 2000 2000 2001 2001 2001 2001 2002 2002

become one of Feiyue’s most important markets. The two US subsidiaries are equipped with an ofﬁce, exhibition hall, and warehouse. Before and after China’s accession into WTO, Feiyue became more aggressive in its overseas business expansion. Five overseas subsidiaries were established in 1999, ﬁve in 2000, four in 2001, and two in 2002. These overseas subsidiaries serve as the centers of sales and consumer services. They help Feiyue market and sell its products in foreign countries and provide Feiyue with information regarding recent development from local consumers. In recent years, Feiyue was very active in attending international expositions to market its products to foreign markets. Table 3.4 presents the 18 exhibitions that Feiyue participated in the period 2000-2001. Some expositions are worth particularly mentioning. The Chinese Export Commodities Fair provides a window for the Chinese commodities to move toward the world. It serves as a bridge of exchange and mutual development with overseas traders. The VDTA (Vacuum & Sewing Dealers Trade Association) is a fair that specializes in domestic sewing machines and vacuum cleaners, and is held yearly in Las Vegas. It has a strong inﬂuence on the sewing machinery ﬁeld of the world. Attending numerous international fairs demonstrates Feiyue’s efforts and determination to expand its international sewing machine market. Good quality wins both domestic and international markets. Feiyue, in the course of ‘walking out,’ makes a great effort to raise its product quality. To improve

WTO and Private Enterprises: A Case Study of China Feiyue

Table 3.4 Year 2001

41

International Expos Feiyue Participated In (2001–2002) Month March March

Exhibition The 5th AAMA-TEX Brazil Sewing Machinery & Accessories Show March International Convention & Exhibition (6) March Italy Domestic Sewing Machine Show May Plovdiv International Consumer and Technology Fair July Feiyue Sewing Machinery Special Show August Bobbin World 2001 2002 February Vacuum & Sewing Dealers Trade Association March The 18th Moscow International Wholesales Fair April China Export Commodities Fair April Hangzhou International Sewing Machine & Parts Exhibition September Turkey Clothing, Embroidery, Machines & Accessories Fair September China International Sewing Machinery & Accessories Show October China Export Commodities Fair Source: Feiyue Path, various issues.

Location Singapore San Paul, Brazil Cairo, Egypt Milan, Italy Bulgaria Tiripur, India Orlando, USA Las Vegas, USA Moscow, Russia Guangzhou, PRC Hangzhou, PRC Izmir, Turkey Shanghai, PRC Guangzhou, PRC

quality, Feiyue strengthens staff’s quality awareness and engages every employee in quality control. Strictly abiding by ISO9000, an internationally accepted quality standard of industrial products, Feiyue sets a testing program into every part of the manufacturing process and never lets one defective machine ﬂow into foreign markets. As an export-oriented company, Feiyue enhances its international competitiveness by developing new models, modifying product structures, setting up new service channels, and pursuing art and personalization in its products. Since April 2002, famous supermarkets, such as CCFA, Wal-Mart, and German ALDI, each have ordered millions of dollars worth of Feiyue products. Conclusions During the past 17 years, Feiyue experienced a miraculous growth. Started in 1986 as a family workshop, it has become an export-oriented ﬂagship in the Chinese sewing machine industry. In 2002, Feiyue employed 1,750 workers and exported 700,000

42

The Chinese Economy after WTO Accession

sets of sewing machines with a value of $101 million. Today, Feiyue produces more than 31 series and 300 products of sewing machine. In an attempt to provide insights on how private and export-oriented enterprises react to WTO, this chapter has investigated two main strategies that Feiyue has adopted. First, Feiyue considers technology as the driving force in its development. It works with Chinese universities and foreign experts to promote technological innovations. Inside the company, Feiyue has a provincial-level R&D center, which consists of standard group, developing group, and information group. It devotes 5 percent of its sales revenues to R&D, far more than its domestic competitors. Second, Feiyue does international business by establishing subsidiary companies in foreign countries and participating in numerous international fairs. By the end of 2002, Feiyue had established 18 subsidiaries in 17 countries and more than 300 international dealerships. In the past two years, Feiyue attended 18 international expositions. Walking-out is Feiyue’s primary strategy in coping with the changing environment after WTO. Today, consumers in more than 100 countries are using Feiyue products. What will be the next step for Feiyue? Feiyue people believe that China’s entry into WTO has brought them more opportunities than challenges. After WTO, China’s textile export has been increasing, creating a stronger demand for more and better sewing machines. Feiyue needs to continue to invest in technology and innovations, analyze the market, keep track of the latest market information, and carry out its globalization strategy. It should consolidate its Latin American market, reinforce its presence in Southeast Asia, and further expand its business in developed countries. References China National Light Industry Website (2003): www.clii.com.cn/hyxx/frj/zt4.htm. Retrieved October 15. China Sewing Machinery Website (2003): www.sewinginfo.com/hydt/hydt05346. htm. Retrieved October 15. Feiyue Group, Feiyue Path, various issues. Feiyue Group, Feiyue News. National Bureau of Statistics of China (1993-2001), China Statistical Yearbook, Beijing: China Statistics Press. Tian, Minyu (2003), ‘Promoting Technological Innovation, Depending on Science and Technology, and Creating a New Stage of the Sewing Machine Industry,’ presidential speech at the 7th board meeting of the Chinese Sewing Machinery Association, Beijing, March 7.

WTO and Private Enterprises: A Case Study of China Feiyue

43

Appendix: Feiyue’s Milestones October 1986

The origin of Feiyue Group-Jiaojiang Second Industrial Sewing Machine Factory was founded. The ﬁrst ‘FEIYUE’ sewing machine came out.

November 1986

Merged with Shanghai sewing machine research institute. It marked the ﬁrst step in combining production and research.

December 1988

The ﬁrst self-designed high-speed overlock sewing machine was produced by Feiyue.

July 1989

The ﬁrst high-speed lockstitch sewing machine was launched and put into mass production.

January 1990

Feiyue set up the ﬁrst overseas branch-Hong Kong (FEIYUE) Sewing Maching Industrial Company.

May 1994

Feiyue Sewing Machine Group Corporation was ofﬁcially formed.

May 1996

Feiyue gave birth to the multi-function domestic sewing machine in China. This was considered a revolution for Chinese domestic sewing machines.

November 1996

Feiyue Imp.& Exp.co.,Ltd. was founded.

May 1997

Feiyue acquired Zhejiang First Industrial Sewing Machine Factory to form Feiyue Group Fifth District; Feiyue Sewing Machine R&D Institute was founded.

October 1997

Feiyue US Los Angles branch was set up; Feiyue invited German and Italian experts to form Beijing Information and Technology Development Center.

November 1997

Feiyue was appointed as an important light industry enterprise by China Light Industry Union and was named as ‘the leader of the business.’

July 1998

Feiyue bought Hangzhou Dubang Sewing Machine Company and formed the biggest manufacturing base for interlocking sewing machines in China.

August 1998

Feiyue invested heavily in buying equipment and technology from the US, Korea and Japan.

October 1998

Feiyue Miami branch was set up; Feiyue’s products got international certiﬁcations such as CE, GS, UL and TÜV; Feiyue’s multi-function domestic sewing machine was ﬁrst sold to Japan, which was the ﬁrst time Chinese sewing machines were exported to Japan.

June 1999

Feiyue hired local experts to form Feiyue Technology Development Center in Japan.

December 1999

‘Feiyue’ was selected as ‘China’s Famous Brand Name’.

August 2000

Shanghai Feiyue Sewing Machine Manufacture Corp. Ltd. was founded.

October 2000

Feiyue Technology Development Center was founded; Feiyue European branch was set up.

44

The Chinese Economy after WTO Accession

December 2000

Feiyue Co., Ltd. was founded.

November 2001

Feiyue 5 models of multifunctional domestic sewing machines obtained certiﬁcation of UL.

December 2001

An ofﬁcial ceremony marked the start of the Feiyue industrial park project. Feiyue launched computerized multifunctional domestic sewing machines.

August 2002

Feiyue’s numerical servo system for industrial sewing machines was put into production and launched into the market. Feiyue holds proprietary intellectual property rights of the core technology of this new system.

December 2002

Feiyue (Brazil) Co. was founded in order to increase Feiyue’s Latin American sales volume. Feiyue successfully developed several new hightech products and put them into production, including an electronic eyelet button hole machine, computer-controlled pattern sewing machine and automatic lockstitch pocket welt sewer etc.

January 2003

Feiyue was titled ‘AA’ class enterprise by the state customs.

June 2003

The central block of Feiyue industry park has been completed, covering 300 thousand square meters, this means another great step towards the future.

Chapter 4

Local Government and Private Sector Development1 Yifan Zhang

Introduction As the young trees make up a healthy ecological environment, new and small enterprises growing alongside older and larger ones create a dynamic economic system. In the context of economic transition, the vibrant new private sector is playing the role of the young trees in Marshall’s (1890) famous forest allegory. Of the two routes to the private sector—privatizing existing ﬁrms (spin-offs) and creating new ﬁrms (start-ups)—many economists began to realize only recently that the latter is at least as important as the former. The formation of new private enterprises is considered a key element of any successful transition (Kornai, 1992; Blanchard, 1997; McMillan and Woodruff, 2002). It is well documented that the post-socialist countries with better new enterprise development tend to enjoy higher growth rates (Johnson et al., 1999). Berkowitz and DeJong (2003; 2005), using a regional data set in Russia, ﬁnd a signiﬁcant relationship between new enterprise formation and growth. In the case of China, the emergence of the private sector is one of the most striking outcomes of the market-oriented reform. In 2002, the nation had more than 2.2 million privately run enterprises representing a total investment of 2.1 trillion Yuan and a labor force of 34 million.2 Table 4.1 illustrates the astonishing growth rate of private sector employment between 1992 and 2002. Economic studies on Chinese enterprises, however, have concentrated almost exclusively on state-owned enterprises (SOEs), township and village enterprises (TVEs) or foreign invested enterprises. The purpose of this chapter is to identify the factors of local governments that drive the development of the private sector. We attempt to explain why some local governments were able to foster the relatively fast growth of new private enterprises while others failed. 1 The earlier version of this chapter was presented at the CES 2003 Annual Conference ‘Chinese Economy after WTO: Opportunities and Challenges for Global Economy’ held at the University of Michigan. I want to thank Daniel Berkowitz, David DeJong, Thomas Rawski, Yang Yao, and Xiaopeng Yin for their valuable comments and suggestions. The usual disclaimer applies. 2 China Statistical Yearbook (2003).

46

Table 4.1

The Chinese Economy after WTO Accession

Private Sector Development in China (1992–2002)

Private Sector Growth Rate of Total Labor Share of Private Labor Force Private Sector Force Sector in Total (Million Persons) Labor Force (%) (Million Persons) Labor Force (%) 1992 2.3 – 594.3 0.39 1993 3.7 60.7 602.2 0.62 1994 6.5 74.0 614.7 1.05 1995 9.6 47.5 623.9 1.53 1996 11.7 22.5 688.5 1.70 1997 13.5 15.3 696.0 1.94 1998 17.1 26.7 699.6 2.44 1999 20.2 18.2 705.9 2.86 2000 24.1 19.1 711.5 3.38 2001 27.1 12.8 730.3 3.72 2002 34.1 25.7 737.4 4.62 Notes: Data include both urban and rural areas. Private sector does not include selfemployed individuals. Source: China Statistical Yearbook. Year

China has witnessed enormous regional heterogeneity in the development of the private sector. Table 4.2 shows the private sector’s share of the labor force in 30 provincial-level regions. It is interesting to see that Shanghai ranks No. 1 while Beijing stays close to the bottom, even though these two cities are of a similar development level. The economic policies regarding the private sector development have also been very different across regions. For example, when coastal regions such as Zhucheng City in Shandong province sold most of their small and medium size state-owned enterprises in the mid-1990s, many western provinces continued to invest heavily in the state sector. The change of ownership can be accomplished overnight, but the transformation of state governance has been much more difﬁcult (Shleifer, 1997). In China, regions differed markedly in local government reform. In most regions, the top-down government downsizing began in 1998. However, some coastal cities such as Shunde in Guangdong province started radical government transformation as early as 1993. The city government of Shunde cut 50 percent of the regular government agencies and 80 percent of the temporary agencies. The reform weakened the government’s power in the economy and shifted the government functions from direct intervention to attracting investment, creating jobs, enforcing contracts, providing local public goods, and managing industrial policy (IFC, 2000). Such regional differences in private sector development and government reform give us an opportunity to test the effects of local government on private enterprises. It is generally agreed in the literature that the government plays a crucial role in determining private sector growth. Existing theories focus primarily on the following four issues: government intervention, law enforcement, government policy, and

Local Government and Private Sector Development

Table 4.2

47

Percentage of Private Sector in Total Labor Force by Region

Rank 1997 1998 1999 2000 2001 Average Nation 1.94 2.44 2.86 3.38 3.72 2.87 Shanghai 1 8.65 11.25 17.17 22.42 28.58 17.62 Zhejiang 2 5.17 5.85 7.21 11.13 12.52 8.38 Tianjin 3 4.29 5.61 8.15 9.34 11.30 7.74 Liaoning 4 3.71 4.43 5.78 6.89 8.60 5.88 Guangdong 5 4.13 4.67 5.19 5.63 6.19 5.16 Jiangsu 6 2.19 3.63 4.89 6.59 8.17 5.09 Hainan 7 2.96 4.39 5.15 5.57 6.03 4.82 Hebei 8 3.38 4.41 4.99 5.86 4.87 4.70 Shaanxi 9 1.88 2.65 3.53 4.60 5.41 3.61 Fujian 10 2.73 3.05 3.57 4.09 4.58 3.60 Shandong 11 2.03 2.69 3.71 4.59 4.54 3.51 Inner Mongolia 12 2.02 2.75 3.81 4.07 4.49 3.43 Qinghai 13 1.39 1.89 2.61 4.69 5.83 3.28 Ningxia 14 1.83 2.86 3.06 3.86 4.57 3.24 Hubei 15 2.34 2.92 3.25 3.17 2.90 2.92 Shanxi 16 2.63 2.98 2.94 2.72 3.10 2.87 Heilongjiang 17 1.89 2.39 3.05 3.28 3.69 2.86 Xinjiang 18 1.76 2.16 2.60 3.32 4.10 2.79 Jilin 19 2.04 2.73 2.58 2.64 3.41 2.68 Chongqing 20 1.68 2.28 2.63 2.98 3.47 2.61 Jiangxi 21 2.02 1.86 2.32 2.44 2.96 2.32 Anhui 22 0.88 1.26 1.66 1.93 2.32 1.61 Sichuan 23 0.98 1.34 1.60 1.81 2.23 1.59 Gansu 24 0.85 1.15 1.75 1.89 2.25 1.58 Beijing 25 1.71 1.38 1.35 1.46 1.40 1.46 Hunan 26 1.22 1.49 1.59 1.33 1.63 1.45 Yunnan 27 0.63 1.16 1.40 1.52 1.97 1.33 Guizhou 28 0.92 1.13 1.15 1.25 1.37 1.17 Henan 29 1.01 1.10 1.19 1.02 1.02 1.07 Guangxi 30 0.67 0.91 1.15 1.21 1.25 1.04 Notes: Data include both urban and rural areas. In our deﬁnition, the private sector does not include self-employed individuals. See the next section. Source: China Statistical Yearbook.

ﬁscal autonomy. This chapter studies these four aspects of local government using a provincial dataset during the period 1997–2001. We ﬁnd empirical support for the theories of government intervention, law enforcement and government policy. The evidence on the inﬂuence of ﬁscal autonomy is relatively weak.

48

The Chinese Economy after WTO Accession

The remainder of this chapter is organized as follows. The next section describes the evolution of the Chinese private sector. Then we develop hypotheses based on transition economic theories. The following section describes the data and deﬁnes the variables. The estimation results are reported next. The ﬁnal section concludes the chapter. Development of the Private Sector in China The Emerging Private Enterprises The broad deﬁnition of the private sector includes both ‘self-employed individuals’ and ‘private enterprises’. Ofﬁcial publications in China (for example, China Statistical Yearbook) distinguish ‘self-employed individuals’ (getihu) from ‘private enterprises’ (siying qiye). The latter have at least eight employees while the former have fewer than eight. Most individual businesses are single person or husband/wife endeavors. Our study uses the narrow deﬁnition of private sector, which excludes individual businesses because they are not formal enterprises. Private enterprises emerged in the early 1980s as a consequence of the rapid expansion of self-employed individual economy. The new private enterprises were intended to play a role that was ‘supplementary’ to the state and collective sectors. They were not ofﬁcially recognized and registered until 1988. It is estimated that by the end of 1988, China already had 500,000 private enterprises (IFC, 2000). In 1989, the private sector suffered a major setback from the aftermath of the Tiananmen Square event. As a result, the number of registered private enterprises decreased by 16 percent. The new wave of reform in 1992, following Deng Xiaoping’s southern tour, provided the private sector with a more hospitable atmosphere. The most rapid expansion of the private sector occurred in the 1990s. Between 1991 and 1997 its annual growth rate of output was as high as 71 percent (IFC, 2000).3 In addition to the role of growth engine, the private sector has also been a major source of job creation, absorbing the workers laid off from the SOEs and TVEs. During the period between 1990 and 1997, new jobs created by the private sector accounted for 37.6 percent of all new formal employment. New employment in the private sector has exceeded the combined total for state, collective, and township and village enterprises between 1995 and 1997 (Rawski, 1999). The private sector not only contributes to the employment and output directly, but also improves the efﬁciency of the economy through indirect channels such as better allocation of resources, increased competition and technology spillovers.4

3 IFC uses a ‘sector-based method’ to estimate the private sector’s output. For details, see IFC (2000). 4 Many of the high-tech enterprises in computer, pharmaceutical and telecommunication industries are private. Even in the traditional industries, many private enterprises are now technology leaders.

Local Government and Private Sector Development

49

The Chinese private sector ﬂourished as ideological barriers have fallen. In the early days of China’s economic reform, private enterprises faced an openly hostile political atmosphere and, as a result, were extremely limited in scale. The 15th Party National Congress in 1997 lifted the status of the private sector from ‘supplementary’ to ‘an important component’ of the economy. The revision of the constitution in 1999 equated the state and non-state sectors further and improved the political environment for private entrepreneurs. The policy shift to allow the private enterprise owners to join the Party showed a new trend and might lay a new foundation for the future development of the private sector. The constitutional amendments from 2004 certainly helped to promote private sector growth and better safeguard private property rights. Newly Privatized Enterprises As in the former Soviet Union and Eastern Europe, China’s private sector includes not only the start-ups, but also the spin-offs. There has been massive privatization of small and medium size SOEs and TVEs since the mid-1990s. China began a quiet reform of restructuring and privatizing the small and medium size SOEs in 1992 (Cao et al., 1999; Li, 2001). The central government formed a policy called ‘keep the large ones and let go the smaller ones’ (zhuada fangxiao) in 1995. As many as 80 percent of the small and medium size SOEs had gone through such transformation by 1998 (Zhao, 1999). A survey by China Reform Foundation (1997) also suggests that 70 percent of the small SOEs in Shandong province were fully or partially privatized in 1996. Consequently, a large number of new private enterprises were born. The TVEs suffered a variety of difﬁculties in the mid-1990s. The privatization of TVEs began a few years later than SOEs (Sonobe and Otsuka, 2001). A survey conducted in 1998 and 2000 shows that 60 percent of TVEs in Zhejiang province and 55 percent in Jiangsu province had been privatized by 1999 (Li et al., 1999). Since most large SOEs are under supervision of central and provincial governments, a majority of the government enterprises at the city, county, township and village levels have been transformed into private enterprises. The start-ups, together with the spin-offs, have changed the landscape of China’s economy. The emerging private enterprises have replaced foreign enterprises and TVEs as the most dynamic force in the economy. Developing Hypotheses Since the reform began in 1979, Chinese local governments have been playing an increasingly important role in economic development. In the early 1980s, most SOEs were delegated by the central government to local governments at provincial, city, and county levels. During the reform period, such decentralization of power went far beyond delegating the existing SOEs. The local governments assumed primary

50

The Chinese Economy after WTO Accession

responsibility for economic development and local public goods provision within their jurisdictions and at the same time enjoyed a broad range of authority. In this section, we develop our hypotheses based on the theories of government behavior in transition economics. Government Intervention The governing force of modern market economy is the rule of law. The rule of law has two primary functions. The ﬁrst function is that it limits the government’s intervention in economic activity and thus protects the private property rights. It is generally agreed that a good government is relatively noninterventionist and protects property rights (Smith, 1776; North, 1981). Frye and Shleifer (1997) describe a ‘grabbing hand’ model of government, in which the government uses its power to extract the rent by imposing various predatory regulations on private businesses. They ﬁnd that the regulatory burden on small shops was much higher in Moscow than in Warsaw, which may explain the sharp contrast in economic performance between Russia and Poland. Johnson et al. (2002a) ﬁnd in a survey of private manufacturing ﬁrms in Romania, Slovakia, Ukraine, and Russia that it was indeed the lack of property rights protection that discouraged the ﬁrms from investing, not the lack of access to credit. IFC (2000) ﬁnd that Chinese local government and ofﬁcials tend to overexpand their duties and focus on rent-seeking opportunities. The roles of government bureaus are often overlapped and ill-deﬁned. Law Enforcement The second function of rule of law is that it regulates the behavior of individuals and ﬁrms and enforces the contracts. Effective law enforcement produces a more transparent, equal, and predictable environment for the businesses (Hayek, 1944). The failure to enforce the contracts is widely acknowledged as one of the primary problems in economic transition (Johnson et al., 1999). China’s legal institutions remain essentially under-reformed and ill-suited to institutions of market economy. Even if the laws were written, it is a question whether the courts are capable of enforcing them (Clark, 1994). Administrative mediation is often used in substitution for courts when disputes occur between two SOEs. However, Chinese private ﬁrms have no corresponding government agency and no effective legal mechanism for the resolution of contract disputes. As a result, other ﬁrms are often unwilling to enter into contract with private ﬁrms. Thus, the contract enforcement is particularly important for private ﬁrms. However, the importance of law enforcement should not be overstated. McMillan and Woodruff (1999) ﬁnd that Vietnam’s private sector boomed in the absence of a legal system. In fact, reputation and relational contracts served as substitutes for courts in Vietnam (Johnson et al., 2002b).

Local Government and Private Sector Development

51

Government Policy Local government policy on state and private enterprises could be a key determinant of private sector development in China. Chinese local governments have the incentive to use their power over private enterprises in order to protect their SOEs (McMillan, 1995). Consequently, Chinese private enterprises still face numerous obstacles such as discrimination of bank credits (Brandt and Li, 2002). To study the government policy on the private sector, we will focus on banks’ credit policy, since Chinese banks are not completely independent of the local governments. It is well known that local branches of the state banks are pressured by the local governments to channel funds to those loss-making SOEs so that unemployment and instability are minimized. Therefore, the availability of external ﬁnance is a good indicator of local government’s attitude towards the private sector (Tsai, 2002). A survey by IFC (2000) reveals that only 12.3 percent of private enterprises in Beijing received loans from banks. This is in sharp contrast with Wenzhou in Zhejiang province, where 70 percent of private enterprises applied for bank loans and 96 percent of the applications succeeded. External ﬁnance itself is important for the private enterprises. If bank credit is not available, private entrepreneurs may not be able to take advantage of investment opportunities. It is found that in transition economies smaller ﬁrms have lower rates of investment because their investment depends on the availability of internal funds (Pissarides et al., 2003; Lizal and Svejnar, 2002). The problem of external ﬁnance is more serious in China than other transition countries. Chinese entrepreneurs started their businesses relying almost exclusively (91 percent) on self-ﬁnancing. This ratio is higher than 66 percent in Russia and 79 percent in Vietnam.5 Table 4.4 illustrates that the average share of the private sector in total bank loans in China was less than 1 percent between 1997 and 2001 (see ‘External Finance’ in Table 4.4). About 40 percent of the private enterprises surveyed by IFC (2000) considered access to ﬁnancing a major constraint. When access to bank credit is denied, many Chinese private entrepreneurs resort to informal or even underground sources of funds (Tsai, 2002). However, when the size of the private sector becomes larger, depending on informal lending can be problematic and unsustainable. Fiscal Autonomy Qian and Roland (1998) construct a model that studies the effects of federalism, Chinese style, on regional economic growth. The so-called ‘market-preserving federalism’ limits local governments’ predatory behavior and at the same time provides greater incentive for the local governments to encourage business formation. Jin et al. (2004) ﬁnd the evidence that ﬁscal decentralization during China’s reform period promotes the growth of non-state enterprises. In contrast, a revenue-sharing relationship between local and regional governments in Russia 5

See IFC (2000), Table 5.4.

52

The Chinese Economy after WTO Accession

hinders local government’s incentive to provide infrastructure for private businesses (Zhuravskaya, 2000). An emerging line of research has attempted to test the relationship between ﬁscal decentralization and economic growth with mixed results. Contrary to traditional wisdom, Zhang and Zou (1998) ﬁnd that a higher degree of ﬁscal decentralization is associated with lower provincial economic growth over the period 1978–1992. Their estimation results suggest that greater ﬁscal decentralization may constrain the ﬁscal capability of the central government, resulting in the suboptimal provision of public goods. However, it is argued that if a set of annual dummy variables is included in their regression, the negative sign of ﬁscal decentralization turns positive (Lin and Liu, 2000; Jin et al., 2005). Hypotheses Based on the above theories, it is hypothesized that those regions with less government intervention, stronger enforcement of law, a more friendly policy on the private sector, and greater ﬁscal autonomy are associated with better private sector development. To test these hypotheses, we also need to control for other factors that may have inﬂuences on the private sector. Economic growth provides more resources and more business opportunities for the entrepreneurs. Thus, private enterprises may geographically concentrate on those fast-growing regions. In a study of Russia, however, growth does not appear as a statistically signiﬁcant variable in the regression of new enterprise formation (Berkowitz and DeJong, 2002). In addition to economic growth, we also include growth related variables in our regression: education, openness and initial income level (Barro, 1991). In the cross-country studies on the former Soviet Union and Eastern European countries, considerable emphasis has also been placed on the achievement of macroeconomic stability (Fischer et al., 1996; Johnson et al., 2000). However, macroeconomic stability was not a serious problem during the period under our study (1997–2001). We also believe that unlike the cross-country studies, macroeconomic stability did not vary very much at the sub-national level in China. Data and Variables Our initial measure of private sector development was the share of private sector in provincial GDP. Unfortunately, in most statistical publications, private enterprises are grouped into a category called ‘other economic elements.’ The only source that lists statistics on the output of the private sector is Yearbook of China Industrial and Commercial Administrative Management. However, as Yao (1999) points out, private enterprises tend to underreport their output to the Bureau of Industrial and Commercial Administration because the latter collects a fee proportional to their sales. Thus, we use the share of the private sector in total labor force as a measure of

Local Government and Private Sector Development

53

private sector development because we believe the labor data from China Statistical Yearbook are more reliable. In our study, private sector and labor force data include those from both urban and rural areas. There is a possibility of underestimating the size of the private sector. Since the beginning of the reform, many private enterprises have been registered as collective enterprises in order to be better protected and get access to more resources. They could obtain a collective license by paying an ‘administration fee’ to a state or collective enterprise or local government organization. This allowed them to receive the approval of the application for registration. Such ﬁrms were called ‘red hat ﬁrms.’ The government issued a directive requiring all the red hat ﬁrms to ‘take off the hat’ in 1998 (IFC, 2000). We have two measures for government intervention. The ﬁrst measure is the government size (GOVSIZE) as deﬁned by the ratio of the expenditure on government administration to GDP. The expenditure on administration is the government’s ‘own’ consumption (Barro, 1991). It is roughly equal to ﬁscal spending minus expenditures on local public goods such as education. The second measure is the provincial government index (GOVINX) complied by Fan and Wang (2001). The government index is part of the broader marketization index constructed by the same authors. There are three components in this government index: (1) ﬁscal revenue as a percentage of GDP; (2) public ﬁnancial burden on farmers. In addition to taxes, local government ofﬁcials in China frequently impose a variety of fees on farmers, most of which are illegal. Public ﬁnancial burden shows the size of the fees relative to farmers’ income; (3) survey of business executives on ‘how much time you have to allocate to deal with the government’. To construct the index, these three variables were transformed into 0-to-10 scale values. Principal component analysis was used to generate the weights. The index shows the relative position of a province in the country. Higher index indicates less government intervention. We believe that GOVINX is a more comprehensive and more direct indicator of government intervention, since it includes both objective and subjective measures. We use the ratio of the number of lawyers to labor force (LAWYER) to proxy law enforcement. This measure approximates the development of a legal environment, because a broad range of legal issues now require the participation of lawyers. The data are from Collection of National Judicial Materials (Quanguo Sifa Ziliao Huibian, unpublished). The external ﬁnance (FINANCE) is measured by the share of the private sector in total bank loans. The data are from China Financial Statistics (Zhongguo Jinrong Tongji, unpublished). We use the common measure of ﬁscal autonomy (FD) in the literature: the ratio of provincial to central government expenditure per capita (Zhang and Zou, 1998; Jin et al., 2004). As in most growth regressions, our measure of economic growth is the annualized growth rate of real GDP per capita between 1997 and 2001. It is argued that the provincial GDP data as well as national GDP data between 1998 and 2000 contain

54

The Chinese Economy after WTO Accession

major exaggerations (Rawski, 2001). However, assuming that the regional growth patterns did not change substantially during this period, our regression results are not sensitive to the use of the growth rate between 1978 and 1997. The variable of education (EDU) is calculated as the number of students enrolled in secondary schools relative to total population. Openness (OPEN) is deﬁned as the percentage of exports in GDP. Initial income is measured by real GDP per capita in 1997. This study adopts a cross-section approach. With the exception of GOVINX, all variables are averaged over 1997–2001. GOVINX is averaged over the period 1997– 99. The basic dataset is reported in Table 4.3. Our dataset for regression includes 28 provincial-level regions. We have excluded Beijing and Shanghai because, from the scatter plots, we ﬁnd that Beijing and Shanghai are outliers. This is not surprising given their special status in the country. All the data are taken from China Statistical Yearbook except GOVINX, LAWYER and FINANCE. The summary statistics of the variables are reported in Table 4.4. Correlation coefﬁcients of the independent variables are shown in Table 4.5. We notice that FD and LAWYER have a very high correlation coefﬁcient (0.849). We also ﬁnd that initial income (INITIAL) is highly correlated with the rest of the variables. To avoid the problem of multicollinearity, INITIAL will not enter the regression and will only be used in 2SLS as an instrument for GROWTH. Results Baseline Regressions We begin our analysis by regressing the percentage private sector in total labor force (PRIVATE) on GOVINX, LAWYER and FINANCE, using ordinary least squares (OLS)6 The resulting estimates are reported in column 1 of Table 4.6. Throughout this chapter, t-statistics are based on White consistent covariance estimates. The asterisks *, **, *** indicate the signiﬁcance level of 0.10, 0.05 and 0.01, respectively. All three independent variables in our regression appear statistically signiﬁcant at the 1-percent level. The R2 statistic we obtain is quite impressive (0.680), showing that our variables have relatively high explanatory power in accounting for private sector development. The quantitative impact of these variables is also substantial: onestandard-deviation increase in GOVINX, LAWYER and FINANCE is associated with an additional 0.64, 1.23 and 0.73 percentage point of the private sector’s share in total labor force. In column 2 of Table 4.6, we replace GOVINX with government size (GOVSIZE). As expected, the sign of GOVSIZE is negative and statistically signiﬁcant at a 56 Since private sector share in total labor force is bounded between 0 and 1, it may not be appropriate to use PRIVATE as the dependent variable. We used its logistic transformation as dependent variable: LOGIT PRIVATE = ln [PRIVATE/ (1 – PRIVATE)]. The results are qualitatively the same.

Local Government and Private Sector Development

Table 4.3

55

Private Sector Development and Regional Characteristics (1997–2001)

Provincial Lawyerto Central Labor Share in Gov. per Force Total Bank capita Ratio (per Loans (%) Spending 10,000 Ratio Persons) Beijing 1.46 7.13 0.76 9.52 0.58 9.09 Tianjin 7.74 5.33 0.57 4.25 0.88 5.40 Hebei 4.70 8.41 0.71 1.23 0.64 1.72 Shanxi 2.87 4.57 1.57 2.01 0.60 1.92 Inner Mongolia 3.43 1.96 1.78 2.06 0.96 2.80 Liaoning 5.88 7.35 0.70 2.68 0.46 3.47 Jilin 2.68 6.09 1.00 2.31 0.46 2.72 Heilongjiang 2.86 5.23 0.79 1.97 0.51 2.77 Shanghai 17.62 6.70 0.48 8.14 0.26 11.95 Jiangsu 5.09 8.44 0.62 1.75 0.43 2.21 Zhejiang 8.38 8.59 0.72 1.91 2.32 2.53 Anhui 1.61 4.84 1.04 1.08 0.54 1.46 Fujian 3.60 5.54 0.71 1.89 1.29 2.72 Jiangxi 2.32 7.38 1.00 1.24 0.74 1.54 Shandong 3.51 7.66 0.72 1.48 0.45 1.97 Henan 1.07 7.70 0.93 1.09 0.46 1.29 Hubei 2.92 7.26 0.83 1.66 0.37 1.73 Hunan 1.45 8.01 0.90 1.33 0.46 1.51 Guangdong 5.16 9.00 0.91 2.44 0.78 4.01 Hainan 4.82 7.61 1.36 2.46 0.37 2.57 Guangxi 1.04 7.22 1.32 1.01 0.71 1.58 Chongqing 2.61 8.34 1.04 1.88 0.56 1.57 Sichuan 1.59 7.42 1.20 1.28 0.53 1.44 Guizhou 1.17 5.82 2.42 0.75 1.00 1.46 Yunnan 1.33 4.84 1.90 1.29 0.84 2.88 Shaanxi 3.61 4.53 1.64 1.53 0.34 1.88 Gansu 1.58 5.57 1.96 1.19 0.73 1.91 Qinghai 3.28 2.82 2.58 1.78 0.56 3.54 Ningxia 3.24 3.09 1.70 2.62 0.84 3.08 Xinjiang 2.79 4.59 1.69 3.17 0.63 3.04 Notes: Data include both urban and rural areas. Private sector does not include selfemployed individuals. Government index is averaged over 1997–1999. Sources: See p.54. Share in Total Gov. Index Labor Force (%)

Gov. Admin. Expend. to GDP Ratio (%)

The Chinese Economy after WTO Accession

56

Table 4.4

Summary Statistics of the Variables

Mean Std. Dev. Minimum Maximum Private 3.298 1.892 1.039 8.377 Government Index 6.256 1.900 1.960 9.003 Government Size 1.225 0.553 0.570 2.577 Lawyer 1.833 0.752 0.750 4.248 External Finance 0.696 0.388 0.337 2.320 Fiscal Decentralization 2.382 0.950 1.289 5.396 Notes: The two outliers, Beijing and Shanghai, are excluded in the analysis.

Table 4.5

Correlation Patterns Gov. Index -0.649 0.000

Gov. Size –

–

–

–

–

–

–

Lawyer

-0.194 0.322

-0.209 0.287

–

–

–

–

–

–

External Finance

0.004 0.984

-0.024 0.903

0.068 0.732

–

–

–

–

–

-0.288 0.138

-0.042 0.832

0.849 0.000

0.191 0.331

–

–

–

–

Growth

0.290 0.135

-0.560 0.002

0.030 0.879

0.123 0.533

0.045 0.820

–

–

–

Education

0.134 0.498

-0.531 0.004

0.121 0.541

-0.031 0.876

0.005 0.981

0.397 0.037

–

–

Openness

0.303 0.118

-0.354 0.065

0.456 0.015

0.270 0.165

0.603 0.001

0.159 0.419

0.223 0.255

–

Initial 0.270 -0.621 0.670 0.281 0.679 0.431 Income 0.165 0.000 0.000 0.147 0.000 0.022 Note: P-values are given under the correlation coefﬁcients.

0.267 0.169

0.695 0.000

Gov. Size

FD

Lawyer Finance

FD

Growth Education Open

percent level. GOVSIZE is also quantitatively signiﬁcant: one-standard-deviation increase in GOVSIZE results in a reduction of 0.60 percentage point of the private sector’s share. We add ﬁscal decentralization (FD) into the regression and the results are summarized in column 3 and column 4 of Table 4.6. FD has a positive sign but appears only marginally signiﬁcant. The p-values of these two speciﬁcations are 0.16 and 0.12, respectively. Regarding the quantitative impact of FD: in both speciﬁcations, one-standard-deviation increase in FD results in an additional 0.58 percentage

Local Government and Private Sector Development

Table 4.6

57

PRIVATE Regression I

Constant Government Index

(1) OLS -3.267 (-3.35)*** 0.334 (2.99)***

(2) OLS 0.745 (0.80) –

(4) OLS 0.847 (1.025) –

-1.236 (-3.25)*** 1.726 1.097 0.705 Lawyer (6.97)*** (1.97)* (1.28) 1.883 1.680 1.69 External Finance (3.50)*** (2.70)*** (2.24)** Fiscal 0.611 0.629 – – Decentralization (1.43) (1.60) R2 0.680 0.667 0.703 0.691 Note: Numbers in parentheses are t-statistics based on the White consistent covariance estimates. * Signiﬁcant at the 0.10 level. ** Signiﬁcant at the 0.05 level. *** Signiﬁcant at the 0.01 level. Government Size

–

-1.082 (-2.42)** 1.396 (5.37)*** 1.896 (2.93)***

(3) OLS -3.678 (-3.59)*** 0.374 (3.34)*** –

point of private sector’s share. The slight increase in R2 statistics indicates that the marginal explanatory power of FD is weak. When FD is included, the t-statistic of LAWYER deteriorates. This is not surprising, given the fact that FD and LAWYER are highly correlated. In the following analyses, only GOVINX will be used as a measure of government intervention. It turns out that our two measures of government intervention yield similar results. Robustness Checks The above analyses may suffer from omitted variable bias, because other relevant variables, especially economic growth, are not controlled for. In the ﬁrst column of Table 4.7, we include GROWTH as a control variable in our regression. All variables have positive signs. Only FD is not statistically signiﬁcant at the 10-percent level. One-standard-deviation increase in GROWTH is associated with an increase of 0.55 percentage point in private sector’s share. Since economic growth and private sector development are simultaneous in nature, we use the two-stage least squares (2SLS) estimation procedures to deal with the potential simultaneity. To generate 2SLS estimates, we ﬁrst regress GROWTH on GOVINX, LAWYER, FINANCE, FD, EDU, OPEN and INITIAL, and then use ﬁtted value of GROWTH in the second stage (PRIVATE) regression. The estimation results are presented in column 2 of Table 4.7. The coefﬁcient of GROWTH increases from

The Chinese Economy after WTO Accession

58

Table 4.7

Constant Gov. Index Lawyer External Finance Fiscal Decent. Growth Education

PRIVATE Regression II (1) OLS -6.381 (-4.58)*** 0.280 (2.52)** 1.158 (2.54)** 1.548 (3.02)*** 0.514 (1.27) 0.436 (2.71)** -

(2) 2SLS -8.19 (-5.98)*** 0.217 (2.103)** 1.138 (2.32)** 1.460 (3.25)*** 0.449 (1.54) 0.728 (3.32)*** -

-

-

(3) OLS -6.242 (-5.29)*** 0.280 (3.06)*** 1.158 (2.54)** 1.544 (3.13)*** 0.501 (1.53) 0.448 (2.84)*** -0.427 (-0.21) -

(4) 2SLS -8.079 (-5.07)*** 0.177 (1.51) 1.351 (2.95)*** 1.375 (3.68)*** 0.293 (0.93) 1.004 (3.64)*** -0.372 (-1.57) -

(5) OLS -6.335 (-4.67)*** 0.275 (1.96)* 1.144 (2.53)** 1.543 (3.00)*** 0.496 (1.14) 0.437 (2.94)*** -

(6) 2SLS -8.052 (-5.64)*** 0.198 (1.21) 1.187 (2.83)*** 1.442 (3.21)*** 0.386 (0.92) 0.734 (3.13)*** -

0.0011 (0.07) 0.777 0.744 0.778 0.676 0.777 Numbers in parentheses are t-statistics based on the White consistent

Openness

R2 Note: estimates. * Signiﬁcant at the 0.10 level. ** Signiﬁcant at the 0.05 level. *** Signiﬁcant at the 0.01 level.

0.0039 (0.23) 0.743 covariance

0.436 to 0.728, indicating a large gain in quantitative signiﬁcance. Other than this, the results do not change substantially. To further check the robustness of our results, we include education (EDU) in our OLS regression. The results are reported in column 3 of Table 4.7. The sign of EDU is negative and statistically insigniﬁcant. The R2 statistic almost remains unchanged. We also conduct 2SLS procedure, adding EDU in the second stage (PRIVATE) regression. Column 4 of Table 4.7 shows the estimation results. Again, the sign of EDU is negative and statistically insigniﬁcant. In column 5 and column 6 of Table 4.7, we add OPEN to OLS and 2SLS regressions. The results are similar to those of EDU, implying that EDU and OPEN are probably irrelevant variables in PRIVATE regression. However, since GROWTH is signiﬁcant, EDU and OPEN may have indirect effect on PRIVATE through GROWTH. In summary, we ﬁnd that GOVINX, GOVSIZE, LAWYER, FINANCE, and GROWTH exhibit a statistically and quantitatively signiﬁcant relationship with private sector development. FD is only marginally signiﬁcant, but since our sample size is relatively small and our time horizon is relatively short, caution should be taken in basing conclusions on this ﬁnding.

Local Government and Private Sector Development

59

Conclusions This chapter examines the relationship between local government and the private sector in China, drawing heavily from the existing literature on economic transition. We use a regional dataset between 1997 and 2001 and ﬁnd that government intervention, law enforcement and government policy all have strong and enduring relationships with the private sector. The evidence on the inﬂuence of ﬁscal autonomy is relatively weak. It is also found that economic growth is positively related to the private sector, indicating a possible two-way relationship between them. In addition, the direct effect of education and openness on the private sector appears insigniﬁcant. It is clear that institutions, regulations, and policies that stiﬂe entrepreneurship will erode the dynamism of the economy. The policy implications are quite straightforward: in order to protect the private property, the government should not be too strong, but it has to be strong enough to enforce the rule of law. This is the dilemma that most Chinese local governments currently face. References Barro, R. (1991), ‘Economic Growth in a Cross Section of Countries’, Quarterly Journal of Economics, Vol. 106, pp. 408–43. Berkowitz, D. and DeJong, D. (2002), ‘Accounting for Growth in Post-Soviet Russia’, Regional Science and Urban Economics, Vol. 32, pp. 221–39. Berkowitz, D. and DeJong D. (2003), ‘Policy Reform and Growth in Post-Soviet Russia’, European Economic Review, Vol. 47, pp. 141–57. Berkowitz, D. and DeJong, D. (2005), ‘Entrepreneurship and Growth in Post-Soviet Russia’, Oxford Bulletin of Economics and Statistics, Vol. 67, pp. 25–46. Blanchard, O. (1997), The Economics of Post-Communist Transition, Oxford University Press, London. Blanchard, O. and Shleifer, A. (2001), ‘Federalism With and Without Political Centralization: China vs. Russia’, IMF Staff Working Papers, Special Issue. Brandt, L. and Li, H. (2002), ‘Bank Discrimination in Transition Economies: Ideology, Information or Incentives’, William Davidson Working Paper, Number 517. Brezinski, H. and Fritsch, M. (1996), The Economic Impact of New Firms in the Post-Socialist Countries, Edward Elgar, Cheltenham, UK. Cao, Y., Qian, Y. and Weingast, B. (1999), ‘From Federalism, Chinese Style, to Privatization, Chinese style’, Economics of Transition, Vol. 7, pp. 103–31. China Reform Foundation (1997), Xianshi De Xuanze: Guoyou Xiao Qiye Gaige Shijian De Chubu Zongjie, Shanghai Yuandong Press, Shanghai. China Statistical Yearbook 2003, China Statistics Press, Beijing. Clark, D. (1994), ‘The Creation of a Legal Strucutre for Market Institution in China’, in J. McMillan and B. Naughton (eds), Reforming Asian Socialism: The Growth of Market Institutions, University of Michigan Press, Ann Arbor.

60

The Chinese Economy after WTO Accession

Fan, G. and Wang, X. (2001), Marketization Index, Economic Sciences Press, Beijing. Fischer, S., Sahay, R. and Vegh, C. (1996). ‘Stabilization and Growth in Transition Economies: The Early Experience’, Journal of Economic Perspectives, Vol. 10, pp. 44–66. Frye, T. and Shleifer, A. (1997), ‘The Invisible Hands and the Grabbing Hands’, American Economic Review Paper and Proceedings, Vol. 87, pp. 354–58. Hayek, F. [1944] (1994), Road to Serfdom, University of Chicago Press, Chicago. International Finance Corporation (2000), China’s Emerging Private Enterprises: Prospects for the New Century, World Bank Publications, Washington DC. Jin, H., Qian, Y. and Weingast, B. (2005). ‘Regional Decentralization and Fiscal Incentives: Federalism, Chinese Style’, Journal of Public Economics, Vol. 89, pp. 1719–42. Johnson, S., McMillan, J. and Woodruff, C. (1999), ‘Contract Enforcement in Transition’, working paper, Graduate school of Business, Stanford University. Johnson, S., McMillan, J. and Woodruff, C. (2000), ‘Entrepreneurs and the Ordering of Institutional Reform’, Economics of Transition, Vol. 8, pp. 1–36. Johnson, S., McMillan, J. and Woodruff, C. (2002a), ‘Property Rights and Finance’, American Economic Review, Vol. 92, pp. 1335–56. Johnson, S., McMillan, J. and Woodruff, C. (2002b), ‘Court and Relational Contracts’, Journal of Law, Economic & Organization, Vol. 18, pp. 221–77. Kornai, J. (1992). The Socialist System: The Political Economy of Communism, Princeton University Press, Princeton. Li, D. (2001), ‘Why do Governments Dump State Enterprises? Evidence from China’, Twelfth Annual East Asian Seminar on Economics, Hong Kong. Li, H., Rozelle, S. and Brandt, L. (1999), ‘Saving or Stripping Rural Industry: An Analysis of Privatization and Efﬁciency in China’, American Agricultural Preconference, Nashville. Li, S., Li, S. and Zhang, W. (2000), ‘Road to Capitalism: Competition and Institutional Change in China’, Journal of Comparative Economics, Vol. 28, pp. 269–92. Lin, J. and Liu, Z. (2000), ‘Fiscal Decentralization and Economic Growth in China’, Economic Development and Cultural Change, Vol. 49, pp. 1–21. Lizal, L. and Svejnar, J. (2002), ‘Investment, Credit Rationing and the Soft Budget Constraint: Evidence from Czech Panel Data’, Review of Economics and Statistics, Vol. 84, pp. 353–70. Marshall, A. (1890), Principles of Economics, McMillan, London. McMillan, J. (1995), ‘China’s Nonconformist Reforms’, in E. Lazear (eds), Economic Transition in Eastern Europe and Russia: Realities of Reform, Hoover Institution Press, Stanford. McMillan, J. and Woodruff, C. (1999), ‘Interﬁrm Relationships and Informal Credit in Vietnam’, Quarterly Journal of Economics, Vol. 114, pp. 1285–1320. McMillan, J., and Woodruff, C. (2002), ‘The Central Role of Entrepreneurs in Economic Transition’, Journal of Economic Perspective, Vol. 16, pp. 153–70. North, D. (1981), Growth and Structural Change, W. W. Norton, New York.

Local Government and Private Sector Development

61

North, D. (1990), Institutions, Institutional Change and Economic Performance, Cambridge University Press, UK. Pissarides, F., Singer, M. and Svejnar, J. (2003), ‘Objectives and Constraints of Entrepreneurs: Evidence from Small and Medium Size Enterprises in Russia and Bulgaria’, Journal of Comparative Economics, Vol. 31, pp. 503–31. Qian, Y. and Roland, G. (1998), ‘Federalism and Soft Budget Constraint’, American Economic Review, Vol. 88, pp. 1143–62. Rawski, T. (1999), ‘The Political Economy of China’s Declining Growth’, working paper, Department of Economics, University of Pittsburgh. Rawski, T. (2001), ‘What’s Happening to China’s GDP Statistics?’ China Economic Review, Vol. 12, pp. 347–54. Shleifer, A. (1997), ‘Schumpeter Lecture: Government in Transition’, European Economic Review, Vol. 41, pp. 385–410. Smith, A. [1776] (1976), An Inquiry into the Nature and Causes of the Wealth of Nations, University of Chicago Press, Chicago. Sonobe, T. and Otsuka, K. (2001), ‘Productivity Effects of TVE Privatization: The Case Study of Garment and Metal Casting Enterprises in the Greater Yangtze River Region’, Twelfth Annual East Asian Seminar on Economics, Hong Kong. Tsai, K. (2002), Back-Alley Banking: Private Entrepreneurs in China, Cornell University Press, Ithaca. Yao, Y. (1999), ‘The Size of China’s Private Sector’, working paper, CCER, Peking University. Zhang, T. and Zou, H. (1998), ‘Fiscal Decentralization, Public Spending, and Economic Growth in China,’ Journal of Public Economics, Vol. 67, pp. 221–40. Zhao, X. (1999), ‘Competition, Public Choice and Institutional Change’, working paper No. C1999025, CCER, Peking University. Zhuravskaya, E. (2000), ‘Incentive to Provide Local Public Goods: Fiscal Federalism, Russian Style’, Journal of Public Economics, Vol. 76, pp. 337–68.

This page intentionally left blank

Chapter 5

Globalization and Privatization: Evidence from China Jian Su

Introduction Globalization and privatization are two of the most important and interesting phenomena in current world economic and political relations. While much research has been done on each of them and on their impacts on other aspects of the world economy,1 no one has examined the interaction between them. This chapter is to ﬁll this gap. Why do we choose the case of China? Many reasons help justify this choice, but the most important reason is that the experience of China offers a quasi-natural experiment on both globalization and privatization and their relationships, and thus provides an excellent opportunity to study the interaction between globalization and privatization. First, in order to conduct valid econometric analysis, there should be sufﬁcient variation in the data on both the globalization and privatization processes. For many other countries, the globalization process began much earlier than the privatization process. Thus, when the privatization process of such countries was initiated, they were already quite globalized and thus further changes in their degree of globalization become quite small and, therefore, the variation needed to conduct valid statistical analysis may not be sufﬁcient, especially when privatization is carried out over a very short period of time. In China, the variations in the data on both the globalization and privatization processes are guaranteed for two reasons: ﬁrst, the globalization and privatization processes were started simultaneously in 1978, when Deng Xiaoping became the top leader of China and the ‘reform and 1 For example, Sachs and Warner (1995) documented the process of global integration and assessed its effects on economic growth in the reforming countries; Rodrik (1998) studied the effects of international trade on the size of governments; Blonigen and Figlio (1998) studied the effects of foreign direct investment on legislator behavior; Baily and Gersbach (1995) studied the relation between global competition and productivity; Florida (1996) studied the relationship between globalization and regional economic transformation; Wei and Wu (2001) studied the effects of globalization on the income inequality using the regional data of China. There are even more papers on privatization.

64

The Chinese Economy after WTO Accession

openness’ policy was announced; also, both globalization and privatization have been carried out smoothly since then. Second, the decentralized nature of the privatization process in China allows us to study the natural relationship between globalization and privatization. As Su and Jefferson (2003) suggest, the privatization of Chinese ﬁrms is decentralized, especially that of the large and medium enterprises, which means that the privatization process of a ﬁrm is usually initiated by the ﬁrm itself. Thus, the privatization of a state-owned enterprise (SOE) in China is in fact a kind of natural evolution of the ﬁrm’s institution. This is opposite to the privatization process of other transitional countries, in which the privatization process has been controlled by government and usually carried out in a very short period of time. In the latter case, the objective of government dominates any other factors that may affect privatization, and thus, although other factors may affect the privatization process of ﬁrms, their effects may be relatively small and usually cannot be identiﬁed. Third, both the globalization and privatization processes of China were conducted gradually and continuously, and unlike other countries that experienced transformation of political power which accompanies the process of privatization and which usually makes data before and after privatization not comparable, China has a relatively stable political system, which allows the continuous collecting of data and the comparability of data before and after the privatization process. During the last several decades, the National Bureau of Statistics of China has continuously collected data, and is able to prepare a ﬁrm-level dataset on the population of the large and medium industrial enterprises of China that spans the period from 1995 to 2001. This unique dataset makes our analysis possible. The chapter identiﬁes three channels of globalization: the outﬂow of goods, the inﬂow of foreign capital (FDI), and the inﬂow of knowledge. This chapter shows that globalization and privatization are mutually reinforcing. The rest of this chapter is organized as follows. The next section is an introduction of the globalization and privatization processes of China. The third section examines the effects of globalization on privatization. The main ﬁnding is that exports, capital inﬂows, and the inﬂow of knowledge all have positive effects on the probability of an SOE’s privatization. In the fourth section, the effects of privatization on globalization are examined. We ﬁnd that privatization has signiﬁcant and positive effects on the growth rates of all three ﬂows mentioned above. The ﬁfth section discusses more about methodology. This section explains why simultaneous equation system and panel data models are not applicable to this research. The sixth section concludes this chapter. Reform and Openness: Twin Themes in the Post-Mao Era of China The Mao era (1949–76) of China has two characteristics. One is the predominance of state ownership of production materials, especially capital. According to the Communist ideology, public ownership, with state ownership as its major form

Globalization and Privatization: Evidence from China

65

of realization, is more advanced and efﬁcient than private ownership because the working class becomes the real master of the country and thus the workers would be more concerned about the performance of their ﬁrms than when they were working for the capitalists. Public ownership is also regarded as the economic foundation of a socialist system; private ownership was regarded as that of the capitalist system and thus a force that potentially threatens the political power of the Communist Party. The second characteristic of the Mao era is the famous policy of ‘Independence and Autonomy’ (duli zizhu), which in fact meant to close China’s door to the outside world. By this policy, the Chinese government tried to do everything by itself, to get rid of China’s dependence on any foreign countries and international organizations. The Chinese government was proud of having neither domestic nor foreign debt. Because of the hostility of the Chinese government against private ownership, no foreign investment was accepted during that period. International trade was relatively small; the ratios of imports and exports to GDP were around 4 percent for each year from 1952 to 1976. Exchange in personnel and academics was also very limited. After a short period of transition following Mao’s death in 1976, Deng Xiaoping became the top leader of China in 1978. After 10 years of the Cultural Revolution, the economic efﬁciency of China was too low and that Chinese economy was close to collapse. In order to help the Chinese economy recover and develop, Deng proposed the policy of ‘reform and openness’, which was passed in 1978 in the Third Plenary Meeting of the Eleventh Congress of the Communist Party of China (CPC). With the proposal and passage of the ‘reform and openness’ policy, both the globalization and privatization processes of China were initiated. The Evolution of Ownership Structure and Privatization in China since 1978 Before the beginning of economic reform in the late 1970s, there were in fact only two ownership types in the Chinese economy: state ownership and collective ownership. Since the beginning of reform and openness to the outside world in the late 1970s, the ownership structure of China changed gradually. The Chinese government began to allow foreign companies and small private companies to co-exist with stateowned and collective-owned enterprises. By 1992, the number of ownership types in China had increased to 21, as shown in Appendix 1. In 1998, a new statistical system concerning ownership classiﬁcations was introduced, and the number of ownership types increased to 23. Appendix 1 also shows the correspondence between the old and new ownership classiﬁcations. From Appendix 1, we can see all the ownership types that existed in China from 1992 to 2001. In this chapter, three ownership types that appear in Appendix 1 are considered as state ownership: state-owned enterprises, state-owned jointly operated enterprises, and wholly state-owned companies. All the other ownership types are classiﬁed as non-state ownership. Rows 2 and 3 of Table 5.1 show the number of SOEs and the percentage of SOEs in large and medium industrial enterprises (LMIEs) from 1995 to 2001, based on the dataset used in this research. From Table 5.1 we can see clearly the steady decrease in the number and proportion of state-owned enterprises and the steady increase for

66

Table 5.1

The Chinese Economy after WTO Accession

Number of LMIEs and SOEs (1995–2000)

1995 1 Total number 22,892 of LMIEs 2 Number of SOEs 15,533 3 Percentage of SOEs 67.9 (100*(2)/(1)) 4 Number of stable – SOEs from the previous year 5 Number of – privatized SOEs Source: NBS dataset.

1996 24,161

1997 24,260

1998 23,752

1999 22,598

2000 20,789

2001 22,878

15,331 63.5

13,100 54.0

12,785 53.8

10,451 46.2

8,499 40.9

8,675 37.9

13,292

13,789

12,184

9,853

8,705

7,471

383

520

1223

628

738

572

the other ownership types. In 1995, 67.9 percent of the LMIEs were state-owned; while in 2001, 37.9 percent were state-owned. The deﬁnition of privatization needs some explanation. The actual privatization process is very complicated. The speed and strategy of a ﬁrm’s privatization process vary signiﬁcantly across countries and ﬁrms because of the policies the countries adopted and other ﬁrm-speciﬁc characteristics. Some SOEs may change to private enterprises quickly and directly, while many others may take several steps and take several years to change to private ownership. For example, an SOE may ﬁrst change to a limited liability company, then a corporation, and ﬁnally take complete privatization if the state sells all its shares. Thus, when should we say that a ﬁrm is privatized: the time when it deviates from full state ownership, or the time when it completes the privatization process, or any other time between them? By now, in most research in this area, a ﬁrm is said to be privatized when private ownership is above some threshold. For example, for Estrin and Rosevers (1999), the threshold is any positive level of private ownership (that is, 0 percent); for Jones and Mygind (2002), it is 50 percent; for Frydman et al. (1999), 33.3 percent; and for Claessens and Djankov (2000), 66.7 percent. In this chapter, privatization is deﬁned as the formal reclassiﬁcation, by China’s National Bureau of Statistics, of a ﬁrm’s ownership. When a Chinese ﬁrm is registered as an SOE in one year and in the next year is registered as another kind of ownership, we assume that the enterprise has been privatized. This deﬁnition is justiﬁed by two reasons. First, the change in ownership classiﬁcation usually indicates the changes in ownership structure of the ﬁrm. Second, in China, the change in ownership type is also important for a ﬁrm, because many policies of the Chinese government are based on the ownership type. For example, a ﬁrm classiﬁed as foreign ownership is eligible to enjoy certain favorable policies of the Chinese government, such as low tax in the ﬁrst several years of operation and the eligibility of entry into some industries. Likewise, an SOE usually can enjoy low-interest loans and other kinds of subsidies.

Globalization and Privatization: Evidence from China

67

The Globalization Process in China While the word ‘globalization’ is being used widely now, it has different meanings to different people. To most people who talk about it, globalization means a process that leads to the increase in economic, cultural, and political relations between people of different countries. The economic side of globalization has three channels: ﬂow of goods and services, ﬂow of productive factors, and ﬂow of knowledge. The implementation of the openness policy leads to rapid globalization of Chinese economy. Goods and servicesThe exports and imports of China have exploded since the ‘reform and openness’ policy was put into practice. During the period from 1949 to 1978, both exports and imports were small in terms of their absolute values and their ratios to GDP.2 In 1978, exports were US$9.75 billion, and imports were US$10.89 billion; their shares in GDP were 4.6 percent and 5.2 percent respectively. In 2000, exports were US$249.2 billion, and imports were US$225.1 billion, their shares in GDP were 23.1 percent and 20.8 percent respectively. During the period from 1978 to 2001, the average annual growth rate of exports and imports were 14.7 percent and 13.8 percent respectively.3 With the entry of China into the WTO, we expect that the exports and imports will expand further in the coming years. CapitalAs a consequence of the implementation of the ‘independence and autonomy’ policy and the hostility of the Chinese government toward private ownership, there was no FDI in China before 1978. With the openness policy implemented, the Chinese government began to allow foreigners to invest in China. Since then, the inﬂow of foreign capital has sky-rocketed. In 1984, FDI to China was US$1.258 billion, while by 2000 it had increased to US$40.715 billion, with an average annual growth rate of 21.7 percent.4 KnowledgeThe inﬂow into China of knowledge about modern science and technology is mainly in the form of learning, especially that from formal education. During the period of Cultural Revolution (1966–1976), the educational system of China was completely destroyed. Colleges and universities were closed, most professors and scientists were sent to factories and the countryside to be reformed by working. In 1977, one year after the death of Mao Zedong and the end of the Cultural Revolution, Deng Xiaoping returned to the political stage of China and restored all levels of education to their normal state, reopened the colleges and universities, and

2 The shares of exports and imports in GDP were around 4 percent during this period. 3 The source of data is China Statistical Yearbook, various issues. The growth rates and percentages were calculated by the author. 4 During this period, many Chinese ﬁrms also invested in other countries. But the quantity is small.

The Chinese Economy after WTO Accession

68

called back the professors and scientists. The ﬁrst batch of college and university students was admitted in late 1977. Since 1977, education at all levels has expanded signiﬁcantly. The number of students admitted to colleges and universities increased year by year. In 1978, the number was 402,000; in 2000, this number was 2.2 million, with an average annual growth rate of 7.7 percent. The universities also began to offer Master’s and Ph. D. programs. The number of graduate admissions increased from 10,708 in 1978 to 128,484 in 2001. Besides the restoration of the domestic educational system, studying abroad became popular. The number of students or scholars studying abroad increased from 860 in 1978 to 38,989 in 2000, an average annual growth rate of 16.6 percent. In total, during this period, 223,356 students went abroad to study, and 77,992 returned to China.5 They have been playing active and important roles in various ﬁelds such as education, scientiﬁc research, ﬁnance, and business. One outstanding example is that, in 1998, of the 629 fellows of the Chinese Academy of Science, 507 had the experience of studying or working abroad (80.6 percent); of the 423 fellows of the Chinese Academy of Engineering, 227 had the experience of studying or working abroad (53.7 percent) (Ministry of Education, 2000). The channels of globalization studied in this chapterBecause any ‘ﬂows’ have two directions, i.e., inﬂows and outﬂows, it would be better to analyze the interaction between ﬂows in both directions and privatization. Unfortunately, ﬁrm-level data for all such ﬂows are not available. As Table 5.2 shows, the ﬂows with data availability for this research are outﬂow of goods and services, inﬂow of capital, and inﬂow of knowledge. This chapter thus examined two issues. The ﬁrst is the effects of the above three ﬂows on the probability of a ﬁrm’s privatization. The second is the effect of privatization on the average annual growth rates of the three ﬂows.

Table 5.2

The Channels of Globalization and Data Availability

Channels of Globalization Goods and services Productive factors

Capital

Labor Knowledge

Inﬂow/Outﬂow Inﬂow Outﬂow

Data availability No Yes

Inﬂow

Yes

Outﬂow

No

Inﬂow

No

Outﬂow Inﬂow

No Yes

Outﬂow

No

5 The ﬁgures here are based on Table 20-8 of China Statistical Yearbook, National Bureau of Statistics, 2001.

Globalization and Privatization: Evidence from China

69

Effect of Globalization on the Probability of Privatization In this section, we examine empirically the effects of the three ﬂows on the probability of a ﬁrm’s privatization: exports, inﬂow of foreign capital, and inﬂow of knowledge. Testable Hypothesis The effect of tradeThe effect of trade stems from the more intense competition a ﬁrm faces when it enters the international market. Generally, international markets are much more competitive than domestic markets. When facing more severe competition, the ﬁrm has to have better production organization to survive. Two papers related to our research are Baily and Gersbach (1995) and Florida (1996). Baily and Gersbach (1995) considered the effects of globalization on the choices of enterprises concerning their production technology. They found a positive causal relation between global competition and productivity of the production process a ﬁrm chooses. As they put it, ‘Trade and direct foreign investment force the adoption of bestpractice production process and facilitate the transfer of technology,’ (pp. 346–347) and ‘Domestic competition alone … is not enough to stimulate operations to achieve the highest productivity’ (p. 346). Florida (1996) studied the relationship between globalization and regional economic transformation and found that the adoption and diffusion of new forms of work and production organization were accelerated by globalization. The channel through which competition shapes institutions may be learning, as speciﬁed by North (1994). North (1994) suggests, ‘The most fundamental long-run source of change [of institutions] is learning by individuals and entrepreneurs of organizations’ (p. 362), while ‘… the rate of learning will reﬂect the intensity of competition among organizations. Competition … induces organizations to engage in learning to survive. … The greater the degree of monopoly power, the lower is the incentive to learn’ (p. 362). North’s statements give a clear picture of how competition affects institutional changes: competition forces ﬁrms to learn, while learning leads to institutional change. Thus, we can get the following testable hypothesis: Hypothesis I: The higher the share of a ﬁrm’s exports in sales revenue, the more probable that it will be privatized.

This hypothesis is also consistent with the results of other research concerning the effect of competition on a ﬁrm’s choice of production organization and/or production process. Baily and Gersbach (1995) found that the more competitive is the market a ﬁrm faces, the more efﬁcient the production process the ﬁrm would choose. By comparing the efﬁciency of service industries of several countries, Baily (1993) supported ‘the idea that competition encourages efﬁciency’ (p. 130). Based on panel data from 1975 to 1987 of 94 manufacturing industries, MacDonald (1994) found that ‘increases in imports competition led to large increases in labor productivity growth in highly concentrated industries’ (p. 721).

70

The Chinese Economy after WTO Accession

The effect of capital inﬂowAccording to ‘The Foreign Enterprise Law of the PRC’ issued in 1986, when the proportion owned by foreigners is or exceeds 25 percent of the total registered capital, the ownership classiﬁcation of the ﬁrm can be changed to foreign enterprise. Thus, higher inﬂow of foreign capital seems certain to lead to a higher privatization probability. But the true story is not that simple. For example, suppose a foreign investor owns a proportion of an SOE, say 24 percent. The foreign owner (or other foreign investors) now needs to decide whether to invest more in this ﬁrm or not. If they do invest more, the SOE may become a foreign enterprise. But they may also sell their share of the ﬁrm to other persons, including the state, if they think this investment is not good. Thus partial ownership by a foreigner by itself does not naturally lead to privatization. As Su and Jefferson (2003) suggest, a ﬁrm’s desired future production organization is better if the efﬁciency of its present production organization is higher. That is, there is a positive feedback mechanism in the evolution of production organization: other factors given, the more efﬁcient is the present production organization, the more efﬁcient is the desired one; once the production organization is improved and gets to a higher level, even better production organization will be desired by the ﬁrm. The economic motive behind this mechanism is the pursuit of proﬁts. As a proﬁtmaximizer, the ﬁrm is willing to improve its production organization any time; thus the more efﬁcient is its present production organization, the more efﬁcient an organization it desires. The inﬂow of capital of other ownership (especially foreign capital), even just a small amount, usually leads to a change in corporate governance. For an SOE with just a small amount of foreign investment, although the ownership classiﬁcation is still state ownership, its corporate governance has changed because of the introduction of new owners. The change in corporate governance will usually lead to improvement of production organization and thus production efﬁciency, especially when FDI is involved, which usually comes with more advanced management. Because state ownership is the least efﬁcient production organization among all the ownership types, a demand for more efﬁcient production organization usually means a higher probability of privatization of the ﬁrm. Thus we get the following testable hypothesis: Hypothesis II: The higher the present level of foreign capital share, the more probable that the ﬁrm will be privatized.

The effect of knowledge inﬂowIntuitively, inﬂow of knowledge should have positive effects on the probability of privatization. With more knowledge about the production process and production organization in other countries, the options for organizing production increase. Knowledge is usually embodied in the employees of a ﬁrm in the form of human capital, which is necessary for the ﬁrm to design the most efﬁcient production organization and carry out the reform process. Thus we have the following testable hypothesis: Hypothesis III: The more human capital a ﬁrm has, the more probable that the ﬁrm will be privatized.

Globalization and Privatization: Evidence from China

71

Econometric Model A Probit model is used in this section. In this model, the dependent variable is coded one if the ﬁrm took ownership change in a year and zero if it did not. The speciﬁcation of the model is: yit =α0 +Gα+Xβ+ μit

(I)

where G is a vector representing the three channels of globalization; X is a vector including some economic variables, year dummies, size dummies, industry dummies and regions dummies. X is included only when robustness of the empirical results is being tested. In order to avoid a possible endogeneity problem, the values of variables on the right hand side are lagged one year, except for year dummies, size dummies, industry dummies, and regions dummies, which are clearly exogenous. The next question concerns how to measure the three channels of globalization. In this section, goods outﬂow is measured by the ratio of a ﬁrm’s exports to its sales revenue; capital inﬂow is measured by the proportion of foreign ownership of the ﬁrm, that is, the ratio of foreign direct investment6 to the total capital of a ﬁrm. As we explained above, the inﬂow of knowledge is embodied in employees as human capital, which is measured here by the ratio of technology development personnel to the total employment of the ﬁrm. Ideally, we need also to include the personnel who studied management, economics or ﬁnance, but unfortunately, we don’t have such data. We assume that a ﬁrm that hires more technology development personnel also hires more personnel of management, economics or ﬁnance. Data The dataset was provided by the National Bureau of Statistics (NBS) of China. This dataset includes the data for China’s LMIEs from 1995 to 2000. The number of LMIEs from 1995 to 2000 is shown in Table 5.1. We can see that the total number of LMIEs in Chinese industry changes by large numbers, ranging from 22,598 to 24,260 during the period from 1995 to 2000. Because the data on foreign capital and exports for 1996 and 1997 are not included in our dataset, the data used in this chapter are just those for 1995, 1998, 1999, and 2000. The dataset used in this section is formed in the following way. First, extract all the SOEs from the 1995 dataset. As Table 5.1 (row 2) shows, there were 15,533 SOEs in 1995. Second, extract those of them that still exist in 1996. The number is in Row 4 of Table 5.1. This number is 13,292 for the 1995–96 pair. Third, check the 1996 ownership type of the ﬁrms from step 2. If the ownership type is the same as in 1995, it means that this ﬁrm did not take ownership reform, which is privatization according to our deﬁnition, in 1996; if the ownership type is different, it means that 6 In this chapter, foreign capital refers to capital from foreign countries and from Hong Kong, Taiwan, and Macao (Aomen), which are regions of China but adopt different economic systems.

The Chinese Economy after WTO Accession

72

this ﬁrm did take ownership reform and was privatized in 1996. Fourth, create a dummy variable for ownership change: let the value of this variable equal 0 if the corresponding ﬁrm did not take ownership reform in 1996 and 1 if it took ownership reform in 1996. Fifth, create the ﬁnal dataset for the 1995–96 pair by integrating this ownership change dummy and the ﬁrms’ 1995 economic data. The purpose of doing so is to avoid a possible endogeneity problem, that is, to avoid the possibility that the values of the explaining variables are affected by the explained variable. Sixth, repeat the above steps for 1998–99, 1999–2000, and 2000–2001 pairs. Seventh, pool them. A year dummy is included to indicate the year when ownership change took place (or not). Eighth, clean the data by dropping observations with unreasonable values. The observations satisfying any of the following criteria are dropped: a. b. c. d.

The value of any one of the right hand side variables is missing; The ratio of exports to sales revenue is greater than 1; The ratio of proﬁts to sales revenue is greater than 1; The ratio of technology development personnel to total employment is greater than 1.

When these observations are dropped, we get a sample of 25,825 observations, in which 1,310 (5.07 percent) have ownership changes. The descriptive statistics for the variables are shown in Table 5.3.

Table 5.3

Descriptive Statistics for Variables Used in the Regressions

Ownership change dummy

Mean 0.051

Std. Dev. 0.219

Observations 25,825

Export-sales ratio

0.068

0.172

25,825

Share of foreign capital

0.010

0.061

25,825

Proﬁt-sales ratio

0.156

0.219

25,825

Subsidy-sales ratio Share of Technology Development Personnel in Total Employment Two-ﬁrm Concentration Ratio

0.011

0.062

25,825

0.042

0.076

25,825

23.082

17.272

25,825

Year Dummy for 1995

0.357

0.479

25,825

Year Dummy for 1998

0.242

0.428

25,825

Year Dummy for 1999

0.213

0.409

25,825

Year Dummy for 2000 Source: NBS dataset.

0.189

0.392

25,825

Globalization and Privatization: Evidence from China

73

Estimation Results Table 5.4 shows the estimation results with the one-year-lag data. Equation (1) is the benchmark model, in which only the three globalization variables are included. In equation (1), the three variables all display positive and signiﬁcant effect on the probability of privatization. The coefﬁcients of exports ratio and the share of technology development personnel are signiﬁcant at the 1 percent level, while the coefﬁcient of the share of foreign capital is signiﬁcant at the 5 percent level. These results suggest that globalization has positive effects on privatization. One possibility is that this result is due to variations in other aspects of the ﬁrm. In order to control for such variations and test the robustness of the above results, more variables are added from equation (2) to equation (8). We can see from Table

Table 5.4

Results with One-year-lag Data 1 0.232a (0.071) 0.465 b (0.186)

2 0.272a (0.073) 0.499a (0.189)

3 0.285a (0.073) 0.516a (0.190)

4 0.278a (0.073) 0.506a (0.190)

5 0.292a (0.073) 0.523a (0.190)

6 0.260a (0.074) 0.530a (0.192)

7 0.277a (0.075) 0.562a (0.192)

8 0.250a (0.081) 0.513a (0.193)

0.653a (0.153)

0.510a (0.155)

0.451a (0.157)

0.445a (0.157)

0.529a (0.158)

0.430a (0.161)

0.473a (0.161)

0.380 b (0.173)

–

–

0.369a (0.110)

–

–

–

0.339a (0.111) -0.701b (0.31)

0.354a (0.111) -0.687b (0.311)

0.326a (0.113) -0.722b (0.317)

0.330a (0.113) -0.689b (0.315)

0.509a (0.124) -0.385 (0.332)

–

–

–

–

-0.003a (0.001)

-0.003a (0.001)

-0.003a (0.001)

-0.003a (0.001)

-

yes

yes

yes

yes

yes

yes

yes

region

-

-

-

-

-

yes

yes

yes

size

-

-

-

-

-

-

yes

yes

industry

-

-

-

-

-

-

-

yes

Exports ratio Share of foreign capital Share of R&D personnel in total employment Ratio of proﬁts to sales revenue Ratio of subsidy to sales revenue Two-ﬁrm concentration ratio Year

LR statistic (57 36.358 372.458 384.568 390.330 406.133 533.099 541.840 717.634 df) Probability 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 (LR stat) McFadden 0.004 0.036 0.037 0.038 0.039 0.051 0.052 0.069 R-squared a b c d Note: signiﬁcant at 1%; signiﬁcant at 5%; signiﬁcant at 10%; signiﬁcant at 15%; the number of observations is 25,829, that with dependent variable=1 is 1,310. The number in parenthesis is the standard deviation.

74

The Chinese Economy after WTO Accession

5.4 that the estimated results for all the three channels of globalization are quite robust in all eight equations. As the ﬁrst trial, year dummies are included in equation (2). Year dummies are added to control for the macroeconomic ﬂuctuations and year-to-year variations in government policies concerning ownership reform. When year dummies are added, the signiﬁcance level of the share of capital inﬂow improves from 5 percent to 1 percent. The standard deviation of the estimated coefﬁcients of all three variables are quite robust, the value of the coefﬁcients of exports ratio and the share of foreign capital increase a little, while that of the share of technology development personnel in total employment decreases a little. In equation (3), the proﬁtability of the ﬁrms is added. Proﬁtability here equals the ratio of sales proﬁts of a ﬁrm to its sales revenue. This is to control for the effects of selection bias: more proﬁtable ﬁrms may be intentionally privatized ﬁrst or last, depending on the objectives of the government or the incentive facing ﬁrms.7 The estimation result here shows that the more proﬁtable ﬁrms are privatized ﬁrst. When proﬁtability is added, the signiﬁcance levels of the estimated coefﬁcients for globalization variables do not change, and the values of the estimated coefﬁcients also change little compared with equation (2). There is a kind of opportunity cost for an SOE to deviate from state ownership. Such opportunity cost includes the subsidy an SOE receives from government, the special favorable policies SOEs enjoy, and other beneﬁts. Such opportunity cost may affect the probability of the ﬁrm’s privatization. In order to control for the effects of such opportunity cost, the ratio of the subsidy an SOE receives to the ﬁrm’s sales revenue is added in equation (4). The estimation result shows that this ratio does have negative impact on the probability of privatization, but just at the 15 percent level of signiﬁcance. The signs and signiﬁcance levels of the coefﬁcients for the three globalization variables keep stable compared with equations (2) and (3). Competition in the domestic market may affect a ﬁrm’s probability of privatization. In equation (5), a new variable is added to control for the effect of domestic competition. This new variable is the two-ﬁrm concentration ratio of an industry: the share of the top two ﬁrms’ sales revenue in the total sales revenue of the 4-digit industry to which the ﬁrms belong.8 We can see clearly that competition is positively related to the probability of a ﬁrm’s privatization. The addition of this variable does not change the sign and signiﬁcance levels of the three globalization variables. The government, both central and local, may have some policies and preferences concerning privatization of ﬁrms in different regions, industries, or of different sizes. In order to control for such effects, dummies for region, size, and industry are added 7 See Gupta et al. (2000) for explanations of the effect of government objectives on the probability of privatization of ﬁrms; see Su and Jefferson (2003) for an explanation of the effect of the incentive of the ﬁrms on its probability of privatization; and see Jefferson et al. (2002) for a discussion of selection bias in the ownership reform (privatization) of China. 8 There are about 550 4-digit industries in our dataset.

Globalization and Privatization: Evidence from China

75

one by one from equations (6) to (8). In equation (6), only region dummies are added. The Chinese government divides China into six big regions, each of them containing several provinces. These six regions are: Northern China, Northeast China, Southern China, Eastern China, Northwestern China, and Southwestern China. The region dummies here exactly reﬂect this division. As equation (6) shows, the addition of region dummies does not change the effects of the three channels of globalization. Size dummies are added in equation (7). In China, the government evaluates the size of ﬁrms according to their capital stock, employment, industry, and other factors. The government divides the LMIEs into two big categories: large enterprises and medium enterprises. Then the large enterprises are again divided into 3 subcategories, and the medium enterprises are divided into 2 sub-categories. Thus there are 5 sub-categories in total, which is reﬂected by the size dummies. Equation (7) shows that, with the addition of size dummies, the signs and signiﬁcance levels of the globalization variables are the same as in other equations. Industry dummies are added in equation (8). In our dataset, the industry to which a ﬁrm belongs is indicated by a four-digit industry code. The industry dummies used here are of two-digit level, from industry 6 to industry 46 (see Appendix 2 for a description of the industry classiﬁcation). Again, the addition of industry dummies does not change the estimated effects of the three channels of globalization. Finally, we further address the question of endogeneity: because only a one-year lag is used in the regressions, there is still a possibility that the prospect of a ﬁrm’s ownership reform in the following year may affect the right-hand-side variables this year. That is, when a ﬁrm expects to take ownership reform in the next year, it may change its behavior this year. Another possibility is that the ﬁnal effect of a variable may be the total effects accumulated over the several years before privatization. Thus, a one-year lag may not be sufﬁcient. In order to examine such possibilities, the ownership change dummy of 2001 is regressed on the four-year averages of all the independent variables that appeared in the above regressions, while the value of the other independent dummy variables are of 2001. The result is shown in Table 5.5. We can see from Table 5.5 that the estimation coefﬁcients for exports rate, the share of foreign capital, and the share of technology development personnel are still positive and signiﬁcant in most cases (the only exception is that the coefﬁcient for the share of technology development personnel is not signiﬁcant when industry dummies are added), which is consistent with results in Table 5.4. In summary, the empirical results above show robust evidence that exports, FDI, and inﬂow of knowledge have positive and signiﬁcant effects on the probability of privatization. Economic Signiﬁcance of the Effect Let’s use equation (1) in Table 5.4 to compute how much difference globalization makes in the probability of privatization. Rows (1) to (3) of Table 5.6 show

76

Table 5.5

The Chinese Economy after WTO Accession

Results with Four-year-average Data

4 5 6 7 0.658a 0.571a 0.596a 0.447c (0.204) (0.211) (0.212) (0.241) Share of foreign capital 1.29b 1.204b 1.235b 1.378b (0.574) (0.589) (0.595) (0.620) Share of R&D personnel 1.312b 1.018c 1.053c 0.716 in total employment (0.55) (0.566) (0.569) (0.657) Ratio of proﬁts to 0.350 0.327 0.356 0.907b sales revenue (0.349) (0.365) (0.367) (0.438) Ratio of subsidy to -2.533d -3.126c -3.004c -2.264 sales revenue (1.559) (1.710) (1.706) (1.931) Two-ﬁrm concentration -0.006b -0.007a -0.007a -0.005c ratio (0.002) (0.003) (0.003) (0.003) Region yes yes yes Size yes yes Industry yes LR statistic (52 df) 22.751 24.576 28.027 33.925 109.807 113.153 182.156 Probability (LR stat) 0.000 0.000 0.000 0.000 0.000 0.000 0.000 McFadden R-squared 0.015 0.016 0.019 0.023 0.073 0.075 0.121 Note: a signiﬁcant at 1%; b signiﬁcant at 5%; c signiﬁcant at 10%; d signiﬁcant at 15%; the number of observations is 2,762, that with dependent variable=1 is 214. The number in parenthesis is the standard deviation. Exports ratio

1 0.640a (0.202) 1.215b (0.569) 1.192b (0.525) -

2 0.660a (0.203) 1.239b (0.570) 1.064b (0.537) 0.434 (0.342) -

3 0.635a (0.203) 1.198b (0.570) 1.027c (0.538) 0.349 (0.344) -2.565c (1.558) -

the results.9 Row (1) reports the probability of privatization for ﬁrms that have zero globalization in a speciﬁc channel (the corresponding globalization variable equals zero); row (2) that for ﬁrms that have full globalization in a speciﬁc channel (the corresponding globalization variable equals 1); and row (3) the percentage difference between these two probabilities. Row (3) shows that the probability of privatization for ﬁrms that export all their output is 58 percent higher than that for ﬁrms that do not export; that the probability of privatization for ﬁrms whose share of foreign capital is 100 percent10 is 140 percent higher than that for ﬁrms that do not use foreign capital; that the probability of privatization for ﬁrms whose share of R&D personnel in total employment is 100 percent is 229 percent higher than that for ﬁrms that do not have R&D personnel. The numbers in row (3) overestimate the effect of globalization because we compared the probability of privatization between ﬁrms that have no globalization 9 The estimated intercept of equation (1) is –1.69 (not shown in Table 5.4). When the effect of one variable is computed, the other variables in equation (1) of Table 5.4 are evaluated with their means, which are shown in Table 5.3. 10 According to the Foreign Enterprise Law of China, a ﬁrm has the option of being registered as an SOE even if it satisﬁes the requirements to be registered as a foreign enterprise. In fact, our data show that some SOEs’ share of foreign capital is as high as 98 percent.

Globalization and Privatization: Evidence from China

Table 5.6

Based on equation (1) of Table 5.4

Based on Table 5.7

77

The Economic Signiﬁcance of the Effects of Globalization on Probability of Privatization

(1) For the ﬁrms with no degree of globalization (2) For the ﬁrms with full globalization (3) Percentage difference between (1) and (2)* (4) For the ﬁrms with no degree of globalization (5) For the ﬁrms with positive degree of globalization (6) Percentage difference between (4) and (5)

Export

FDI

0.0485

0.0495

R&D Personnel 0.0475

0.0764 57.5

0.1190 140.4

0.1562 228.8

0.0462

0.0500

0.0484

0.0601

0.0638

0.0515

30.1

27.6

6.4

and those that have full globalization. Anyway, very few ﬁrms have full globalization. At least no ﬁrms’ share of R&D personnel in total employment can be 100 percent. A more reasonable way to see the economic signiﬁcance of the effect of globalization on the probability of privatization is comparing the probabilities of privatization between ﬁrms that do not have any degree of globalization and those that have some degree of globalization. In order to do this, we created three dummy variables for the three channels of globalization; the dummies take a value of 1 if the corresponding ﬁrm has some degree of globalization in the corresponding channels, 0 otherwise. For example, the dummy for exports equals 1 if a ﬁrm’s exports are greater than 0, 0 if the exports are 0. Then, a Probit model similar to equation (1) of Table 5.4 is estimated, Table 5.7 shows the result. The only difference between Table 5.7 and equation (1) of Table 5.4 is that the globalization variables in Table 5.7 are dummy variables. Based on Table 5.7, we can compute the economic signiﬁcance of the effect of globalization on the probability of privatization. Rows (4) to (6) of Table 5.6 show the probabilities and their percentage differences. For ﬁrms that do not export, their probability of privatization is 0.0462; for those that export something, 0.0601, a difference of 30.1 percent. The difference in the probabilities for ﬁrms that do not use foreign capital and those that use some foreign capital is 27.6 percent, that for ﬁrms that have no R&D personnel and those that have some R&D personnel is 6.4 percent. In the extreme case, for the ﬁrms that do not export, have no foreign capital, and have no R&D personnel, the probability of privatization is 0.0436; for those that export, have foreign capital, and have R&D personnel, 0.077, a difference of 77.6 percent. Thus there is substantial difference in probability of privatization between ﬁrms that have no globalization and those that have some degree of globalization.

The Chinese Economy after WTO Accession

78

Table 5.7

Regression Result with Dummy Variables of Globalization

0.129 (0.029) 0.124 Dummy for use of foreign capital (0.056) 0.029 Dummy for R&D personnel (0.028) -1.707 Constant (0.023) LR statistic (3 df) 34.138 Probability (LR stat) 0.000 McFadden R-squared 0.003 Note: The total number of observations is 25,829, the total number of observations with dependent variable=1 is 1,310. Dependent variable: Dummy for ownership change. The number in parenthesis is the standard deviation. Dummy for export

Effect of Privatization on Globalization This section examines the effect of privatization on the three channels of globalization: export, inﬂow of foreign capital, and inﬂow of knowledge. Speciﬁcation Privatization should have a positive effect on globalization. The major reason is that the incentive structure of the ﬁrm changes in substantial ways. The objective function of the ﬁrm changes because private agents as owners are much more proﬁtoriented than state-owners. As a result, owners will implement more supervision, and thus the incentive of the managers is improved. A ﬁrm that maximizes proﬁts will try to have access to bigger product markets and factor markets. Bigger product markets make it possible to expand output and make use of more efﬁcient technologies. Access to bigger factor markets also helps them to ﬁnd the cheapest factors available to them. Also, they can get more and better business information from participating in bigger markets. Thus, a privatized ﬁrm should have more incentive to make use of international product and factor markets, and, therefore, privatization should have a positive effect on the globalization process. Based on the above discussion, we estimate the following equations: Annual growth rate of Z from t-1 to 2001= α0+α1 CONVERSIONt + α2 ln(Zt-1)+αYEAR + ut

(II)

where Z is a vector representing the three ﬂows: exports, inﬂow of foreign capital, and inﬂow of knowledge; CONVERSION is a dummy variable indicating whether an SOE takes ownership conversion in year t or not: it equals 1 if yes, 0 if no; ln(Zt-1)

Globalization and Privatization: Evidence from China

79

is the log of Z at t-1, which is included to control for the difference in the initial value of Z; YEAR is a vector of dummy variables indicating whether the ownership conversion took place in year t or not; ut is the error term. Intuitively, a privatized ﬁrm should be more interested in exploiting international goods and factor markets. Thus, the conversion should have a positive effect on all three ﬂows mentioned above. If the coefﬁcient of CONVERSIONt, α1, is positive and signiﬁcant, we can conclude that privatization is globalization-enhancing, otherwise not. The dataset is again based on our LMIE dataset. All the observations with unreasonable or missing values of any of the three ﬂows are dropped. Regression Results Table 5.8 reports the regression results. The results show that change in ownership type does have positive and signiﬁcant effects on all three ﬂows. The signiﬁcance level for the inﬂow of foreign capital is 1 percent, that for exports is 5 percent, and that for inﬂow of knowledge is 15 percent. Based on the estimated coefﬁcients, we can compute the difference in the growth rates of the three ﬂows between ﬁrms that are privatized and those that are not privatized. For the average growth rate of the exportsales revenue ratio, the privatized ﬁrms are higher than those not privatized by 10.6 percent; for the average growth rate of the share of foreign capital, 11.7 percent; for the growth rate of the ratio of R&D personnel to total employment, 6.7 percent.

Table 5.8

The Impact of Privatization on Globalization

Exports Ratio Constant Dummy for ownership change Ln(Exports ratio)t-1

-0.962 a (0.019) 0.101 b (0.040) -0.068 a (0.001) -

Growth Rate of the: Share of Foreign Share of R&D Personnel Capital in Total Employment -1.836 a -1.006 a (0.021) (0.018) 0.111 a 0.065d (0.022) (0.041) -

Ln(Share of foreign -0.110 a capital)t-1 (0.001) -0.124 a Ln(Share of R&D personnel (0.001) in total employment)t-1 YEAR dummy yes yes yes Adjusted R-squared 0.17 0.26 0.23 F-statistic 980.39 1677.75 1473.49 Included observations: 23,389 23,549 24,048 a Note: signiﬁcant at 1%; b signiﬁcant at 5%, c signiﬁcant at 10%; d signiﬁcant at 15%. The number in parenthesis is the standard deviation.

80

The Chinese Economy after WTO Accession

We have shown in the last section that globalization has positive and signiﬁcant effects on the probability of privatization. Thus the two independent variables in equation (II), COVt and LnZt-1 may be linearly correlated, that is there may be problem of multicollinearity. By checking the goodness of ﬁt of equations (I) and (II), we found that the McFadden R2 values of equation (I) are less than the R2 values of equation (II). As a rule of thumb, we can conclude that multicollinearity should not be a serious problem here. More about Methodology In the above, we used pooled data for 1995–96, 1998–99, 1999–2000, and 2000–2001 pairs to analyze the interaction between the globalization and privatization processes. We do two independent regressions: one is the effect of the globalization variables on the probability of privatization, the other is the effect of privatization on the growth rates of the globalization variables. Thus, two questions naturally arise. First, given the nature of our dataset, which is approximately 22,000 LMIEs and that spans the years from 1995 to 2001, why don’t we use a panel data approach, which has the beneﬁt of controlling for ﬁrm-speciﬁc characteristics? This is due to the nature of the dependent variable, the ownership change dummy. Let’s illustrate this with a hypothetical ﬁrm. Suppose this ﬁrm was an SOE in 1997, and it changed to foreign enterprise in 1998 and kept this ownership type till 2001. This means that it was privatized in 1998. So the value for the ownership change dummy is 1 for 1998, 0 for other years. Suppose the ﬁrm’s shares of exports in total sales revenue are 0.20, 0.25, 0.30, 0.35, and 0.40 for 1997, 1998, 1999, 2000, and 2001, respectively, as shown in the Table 5.9. Table 5.9 shows that with the steady increase of the share of exports in sales revenue, the value of the ownership change dummy increases from 0 for 1997 to 1 for 1998, then decreases to 0 for 1999 and remains stable thereafter. Thus if we regress the ownership change dummy on the share of exports in total sales revenue, we will get a negative relation. This is in contradiction with both our intuition and the cross-section results we get above. The same argument applies to our analysis of the effect of privatization on globalization, because this dummy is the major independent variable in that regression. Second, now that we are analyzing the interaction between the globalization and privatization processes, why don’t we use simultaneous equation systems? This is due to the nature of the issue we are discussing. In equation (I), the probability of privatization is determined by the past or contemporary values of the globalization variables. While in order to analyze the effect of privatization on globalization, we need to compare the behavior of the globalization variables before and after privatization, thus, the average annual growth rate of the globalization variables (or the 2001 values of these variables) should be used as dependent variables. Therefore, the dependent variables in equation (II) cannot be the same as the independent variables of equation (I); thus, the simultaneous method does not apply here.

Globalization and Privatization: Evidence from China

Table 5.9

81

A Hypothetical Firm

Year

Ownership type

1997 1998 1999 2000 2001

State-owned Foreign ownership Foreign ownership Foreign ownership Foreign ownership

Dummy for ownership change 0 1 0 0 0

The ratio of exports to sales revenue 0.20 0.25 0.30 0.35 0.40

Conclusion This research empirically studied the interaction between globalization and privatization processes. The evidence from China suggests that they are mutually reinforcing. We found that three channels of globalization, namely, outﬂow of goods and services, inﬂow of foreign capital, and inﬂow of knowledge, have positive and signiﬁcant effects on privatization. Also, privatization is found to have positive and signiﬁcant effects on the three channels of globalization. More generally, the evidence in this chapter offers support to the idea that reform, with privatization as a part, and openness cannot be separated, and to the policy of China in the last quarter of a century that combines openness and reform. This positive feedback mechanism can also be seen in the role of China’s special economic zones (SEZs). In the beginning of the ‘reform and openness’ era, China set up several special economic zones that enjoyed more openness than the rest of China and were regarded by the Chinese government as ‘windows of learning’. The inﬂow of foreign capital into these SEZs was accompanied by the inﬂow of more advanced and efﬁcient management, which was learned by the Chinese employees in these foreign enterprises, by the neighboring Chinese ﬁrms, by the Chinese business partners of these ﬁrms, and by their Chinese competitors. Such learning promoted the reform process, which in turn improved the efﬁciency of the Chinese economy. As the efﬁciency of the Chinese economy improved, Chinese ﬁrms needed bigger product and factor markets, which led to further openness. This evidence has policy implications to other transitional countries and the countries that may be in transition in the future, such as North Korea and Cuba. The positive feedback between openness and reform suggests that, in order to accelerate the reform process and to improve the economic efﬁciency of a country, openness and reform should be combined and implemented simultaneously. This evidence also has policy implications to the western democratic countries, such as the US. For several decades since the beginning of the Cold War, the US government had tried various policies to provoke a change or even a collapse of the Communist regimes in the world. Such policies range from engagement to direct military confrontation. The evidence in this chapter supports the effectiveness of engagement.

82

The Chinese Economy after WTO Accession

References Baily, Martin Neil (1993), ‘Competition, Regulation and Efﬁciency in Service Industries’, Brookings Papers on Economic Activity: Microeconomics, Vol. 1993, No. 2, pp. 71–30. Baily, Martin Neil and Gersbach, Hans (1995), ‘Efﬁciency in Manufacturing and the Need for Global Competition’, Brookings Papers on Economic Activity: Microeconomics, Vol. 1995, pp. 307–47. Blonigen, Bruce A. and Figlio, David N. (1998), ‘Voting for Protection: Does Direct Foreign Investment Inﬂuence Legislator Behavior?’ American Economic Review, Vol. 88, No. 4 (Sep.), pp. 1002–14. Claessens, Stijn and Djankov, Simeon (2000), ‘Government Regulation and Privatization Beneﬁts in Eastern Europe’, working paper, World Bank. Djankov, Simeon and Murrell, Peter (2002), ‘Enterprise Restructuring in Transition: A Quantitative Survey’, Journal of Economic Literature, Vol. XL (Sept. 2002), pp 739–92. Estrin, Saul and Rosevers, Adam (1999), ‘Enterprise Performance and Ownership: The Case of Ukraine’, European Economic Review, Vol. 43, No. 4, pp. 1125–36. Florida, Richard (1996), ‘Regional Creative Destruction: Production Organization, Globalization, and the Economic Transformation of the Midwest’, Economic Geography, Vol. 72, No. 3, July, pp. 314–34. Frydman, Roman, Hessel, Marek, and Rapaczynski, Andrej (1999), ‘Why Ownership Matters? Entrepreneurship and the Restructuring of Enterprises in Central Europe’, working paper, Economics Department, New York University. Gupta, Nandini, Ham, John C., and Svejnar, Jan (2000), ‘Priorities and Sequencing in Privatization: Theory and Evidence from the Czech Republic’, Working Paper Number 323, The William Davidson Institute at the University of Michigan Business School. Jefferson, Gary H. and Rawski, Thomas G. (1994), ‘Enterprise Reform in Chinese Industry’, The Journal of Economic Perspectives, Vol. 8, No. 2. (Spring), pp. 47–70. Jefferson, Gary H., Su, Jian, Jiang, Yuan, and Yu, Xinhua (2002), ‘The Impact of Shareholder Reform on Chinese Industry, 1995-2001’, working paper, Brandeis University; presented at the conference ‘Distributional Consequences of Privatization,’ Center for Global Development and The Inter-American Development Bank, February 24–25, 2003. Jones, Derek, and Mygind, Niels (2002), ‘The Economic Effects of Privatization: Evidence from a Russian Panel’, Comparative Economic Studies, Vol. 40, No. 2, pp. 75–102. Macdonald, James M. (1994), ‘Does Import Competition Force Efﬁcient Production?’, The Review of Economics and Statistics, Vol. 76, No. 4, pp. 721–7. Megginson William L., and Netter, Jeffry M. (2001), ‘From State to Market: A Survey of Empirical Studies on Privatization’, Journal of Economic Literature, Vol. XXXIX (June), pp. 321–89.

Globalization and Privatization: Evidence from China

83

Ministry of Education (2000), ‘Notiﬁcation of the Ministry of Education Concerning the Distribution of ‘Summary of the Meeting on the Foreign Affairs in Education’’ (jiaoyubu bangongting guanyu yinfa ‘quanguo jiaoyu waishi gongzuo huiyi jiyao’ de tongzhi), Foreign Affairs Ofﬁce of the Ministry of Education [2000] No. 9, May 10, 2000. National Bureau of Statistics, China Statistical Yearbook, various issues. National Bureau of Statistics (2002), China Statistical Yearbook of Science and Technology. Naughton, Barry (1994), ‘Chinese Institutional Innovation and Privatization From Below’, American Economic Review, Vol. 84, No. 2, pp. 266–70. North, Douglass (1994), ‘Economic Performance through Time’, American Economic Review, Vol. 84, No. 3, pp. 359–68. Rodrik, Dani (1998), ‘Why Do More Open Economies Have Bigger Governments?’, The Journal of Political Economy, Vol. 106, No. 5. (Oct.), pp. 997–1032. Sachs, Jeffrey D. and Warner, Andrew (1995), ‘Economic Reform and the Process of Global Integration’, Brookings Papers on Economic Activity, 1. Song, Ligang, and Yao, Yang (2003), ‘Impacts of Privatization on Firm Performance in China’, Working Paper, China Center for Economic Research. Su, Jian and Jefferson, Gary H. (2003), ‘A Theory of Decentralized Privatization: Evidence from China’, manuscript, GSIEF, Brandeis University. Wei, Shangjin and Wu, Yi (2001), ‘Globalization and Inequality: Evidence from within China’, NBER Working Paper 8611.

The Chinese Economy after WTO Accession

84

Appendix Table 5.A-1 Concordance of Ownership Classiﬁcations (1992–2001)

Stateowned

Collectiveowned

Hong Kong, Macao, Taiwanowned

Foreignowned

Shareholding

Private

1992-1997 Code Ownership types 11 State-owned enterprises 12 State owned jointly operated enterprises 11 Wholly state-owned companies 21 Collective-owned enterprises – – 22 Collective jointly operated enterprises 81 Overseas joint ventures 82 Overseas cooperatives 83 Overseas whollyowned enterprises – – 71 72 73 –

Foreign joint ventures Foreign cooperatives Foreign whollyowned enterprises –

62

Limited liability company

61

Shareholding limited companies Private whollyowned enterprises Private cooperative enterprises Private limited liability companies –

31 32 33 –

1998-2001 Code Ownership types 110 State-owned enterprises 141 State owned jointly operated enterprises 151 Wholly state-owned companies 120 Collective-owned enterprises 130 Shareholding cooperatives 142 Collective jointly operated enterprises 210 Overseas joint ventures 220 Overseas cooperatives 230 Overseas whollyowned enterprises 240 Overseas shareholding limited companies 310 Foreign joint ventures 320 Foreign cooperatives 330 Foreign whollyowned enterprises 340 Foreign shareholding limited companies 159 Other limited liability companies 160 Shareholding limited companies 171 Private whollyowned enterprises 172 Private cooperative enterprises 173 Private limited liability companies 174 Private shareholding companies

Globalization and Privatization: Evidence from China

Table 5.A-1 Continued

Other domestic

1992-1997 Code Ownership types 51 State-collective jointly operated enterprises 52 53 54

9 Source: NBS dataset.

State-private jointly operated enterprises Collective-private jointly operated enterprises State-collective-private jointly operated enterprises Other enterprises

1998-2001 Code Ownership types 143 State-collective jointly operated enterprises 149 Other jointly operated enterprises -

-

-

-

190

Other enterprises

85

The Chinese Economy after WTO Accession

86

Table 5.A-2 Industrial Code Industry Code 6

Industry Mining

Industry Code 26

Industry

7

Oil mining

27

8

Ferrous metal mining

28

Raw chemical materials and chemical products Medical and pharmaceutical products Chemical ﬁber

9

Non-ferrous metal mining

29

Rubber products

10

Non-metal mining

30

Plastic products

11

Other mining

31

Nonmetal mineral products

12

Lumbering and bamboo mining

32

13

Food processing

33

14

Food production

34

Smelting and pressing of ferrous metals Smelting and pressing of nonferrous metals Metal products

15

Beverage production

35

16

Tobacco processing

36

17

Textile Industry

37

18

Garments and other ﬁber products Leather, furs, down and related products Timber, bamboo, cane, palm ﬁber and straw products Furniture manufacturing

40

19 20 21 22

Paper-making and paper products 23 Printing and record medium reproduction 24 Cultural, educational and sports goods 25 Petroleum processing and coking Source: NBS dataset.

41 42 43 44 45 46 -

Ordinary machinery manufacturing Special purpose equipment manufacturing Transport equipment manufacturing Electric equipment and machinery Electronic and telecommunications equipment Instruments, meters, cultural and ofﬁce machinery Other Manufacturing Production and supply of electricity, steam, and hot water Production and supply of coal gas Water production and supply -

PART II Fiscal Policy Reform and Financial Development

This page intentionally left blank

Chapter 6

China’s Capital Tax Reforms in an Open Economy1 Shuanglin Lin

Introduction China now has a dual tax system for corporate income, i.e., foreign investors are subject to a lower corporate income tax rate than Chinese domestic enterprises in speciﬁed areas. Recently, domestic industries and tax administrators have been demanding the abolition of tax concessions to foreign enterprises and to foreigninvested enterprises in favor of an equal tax burden on all enterprises. However, only limited analyses of the effects of the tax reform on the domestic, as well as on the interdependent economies, have been undertaken. This chapter examines the effects of

capital tax reforms that unify the corporate income tax rates faced by domestic and foreign investors on the capital-labor ratio, the output-labor ratio, foreign investment and the trade balance in China.2 The effect of capital taxation is debated in the literature. In an overlapping generations (OLG) growth model, Diamond (1970) demonstrates that, with the tax revenue being rebated to individuals who pay the tax, an increase in the capital tax rate will decrease the real wage rate and lower the representative individual’s utility if the after-tax interest rate is greater than the growth rate. Summers (1981) develops a life-cycle equilibrium model and demonstrates that a shift from capital taxation to consumption taxation would increase welfare substantially. With the tax revenue being refunded to tax payers in a lump sum manner, Chamley (1981, 1986) and Judd (1985) show that the welfare-maximizing capital income tax should be zero in the steady state. In addition, Lucas (1990) argues that capital should not be taxed at all. 1 The author thanks the participants of CES conferences in Hong Kong and at the University of Michigan for their helpful comments and Jackie Lynch and Ke Yang for their research assistance. The author especially thanks John P. Bonin and two anonymous referees for their insightful comments and helpful suggestions. Reprinted with permission from the Association of Comparative Economic Studies and Elsevier, ‘China’s Capital tax reform in an open economy,’ Journal of Comparative Economics, Vol. 32 (2004), pp. 128–147. Minor revisions were made in reformatting. 2 Since the opportunity cost of capital supplied by shareholders is not deductible from taxable income, the corporate income tax is a tax on capital as Rosen (2002) argues.

90

The Chinese Economy after WTO Accession

The integration of the world economy has increased interest in the effect of capital taxation in an open economy. Two types of capital tax systems exist in an open economy, namely, a territorial tax system and a residential tax system. Under the territorial tax system, the capital income generated in the territory is taxed uniformly regardless of the residence of those who receive the capital income. In the residential tax system, all the capital income of all domestic residents is taxed regardless of where the income is earned.3 In a model in which tax revenue is used to ﬁnance government consumption, Bovenberg (1989) shows that decreasing capital income taxation is particularly attractive to open economies. Gordon (1990), Gordon and Bovenberg (1996), Frenkel, Razin, and Sadka (1991), and Mintz (1992) argue that, in theory, capital taxes might not be sustainable in an open economy with perfect capital mobility. Because capital is important for output growth, all countries have incentives to attract foreign capital and keep domestic capital. Since investors are motivated by after-tax returns to capital, higher capital income tax rates result in lower after-tax returns to capital and less capital. In theory, competition among countries to attract foreign capital drives the capital tax rate to zero. However, foreign capital taxation is widespread in the real world. Developed countries, with free capital mobility, usually assess a uniﬁed capital tax rate on both domestic and foreign enterprises. On the other hand, many developing countries have a dual capital tax system in order to attract more foreign investment; they provide tax incentives to foreign investors, e.g., tax holidays, reduced tax rates, investment allowances, tax credits, and accelerated depreciation. Several reasons justify tax concessions for foreign investors. First, foreign direct investment (FDI) is a transfer of capital from the rich countries to the poor countries and capital is important to both economic growth and employment creation. Second, foreign enterprises and foreigninvested enterprises create positive externalities through transfers of technologies and skills. Third, foreign investors usually face disadvantages when investing in the host countries, e.g., high transaction costs, exchange-rate risk, asymmetric information, and political uncertainty. Tax incentives can be used to offset these disadvantages. The effectiveness of tax incentives in attracting foreign capital and increasing domestic capital formation is still a controversial issue despite a long history of debate. Many economists argue that tax incentives played an insigniﬁcant role in attracting foreign investment, e.g., Tanzi and Zee (2000). The existence of natural resources, political and economic stability, the transparency of the legal and regulatory systems, the adequacy of supporting institutions, such as banking, transportation, communication, and other infrastructure, and the skilled labor force are usually much more important in determining suitable investment locations. However, Hines (1996, 1998) ﬁnds that tax incentives affect signiﬁcantly the pattern of FDI in the United States and that

3 The territorial tax system exists in Hong Kong and Macau, while the residential tax system exists in countries such as the US, the UK, Japan, and Singapore.

China’s Capital Tax Reforms in an Open Economy

91

tax sparing inﬂuences the level and location of FDI.4 Analyses of the effects of tax incentives based on dynamic equilibrium models are still limited in the literature. This chapter will develop an equilibrium model of overlapping generations with a dual tax system to study the effect of China’s capital tax reforms on the rate of returns to capital, the capital-labor ratio, per capita output, foreign investment, and the trade balance. Two tax experiments are analyzed, namely, an increase in the capital tax rate for foreign-invested enterprises and a decrease in the capital tax rate for domestic enterprises. We show that the two types of tax reforms have different effects. Our results apply not only to China but also to other countries with a dual capital tax system that are considering a capital tax reform. In addition, the chapter contributes to the literature on capital taxation in a small open economy by introducing a dual capital tax system.5 The structure of the chapter is as follows. The next section discusses the current dual-track tax system in China. The third section develops an overlapping generations model of a small open economy with a dual tax system for capital income. The fourth section derives analytical results for various capital tax reforms. The ﬁnal section concludes the chapter with policy implications. Corporate Income Taxation in China In 1978, China started to reform state enterprises and the tax system. The reforms resulted in a dual capital income tax system in which foreign enterprises and foreigninvested enterprises are subject to differential tax rates. Based on the State Council’s regulations (1993), state enterprises, collective enterprises, private enterprises, and any other enterprises pay an income tax of 33 percent. Interest expenses incurred during the process of production and business operations, wages and salaries, employee union expenses and donations can be deducted from total income. Investment in capital is not deductible from the taxable income. Thus, the corporate income tax is essentially a tax on capital income. The loss in one year can be offset by the proﬁts in the following years, up to ﬁve years. Chinese enterprises must also pay residential capital taxes, i.e., they need to pay capital income taxes on their worldwide income to the Chinese government. Taxes on foreign income paid to foreign governments are deductible. However, the 4 Tax sparing is the practice of adjusting the home country taxes of foreign capital income to permit investors to receive the full beneﬁts of the host country’s tax reductions as Hines (1998) explains. 5 Most studies of capital taxation in open economies focus on the case of a uniform capital tax rate, i.e., foreign capital and domestic capital are subject to the same tax rate [Bovenberg (1989), Gordon (1990), Gordon and Bovenberg (1996), Frenkel, Razin, and Sadka (1991), Nielsen and Sorensen (1991), and Mintz (1992)]. The assumption of a uniform capital tax is consistent with reality in most developed countries, but inconsistent with reality in many developing countries. Introducing a dual capital tax system provides analytical results that are relevant to the developing countries.

92

The Chinese Economy after WTO Accession

deductible amount cannot exceed the amount of income tax otherwise payable under the provision of domestic tax laws. China’s residential tax laws are similar to those of the US and Singapore. Some other regions, such as Hong Kong and Macau, adopted the territorial tax principle, i.e., only corporate income on the domestic territory is taxed. In addition to the corporate income tax, enterprises in China must pay a valueadded tax (VAT) or a business tax (BT). The VAT rate is 17 percent for most products and 13 percent for certain products, e.g., agricultural products. Imports to China are subject to VAT. Exporting enterprises receive VAT refunds as an export incentive, with refund rates ranging from 9 percent to 17 percent.6 BT is charged at applicable rates, ranging from 3 percent to 7 percent, on service turnover, including rental and interest income.

In September 1980, China promulgated The Law of the People’s Republic of China on the Income Tax of the China-Foreign Joint Ventures setting an overall tax rate of 33 percent. The government provided detailed regulations on how to calculate corporate income. For example, interest payments on capital cannot be excluded from the tax base. Tax holidays were provided to joint-venture enterprises. In the 1980 law, joint-venture enterprises that promise to operate in China for at least 10 years enjoy a one-year tax exemption and a two-year tax reduction consecutively from the ﬁrst proﬁtable year. The law was changed to a two-year exemption and a three-year reduction in 1983 as Li (1988) reports. For a long period of time, foreign enterprises and foreign-invested enterprises were required to pay the consolidated industrial and commercial tax and also other taxes.7 After the 1984 tax reform, domestic enterprises began to pay product tax, business tax, and value-added tax and stopped paying the industrial and commercial tax. However, foreign enterprises and foreign-invested enterprises continued to pay the consolidated industrial and commercial tax. This caused unequal tax treatment of foreign and domestic enterprises with the burden falling on foreign enterprises. In 1993, VAT was applied to all enterprises in an attempt to equalize tax burdens. The Chinese government has also signed bilateral tax treaties with many countries, specifying the tax rates for corporate income, individual income, and capital gains. Since these tax treaties usually provide more generous treatment to foreign 6 Other countries, e.g., European countries having VAT, also rebate VAT on exported goods. 7 The consolidated industrial and commercial tax was established in 1958; see State Council, Provisions on the Consolidated Industrial and Commercial Tax (Draft), September 13, 1958. Products were classiﬁed into more than 100 types, with tax rates ranging from 1.5 percent on cotton clothing of lowest quality to 69 percent on cigarettes of highest level of the product value. For most products, the tax rate is about 10 percent. In 1973, the government combined the consolidated industrial and commercial tax and other taxes into the industrial and commercial tax. As a result, the consolidated industrial and commercial tax rates for about half of the listed products were higher than previous industrial and commercial tax rates by themselves. Thus, foreign enterprises and foreign-invested enterprises paid higher taxes than domestic enterprises producing the same product.

China’s Capital Tax Reforms in an Open Economy

93

enterprises, they override China’s domestic tax laws in that, if the provisions of domestic tax laws are not consistent with a treaty, the treaty prevails. Solely foreign-owned enterprises were legalized in 1986.8 In 1991, the National People’s Congress passed the corporate income tax law for enterprises with foreign investment and foreign enterprises.9 This law speciﬁed that foreign and foreigninvested enterprises should pay taxes on their income derived from production, business operations and other sources within the territory of China. Taxable income is the amount remaining from gross income in a tax year after costs, expenses and losses have been deducted. The law indicates clearly that the interest on capital is not deductible from taxable income. Based on this law, the central government income tax rate is 30 percent of taxable income; the local tax rate is 3 percent. However, the tax rate for foreign investment and foreign enterprises established in special economic zones and engaged in production or business operations or enterprises with foreign investment of a production nature in economic and technological development zones is levied at a reduced rate of 15 percent. The income tax on enterprises with foreign investment of a production nature established in coastal economic open zones, or in the old urban districts where special economic zones or the economic and technological development zones are located is levied at a reduced rate of 24 percent. The tax rate on these above enterprises and those in other regions involved in energy, communications, harbor, wharf or other projects encouraged by the state may have this levy reduced to 15 percent. Any enterprise with foreign investment of a production nature scheduled to operate for a period of not less than 10 years is exempted from income tax in the ﬁrst and second years after the year in which it begins to make proﬁts and allowed a 50 percent reduction in the third to ﬁfth years. However, the State Council regulates enterprises with foreign investment engaged in the exploitation of resources such as petroleum, natural gas, rare metals and precious metals. Enterprises with foreign investment that have actually operated for a period of less than 10 years must repay the amount of already exempted or reduced tax. In recent years, the government has provided tax beneﬁts to stimulate the development of the Western region. In 2000, the State Council ofﬁcially granted a number of preferential tax policies to the West of China.10 According to this law, foreign-invested enterprises that are located in the Western Region and invest in industries that are encouraged by the state government enjoy a preferential rate of corporate income tax of 15 percent; foreign-invested enterprises located in the Western region and investing in industries that are encouraged by the state government have a half exemption for three years after graduating from the basic tax holiday with a minimum tax rate of 10 percent. Foreign enterprises located in the 8 National People’s Congress (1986). 9 National People’s Congress (1991). 10 By the ofﬁcial deﬁnition, the West consists of the following ten provinces: Sichuan, Gansu, Guizhou, Yunan, Qinhai, Shaanxi, the municipality of Chongqing, and autonomous regions of Ningxia, Xinjiang and Tibet.

Table 6.1

Types of Revenue as Percentage of Total Central Government Revenue

Taxes on income, proﬁts, and cap. Gains US 00 57.45 Canada 00 53.30 Australia 98 67.14 Japan 93 36.16i New Zealand 00 61.28 France 97 19.34 Germany 98 14.50 Greece 98 38.49 UK 99 39.17 China 99 6.38 India 00 25.35ii Indonesia 99 59.48 Korea 97 26.41 Malaysia 97 26.32ii Philippines 00 39.71 Singapore 00 25.51 Thailand 00 29.92 Hungary 00 20.63 Poland 00 18.79 Russia 00 11.28 Ukraine 00 12.73 Argentina 00 17.02 Brazil 98 19.44 Chile 00 18.35 Mexico 99 36.49 Country Year

Individual Corporation 47.61 39.16 49.83 23.72i 44.80 14.34 12.56 19.47 29.13 0.01 11.18ii 7.84 16.16 10.20ii 15.93 … 11.11 14.52 10.88 1.62 … 5.34 1.26 … …

9.84 12.27 16.60 12.45i 11.93 4.94 1.94 12.96 10.58 6.36 14.17ii 50.15 10.25 25.93ii 16.69 … 17.46 6.11 7.91 9.65 12.73 11.68 4.99 … …

Domestic Social Taxes on Taxes on taxes on security payroll and property goods and contribution work force services 30.72 … 1.38 3.18 20.51 … … 16.25 …. 2.26 …. 20.24 26.48i …i 4.05i 14.43i … 0.89 0.06 28.40 41.16 1.41 1.97 28.23 47.58 … 0.03 20.04 2.19 … 4.27 54.78 17.02 … 7.21 31.02 … … … 74.49 … … 0.06ii 26.40ii 1.94 … 0.31 27.91 9.25 … 1.73 33.32 1.23ii … 0.65ii 26.39ii … … 0.09 27.15 … … 3.32 15.86 3.03 … 0.45 43.08 26.89 0.32 1.14 35.71 30.33 0.68 … 37.54 28.18 … 0.25 30.20 31.37 1.91 … 33.70 23.39 … 3.18 42.58 32.68 4.18 0.09 20.35 6.43 … … 45.83 10.88 … … 57.78

General sales turnover or VAT … 12.35 9.92 6.99i 19.74 19.16 10.79 33.25 18.81 59.50 0.06ii 16.65 21.19 8.80ii 10.36 4.38 19.66 24.09 24.45 20.15 25.00 29.40 7.89 35.72 23.83

Source: International Monetary Fund, 2001, Government Finance Statistics Yearbook, pp. 4–5. Notes: i. The data for Japan are preliminary or provisional. ii. The data for India and Malaysia are forecasted or projected.

Taxes on Entrep. and internatl. Non-tax property Other taxes trade and revenue income transact. 1.02 … 6.23 3.29 1.25 … 8.65 6.49 2.56 … 6.33 3.34 1.24i 1.60i 15.45i 8.58 i 1.85 … 6.84 4.10 … 0.63 6.07 1.34 … … 15.94 … 0.06 3.63 6.98 4.02 … … 4.89 2.94 9.5 3.65 5.96 4.52 18.96ii 0.12ii 25.30ii 21.91ii 2.53 0.08 7.72 5.18 6.30 8.26 13.38 6.12 12.63ii 4.68ii 17.92ii 10.44ii 18.57 4.09 9.49 2.99 1.25 4.90 35.37 23.05 11.09 0.45 11.32 6.09 2.87 0.68 11.19 3.43 2.38 0.01 9.75 3.12 12.52 0.13 14.81 4.36 4.27 … 15.94 2.68 4.89 0.12 8.69 3.62 2.78 … 16.29 7.79 6.11 3.42 19.69 3.11 4.30 1.45 11.20 9.17

China’s Capital Tax Reforms in an Open Economy

95

Western region and investing in transportation, electricity, water conservancy, postal services, and broadcasting have a full exemption for two years and a half exemption for three years. Revenues from corporate income taxation are relatively low in China. Table 6.1 shows the various tax revenues as percentages of total central government revenue in selected countries. In 1999, China’s corporate income taxes accounted for about 6 percent of total central government tax revenues. Revenues from personal income taxes accounted for only 0.01 percent of total tax revenues which is the smallest percentage for these taxes in the world. China depended mainly on taxes on goods and services; at 74.5 percent of total tax revenues, this is the highest percentage in the world for this class of taxes. Tax revenues from foreign enterprises and foreign-invested enterprises

vary signiﬁcantly from province to province. Table 6.2 shows foreign investment and corporate income taxes from foreign and foreign-invested enterprises in 1998. Although Beijing, Shanghai, Guangdong, and Tianjin collected a large amount of tax revenues from foreign capital, Tibet and Qinghai collected very little. Capital tax revenues from foreign-invested enterprises, as well as from domestic enterprises, are increasing. In 2001, the total tax revenue was 1,517.2 billion yuan, $182.8 billion, an increase of 19.8 percent or 251.1 billion yuan, $30 billion, over the previous year. Of total tax revenue, 545.2 billion yuan, $66 billion, was domestic VAT, up 16.8 percent compared with the previous year and 212.3 billion yuan, $26 billion, was income tax from domestic enterprises, an increase of 47 percent. In addition, 51.1 billion yuan, $6.2 billion, was income tax from foreign-funded enterprises, up 57 percent and 96.6 billion yuan, $12 billion, was personal income tax, an increase of 51 percent. Taxes paid by foreign companies accounted for 3 percent of China’s total tax revenue.11

Tax incentives provided in the early 1990s stimulated foreign investment in China greatly. For more than 20 years prior to 1979, China did not have any foreign investment. China’s foreign capital inﬂow has increased dramatically since 1979 and especially since 1991. Figure 6.1 depicts the foreign funds actually utilized by China during the period from 1983 to 2001. In 2002, FDI to China by itself reached more than $50 billion to surpass the US as the largest foreign investment recipient in the world. Among possible explanations for the sharp increase in FDI after 1991 are Deng’s speeches during the 1992 southern tour that assured the world that China would still pursue reform policies and would open more special economic zones including Shanghai.12 However, the preferential tax policy aimed at foreign investment is the most important factor because the main difference between the special economic zones and other areas is in tax policy toward foreign investment. Table 6.3 depicts the sources of foreign investment to China; Hong Kong is the single largest external investor. In 2000, foreign investment from Hong Kong was 11 China Daily, January 11, 2002. 12 After the 1989 event, a wave of anti-openness and a push for reform emerged in China. Deng Xiaoping, the initiator and designer of China’s economic reform, traveled to the southern part of China, praising the success of economic reforms and encouraging further openness. He also expressed his regret for not opening up Shanghai in the 1980s.

96

Table 6.2

Provinces (Municipalities, Autonomous Regions) Beijing Tianjin Hebei Shanxi Neimonggu Laoning Jilin Helongjiang Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong Henan Hubei Hunan Guangdong Guangxi Hainan Chongqing Sichuan Quizhou Yunan Tibet Shaanxi Gansu Qinghai Ningxia Xinjiang

The Chinese Economy after WTO Accession

Tax Revenues from Foreign-Investment Enterprises and Foreign Investment in 1998

Total Tax Revenue (10,000 yuan) 2294497 1013963 2067587 1041941 776654 2646201 936374 1572703 3806961 2965804 1981028 1591862 1879249 971561 3523912 2081962 1689508 1567662 6407547 1196720 336719 711287 1972882 653426 1682347 36393 933309 540253 127718 177525 653928

Domestic Corporate Income Tax (10,000 yuan) 269813 127852 270711 89355 49581 254045 72173 70352 508818 314476 360983 176677 144074 69235 468054 226756 152467 84430 704950 80076 18098 42437 225418 43163 144691 9855 56160 44537 11939 12845 47335

Foreign Enterprises and Foreign-Invested Enterprises Corporate Income (10,000 yuan) 210670 75723 21964 2254 4577 47192 5914 8853 329437 81901 52021 5298 52198 4260 46780 24140 14211 5245 342312 8012 6121 12492 17990 576 6451 4 9330 1289 24 1376 1528

Actually Utilized FDI ($10,000) 216800 211361 142868 24451 9082 219045 40917 52639 360150 663179 131802 27673 421211 46496 220274 61654 97294 81816 1201994 88613 71715 43107 37248 4535 14568 30010 3864 1856 2167

Source: Statistical Yearbook of China, 1999, p. 277 and p. 599.

$15.5 billion, accounting for 38 percent of China’s total foreign investment. When Singapore, Taiwan, and Macao are included, FDI from these regions of Greater China alone was $20.3 billion, accounting for 50 percent of the total foreign investment in China. In the same year, foreign investment from the US was $4.4 billion, Japan $2.9 billion, Britain $1.2 billion, Germany $1 billion, and France $0.85 billion. These ﬁve countries together accounted for 25 percent of China’s total foreign investment.

China’s Capital Tax Reforms in an Open Economy

Figure 6.1

97

Foreign Capital Actually Utilized by China

Source: Statistical Yearbook of China, 1996, 2002.

At the same time that it is making efforts to attract foreign capital, China has an extremely high savings rate of around 40 percent. In contrast, most developing countries have low savings rates and need foreign funds to ﬁnance investment. Even with a high savings rate, China needs to attract foreign capital for two major reasons. First, China’s non-banking ﬁnancial system is not well developed and most of the savings are in banks. China’s banks are owned by the state and have been burdened with non-performing loans to the state-owned enterprises. Private enterprises have been discriminated against in lending practices by these banks. In addition, state banks are now reluctant to lend money even to state enterprises so that they use their savings deposits to purchase government bonds and other securities. Second, foreign investment brings needed technologies, equipment, management experiences, and modern business cultures to China. The dual capital tax system has existed for more than 20 years in China. It seems that once every ten years, China has undertaken a major tax reform, i.e., 1953, 1963, 1973, 1983/84, and 1994. Merging the capital tax rates on domestically owned and foreign owned enterprises is on the government’s agenda. It is important to conduct a rigorous analysis of the impacts of the capital tax reform before it actually takes place. A Model of Dual Capital Tax Systems A small open economy model with three sectors, namely, ﬁrm, consumer, and government, is considered. The representative ﬁrm produces a single good with a constant-returns-to-scale production technology and faces competitive resource markets. Individuals are identical within and across generations. They live for two

The Chinese Economy after WTO Accession

98

Table 6.3

Sources of Foreign Investment to China (US$ million) 1985

By Source Country (Region) Hong Kong and Macao Hong Kong*

1990

1995

Foreign Loans

FDI and Others

Foreign Loans

FDI and Others

Foreign Loans

60.69

955.68

313.20

2118.48

-

FDI and Others -

2000 Foreign FDI Other Investment -

-

-

-

-

-

216.72

Macao*

-

-

-

-

0

439.82

347.28

35.11

Taiwan

0.00

0.00

0.00

224.26

0

3,165.16

2,296.58

204.44

Indonesia

0.00

0.08

0.00

1.00

0.00

111.63

129.17

0.00

1275.93

315.07

2,500.22

520.48

1,900.85

3,212.47

2,915.85

145.32

Japan

20,185.11 15,499.98 1,129.35

Philippines

0.00

3.11

0.00

1.67

0.00

105.78

111.12

0.00

Singapore F. R. Germany France United Kingdom Canada

0.01

10.13

22.57

53.28

0.00

1,860.61

2,172.20

0.41

133.13

24.14

329.10

69.38

136.93

390.53

1,041.49

0.84

46.50

32.54

555.90

23.44

429.24

287.02

853.16

0.00

26.73

71.35

499.67

19.90

94.11

915.20

1,164.05

0.00 33.71

6.57

9.40

232.03

8.93

362.62

257.04

279.78

United States

24.45

357.19

134.93

461.21

50.93

3,083.73

4,383.89

0.63

Australia

16.81

14.36

36.11

25.15

110.35

232.99

308.88

0.00

New Zealand Republic of Korea Russia

0.00

0.01

0.00

8.88

0.00

20.72

18.22

0.00

0.00

0.00

0.00

0.00

143.43

1,047.10

1,489.63

10.81

World Bank Intn’l Fund for Agricultural Development Asia Development Bank Others Total

0.00

0.00

0.00

0.00

0.00

22.90

16.23

0.00

584.91

0.00

1,010.61

0.00

2,149.45

0.00

-.

-.

19.39

0.00

4.34

0.00

16.22

0.00

-.

-.

0.00

0.00

50.64

0.00

541.42

0.00

-.

-.

240.32

98.30

151.56

110.20

2,467.24

660.58

-

-.

2,688.02

1,958.72

10,327

37,806

10,327.00 37,805.69 40,714.81 8,641.46

Source: Statistical Yearbook of China, 1999, 2001.

periods, working in the ﬁrst period and retiring in the second period. Each supplies one unit of labor inelastically during the working period. The government exists forever and collects taxes to ﬁnance spending. Capital controls are imposed so that domestic individuals cannot invest in a foreign country, although foreign investors can invest in the domestic capital market.13 A dual capital tax system in which the domestic government imposes a higher tax rate on domestically owned capital 13 China’s currency is not freely convertible so that domestic residents cannot freely shift their savings abroad. Many OECD countries, including France and Italy until 1986,

China’s Capital Tax Reforms in an Open Economy

99

than on foreign owned capital is in place. Given capital controls, capital taxation is essentially territorial capital taxation. Our model differs from other studies of capital taxation in small open economies that assume perfect capital mobility, for example, Bovenburg (1986 and 1992) and Nielsen and Sorensen (1991). Let kt be the domestic capital-labor ratio in period t. Domestic output per worker, yt, is given by: yt = f (kt ) .

(1)

The production function exhibits constant returns to scale in capital and labor. Following Persson (1985), Sibert (1990), and Lin (1998), we assume that capital does not depreciate and that each factor is paid its marginal product. Thus, rt = f ’(kt ) and

(2)

wt = f (kt ) - kt f ’(kt ) ,

(3)

where rt is the rate of return on capital in the domestic country, or the before-tax interest rate, in period t and wt is the wage rate in period t. The labor force in period, Lt, grows at a constant rate, n > 0, i.e., Lt+1 = Lt(1+n). 1 2 The representative domestic agent maximizes u = u (ct , ct ) subject to the following constraints: ctt £ wt - t t - kth+1 (1 + n) and ctt+1 £ [1 + (1- pth+1 )rt +1 ]kth+1 (1 + n) , where t t is the lump-sum tax, or transfer if it is negative, in period t, which represents h all other taxes in the economy. In addition, pt is territorial tax rate on capital income t t for domestic individuals in period t, rt+1 is the rate of return to capital, ct (ct +1 ) h is consumption while young (old) by a member of generation t, and (1 + n)kt +1 is domestic capital held by a domestic individual and denoted domestic capital at the end of t. The utility function u is twice differentiable, strictly quasi-concave, and t t increasing in ct and ct +1 . Regarding marginal utilities, u1 (0, ct2 ) = u2 (ct1 , 0) = ∞ 2 1 and u1 (∞, ct ) = u2 (ct , ∞) = 0 . These conditions ensure that the budget constraints hold with equality and that an interior solution will be obtained for ctt and ctt+1 . Solving the domestic representative consumer’s maximization problem yields: ctt = ctt [ wt , τ , (1− π th+1 )rt +1 ] ,

(4)

ctt+1 = ctt+1[ wt , τ , (1− π th+1 )rt +1 ] , and

(5)

st = wt − τ t − ctt = s[ wt ,

(6)

τ , (1− π th+1 )rt +1 ] .

imposed capital controls that prevented domestic residents from shifting their savings abroad (Gordon and Bovenberg, 1996).

The Chinese Economy after WTO Accession

100

Equations (4), (5), and (6) are the consumption and savings functions. Government spending is ﬁnanced by capital and other taxes. Without loss of generality, we assume that other taxes are lump-sum. The government budget constraint is given as: gt = π th rt kth + π tf rt kt f + τ t ,

(7)

where gt is the government spending to labor ratio in period t and ktf is the domestic capital held by foreign investors and denoted foreign capital. By deﬁnition, h h f f kt = kth + kt f . Note that π t rt kt is the tax revenue from domestic capital and π t rt kt is the tax revenue from foreign capital. Foreign investors can invest in both foreign countries and the domestic country but domestic investors can invest only in the domestic country due to the capital controls. We use a circumﬂex to indicate a foreign variable. The small open economy takes the world interest rate, rˆt+1, and the foreign tax rate, πˆ ft+1, as given. Hence, we have: (1 – πft+1)rt+1 = (1 – πˆ ft+1) rˆt+1.

(8)

Equation (8) is the arbitrage condition, indicating that the rate of return to capital in the domestic country equals the world rate in equilibrium. A competitive equilibrium can be characterized as a set of sequences { yt , rt , wt , kt , ctt , ctt+1 , gt , st , πth } satisfying equations (1) through (8) and the following condition: s[ wt , τ t , (1− π th+1 )rt +1 ] = kth+1 (1 + n) .14

(9)

Equation (9) indicates that total savings in the domestic country must equal the total capital stock held by the domestic investor in equilibrium. In the two-period overlapping generations model (OLG), the young use all their savings to purchase capital from the old at the end of each period. Thus, investment equals the capital stock. The trade surplus in period t, X t , is deﬁned as the excess of the value of exports over imports, or the excess of total output over domestic absorption, which is the sum of private and government consumption plus the increase in the capital stock. Deﬁning xt ≡ X t / Lt , the trade surplus per worker is: xt = f (kt ) − ctt − ctt −1 /(1 + n) − [kt +1 (1 + n) − kt ] − g

(10)

t−1

where ct is the second-period consumption of a member of generation t-l. The economy is in steady-state equilibrium if all endogenous variables are invariant over

14 With domestic capital controls, kth + ktf = kt. Also recall that rt = f '(kt) and wt = f (kt) – kt f '(kt).

China’s Capital Tax Reforms in an Open Economy

101

time, i.e., yt = y , wt = w, rt = r , kt = k , kt = k , st = s , gt = g and xt = x . The steady-state equilibrium can be characterized by the following equations: y = f (k ) ,

(11)

r = f '(k ) ,

(12)

w = f (k ) − kf '(k ) ,

(13)

g = π h rk h + π f rk f + τ ,

(14)

(1 – πf )r = (1 – πˆ f )rˆ,

(15)

s[ w, τ , (1− π h )r ] = k h (1 + n) , and

(16)

x = f (k ) − c1 − c2 /(1 + n) − nk − g .

(17)

Equations (11) through (17) are the steady-state versions of equations (1), (2), (3), (7), (8), (9) and (10), respectively. Given these equilibrium equations, we can analyze the effects of capital tax reforms. Comparative Steady-State Analysis of Capital Tax Reforms In the overlapping generations model, taxing capital is equivalent to taxing investment. We consider two tax reforms that unify the capital tax system, namely, a decrease in the tax rate on domestic capital and an increase in the tax rate on foreign capital. The government adjusts expenditures to meet the budget constraint. We analyze the effect of the tax reforms on the interest rate, the capital-labor ratio, and the output-labor ratio, foreign investment, and the trade balance. We ﬁrst analyze the situation in which the government changes the tax rate on domestic capital and keeps the tax rate on foreign capital ﬁxed. Since the world aftertax interest rate is given and the tax rate on foreign capital remain unchanged, the before-tax interest rate in the domestic country, r, remains unchanged from equation (15), i.e., dr / d π h = 0 . Hence, dy / d π h = 0 and dk / d π h = 0 from equations (11) and (12). These results are summarized in the following proposition. Proposition 1 A decrease in the tax rate on domestic capital has no effect on the domestic beforetax interest rate, the capital-labor ratio, or the output-labor ratio. Although not altering the before-tax interest rate, a change in the tax rate on domestic capital does change the after-tax interest rate, i.e., (1− π h )r , faced by domestic investors. Studies of small open economies focus on small industrial

The Chinese Economy after WTO Accession

102

economies having a uniﬁed capital tax system, e.g., Bovenburg (1986 and 1992). If foreign capital and domestic capital are subject to the same tax rate, a change in the capital tax rate changes the before-tax interest rate in a small open economy. Our consideration of a dual capital tax system leads to different conclusions concerning the effect of capital taxation in a small open economy. To determine how a change in the tax rate affects domestic capital and foreign capital, differentiate equation (16) totally. Deﬁning sr ≡ ∂ s / ∂ [(1− π h )r ] ≥ 0 , sw ≡ ∂ s / ∂ w , assuming sr ≥ 0 and 0 < sw < 1 , using dw / dr = −k , and holding τ constant, we obtain: sw (−k )dr + sr [(1− π h )dr − rd π h ] = (1 + n)dk h or [ sr (1− π h ) − sw k ]

dr dk h . − s r = ( 1 + n ) r d πh d πh

Since dr / d π f = 0 from Proposition 1, we have: −s r dk h = r <0 . h 1+ n dπ

(18)

The denominator on the right side of equation (18) is positive and the numerator is negative. Thus, dk h / d π h < 0. Since k h + k f = k , the change in foreign capital is the difference between the change in total capital in the domestic country and f h the change in domestically owned capital, i.e., dk = dk − dk . Given that the h h dπ dπ d πh change in domestic capital is zero, i.e., dk / d π h = 0 , we have: dk f dk h = − >0. d πh d πh

(19)

In the special case in which savings are interest-inelastic, i.e., sr = 0 15 dk h / d π h = dk f / d π h = 0 from equations (18) and (19). Proposition 2 summarizes the above results. Proposition 2 A decrease in the tax rate on domestic capital increases domestic investment and decreases foreign investment. If savings are completely interest-inelastic, a decrease in the tax rate on domestic capital has no effect on domestic and foreign investment. Intuitively, a decrease in the tax rate on domestic capital increases the domestic aftertax rate of return to capital, given the world interest rate. An increase in the after-tax 15 If the utility function is Cobb-Douglas, savings are perfectly interest-inelastic. Many empirical studies ﬁnd no signiﬁcant relationship between savings and the interest rate.

China’s Capital Tax Reforms in an Open Economy

103

rate of return to capital increases domestic savings and investment. Since total capital in the domestic country remains unchanged, the increase in domestic investment implies a decrease in foreign investment. If savings are completely interest-inelastic, an increase in the after-tax rate of return to capital due to the tax cut has no effect on domestic savings. Therefore, in this case, domestic investment and foreign investment are unchanged. In the standard literature of a small open economy with a uniﬁed capital tax rate, a decrease in the capital tax rate in the domestic country lowers the interest rate in the domestic country because the small country takes the world after-tax interest rate as given, increases the capital labor ratio, and may increase both domestic and foreign capital. Next we examine the effect of tax reform on the trade balance.16 Substituting equation (14) into equation (17) yields: x = f (k ) − c1 − c2 /(1 + n) − nk − (π h rk h + π f rk f + τ )

(20)

Differentiating equation (20) with respect to π h , holding τ and π f ﬁxed and using dw / dr = −k , gives the effect of a change in the tax rate on domestic capital on the trade account as: ∂c ∂ c k dr dx dk dr dr dk dr =r + 1k + 2 −n dr d π h ∂ w d π h ∂ w 1+ n d πh dr d π h d πh ⎡ dk h dr dk f dr ⎤⎥ . f − ⎢ rk h + π h r h + π h k h + π r +πf k f h h ⎢ dπ dπ dπ d π h ⎥⎦ ⎣

(21)

f h With dr / d π h = 0 and dk = − dk from Proposition 1 and equation (19), equation h dπ d πh (21) becomes: f ⎛ h dk h dx h f dk ⎜ r + π h = −rk − ⎜π r ⎜⎝ dp dr d πh

⎞⎟ ⎡ dk h ⎤ ⎟⎟ = −r ⎢ k h + (π h − π f ) h ⎥ . ⎢ d π ⎥⎦ ⎠⎟ ⎣

(22)

We know that π h − π f > 0, i.e., domestic capital is subject to a higher tax rate, and dk h < 0 from equation (18). It is plausible to assume that the decrease in domestic d πh dk h capital is smaller than the existing stock of domestic capital, i.e., k h > . Thus, d πh dx < 0. If savings are completely interest-inelastic, dk h / d π h = 0 from equation d πh (18) and dx / d π h = −rk h < 0 unambiguously from equation (22). Hence, we have the following proposition.

16 Persson (1985) studies the effect of government debt on the trade surplus of a small open economy.

The Chinese Economy after WTO Accession

104

Proposition 3 A decrease in the tax rate on domestic capital increases the domestic trade surplus. Intuitively, the domestic interest rate being unaffected by the change in the tax rate on domestic capital so that the domestic output and consumption are unchanged. Thus, only the change in government spending affects the trade surplus. An increase in the tax rate on domestic capital increases government revenues and, consequently, government spending. Hence, the trade surplus decreases. A decrease in the tax rate on domestic capital has the opposite effect. We turn now to the case in which the domestic government increases the tax rate on foreign capital and adjusts its spending to meet the budget constraint. Totally differentiating equation (15) with respect to r and π f , holding rˆ and πˆ f ﬁxed, yields: (1− π f )dr − rd π f = 0 or

dr r = >0. f dπ 1− π f

Differentiating equation (12) with respect to π f gives the effect of an increase in the tax rate on the capital-labor ratio as: dr dk dk dr r 1 1 or = f "(k ) = = <0. f f f f f dπ dπ dπ d π f "(k ) 1− π f "(k )

(23)

Hence, an increase in the tax rate on foreign capital decreases the capital-labor ratio. Since the capital-labor ratio and the output-labor ratio are positively related from equation (11), a decrease in the capital-labor ratio results in a decrease in the outputlabor ratio. These results are summarized in the following proposition. Proposition 4 An increase in the tax rate on foreign capital increases the before-tax interest rate and decreases the capital-labor ratio and the output-labor ratio. Intuitively, in a small open economy, the after-tax world interest rate is given. When the domestic country raises the tax rate on foreign capital, the domestic beforetax interest rate must increase to equate the domestic after-tax interest rate with the world after-tax interest rate. As the domestic before-tax interest rate increases, both the capital-labor ratio and the output-labor ratio decrease. The effect of an increase in the tax rate on foreign capital on both domestic capital and foreign capital is given in the following proposition.

China’s Capital Tax Reforms in an Open Economy

105

Proposition 5 If savings are completely interest-inelastic and the production function is CobbDouglas, an increase in the tax rate on foreign capital decreases both domestic capital and foreign capital. The proof follows. Totally differentiate equation (16), deﬁning sr ≡ ∂ s / ∂ [(1− π h )r ] and sw ≡ ∂ s / ∂ w and using dw / dr = −k , holding τ and π h ﬁxed, to obtain: −sw kdr + sr (1− π h )dr = (1 + n)

dk h −sw k + sr (1− π h ) dk h . = dr , or 1+ n dr dr

(24)

The denominator of equation (24) is positive and the sign of the numerator is ambiguous. If savings are completely interest-inelastic, the sign of the numerator is negative so that dk h / dr < 0 . Since dr / d π f > 0 , dk h / d π f < 0 , i.e., an increase in the tax rate on foreign capital decreases capital held by domestic investors. Differentiating the identity given by kth + kt f = kt and using equations (23) and (24), we obtain the change in foreign capital after an increase in the tax rate on foreign capital as follows: dk f dk dk h dk dr dk h dr = − = − dr d π f dr d π f dπ f dπ f dπ f ⎡ r −sw k + sr (1− π h ) ⎤⎥ dr 1 . =⎢ − ⎢1− π f f "(k ) ⎥ dπ f 1+ n ⎣ ⎦

(25)

r 1 < −sw k + sr (1− π h ) . 1+ n 1− π f f "(k ) > If the production technology is Cobb-Douglas, e.g., y = f (k ) = Ak α and 0 < α <1,

Thus, dk f / d π f

< >

0 as

r = f '(k ) = Aαk α−1 , f "(k ) = Aα (α −1)k α−2 , and sw k sr (1− π h ) r k 1 . = − < − + 1+ n 1+ n 1− π f f "(k ) (1− π f )(1− α ) Recall that 0 < sw < 1 and sr ≥ 0 . Thus, dk f / d π f < 0 , which completes the proof. Intuitively, an increase in the tax rate on foreign capital increases the domestic before-tax interest rate and decreases the domestic wage rate. The increase in the interest rate increases domestic savings, while the decrease in the wage rate decreases domestic savings. If savings are completely interest-inelastic, domestic savings decreases unambiguously because only the second effect applies. A decrease in domestic savings leads to a decrease in capital held by domestic investors. The change in the foreign

The Chinese Economy after WTO Accession

106

capital depends on the magnitude of the change in the capital held by domestic investors. If the production function is Cobb-Douglas, the decrease in capital held by the domestic investors will be smaller than the decrease in total capital. Therefore, foreign capital must decrease. Proposition 5 predicts that an increase in the tax rate on foreign capital decreases foreign investment and a decrease in the tax rate on foreign capital does the opposite, which is consistent with the evidence from China in the 1990s. In 1991, China announced a low tax rate for foreign enterprises in special areas. FDI increased immediately from $4.4 billion in 1991 to $11 billion in 1992, and reached more than $40 billion in 1996. Thus, our analysis predicts that, if China reforms the capital tax system by increasing the tax rate on foreign capital, FDI to China is likely to decrease, other things being equal. In addition, Proposition 5 predicts that an increase in the tax rate on foreign capital reduces domestic investment, and a decrease in the tax rate on foreign capital increases domestic investment. This is also consistent with the evidence in the early 1990s. After the 1991 tax cut on foreign capital, China’s domestic investment also increased dramatically, causing an overheating of the economy for several years. The effect of a change in the tax rate on foreign capital on the trade surplus is summarized by the following proposition. Proposition 6 If the change in government spending is positive and the effect on net output dominates the effect on consumption, an increase in the tax rate on foreign capital decreases the trade surplus. Differentiating equation (20) with respect to π f , holding τ and π h ﬁxed, yields the effect of a change in tax rate on the trade account as follows: ∂c ∂ c k dr dx dk dr dr dk dr =r + 1k + 2 −n dr d π f ∂w dπ f ∂ w 1+ n d π f dr d π f dπ f ⎡ dk h dr dk f dr − ⎢ rk f + π h r + πhk h +πf r +πf k f f f ⎢ dr dπ dπ dπ f ⎣ dk dr ⎛⎜ ∂ c1 ∂ c2 1 ⎞⎟ dr = ( r − n) −⎜ + ⎟⎟ (−k ) f ⎜ ⎝ ∂w dr d π ∂ w 1+ n ⎠ dπ f ⎡ dk h dk f dr ⎤⎥ . −rk f − ⎢ π h r +πf r + (π h k h + π f k f ) f ⎢ dr dπ d π f ⎥⎦ ⎣

⎤ ⎥ ⎥ ⎦

(26)

Capital taxation affects the trade surplus through output, consumption, and government spending. The ﬁrst term in equation (26) is the net change in output, i.e., the increase in output minus the output needed for the increased population, and is negative if r > n. This effect decreases the trade surplus. The second term is the change in consumption and it is negative, i.e., an increase in the tax rate on foreign capital decreases domestic consumption. This effect increases the trade surplus. The

China’s Capital Tax Reforms in an Open Economy

107

remaining terms correspond to the change in government revenue or spending, which can be either positive or negative. An increase in the tax rate on foreign capital does not necessarily increase government tax revenue. As the tax rate increases, foreign investment may decrease and the effect of this decrease in foreign investment on tax revenue may dominate the effect of the increased tax rate. Furthermore, the decrease in capital held by domestic investors also tends to reduce total tax revenue. Hence, the net effect on the trade surplus is ambiguous and depends on the magnitude of each effect. Bovenburg (1986 and 1992) shows that an increase in territorial capital taxation worsens the trade balance. In the present model, if r > n, if the change in government spending is positive, and if the effect on net output dominates the effect on consumption, an increase in the tax rate on foreign capital reduces the trade surplus. The above analysis shows that the two ways of merging the capital tax rates have different effects on the economy. A decrease in the tax rate on domestic capital has no effect on the domestic interest rate, the capital-labor ratio or the output-labor ratio, but it increases domestic capital, decreases foreign capital and increases the trade surplus. An increase in the tax rate on foreign capital increases the domestic interest rate, decreases the capital-labor ratio, decreases the output-labor ratio, and decreases domestic capital; it may also decrease foreign capital and the trade surplus if savings are perfectly interest-inelastic. These results imply that lowering the tax rate on domestic capital is better for the domestic economy than raising the tax rate on foreign capital. The current model can be extended in two interesting directions. First, if the government transfers the capital tax revenue to workers, the way in which capital taxation affects the economy through intergenerational redistribution, in addition to other channels, can be analyzed. Second, introducing public capital to the model allows a consideration of how capital taxation affects the economy through infrastructure development. In this case, the efﬁciency of government expenditure is crucial in determining the effect of capital tax reforms. If public spending is less efﬁcient than private investment, our analytical results remain unchanged qualitatively. Concluding Remarks Inevitably, China will unify the tax rates on capital income for domestic and foreign-invested enterprises. The dual capital tax system did encourage foreign investment, which plays an important role in China’s economic growth. However, the disadvantages of the system are substantial. First, the unequal tax treatment distorts investment incentives for domestically owned ﬁrms. Second, the dual rates create a strong incentive for Chinese ﬁrms to send funds abroad and then repatriate them as FDI so as to qualify for more favorable tax treatment. The large ﬂow of funds from Hong Kong is consistent with such behavior. Third, the dual system may cause capital ﬂight out of China. Capital controls combined with a higher tax rate on domestically owned capital induce Chinese ﬁrms to move capital out of China

108

The Chinese Economy after WTO Accession

illegally. The extent of this capital ﬂight is unknown but some estimates indicate that it has been rather large.17 The policy implication of our model is that, to maintain a high level of per capita output and a high level of wage rate, and to improve the trade balance, the government should lower the tax rate on domestically-owned capital. However, many people anticipate that China will reform the dual capital tax system by both lowering the domestic capital tax rate and increasing the foreign capital tax rate. China’s capital tax reforms will have different effects on source countries. Foreign investment to China is highly concentrated, mainly from the greater China region, Japan, the US, and several major European countries. The impact of China’s tax reforms depends also on the tax systems of the interdependent economies. If these economies, such as Hong Kong and Macau, have a territorial capital tax system, the change in China’s capital tax system will reduce unambiguously the after-tax proﬁts and investment incentives of the ﬁrms from these economies. If these economies have a residential capital system and tax sparing agreements, such as Japan and the UK, a change in China’s capital tax system will also reduce the after-tax proﬁts and investment incentives of the ﬁrms from these economies. Alternatively, if these economies have a residential capital system but don’t have tax sparing agreements, such as the U.S., a change in China’s capital tax system will have no effect on the investment decision of the ﬁrms from these countries. Capital tax reforms in China will have differential impacts on different regions because foreign investment is highly concentrated in the coastal areas. In 2001, Guangdong received $11.9 billion, accounting for 25.5 percent of total foreign investment; Jiangsu, Shanghai, and Fujian received $6.9 billion, $4.3 billion and $3.9 billion, respectively.18 Taken together, these four of thirty-one provinces and municipalities in China received 58 percent of the total foreign investment. Thus, capital tax reforms have more impact on the regions with more foreign capital. Also, the uniﬁcation of capital tax rates will eliminate the tax advantages obtained recently by the less developed regions, i.e., the middle and western parts of China. For years, the government has provided preferential tax policies to relatively more developed areas, i.e., special economic zones and economic and technological development zones assuming that the early-developed areas could help the less developed areas. However, this strategy was based on a planned economy system while the Chinese economy has become more market-oriented. The government’s ability to force the rich areas to help the poor areas has been weakened so that the poor areas are now in need. If the preferential tax policies are eliminated, the less-developed regions will be hurt again. Therefore, other preferential policies should be provided to the lessdeveloped regions once capital taxes are uniﬁed.

17 For example, China Daily (August 13, 2003) estimates that capital ﬂight from China amounted to $52 billion in 1997–99, http://www.china.org.cn/english/China/72319.htm. 18 China Statistical Bureau (2002).

China’s Capital Tax Reforms in an Open Economy

109

References Bovenberg, A. Lans (1986), ‘The effects of capital income taxation on international competitiveness and the trade ﬂows’, American Economic Review, Vol. 79, No. 5, pp. 1045–64. Bovenberg, A. Lans (1989), ‘Capital income taxation in growing open economies’, Journal of Public Economics, Vol. 31, No. 3, pp. 347–76. Bovenberg, A. Lans (1992), ‘Residence- and source based taxation of capital income in an overlapping generations model’, Journal of Economics, Vol. 56, No. 3, pp. 267–95. Chamly, Christopher (1981), ‘Welfare cost of capital taxation in a growing economy’, Journal of Political Economy, Vol. 89, No. 4, pp. 469–96. Chamly, Christopher (1986), ‘Optimal taxation of capital income in general equilibrium with inﬁnite lives’, Econometrica, Vol. 54, No. 3, pp. 607–22. China Daily, August 13, 2003. Available at http://www.china.org.cn/english/ China/72319.htm. China Statistical Bureau (1999, 2001, 2002), Statistical Yearbook of China, Statistical Publishing House, Beijing. Diamond, Peter A. (1970), ‘Incidence of an interest income tax’, Journal of Economic Theory, Vol. 2, No. 3, pp. 211–24. Frenkel, Jacob A., Razin, Assaf and Sadka, Efraim (1991), International Taxation in an Integrated World, MIT Press, Cambridge, Mass. Gordon, Roger H. (1990), ‘Can capital income taxes survive in open economies?’ Working Paper 3416, National Bureau of Economic Research, Cambridge, Mass. Gordon, Roger, Bovenberg, A. Lans (1996), ‘Why is capital so immobile internationally? Possible explanations and implications for capital income taxation’, American Economic Review, Vol. 86, No. 5, pp. 1057–75. Hines, James (1996), ‘Altered states: taxes and the location of foreign direct investment in America’, American Economic Review, Vol. 86, No. 5, pp. 1076– 94. Hines, James (1998), ‘Tax sparing and direct investment in developing countries’, NBER Working Paper 6728. International Monetary Fund (2001), Government Finance Statistics Yearbook, Washington D.C. Judd, Kenneth L. (1985), ‘Redistributive taxation in a simple perfect foresight model’, Journal of Public Economics, Vol. 8, No. 1, pp. 59–83. Li, Bichang (1988), China’s External Tax Law System (Zhongguo Duiwai Shuishou Falu Zhidu). Law Publishing House, Beijing. Lin, Shuanglin (1998), ‘Labor income taxation and human capital accumulation’, Journal of Public Economics, Vol. 68, No. 2, pp. 291–302. Lucas, Robert E. (1990), ‘Supply-side economics: an analytical review’, Oxford Economic Papers, Vol. 42, No. 2, pp. 293–316.

110

The Chinese Economy after WTO Accession

Mintz, Jack M. (1992), ‘Is there a future for capital income taxation?’, Working Paper 108. Organization for Economic Co-operation and Development, Paris. National People’s Congress (1986), Law of the People’s Republic of China on Foreign-Capital Enterprises, Promulgated by Order No. 39 of the President of the People’s Republic of China and effective as of April 12, 1986. National People’s Congress (1991), Income Tax Law of the People’s Republic of China for Enterprises With Foreign Investment and Foreign Enterprises, April 9. Nielsen, Soren Bo, Sorensen, Peter Birch (1991), ‘Capital income taxation in a growing open economy’, European Economic Review, Vol. 34, No. 1, pp. 179– 97. Persson, Torsten (1985), ‘Deﬁcits and intergenerational welfare in open economies’, Journal of International Economics, Vol. 19, Nos 1–2, pp. 67-84. Rosen, Harvey S. (2002), Public Finance, Irwin, Chicago. Sibert, Ann (1990), ‘Taxing capital in a large open economy’, Journal of Public Economics, Vol. 41, No. 3, pp. 297–317. State Council (1993), Provisional Regulations of the PRC on Enterprise Income Tax, December 13. Summers, Lawrence H. (1981), ‘Capital taxation and accumulation in a life cycle growth model’, American Economic Review, Vol. 71, No. 4, pp. 533–44. Tanzi, Vito, and Zee, Howell (2000), ‘Tax policy for emerging markets: developing countries’, National Tax Journal, Vol. 53, No. 2, pp. 299–322.

Chapter 7

On the Intertemporal Sustainability of Fiscal Debt1 Ying Wu

Introduction The growing government budget deﬁcits in China have recently attracted the increasing attention of economists not only because China’s economic growth has recently outperformed its counterparts in the world but also because lasting ﬁscal expansions have been a main engine for the recent robust economic growth in China. Is the pattern of China’s growing government debt and ﬁscal deﬁcits sustainable into the future? Many existing studies of the ﬁscal risk in China underline deﬁcient ﬁscal revenues (Bahl and Wallace, 1995; Brean, 1998; Lin, 2000) or highlight the potential threat that stems from the large contingent public debt such as state-owned banks’ non-performing loans extended to other state-owned enterprises, and ﬁscal subsidies to the pay-as-you-go type social security system (Lardy, 1998; IMF, 2002; Krumm and Wong, 2002; Lin, 2003). Surprisingly few studies, however, have attempted to analyze the ﬁscal sustainability of China’s explicit public debt at the government level by examining the state budgetary data in the present value terms. The basic analytical framework in this study is therefore simply to follow the law of motion of government debt over time and thus assess the imbedded ﬁscal risk and sustainability in present value terms using the current and historical budgetary data.2 Speciﬁcally, this chapter is based on two strands of the existing literature on the assessment of ﬁscal sustainability. The ﬁrst strand involves theoretical studies that 1 The paper was presented at the CES 2003 Conference held at the University of Michigan, Ann Arbor, August 2–3, 2003. 2 Indeed, the political economy aspects of government debt have also recently received increasing attention; see Persson and Svenson (1989), Rogoff (1990), and Tanzi (1994), among others. This literature challenges the conventional assumption that the government acts as a benevolent social planner pursuing efﬁcient ﬁscal policies by pointing to an inherent bias of excessive ﬁscal debt: as a strategic player of ﬁscal debt, the government seeks higher expenditure or tax cuts to boost its reelection prospects or to manage to constrain the actions of successor regimes. Unlike the simple and direct assessment approach using the present value budget constraint, the political-economy approach sheds light on a new way of thinking of the debt formation mechanism and underlined behavior of policy makers.

112

The Chinese Economy after WTO Accession

center on the transversality condition that ﬁscal debt grows no faster than the interest rate to respect the present value budget constraint and exclude an unbounded ratio of debt to GDP (the debt ratio, for short). Nevertheless, McCallum (1984) notes that the transversality condition can be violated, as the debt could grow faster than the economy in a prolonged period of high interest rates. Therefore, Barro (1989) and Kremers (1989) argue for a necessary condition that constrains the size of primary ﬁscal surpluses and thus serves on the upper bound for the debt ratio if the interest rate is greater than the growth rate. This strand of literature judges on ﬁscal sustainability by essentially examining the relative magnitude of the interest rate and the growth rate. The second strand of the literature focuses on the time-series properties of ﬁscal debt and primary ﬁscal balances. Speciﬁcally, it intends to answer whether the historical process that generates ﬁscal data is likely to result in the violation of the present value budget constraint. If so, ﬁscal policy should be regarded as unsustainable. In this strand, Hamilton and Flavin (1986) suggest a sufﬁcient condition for sustainability in the sense of present value budget constraint: if debt is stationary, the primary ﬁscal balance will be sustainable. In contrast, while showing that the debt in the United States is nonstationary, Trehan and Walsh (1988) argue for a necessary and sufﬁcient condition for sustainability (again, in the sense of present value budget constraint): debt and primary balances are cointegrated with a cointegrating vector (i, 1) if debt and deﬁcits are integrated of order one and interest rates are constant. Alternatively, Trehan and Walsh’s necessary and sufﬁcient condition for sustainability can also be stated as the following: overall government expenditures (inclusive of interest) and tax revenues must be cointegrated with the cointegrating vector (1, -1) if the same conditions hold as above. Intuitively, under a balanced budget in the prevent value terms (i.e., ﬁscal sustainability), overall government expenditures and taxes are supposed to move together in a one-toone fashion: if one increases by a certain amount, so does the other. In turn, this pattern of cointegration guarantees the proportional comovement between debt and primary ﬁscal surpluses: if debt increases by i, primary ﬁscal surpluses will increase by one dollar. Bravo and Silvestre (2002) applied the Trehan-Walsh condition to eleven member states of the European Union and identiﬁed the possibly sustainable budgetary paths for ﬁve countries in the group. Without actually relying on debt data per se, this chapter employs two methods concerning only budgetary data to analyze the long-run dynamics of China’s government ﬁscal debt from 1979 to 2001. The ﬁrst concerns traditional difference approach. It begins with the government’s intertemporal budget constraint and examines the sustainability condition under which the debt-GDP ratio converges to its intertemporal equilibrium level. The second approach applies the concept of the present value budget constraint – the expected present value of net ﬁscal surpluses in the future has to meet the difference between the initial debt stock and the present value of the terminal debt stock (Hamilton and Flavin, 1986) – and undertakes the approach by Trehan and Walsh (1988) to perform a cointegration analysis for China’s government expenditures and tax revenues.

On the Intertemporal Sustainability of Fiscal Debt

113

The chapter is organized as follows. The next section derives the ﬁrst-order difference equation of debt ratio from the intertemporal budget constraint and analyzes the convergence condition and some comparative static results. The third section provides a brief review of the present value budget constraint. The fourth section is devoted to the cointegration analysis of government expenditures and tax revenues, and the estimations of the restricted and unrestricted vector error correction (VEC) models. The ﬁfth section concludes the chapter. A Traditional Difference Analysis Before turning to the modern approach to the present value budget constraint, it might be useful to begin with a more traditional difference analysis as a supplementary assessment of ﬁscal sustainability, which essentially does not need to use debt data. Starting with a setting of government intertemporal budget constraint, this section derives a ﬁrst-order difference equation on the debt ratio and then analyzes its dynamic properties. Denote ﬁscal debt by B, government expenditures net of interest payment (primary expenditures) by G, tax revenues by T, and the nominal interest rate by i. All the upper-case letters represent variables in nominal terms, and the subscript t is a time index. The government intertemporal budget constraint for period t is: Bt − B = Gt − Tt + it B . t−1 t−1

(1)

Let P represent the GDP deﬂator and Y real GDP. Dividing both sides of (1) by the nominal GDP for period t, PtYt, yields: b ib t−1 = g − τ + t t−1 , bt − t t 1+ π + η ' 1+ π + η '

(2)

where π is the inﬂation rate between t-1 and t, η′ is the growth rate of real GDP between t-1 and t, τ is the tax-to-GDP ratio, and the lower-case letters b and g are the debt ratio and government expenditure-GDP ratio, respectively. A further manipulation and rearrangement of (2) produces a ﬁrst-order difference equation for the debt ratio below (taking the interest rate, inﬂation rate, and the growth rate as either exogenous or pre-determined):3 1+ i b bt = +z , 1 + η t−1

(3)

3 Given the fact that the nominal interest rate in China, which is, to a great extent, under the control of monetary-ﬁscal authority, ﬂuctuates only limitedly, it may also be appropriate to assume that the interest rate is stationary around a mean (Bravo and Silvestre, 2002).

114

The Chinese Economy after WTO Accession

where z is the ratio of primary budget deﬁcit to GDP (z ≡ g-τ >0), and η is the growth rate of nominal GDP between t-1 and t (η=η′+π). Denote the initial debt ratio by b0. It then immediately follows that the time path of debt ratio determined by (3) can be expressed as: ⎡ ⎤ ⎢ ⎥ t ⎢ z z ⎥⎥ ⎛⎜ 1 + i ⎞⎟ ⎢ . ⎟ + bt = ⎢b − 0 ⎛ η − i ⎞ ⎥ ⎜⎜⎝1 + η ⎟⎟⎠ ⎛ η − i ⎞ ⎟⎟ ⎟⎟ ⎥ ⎢ ⎜⎜ ⎜⎜ ⎢ ⎜⎝1 + η ⎟⎟⎠ ⎜⎝1 + η ⎟⎟⎠ ⎥⎥ ⎢⎣ ⎦

(4)

Suppose that the growth rate of nominal GDP is positive. The necessary condition for ﬁscal sustainability requires that nominal GDP grow at a rate not lower than the nominal interest rate. In particular, if the growth rate exceeds the interest rate and the initial debt ratio is greater than the intertemporal equilibrium level of the debt z ratio, i.e., , the debt ratio will smoothly decline in a non-oscillatory ((η − i) (1 + η )) way over time to approach the intertemporal equilibrium level of the debt ratio. If the growth rate exceeds the interest rate but the initial debt ratio is less than its intertemporal equilibrium level instead, then the debt ratio will smoothly rise in a non-oscillatory way over time to approach (from below rather than from above) the intertemporal equilibrium level of the debt ratio. The convergence to equilibrium hinges upon the relative magnitude of the nominal interest rate and the growth rate of nominal GDP, or the relative magnitude of the real interest rate and the growth rate of real GDP. In the last decade, China’s growth rate was uniformly higher than the interest rate except only one year, as shown in Table 7.1. The evidence strongly favors the convergence of the debt ratio to its intertemporal equilibrium level. In the above analysis, although the intertemporal equilibrium level of the debt ratio per se is greater if the inﬂation rate is higher or if the deﬂation rate is lower, the impact of inﬂation on the time path of the debt ratio depends on whether the initial debt ratio is greater than its intertemporal equilibrium level. Suppose that the growth rate exceeds the interest rate. If the initial debt ratio is less than its intertemporal equilibrium level (the most likely case, since an economy with large ﬁscal debt often has a large amount of primary budget deﬁcit plus interest payment), a lower inﬂation rate raises the debt ratio, and vice versa. Otherwise, a sufﬁcient condition for lower inﬂation to reduce the debt ratio is that the time elapse from the initial period is sufﬁciently long (see (A1) in the Appendix); in this case, the same sufﬁcient condition also applies for higher inﬂation to raise the debt ratio. A comparable result holds for the impact of real GDP growth on the dynamics of the debt ratio. Again, suppose that the economy grows at a rate higher than the interest rate. If the initial debt ratio is less than its intertemporal equilibrium level, then a higher real GDP growth will eventually lead to a lower debt ratio in the long run; otherwise, the result always holds (see (A2) in the Appendix). Unlike inﬂation, real GDP growth relieves debt burden by improving the economy’s ability to pay

On the Intertemporal Sustainability of Fiscal Debt

Table 7.1

115

Interest Rates and GDP Growth Rates in China (1990–2000)

Years Interest Ratesa (%) GDP Growthb (%) 11.2 7.92 1990 16.1 7.20 1991 21.5 7.20 1992 33.3 10.08 1993 35.3 10.08 1994 25.3 10.44 1995 16.7 9.00 1996 9.60 8.55 1997 6.62 4.59 1998 2.75 3.24 1999 8.95 3.24 2000 a. The interest rate is the bank rate based on the end-of-period annual data. b. The GDP growth rate in the table measures the growth rate of nominal GDP. Source: International Financial Statistics (Washington D.C.: International Monetary Fund).

without deteriorating the terms of debt payment. Furthermore, as shown in (A3) of the Appendix, an increase in the primary deﬁcit not only raises the intertemporal equilibrium level of the debt ratio but also ‘shifts’ up the time path of the debt ratio. A Brief Review of the Present Value Budget Constraint Although the previous difference analysis of the debt ratio is useful in determining the debt-convergence conditions and the roles of other economic variables in such a process, it either considers the budget deﬁcit as pre-determined or treats the budget deﬁcit only as a shifting factor in inﬂuencing government debt. Without incorporating the present value budget constraint, the sustainability analysis will lose an important dimension in assessing the future ﬁscal path and allowing more options to achieve ﬁscal sustainability. The purpose of this section is to review Hamilton and Flavin (1986) and Trehan and Walsh (1988), the two key studies of the present value budget constraint. It mainly involves using the concept of intertemporal budget constraint to derive the present value budget constraint and transform it into a testable version for ﬁscal sustainability. For the analytical convenience below, the review begins with time index j rather than t. Adding Bj-1 on both sides of (1) and dividing across the equation by the discount factor (1+i)j produces a budget constraint in present value terms: Bj (1 + i )

j

=

G j −T j j−1 . + j−1 (1 + i ) j (1 + i ) B

(5)

116

The Chinese Economy after WTO Accession

(5) says that, according to the present value budget constraint, the present value of the primary deﬁcit reﬂects an increase in the public debt. Considering a period beginning from an initial time j=0 to an arbitrary future time j=s, in which (5) holds for each time slot. Adding up (5) over all the time slots within the period and then dividing both sides by 1+i yields: s T j −G j Bs . B = ∑ + −1 j=0 j +1 (1 + i ) s+1 (1 + i )

(6)

Moving time slots forward by t+1 turns (6) into: B s Tt +1+ j − Gt +1+ j t +1+s . + Bt = ∑ j=0 (1 + i )t +2+ j (1 + i )t +2+s

(7)

Let s→∞, (7) becomes: B ∞ Tt +1+ j − Gt +1+ j t +1+s . Bt = ∑ + lim s→∞ (1 + i )t +2+s j=0 (1 + i )t +2+ j

(8)

According to (8), since the asymptotic value of the present value of the debt is worth the sum of all past primary deﬁcits (the ﬁrst term on the RHS of (8) with the sign reversed) and the initial debt (Bt), ﬁscal sustainability in the sense of the present value budget constraint does not necessarily exclude either large primary deﬁcits or high debt; instead, it only requires that future primary surpluses outweigh future primary deﬁcits by a sufﬁcient amount to cover the gap between the initial debt stock and the present value of the terminal debt stock. A stronger condition for ﬁscal sustainability concerns the transversality condition that the present value of government debt goes to zero in the long run:4 B t +1+s = 0 lim . s→∞ (1 + i )t +2+s

(9)

Now, taking ﬁrst-order difference of equation (8) and substituting (1) into the LHS of the resulting difference equation yields: ΔB ∞ ΔTt +1+ j − ΔGt +1+ j t +1+s . + lim (Gt + iB ) − Tt = ∑ t−1 →∞ s 2 t + + j j=0 (1 + i )t +2+s (1 + i )

(10)

4 In general, the sustainability condition does not require that the present value of the terminal debt stock go to zero; however, in the representative agent model, this could cause difﬁculty to the government’s strategy of rolling over its debt.

On the Intertemporal Sustainability of Fiscal Debt

Table 7.2

117

Augmented Dickey-Fuller Test for Unit Roots*

ADF Test Statistic on T: 3.226 lag: 1 with constant and trend

1% critical value: 4.469 5% critical value: 3.645 1% critical value: 3.260

ADF Test Statistic on ΔT: 3.080 lag: 2 without constant and trend

1% critical value: 2.697 5% critical value: 1.960 1% critical value: 1.625

ADF Test Statistic on G: 3.150 lag: 1 with constant and trend

1% critical value: 4.469 5% critical value: 3.645 1% critical value: 3.260

1% critical value: 2.697 ADF Test Statistic on ΔG: 3.215 5% critical value: 1.960 lag: 2 1% critical value: 1.625 without constant and trend * The critical values are MacKinnon critical (absolute) values for rejection of hypothesis of a unit root. The symbol Δ is the ﬁrst-difference operator so that and ΔT≡Tt-Tt-1, and ΔG ≡

If all the variables are integrated of order one, the variables on the RHS of (10) are stationary. Then, the LHS of (10) must also be stationary to satisfy the present value budget constraint. Hence, if the overall government expenditure (Gt+iBt-1) and taxes (Tt) are respectively integrated of order one, they must be co-integrated with the co-integrating vector (1, -1). It therefore follows that the assessment of ﬁscal sustainability over the long run boils down to a stationarity test of overall government expenditures and tax revenues and the consequent cointegration analysis of these ﬁscal data. Cointegration Analysis with the Data on China The annual data used in the empirical analysis in this section are China’s consolidated governments’ tax revenues and overall government expenditures published in the International Financial Statistics, with the sample period beginning from 1979 and ending in 2001. Table 7.2 shows the results of the Augmented Dickey-Fuller test of unit roots for the ﬁscal data. The inspection of the unit roots tests suggests that the series of tax revenue and overall government expenditure are integrated of order one, i.e., I(1). Given the unit-root property of the series, to determine whether there is a long-run equilibrium relationship between tax revenues and overall government expenditures, the empirical analysis below proceeds with the Johanson cointegration test.5 The Johanson test assumes no trend in the series with a restricted intercept in the cointegration equation and uses one lag in differences (two lags in levels). As shown 5

For the classical work on the Johanson test, see Johanson and Juselius (1990).

The Chinese Economy after WTO Accession

118

Table 7.3

Johanson Cointegration Test

-

Number of Eigenvalue Cointegration Equations λtrace test None 0.632 At most 1 0.505 λmax test None 0.632

H0

H1

Statistic

1% c.v.

r=0 r≤1 r=0

r>0 r>1 r=1

35.78 14.78 20.99

24.60 12.97 20.20

in Table 7.3, both the λtrace and λmax tests in the Johanson methodology suggest that the two series are cointegrated. Although both the trace test and the max-eigenvalue test indicate that there are two cointegrating equations at both the 5 per cent and 1 per cent levels, it is clear that there can actually be only one linearly independent cointegrating vector since only two endogenous variables are involved. Therefore, the Johanson cointegration test reported in Table 7.3 literally suggests that there is a unique linearly independent cointegration relationship and thus allows the analysis below to proceed with the unique cointegration relationship.6 The estimated cointegration equation is: Tt = 12.40 + 0.92G*t (2.48) (34.26)

(11)

In (11), Gt* is the overall government expenditure inclusive of interest, and the ﬁgures in the parentheses are t-values, which are statistically signiﬁcant at the 1 per cent level. Thus, the long-run equilibrium ﬁscal path in China suggests that as the government expenditure containing interest payment increases by one yuan (a unit of Chinese currency: RMB), tax revenue will increase by about 92 cents accordingly. The rest of the cointegration analysis in this section focuses on two parts: 1). Examine the unrestricted vector error correction (VEC) model and its estimations; 2) Test the VEC restrictions and estimate the restricted VEC model. Two methods are used in selecting the length of lags for the model: the Akaike information criteria (AIC) and the Schwarz criteria (SC). The two information criteria 6 The result of the Johanson cointegration test can also be supplemented by running an alternative and quicker test called Cointegrating Regression Durbin-Watson (CRDW) test. In the CRDW test, the null hypothesis of cointegration is that the Durbin-Watson d value from the cointegration regression is zero; if the computed d value is smaller than the critical value, the hypothesis of cointegration is rejected. Running the spurious regression with the cointegration equation yields the Durbin-Watson statistic d with a value of 0.729, which exceeds 0.511, the 1% critical value for rejection of the hypothesis of cointegration. Therefore, according to the CRDW test, there exists a co-integration relationship between taxes (T) and overall government expenditure (G*).

On the Intertemporal Sustainability of Fiscal Debt

119

consider the model with the lowest AIC or SC value as desirable. Consequently, the model with one-year lag is chosen as the most appropriate model. Below is the estimated VEC model with one-year lag:

(

)

+ 1.1ΔT − 0.12ΔG* + εT ,t ΔTt = −0.88 T −12.40 − 0.92G* t−1 t−1 t−1 t −1 (−2.29) (1.91) (−0.23)

(

)

(12)

ΔG*t = 0.15 T −12.40 − 0.92G* − 0.21ΔT + 1.36ΔG* + ε t−1 G,t t−1 t−1 t− −1 (0.33) (−0.31) (2.17) Again, the ﬁgures in the parentheses are t-values. The estimated short-run adjustment coefﬁcient for the tax equation is signiﬁcant at the 1 per cent level with the correct sign, indicating that tax revenue will correct itself once it deviates from its longrun equilibrium path in the previous period. Speciﬁcally, if tax exceeds its longrun equilibrium level (i.e., equation (11)) by one yuan, it tends to revert itself by eighty-eight cents of the deviation; the resulting ﬁscal adjustment helps restore the long-run equilibrium, as shown in Figure 7.1. The similar analysis applies to the case when tax is below its co-integration relationship. This ﬁnding supports the existence of the long-run cointegration relationship between tax revenues and overall government expenditures. Furthermore, the short-run adjustment coefﬁcient for the government-expenditure equation exhibits a positive sign though the t-value itself is not signiﬁcant. Given the large magnitude of the short-run adjustment coefﬁcient for the tax equation, it follows that the follow-up changes in tax revenues bear the main adjustment burden in the long-run equilibrium achievement. Figure 7.1 depicts the long-run equilibrium path for T and G*, which is characterized by the unrestricted co-integration equation, and the two variables’ short-run adjustment paths. In contrast with the government-expenditure adjustment coefﬁcient, the tax adjustment is not only statistically signiﬁcant but also more responsive in terms of the magnitude of the coefﬁcient; therefore, the solid and long arrows pinpoint the direction, magnitude, and signiﬁcance of the tax adjustment whereas the dashed and short arrows provide the similar information for the expenditure adjustment.

Figure 7.1

The Long-run Equilibrium Relationship and Short-run Adjustment (Unrestricted Co-integration)

The Chinese Economy after WTO Accession

120

The analysis of intertemporal ﬁscal sustainability hinges upon not only the existence of a cointegration relationship between tax revenues and overall government expenditures but also the fact that the value of the cointegrating vector is (1, -1). Therefore, a natural step is to test the VEC restrictions by reestimating the VEC model under the hypothesis that the cointegrating vector is (1, -1). The estimation of the restricted VEC model involves an iterative process using the switching algorithm (See EViews manual, 2000, chapter 20). The process achieves its convergence after ﬁve iterations in my study. With the p-value equaling 0.21, the resulting likelihood ratio statistic (LR) fails to reject the null hypothesis that the cointegrating vector is (1, -1). The test result here suggests that the estimated unrestricted VEC model is not signiﬁcantly different from the restricted VEC model, which is presented below:

(

)

ΔTt = −1.09 T − 7.21− G* + 1.19ΔT − 0.75ΔG* + εT ,t t−1 t−1 t−1 t −1 (−3.99)

(

(−1.46)

(2.84)

)

(−1.56)

(13)

+ 0.55ΔG* + ε ΔG*t = −0.42 T − 7.21− G* + 0.35ΔT t−1 G,t t−1 t−1 t−1 (−1.13) (−1.46) (0.62) (0.83) Based on the insigniﬁcant estimates for the government expenditure equation, it is tax revenue that plays a key role in maintaining the co-integration relationship of ﬁscal budget and thus the sustainability of public debt. In particular, a pairwise Granger causality test with the series of taxes and government expenditure rejects the hypothesis that government expenditure does not Granger-cause tax revenue at the 1 per cent level; but it fails to reject the hypothesis that tax revenue does not Granger-cause government expenditure. Therefore, the analysis below focuses on the tax equation in (13). In the restricted VEC model (13), the short-run tax adjustment with respect to its deviation from the long-run equilibrium path remains convergent based on its negative coefﬁcient, and the speed of adjustment is even faster than the unrestricted model (-1.09 vs. –0.92). In contrast, government expenditure, like in the unrestricted VEC model, appears not to be signiﬁcantly responsive to the deviations in tax from the long-run equilibrium path. For the VEC dynamic system to be compatible to the cointegration relationship, tax must assume the burden of adjustments by exhibiting higher and more statistically signiﬁcant speed of adjustment than government expenditure. As far as the autoregressive part of the VEC model is concerned, changes in taxes are cumulative, that is, the current year’s ﬁrst-order change in taxes tends to reserve the previous year’s pattern in terms of both the direction and the magnitude of such ∂ΔTt a change ( = 1.19 ), and the estimate is signiﬁcant at the 1 per cent level. In ∂ΔT t−1 addition, (13) also suggests that if the change in government expenditure increases by one yuan in the last year relative to its level in the year before, the change in

On the Intertemporal Sustainability of Fiscal Debt

Figure 7.2

121

The Long-run Equilibrium Relationship and Short-run Adjustment (restricted co-integration)

taxes will fall by seventy-ﬁve cents this year relative to its level in the last year. Combining this result with the autoregressive tax response, a balanced budget in year t-1 (ΔTt-1=ΔG*t-1) will lead to a net tax increase of forty-four cents in year t, i.e., ΔTt=1.19-0.75=0.44. It follows that even if budget deﬁcits may exist in the present and some future periods, the long-run tax dynamic exhibits its potential in reverting ﬁscal deﬁcits and thus favoring tax saving. Figure 7.2 characterizes the long-run ﬁscal equilibrium for T and G* and their shortrun adjustment in the VEC model with the restricted cointegration relationship. The prevailing feature associated with both Figure 7.1 and Figure 7.2 is the convergence of the vector error correction process. Tax declines if it exceeds the level determined by the cointegration equation and rises if it falls short of its equilibrium level. It is tax rather than government expenditure that assumes the main role of error corrections in the adjustment toward the cointegration relationship. Concluding Remarks This chapter uses two approaches to analyze government debt and the state budget ﬁscal sustainability in China. The ﬁrst approach derives a ﬁrst-order difference equation of the debt ratio from the intertemporal budget constraint and shows that the convergence of the debt ratio (ﬁscal sustainability) to its equilibrium depends on a higher growth rate relative to the interest rate. The ﬁrst approach also shows that, with the growth rate higher than the interest rate and the initial debt ratio less than its equilibrium level, the debt ratio in the long run is negatively related to the inﬂation rate and the growth rate but positively related to primary deﬁcit itself. The second approach applies the concept of the present value budget constraint to China’s data and investigates the cointegration properties of public expenditures and tax revenues. The ﬁndings are two-fold. First, China’s government expenditures (inclusive of interest) and tax revenues are cointegrated and the cointegration vector is signiﬁcantly close to the theoretical value for ﬁscal sustainability. Second, it is the tax revenue rather than government expenditure in China that plays a key role in sustaining ﬁscal debt within the framework of the present value budget constraint. The main ﬁnding obtained in this chapter suggests that the government’s ﬁscal sustainability at this stage does not appear to be a pressing issue in China.

122

The Chinese Economy after WTO Accession

Nevertheless, the data used in the study cannot reﬂect the government’s substantial off-budget and extra-budget expenditures during the period from 1979 to 2001, though it also leaves out off-budgetary and extra-budgetary revenues.7 As the ﬁscal activity in China extends well beyond the ofﬁcial budget, especially at the level of local governments, the implicit and contingent as well as explicit public debt in the future may present considerable exposure of ﬁscal risk to China. The assessment of implicit and contingent liabilities to China’s government represents a direction of future studies of ﬁscal sustainability. In this regard, IMF (2002) argues that a strong economic growth alone might even not be sufﬁcient to meet the challenge of ﬁscal sustainability when contingent liabilities are taken into account. Under these circumstances, satisfying the government’s intertemporal budget constraint in the present value terms does call for a series of structural reform, such as the reforms in state-owned enterprises and in the pension system, as well as prudent ﬁscal measures in the medium term. In this sense, the ﬁndings in this chapte should not justify complacency without qualiﬁcations. References Bahl, Roy and Sally Wallace (1995), ‘Intergovernmental Fiscal Relations in China,’ in the Proceedings of the Eighty-Eighth Annual Conference on Taxation held under the auspices of the National Tax Association at San Diego, California, October 8–10, 1995, Columbus: National Tax Association-Tax Institute of America, pp. 110–14. Barro, Robert J. (1989), ‘The Ricardian Approach to Budget Deﬁcits,’ Journal of Economic Perspectives, Vol. 3, pp. 37–54. Bravo, Ana B. S. and António L. Silvestre (2002), ‘Intertemporal Sustainability of Fiscal Policies: Some Tests for European Countries,’ European Journal of Political Economy, Vol. 18, pp. 517–28. Brean, Donald (1998), ‘Financial Perspectives on Fiscal Reform,’ in Trish Fulton, Jinyan Li, and Dianqing Xu (eds), China’s Tax Reform Options, World Scientiﬁc, Singapore and New Jersey, pp. 47–56. Eckaus, R. S. (2003), ‘Some Consequences of Fiscal Reliance on Extrabudgetary Revenues in China,’ China Economic Review, Vol. 14, pp. 72–88. EViews 4.0 User’s Guide (2000), Quantitative Micro Software, LLC. Gujarati, Damodar N. (1995), Basic Econometrics, McGraw-Hill, New York. Hamilton, James D. and Marjorie A. Flavin (1986), ‘On the Limitations of Government Borrowing: A Framework for Empirical Testing,’ American Economic Review, Vol. 76, pp. 809–19. International Monetary Fund (IMF) (2002), World Economic Outlook (April), International Monetary Fund, Washington D.C. 7 Whereas the existing literature has widely indicated the existence of off-budgetary expenditures and contingent liabilities in China, relatively few studies discuss the extrabudgetary revenue in China; for the latter, see Eckaus (2003).

On the Intertemporal Sustainability of Fiscal Debt

123

Johansen, Soren and Katerina Juselius (1990), ‘Maximum Likelihood Estimation and Inference on Cointegration with Application to the Demand for Money,’ Oxford Bulletin of Economics and Statistics, Vol. 52, pp. 169–209. Kremers, Jeroen (1989), ‘U.S. Federal Indebtedness and the Conduct of Fiscal Policy,’ Journal of Monetary Economics, Vol. 23, pp. 219–38. Krumm, Kathie L. and Christine P. Wong (2002), ‘Analyzing Government Fiscal Risk Exposure in China,’ in Hana Polackova Brixi and Allen Schick (eds), Government at Risk: Contingent Liabilities and Fiscal Risk, World Bank, Washington D.C., pp. 235–49. Lardy, Nicholas R. (1998), China’s Unﬁnished Economic Revolution, Brookings Institution Press, Washington D.C. Lin, Shuanglin (2000), ‘The Declines of China’s Budgetary Revenue: Reasons and Consequences,’ Contemporary Economic Policy, Vol. 18, pp. 477–90. Lin, Shuanglin (2003), ‘China’s Government Debt: How Serious?’ China: An International Journal, Vol. 1, pp. 73–98. McCallum, Bennett T. (1984), ‘Are Bond-Financed Deﬁcits Inﬂationary? A Ricardian Analysis,’ Journal of Political Economy, Vol. 92, pp. 123–35. Persson, Torsten and Lars E.O. Svensson (1989), ‘Why a Stubborn Conservative Would Run a Deﬁcit: Policy with Time-Inconsistent Preferences,’ Quarterly Journal of Economics, Vol. 104, pp. 325–45. Rogoff, Kenneth (1990), ‘Equilibrium Political Budget Cycles,’ American Economic Review, Vol. 80, pp. 21–36. Tanzi, Vito (1994), ‘The Political Economy of Fiscal Deﬁcit Reduction,’ in William Easterly et al. (eds), Public Sector Deﬁcits and Macroeconomic Performance, Oxford University Press for the World Bank, Oxford and New York. Trehan, Bharat and Carl E. Walsh (1988), ‘Common Trends, the Government Budget Constraint, and Revenue Smoothing,’ Journal of Economic Dynamics and Control, Vol. 12, pp. 425–44.

The Chinese Economy after WTO Accession

124

Appendix Differentiating (4) with respect to the inﬂation rate, π, results in: ⎧⎪ ⎪

⎡

⎤ ⎫⎪ ⎪

⎢ ⎥ ⎪ t−1 ⎪ ⎪⎪ ⎡ (1 + i )(1 + η ) ⎤ ⎢ ⎥ ⎪⎪ ∂bt ⎛⎜ η − i ⎞⎟⎛⎜ 1 + i ⎞⎟ z ⎥ ⎢ ⎥ t⎬ + z ⎟⎟⎜ ⎟⎟ ⎨− ⎢ = ⎜⎜ z + ⎢b0 − ⎥ ⎢ ⎥ ⎜ 2 2 ⎟ ⎛ ⎞ ⎪⎪ ⎢ (η − i ) ∂π ⎜⎝ (1 + η ) ⎟⎠⎝1 + η ⎟⎠ η i − ⎥⎦ ⎟⎟ ⎥ ⎪⎪⎪ η − i ⎜⎜ ⎢ ⎪⎪ ⎣ ⎢ ⎜⎝1 + η ⎟⎠ ⎥ ⎪⎪ ⎪⎪⎩ ⎣ ⎦ ⎪⎭

(A1)

Under the assumptions that the growth rate is higher than the interest rate and that the initial debt ratio falls short of its intertemporal equilibrium level, (A1) is negative; if the growth rate is higher than the interest rate but the initial debt ratio exceeds its intertemporal equilibrium level, (A1) will become positive as t gets sufﬁciently large. Next, differentiating (4) with respect to the real growth rate, η’, produces:

∂bt ∂· '

⎡ ⎤ ⎢ ⎥ t ⎢ z ⎥ ⎛⎜ 1 + i ⎞⎟ ⎢ ⎥ = − b0 − ⎟ ⎢ ⎛ η − i ⎞⎟⎥ ⎜⎜⎝1 + η ⎠⎟ ⎜⎜ ⎢ ⎥ ⎟ ⎜⎝1 + η ⎟⎠⎥ ⎢⎣ ⎦

⎛ 1 + i ⎞⎟ ⎡ ⎛ 1 + i ⎞t ⎤ ⎜⎜ ⎟⎟ ⎥ ⎟⎟ ⎢⎢1 − ⎜⎜ − ⎥ (1 + η ) ⎛ η − i ⎞2 ⎜⎝ (1 + η ) 2 ⎟⎠ ⎢ ⎜⎝1 + η ⎟⎠ ⎥ ⎣ ⎦ ⎟⎟ ⎜⎜ ⎜⎝1 + η ⎟⎠ t

z

(A2)

The ﬁrst term on the RHS is negative as long as the initial debt ratio is greater than its intertemporal equilibrium level, and the second term above is also negative if the growth rate is greater than the interest rate. Even if the initial debt ratio is less than its intertemporal equilibrium level, the second term on the RHS will dominate the ﬁrst as t gets sufﬁciently large. Finally, differentiating (4) with respect to the primary budget deﬁcit, z, yields: t ⎛1 + η ⎞⎟ ⎡⎢ ⎛ 1 + i ⎞⎟ ⎤⎥ ⎜ ⎜ ⎟ ⎟ =⎜ ⎢1− ⎜ ⎥, ∂z ⎜⎝ η − i ⎟⎟⎠ ⎢ ⎜⎝1 + η ⎟⎟⎠ ⎥ ⎣ ⎦

∂bt

which is positive if the growth rate exceeds the interest rate.

(A3)

Chapter 8

Sequencing Domestic Financial Reform: Country Experiences and China’s Roadmap Yanqing Yang

Introduction As a result of continuous economic reform starting from 1978, China presents a great pro forma economic success. Among a number of economic performance explanatory indicators,1 a high savings rate,2 large scale of public expenditures (especially in recent years), abundant foreign direct investment (FDI),3 and buoyant exports, a relatively climbing total factor productivity (TFP)4 contributes to the average annual GDP growth rate, which is over 8 percent during the last 20 years. Arguably, this pro forma economic growth has been combined with mounting structural problems that emerged and hid rotationally during different phases of economic cycles. The malaise is attributable to the incremental/gradualist reform 1 The ofﬁcial data released by the NBS (National Statistical Bureau) is alleged to be overestimated: According to Thomas Rawski (2001), by using income-side data from the 1999 edition of China’s annual statistical yearbooks, nominal growth rate of 1997–98 is 4.6 percent, after adjusting for price deﬂation, real growth of that is 5.7 percent, while the ofﬁcial data from NBS is 7.8 percent. Woo (1998) estimated that for the period of 1979–93, and 1985–93, industrial output has been overestimated in a range of 0.5-0.7 percent and 0.91.2 percent, respectively. Some cross validation methods used to verify the measurement of GDP appear to be more or less consistent with NBS data (Liu, 2002). Other researchers warn that one should be cautious when analyzing quantitative trend by using data published by NBS, e.g., Keidel (2001a) pointed out important issues: the changing seasonality of some of China’s monthly and quarterly data; the sensitivity of household savings to real deposit rate; the importance of China’s treatment of housing services in housing statistical and national income accounts. 2 It is close to 40 percent, which enhanced a more than 35 percent capital formation in the 1990s. 3 FDI played an important role in Gross Fixed Capital Formation, which accounts for over 10 percent in the last half of 1990s. 4 According to Woo (1998), estimated net TFP growth ranges from 1.1 to 1.3 percent from 1979–93.

126

The Chinese Economy after WTO Accession

approach China has undertaken. The beneﬁt of this approach is the avoidance of economic turmoil and resultant social unrest; however, the costs of it are the delay of the seemingly most difﬁcult reforms, the inconsistency of reform paces in different sectors, and the incompatibility of different aspects of the reform in a sector. Notably, when the structural problems are intertwined and need to be addressed in parallel, ﬁnding a comprehensive solution can be difﬁcult. China’s lagging behind ﬁnancial sector reform illustrated the cost of such an incremental approach. Burdened with the large scale of Non Performing Loans (NPLs), the banking sector is the ‘weakest link’ in the economy, and at the same time, this technically insolvent sector plays the role of a fulcrum point upon which other crucial reforms hinge, namely, SOE reform, monetary reform, foreign exchange (forex) reform, capital account liberalization, and so on. Entering into the WTO imposes a daunting pressure on China’s ﬁnancial reform agenda. The commitment and the timetable are there, the best scenario is that the ﬁnancial system will be ready for the full-ﬂedged competition with international rivals at the beginning of 2007; the worst scenario is that we will still not have solved the deep-rooted pitfall in China’s ﬁnancial system, e.g., large scale NPLs, low capital adequacy ratio, banking sector incompetence, insufﬁcient prudence supervision, poor corporate governance, ineffective macro monetary and forex policies. The likely future is that we will arrive at a point between these two scenarios. Indeed, a series of reforms has been initiated in the last few years, aiming to restructure China’s ﬁnancial system and reenergize its lax banking sector. However, when looking at competition between domestic banks and their foreign rivals under the new WTO regime, China’s banking sector remains incredibly vulnerable and this must be tackled in the near future. Meanwhile, resulting from China’s integration into the world economy and ongoing ﬁnancial liberalization, current capital controls are becoming inefﬁcient. This will absolutely accelerate China’s pace of capital account convertibility. Nevertheless, an open ﬁnancial environment will necessitate a tied-in macro and forex policy combination, in which a ﬂoating exchange rate regime may enhance the effectiveness of monetary policy.5 All of these aspects are interlaced, and thus need to be taken into account and addressed in an across-the-board fashion. This chapter argues that sequencing matters. The point where we will arrive between the best and the worst scenarios is a function of ﬁnancial system reform and particularly a function of reform sequencing. And notably, a poorly managed reform may give rise to inefﬁciency and risk a ﬁnancial collapse or crisis. In a comprehensive ﬁnancial system reform plan, domestic and international ﬁnancial reform6 should be tackled together, however, in order to make 5 According to Mundell (1968), monetary policy is ineffective in a small economy with a ﬁxed exchange rate, open capital markets, and perfect substitutability between domestic and foreign assets. 6 Domestic ﬁnancial reform refers to interest rate liberalization and entry opening, while international ﬁnancial reform refers to capital account liberalization and a more ﬂexible foreign exchange regime.

Sequencing Domestic Financial Reform

127

discussions narrow and centered, this chapter focuses only on one issue: domestic ﬁnancial reform. Combined with comparative analysis of country experiences, a roadmap for sequencing China’s domestic ﬁnancial reform is articulated brieﬂy as follows: Step 1 Dealing with the NPLs problem, bank restructuring, together with institutional building. Step 2 Interest rate deregulation and entry opening. The next section will focus on NPLs resolution and bank restructuring, after illustrating several avenues to solve China’s NPL problem and build banking capital adequacy, this section suggest several viable options. Comments on ongoing reforms also will be found in this section. The third section will elaborate how to build up a framework of institutions for a well-functioned ﬁnance sector, such as internal governess, prudential supervision, proﬁtability, legal and accounting codes, etc. The fourth section will discuss different approaches in domestic ﬁnancial reform in light of country experiences, and generally sequence China’s reform plan. The ﬁfth section is the conclusion. The gradual approach in China’s ﬁnancial reform process has its own special characteristics, inter alia, one is the incompatibility of different dimensions in the reform process. Financial sector development / reform can be conceived as a balanced development of the triangle of organizations, instruments, and the market. Application of this triangle analysis to China’s ﬁnancial system indicates that most attention has been given to the organization-building, as a result, a multi-tiered ﬁnancial structure has been in place; the process of market development – including ﬁnancial liberalization – has received the least attention; while the development of new ﬁnancial instruments has been largely conﬁned to the capital markets (Mehran et al., 1996). Step 1 (A): NPLs and Bank Restructuring To an extent, the lag in market building and the slow pace in state-owned enterprise (SOE) restructuring and privatization explains the large scale of NPLs already sitting on the State-owned Commercial Banks’ (SOCBs) balance sheet and the dilemmas China’s banking sector has been facing for years. On the one hand, SOEs, which show little sign of proﬁt-making prospects, still make up the bulk of SOCBs’ borrowers;7 on the other, facing the growing but volatile private sector, SOCBs have neither the expertise nor the experience to conduct customer research and risk analysis. It is of critical importance for China to have a well–functioning and effective ﬁnancial intermediary system, in which the ‘market’ plays a central role, i.e., a marketbased interest rate is in place. This would help to achieve long-term economic growth as 7 Recent statistics indicate that loans to the SOEs account for almost 70 percent of new loan commitments and a higher percentage of their outstanding loan portfolio (Bottelier, 2002b).

128

The Chinese Economy after WTO Accession

well as maintain economic and social stability.8 However, Country experiences highlight the importance of addressing problems in ﬁnancial system soundness, particularly in the banking sector, early in the reform process. To undertake deregulation reforms before addressing the banking sector’s weak institutional structure is risky, and this weakens the effectiveness of reforms in improving resource allocation, and risks a banking crisis that could seriously disrupt stability and economic growth (Johnston and Sundararajan, 1999). In China’s case, as access deregulation is under way in accordance with the WTO commitment,9 the priority at this moment is to address the large scale of NPLs already sitting on the SOCB’s balance sheet and the institutional deﬁciency, in order for the whole sector to have a better footing to face ﬁerce competition with foreign rivals in the coming years. Therefore, before embarking on the arduous ﬁnancial liberalization process to head for a ‘perfect ﬁnancial market’, the ﬁrst sequencing step brought forward in this chapter, is to remove the vast NPLs stock from the balance sheet of the banking sector, together with a pari passu institutional building to avoid the reemergence of new NPLs. Measuring NPLs in China’s Banking Sector The de facto magnitude of NPLs in China’s banking sector proves to be hard to measure. Before ofﬁcial data released in 2000, the proportion of NPLs estimated by several scholars and organizations ranged from 24 percent to 29 percent. In 1999 and 2000, US$169bn (RMB 1.4 tn) worth of NPLs were carved out and transferred to the four asset management corporations (AMCs) at par.10 As at the end of 2003, it is ofﬁcially reported that the four AMCs have altogether disposed of US$61.5 bn (RMB509.37 bn) worth of non-performing assets, and recovered an aggregate of US$12 bn (RMB99.41 bn) in cash, registering a cash recovery ratio of 19.52 percent. Ofﬁcial data, which are based on the recently implemented international 5-category loan classiﬁcation system, showed that the average NPL/ Loan portfolio ratio of the four SOCBs has been edging down for the past years. According to the latest statistics of the China Banking Regulatory Commission (CBRC), the outstanding balance and ratio of the NPLs of the SOCBs were reduced 8 Empirical research conﬁrmed the thinking of many development economists that ﬁnancial deepening is an important characteristic of the growth process. Capirio, Atiyas, and Hanson (1994) indicated that a 1 percent increase in real per capita income is typically associated with approximately a 1.5 percent increase in the various ‘ﬁnancial deepening’ measures. 9 In the banking sector, China has committed to full market access in ﬁve years, both geographic and customer restrictions would be removed within ﬁve years after accession. Foreign banks will be able to conduct local currency business with Chinese enterprises starting two years after accession and local currency business with Chinese individuals by 2006, and foreign banks will have national treatment within designated geographic areas. 10 The four AMCs are Dongfang (BOC), Xinda (CBC), Huarong (ICBC) and Great Wall (ABC), respectively.

Sequencing Domestic Financial Reform

Table 8.1 AMC Orient Great Wall Cinda Huarong Total

129

NPL Disposition and Asset Recovery (December 2001, RMB bn) Face value of NPLs 267.4 345.8 373.0 407.7 1,393.9

Face value of Cash Recovery Annual Interest Cash Recovery deposed assets payment rate (%) 18.3 4.4 5.8 24.2 53.1 3.7 7.6 6.9 29.9 10.5 8.2 35.1 23.2 7.6 8.9 32.5 124.5 26.2 30.5 21.0

Source: Ma and Fung, 2002.

respectively by US$20 bn (RMB171.3 bn) and 5.85 percentage points from the beginning of the year to US$232bn (RMB1916.8 bn) and 20.36 percent at the year end. Early this year, the State Council ofﬁcially announced that the government used US$45bn from foreign currency reserves to recapitalize China Construction Bank (CCB) and Bank of China (BOC) at the end of last year. Some researchers and organizations hold the opinion that the ofﬁcial data underestimated China’s NPLs problem. Bonin and Huang (2002) estimate an average 35 per cent NPLs in SOCBs, considering the huge unrecognized bad assets. Additionally, SOCBs are by no means the only headstream that generates NPLs in China’s ﬁnancial system, many other public ﬁnancial institutions also face portfolio quality problems. According to CBRC, the outstanding balance and the ratio of the NPLs of the major Chinese banking institutions11 were reduced respectively by US$23 bn (RMB190.6 bn) and 5.32 percentage points from the beginning of the year to US$295 bn (RMB2.44 trillion) and 17.8 percent. However, this ﬁgure doesn’t take into account over 40000 Rural Credit Cooperatives (RCCs), which account for about 10 percent of total banking system loans, it is widely believed that RCCs also face severe NPL problems (Bottelier, 2002b). Some estimates put the NPL level within the Chinese system, both carved out and remaining, at around 40 percent of the total loans outstanding (Lardy, 1998, Standard and Poor, 2001). This ﬁgure may have overestimated China’s banking risks, as the aggregate loan portfolio of all banks in China is much larger than GDP.12 Another estimation, which believed that the overall NPL ratio in China’s banking sector could reach 25 percent of the aggregate banking sector loan portfolio (Bottelier, 2002b), is more likely to near the real magnitude of NPLs. Based on all these disclosures and estimations, we calculate the system-wide NPL ratio in this way, as major Chinese banking institutions accounted for 80.74 percent of the total loans made by ﬁnancial institutions in China, given that weighted NPL ratio in other ﬁnancial institutions except major chinese banking institutions is half of that 11 Major Chinese banking institutions includes four wholly State-owned commercial banks, three policy banks, eleven joint-stock commercial banks. 12 In 2002, the aggregate loan portfolio amount to RMB13.9803 trillion, while GDP is RMB10.2398 trillion.

The Chinese Economy after WTO Accession

130

Table 8.2

China’s Central Budgetary Revenue/Expenditure (% of GDP)

Year 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Source: CEIC.

Total Revenue

Total Expenditure

Balance

25.7 24.2 22.9 22.9 22.8 22.3 20.8 18.4 15.8 15.8 15.8 14.5 13.1 12.6 11.2 10.9 11.1 11.8 12.8 14.2 15.2

27.2 23.4 23.2 23.7 23.6 22.3 21.6 18.9 16.7 16.7 16.6 15.6 14.0 13.4 12.4 11.9 11.9 12.6 14.0 16.4 18.0

-1.5 0.8 -0.3 -0.7 -0.8 0.0 -0.8 -0.5 -0.9 -0.9 -0.8 -1.1 -1.0 -0.8 -1.2 -1.0 -0.8 -0.8 -1.2 -2.2 -2.8

in SOCBs, according to ofﬁcial data, the system-wide NPL ratio in loan portfolio is around 17 percent, which amounts to US$ 334 bn (RMB2767 bn). Taking into account the US$169 bn (RMB1.4 trillion) already transferred to the four AMCs and US$45bn injected from foreign currency reserve, given that the blended recovery rate of the NPLs is 15 percent,13 system-wide loan losses could reach 37 percent of 2001 GDP, which amounts to US$433bn (RMB3585 bn). Measuring China’s Fiscal Sustainability As a result of central-local decentralization reform in 1980s, China’s central ﬁscal capacity, which sustained a downward trend in total revenue as a percentage of GDP, had been a big concern for both decision-makers and researchers in China. The tax reform aiming to halt and reverse the trend was solicited in 1994. Nevertheless, the 13 The recovery rate achieved so far is believed to have averaged well below 10 percent, while AMCs expect that the recovery rate is 20 percent (see Table 8.1), 20 percent is also the cap of the price range paid by foreign investors for the acquisition of impaired assets form Huarong Asset management (CLSA, 2002), the recovery rate used in this chapter is an average between this two.

Sequencing Domestic Financial Reform

Table 8.3

131

Contingent Liabilities of China’s Government

Accumulated Public Debts Special T-bonds in 1998 for Recapitalization Estimated Costs for Bank Restructuring Estimated Costs for Social Security Funds Municipal Government Contingent Debt External Debts Total Source: Huang, 2003 and author’s estimate.

RMB bn 1550 270 3585 2500 700 1500 10820

% of 2001 GDP 16.0 2.8 37.0 25.8 7.2 15.5 104.3

central government’s budgetary revenue as a percentage of GDP didn’t start to rise until 1996 (see Table 8.2). China’s ﬁscal capacity takes a crucial part in its transition and reform. Firstly, all ‘transition debt’ (Bottelier, 2002b) aroused in the reform process is supposed to be paid by ﬁscal revenue as a last resort; Secondly, the central ﬁscal capacity backs up the implicit deposit guarantee that is regarded as a vital condition preventing China’s ﬁnancial system from meltdown. Now the question is, is China’s ﬁscal capacity sustainable in light of all transition debts? China’s formal domestic sovereign debts accounted for only 16 percent of GDP at the end of 2001. When taking into account the estimated costs for restructuring the banking system and the potential expenditure to ﬁnance the pension fund, 37 and 25.8 percent of GDP,14 respectively, plus the external debt and municipal government contingent debt, the total liabilities come to over 100 percent of 2001 GDP (see Table 8.3). Indeed, this ratio is quite high according to any international standards, but still does not pose an imminent threat to China’s ﬁscal sustainability in the foreseeable future. First, municipal government contingent debt does not necessarily translate into central government debt, unless municipal government ﬁscal crises take place, which is very unlikely concerning the current de jure compulsory ﬁscal balance requirement in local government. Second, though the estimated cost of pension funds is very high, it is not, however, a sunk cost (Bottelier, 2002b) and therefore can be paid off over a very long period rather than being ﬁnanced by a one-off payment in the near future. Furthermore, if the capital market rallies, being an institutional investor, the pension fund’s ﬁnancial health will improve. Third, according to standard ﬁscal sustainability conditions,15 if growth rate is higher than the real interest rate, the ﬁscal sustainability can be achieved, which means if China GDP grows at a 8 percent, issuing bonds for the purpose of NPL resolution at a real interest rate of lower than 8 percent, say 7 percent, is ﬁscally sustainable. 14 This is based on the Citi Group’s estimate. 15 The ﬁscal sustainability condition is Δd(t)=p(t) + (r--g) / (1+g) d (t-1),where d(t) is the debt ratio as a share of GDP at time t, p(t) is the primary deﬁcit, r is real interest rate, and g is real growth rate. If primary deﬁcit is zero, and debt is declining thus sustainable only if r< g.

The Chinese Economy after WTO Accession

132

Table 8.4

China’s Liabilities versus Assets

Liabilities(RMB bn) AMC bond

Assets(RMB bn) 800 Market value of government-owned shares in listed 3600 SOEs Policy Bank bond 600 Market value of SOEs to be listed in next 5 years 2300 NPL (remaining) 2800 Other marketable asstes 400 Debt to Central bank 600 Cash value of NPL/AMC sales 300 SOE recapitalization 1200 Local government debt 800 Estimated present value 3800 Estimate present value 5400 Some estimates here are inconsistent with data in this chapter. Source: Bottelier, 2002b.

Fourth, although China’s government has a very high contingent liability, there are also vast amounts of assets at the same time that can be exploited to pay the transition debt (Bottelier, 2002b).16 Finally, the Chinese economy has been growing at remarkable rates, and ﬁscal revenues have been increasing at 15 to 20 percent per annum while ﬁscal deﬁcits have been kept below 3 percent of GDP in recent years. The share of ﬁscal revenue in GDP is still below 20 percent and has the potential to rise further. To sum up, making use of innovative measures, such as ABS (Assetbacked securitization) et cetera, it would not be difﬁcult for China to absorb the existing contingent liabilities over an extended period. Growing Out or Radical Approach? Studies show that as governments and central banks have dealt with crises and restructured their banking systems, the cost borne by the countries has varied. There is no single solution. What needs to be done depends very much on circumstances. But some ingredients of successful programs can be discerned, such as privatization, transferring bad assets to a special resolution vehicle, and so on. China’s present strategy on resolving NPLs appears to have broadly resembled the Swedish model of separate and decentralized NPL management. The Ministry of Finance (MOF) provided each AMC with an initial equity capital injection of US$1.2 bn (RMB10 bn). Each of the four AMCs paired up with one of the ‘big four’ in China and has been mandated to maximize asset recovery over 10 years. In 1999–2000, the ‘Big four’ transferred their NPLs to their respective linked but independent AMC (see Table 8.5). Indeed, the more worrying problem is the NPLs stock still on the banks’ balance sheets ex-AMCs and the ﬂow of new NPLs. China’s government once took a growingout approach that there would be no more transfers of NPLs to the AMCs nor any capital injections, and that the banks should aim to reduce the NPL ratios themselves.

16 See Table 8.4: China’s Liabilities versus Assets.

Sequencing Domestic Financial Reform

Table 8.5

133

China’s AMCs: Policy-based NPL Transfers (1999–2000)

AMC

Matched bank Assets transferred Share of bank loan (RMB bn) outstanding (% at end-1998) Orient Asset Management BOC 267.4 20.4% Great Wall Asset ABC 345.8 24.6% Management Cinda Asset Management CCB 373.0 21.7% Huarong Asset ICBC 407.7 17.9% Management Total 1,393.9 20.7% Note: In calculating the CCB loan shares, the table takes into account that RMB100 billion of the assets transferred to Cinda are from China Development Bank. Source: Ma and Fung, 2002.

A very ambitious target was set up in 2002 for the SOCBs to cut NPL ratios by 2 to 3ppts every year for the next ﬁve years. Allowing banks to work their own way out of the serious problem of NPLs is politically attractive,17 however, this approach has dangers. In China’s case, with the growing out approach, banks will always have the excuse of being burdened with NPLs (Dornbusch and Givazzi, 1999). More tellingly, facing the harsh target of reducing NPLs, managers in banks tend to have two contradictory reactions: on the one hand, they may become very conservative and tend to contract loans, only lending to big infrastructure projects that have a very low risk and hold a large sum of government bonds. This gives a negative impact on banking proﬁtability and economic growth. On the other hand, because the total of outstanding loans increases, the ratio of bad loans decreases assuming that new NPLs do not emerge in the short-run. Managers may tend to increase lending, which risks ﬁnancial overheating. Another issue is low Capital Adequacy Ratio (CAR) in China’s banking sector. In 1998, the government injected RMB270 billion into the four SOCBs, through issuance of special Treasury bonds, to raise their average CAR from 4.6 per cent to above 8 per cent. By the end of 2001, however, the average CAR had come down again to around 5 percent (see Table 8.6). According to CLSA (2002), since China’s CAR criteria differ from that of BIS, reported CARs in SOCBs are overstated (see Box 8.1). There are renewed calls for a large-scale up-front recapitalization for the Chinese banks. The experience of TSE illustrated that the key pitfall of recapitalization is the failure to recapitalize adequately. Partial recapitalization, as in the Hungarian case, could undermine the incentive for a change of bank behavior by increasing the possibility of future recapitalization (Lardy, 1998). However, as stated by some researchers, resolution of institutional problems is a necessary condition for any massive capital injection, because a full-scale recapitalization program will work only if the governance problem has been 17 Actions to restructuring bank portfolios tend to get delayed in most countries, because of high ﬁscal cost. And often, regulators lack incentives for prompt action.

The Chinese Economy after WTO Accession

134

Table 8.6

Percent of Reported CAR in SOCBs (1998–2004)

Bank 1998 1999 ICBC 5.68 4.57 ABC 6.67 1.44 BOC 18.76 8.50 CBC 5.22 3.79 Average 7.94 na Source: CLSA, 2002 and author.

2000 5.38 na na 6.51 5.7

2001 5.76 na 8.30 6.88 na

2003 na na 8.15 6.91 na

2004 (after injection) Over 8 Over 8 -

solved. Otherwise, the government will never be able to stop injecting money, as is clearly shown by the experiences of many other countries (Huang, 2002b). Early in 2003, the State Council ofﬁcially announced that the government used US$45bn from foreign currency reserves to recapitalize CCB and BOC at the end of last year. A new entity, central Huijin Investment Corporation Ltd, which is jointly owned by the Ministry of Finance (MOF), People’s Bank of China (PBOC), and the State Administration of Foreign Exchange (SAFE), was created and received a capital injection from the PBOC to conduct the recapitalization. After the injection, as proﬁt and equity in two banks will be fully utilized to write off NPLs, NPL ratio in two banks will be lowered to around 4 percent; newly injected US$22.5bn becomes net asset in CCB and BOC respectively, and the CAR will maintain a level of above 8 percent. In the meantime, the restructuring of two banks aiming for listing has also been initiated. This recapitalization model received positive responses once released,

Box 8.1 Two-tiered Capital Structure for CAR Calculation in China Tier 1: Paid-in capital • Capital reserve • Surplus reserve Tier 2: Provision for doubtful loan • Provision for bad debt • Provision for investment risk • long-term bond with maturity of ﬁve years and above Tier 3: None Deductions: Capital investment in other bank • Capital investment in other non-banking ﬁnancial institutions • Capital investment in industrial and commercial enterprise • Investment in real estate asset not used by the bank • Bad-asset that has not been deducted Total eligible capital for CAR Tier 1+ Tier 2–Deductions Source: CLSA, 2002.

Sequencing Domestic Financial Reform

135

the advantages of this model include: the bypass of the complicated authorization process for ﬁscal injection; the prevention of pressure being put on base money, and the avoidance of the possibility of the MOF having an insufﬁcient budget. Although a case-by-case approach is needed to solve the NPL problem in other SOCBs and joint-stock banks, generally, this chapter argues that China still needs a radical approach to ﬁx its NPL problems and fully recapitalize its banking sector, because the cost of a growing out strategy is much higher than its beneﬁts. New investment and depositors have to pay a ‘tax’ to cover existing losses, which may reduce growth (Capirio, Atiyas and Hanson, 1994). In an interest rate–liberalized market, it is even more dangerous because banks may well bid up deposit rates and invest in overly risky assets.18 Moreover, banks may conceal their problems by rolling over bad loans, leaving a time bomb that can go off in the future. Many of these problems have been illustrated by Chile and the US, which suggest that given the dangers, strong, early intervention is probably the least risky – and lowest cost – solution to the NPL problem. Furthermore, the presence and persistence of large weaknesses in bank balance sheets has also affected the effective and uniform enforcement of prudential norms, further delaying the process of strengthening banking supervision (Johnston and Sundararajan, 1999). Notably, restructuring should proceed hand in hand with institution building, which is of crucial importance for the success of the restructuring. To recapitalize and restructure the banking sector, some options can be utilized, as illustrated below. Capital injection by central governmentCapital injection by central government to resolve the NPL problem in the banking sector in China seems to be unavoidable for a radical resolution approach. There are two reasons: ﬁrstly, NPLs are actually the ‘reform cost’ borne by the banking sector derived from the economic reform, especially the SOE reform; Secondly, even if all NPLs can be transferred to AMCs, after bad asset dealing, about 80–85 percent of NPLs will make up part of the losses in the ﬁnancial system. Only PBOC and MOF can play the role of such a capital injector. However, as the main asset of PBOC is loans extended to policy banks and other ﬁnancial institutions instead of government bonds, its ﬁnancial capacity is very weak. Although the foreign currency reserve is quite ample, it couldn’t be utilized as a universal model to solve banking problems, in the case of ICBC and ABC resolution, foreign currency reserve injection may be viable, but it’s impossible for any further expropriation of foreign currency reserve to solve the system-wide NPL problems, as foreign reserve injection may have a negative impact on central bank’s independence and on China’s sovereign rating.

18 It is hinted that it is very risky to step up efforts to liberalize interest rates and entry into the banking sector at this stage. In Turkey, the freeing of rates in the early 1980 led to bidding up of real interest rates and some ﬁnancial distress. Eventually, deposit rates were recontrolled and entry restricted.

136

The Chinese Economy after WTO Accession

Let’s come back to MOF, a bond-bad-debt swap is an option. The simple way to do this is to follow the course pursued by several transitional economies in Eastern Europe. MOF injected special government bonds in an amount equal to the value of the NPLs being written off. The so-called ‘bond-bad-debt swap’ differs from the injection of paper assets, the bond would be a real asset, so an injection of bonds would provide banks with income earning assets. Over a period of time, the bonds could be sold on the secondary market, making the value of injection transparent (Lardy, 1998). To fund capital to recapitalize the banking sector, another possible solution is for the PBOC simply to print more currency to cover the bad loans. As Chow (2002) calculated, if the income elasticity of demand for real money balance is 1.2, that means that when real output or real income in China increased by 10 percent, the demand for real money balance will increase by 12 percent. The government can increase money supply by 12 percent without causing inﬂation. The quantity of currency in circulation in China at the end of 2002 is RMB1727.8bn, the income elasticity of demand for money is about 1.16 (Chow, 2002). This means without inﬂation, China’s money stock could be increased by an exponential rate of 1.16 times the exponential rate of increase in GDP. The latter is about 0.08. The exponential rate of increase of 0.093 (1.16 multiply 0.08) from RMB1728bn allows an increase in money supply of RMB160bn. If 2 percent of inﬂation is allowed, the exponential rate of increase in money stock can be about 0.115(1.02* 1.093-1), yielding a money supply of RMB200bn. Comparing with RMB3500bn NPLs, in the next 10 years, if newly-printed money is used to solve the NPL problem, about 57 percent of the NPLs sitting on the banking sector will be eliminated. However, this approach risks inﬂation and could be harmful for the economy in the long run, thus it needs to be considered with great caution One alternative to an injection of fresh capital by the state could be through the sale of equity shares in SOEs or through land use rights (Lardy, 1998). Another ﬁnancial instrument that could be used is securitization. By issuing ABS (Assetbacked Securitization, backed by equity shares in SOEs and by land use rights), the government will have the capacity to deal with the impaired assets in the banking sector without unduly depressing the current market value of state assets. But this approach requires legislative changes to ensure that ABS are in fact freely tradable and that the holders of such debts have quick and reliable access to underlying assets in case of default (Bottelier, 2002b). Internal restructuring avenuesApart from central government capital injections, internal restructuring in banks is also critical for the banking sector to ﬁx NPLs and CAR problems. One approach for internal restructuring is the issuance of convertible bonds, i.e., to allow individual banks to issue their convertible bonds for the amount of NPLs they need to carve out. The public chooses either to hold their bonds to maturity or to convert them into equity shares. The bondholders can be domestic and foreign, small and institutional investors. This approach actually completes two tasks with one instrument: carving out the NPLs from the banking system as well as de facto diversiﬁcation of ownership (Liu, 2002).

Sequencing Domestic Financial Reform

137

Strategic investors and access to the capital market seem to be indispensable for next step banking reform. Foreign equity investment in domestic ﬁnancial institutions is an effective method to recapitalize the problematic banks with a beneﬁciary by-product of governance amelioration. A gradual sale may be more politically acceptable, and a public ﬂoat may be attractive in terms of broadening share ownership but may leave effective control of the bank in the hands of the existing management. The ceiling of foreign shares in domestic banks had been 15 percent, with the foreground that the ceiling will be enlarged to 25 percent (CLSA, 2002); foreign strategic investors will play an instructive role in restructuring China’s banking sector. Acquisition by the domestic private sector is an alternative, but will face higher risks. A particularly thorny question that often arises in selling off SOBs is the treatment of loans that could well turn bad at some future date. Potential buyers usually request some form of guarantee. Brazil and Korea have both used mechanisms that allow buyers to sell back assets found to be bad during the ﬁrst months of ownership. Allowing banks to tap domestic and international equity markets to raise the capital to write off their bad debts is also possible, but the banks have to solve the governance demerit and ensure that risk management skills are in place (BIS, 1999). Domestic banks could also look to strengthen their capital bases by issuing long-term subordinated debt. ICBC is reported to submit PBOC an application for issuance of subordinated debt, and more banks are likely to take the same action. Tier-219 capital has seldom been used in China and there is therefore ample room for growth in tier-2 issuance (CLSA, 2002). In practice, the ﬁnal solution could include either a combination of these recapitalization techniques or all of them. This chapter suggests a combination of bond-bad-debt swap, foreign strategic investors/ IPO and ABS. The ﬁrst two techniques are comparably easy and viable to carry out with lower risk of moral hazard. The latter one, by introducing a useful ﬁnancial instrument, is very likely to enable China to have a mature ﬁnancial market in the long run, notwithstanding that the application is quite complicated. Step 1 (B): Institution Building, 2003–2004 Though many developing countries have undertaken various restructuring measures to deal with domestic ﬁnancial weakness, few have proved to be successful.20 After observing the performance of 64 recapitalization schemes, Klingebiel and Caprio (1996) found that the better outcomes have been when restructuring is accompanied by successful macro reform, performance monitoring by outside auditors, tougher 19 See Box 8.1 Two-Tiered Capital Structure for CAR Calculation in China. 20 Klingebiel and Caprio (1996) judged the performance of 64 recapitalization schemes on whether they led to ﬁnancial deepening, moderate growth in real credit, moderate positive real interest rates and no subsequent banking crisis. On this basis, they conclude there have been few clear successes. Chile (1981–83) and Malaysia (1985–88) and to a lesser extent Philippines (1981–87) and Thailand (1983–87) were the best among the emerging economies.

138

The Chinese Economy after WTO Accession

(and enforced) accounting and capital standards, halted lending to defaulting borrowers, and the replacement of senior managers. Parallel with this rationale, this chapter argues that the lack of institution building that goes hand in hand with bank restructuring is one key factor leading to the failure of bank restructuring. It is a must that banking recapitalization be facilitated with institution building in the ﬁnancial system and in the real sector. It has been suggested that the development of a sound banking system and well-functioning prudential supervision arrangement should precede the adoption of market-based monetary arrangements (Villanueva and Mirakhor, 1990; Mathieson and Haas, 1994). And according to Johnston and Sundararajan (1999), the growing globalization of ﬁnancial markets has underscored the importance of strengthening prudential supervision and related information systems to deal effectively with interest rate and exchange rate risks, and other banking risks, particularly in the context of capital account liberalization. Therefore, concerning sequencing of further ﬁnancial reform, the institution building that will be discussed in this section also plays a critical role in later domestic and international ﬁnancial reforms. In practice, some of the institution building could be applied to facilitate NPL resolution, while others may phase in to support interest rate deregulation and other liberalizations. Notably, since ﬁnancial institutions take time to establish. The time framework for building them will be considerably longer. It may proceed from the outset of bank restructuring to the completion of international ﬁnancial liberalization. Internal Governance and Reforms21 Some measures have been taken to strengthen internal governance in China’s banking sector. The blueprint of further reforms is being conceived. To stop the ﬂow of NPLs, recently the international standard ﬁve-category loan classiﬁcation system has been implemented in the whole banking system. According to some estimates (Ma, Qing and Li, 2001), an effective use of credit information system and quantitative credit risk management technologies could reduce new NPLs by 30 to 40 percent. At the same time, the PBOC devised the requirements for loan provisions and demanded all the banks to achieve 8 percent CAR in 2007. The further reform of the SOCBs is conceived for implementation through a three-step approach.22 Breaking up some big banks has also been proposed.23 21 Some of the content of this sub-section may overlap the next sub-section. 22 Step one continues the ongoing reforms improving internal management systems, including the implementation of the accounting, auditing and risk assessment systems. Step two includes both corporatization of the banks and introduction of strategic investors. The government intends to implement the share-holding system to the major banks within 2-3 years, initially with the Ministry of Finance (MOF) as the sole owner. Then the MOF will quickly sell some equities to domestic and foreign strategic investors. Finally, step three aims at the public listing of the large banks in about ﬁve years (Huang, 2002b). 23 According to Huang (2002b), breaking up at least some of the large banks is necessary in order to avoid the ‘too big to fail’ problem, to increase the effectiveness of reforms, to help

Sequencing Domestic Financial Reform

139

Prudential Supervision Concerning ﬁnancial supervision, one of the big concerns is the independence of the central bank in China.24 Despite the progress, including the establishment of four policy banks and a big move at the February policy conference in 2003—the creation of a relatively independent body in charge of ﬁnancial supervision within the PBOC, CBRC, PBOC in China still falls short of considerable autonomy vis á vis other central banks such as Fed in the US, as its power is under the leadership of the SC (State Council). Three laws that regulate PBOC, SOCB and CBRC respectively passed last year and took effect this year. There is still a long way to go. Box 8.2 illustrates other aspects of prudential supervision towards which China should proceed in its ﬁnancial supervision. Building Bank Proﬁtability Tax and regulatory distortions in the ﬁnancial sector worsen allocation and bank proﬁtability (Hanson and Rocha, 1986). In China, the 33 percent proﬁt tax and the 8 percent business tax impose a signiﬁcant burden on banks. In April 2002, the government decided to reduce the business tax on ﬁnancial institutions from 8 percent to 5 percent in the next 3 years, with a cut of 1 percent each year. Though still insufﬁcient, the tax cut will boost the bank’s proﬁt to an extent. Additionally, to improve the banks’ ﬁnancial position to further facilitate NPL workouts, the PBOC drafted ‘management rules for service-based fee charging of commercial banks’.25 Another important move to enhance proﬁtability is to reduce non-interest costs by trimming banks’ staffs and branches. Currently, Chinese banks non-interest cost-toincome ratio is about 30 percent higher than that of their foreign rivals (Ma, Qing and Li, 2001). Assortative Macroeconomic Environment As the ﬁnancial liberalization of the 1980s showed, liberalization in the absence of a consistent and adequate macroeconomic policy framework may be unsuccessful or even counterproductive (Gelb and Honohan, 1989). High inﬂation, the accompanying devaluation and high real interest rates will create uncertainty, a large wealth loss in the economy. Open capital accounts that permit volatile capital ﬂows will exacerbate the situation (Sheng, 1996). It suggests that during bank restructuring, ﬁght the liquidity trap, to facilitate the signing up of strategic investors and public listing, and to smooth the ﬂow of contingent liabilities. Bottelier (2002) also suggested an analogous plan. 24 Lardy (1998) deﬁnes the independence of the central bank: First, it must be able to insulate banks from demands from political leaders to extend loans to projects that do not meet commercial lending standards. Second, it must ensure that any subsidies for policy lending are ﬁnanced through the budget, rather than borne by the banking system. 25 Currently, fees account for less than 10 percent of revenues for most Chinese banks. But in many other countries, this can be as high as 30-40 percent (Huang, 2002b).

140

The Chinese Economy after WTO Accession

Box 8.2 Core Principles of Banking Supervision •

Banking supervisors must be satisﬁed that banks have management information systems that enable management to identify concentrations within the portfolio and supervisors must set prudential limits to restrict bank exposures to single borrowers or groups of related borrowers.

•

In order to prevent abuses arising from connected lending, banking supervisors must have in place requirements that banks lend to related companies and individuals on an arm’s-length basis, that such extensions of credit are effectively monitored, and that other appropriate steps are taken to control or mitigate the risks.

•

Banks must have adequate policies and procedures for identifying, monitoring and controlling country risk and transfer risk in their international lending and investment activities, and for maintaining appropriate reserves against such risks.

•

Banking supervisors must be satisﬁed that banks have in place systems that accurately measure, monitor and adequately control market risks; supervisors should have powers to impose speciﬁc limits and/or a speciﬁc capital charge on market risk exposures, if warranted.

•

Banking supervisors must be satisﬁed that banks have in place a comprehensive risk management process (including appropriate board and senior management oversight) to identify, measure, monitor and control all other material risks and, where appropriate, to hold capital against these risks.

•

Banking supervisors must determine that banks have in place internal controls that are adequate for the nature and scale of their business. These should include clear arrangements for delegating authority and responsibility; separation of the functions that involve committing the bank, paying away its funds, and accounting for its assets and liabilities; reconciliation of these processes; safeguarding its assets; and appropriate independent internal or external audit and compliance functions to test adherence to these controls as well as applicable laws and regulations.

•

Banking supervisors must determine that banks have adequate policies, practices and procedures in place, including strict ‘know-your-customer’ rules, which promote high ethical and professional standards in the ﬁnancial sector and prevent the bank being used, intentionally or unintentionally, by criminal elements.

•

Banking supervisors must set prudent and appropriate minimum capital adequacy requirements for all banks. Such requirements should reﬂect the risks that the banks undertake, and must deﬁne the components of capital, bearing in mind their ability to absorb losses. At least for internationally active banks, these requirements must not be less than those established in the Basle Capital Accord and its amendments.

•

An essential part of any supervisory system is the evaluation of a bank’s policies, practices and procedures related to the granting of loans and making of investments and the ongoing management of the loan and investment portfolios.

•

Banking supervisors must be satisﬁed that banks establish and adhere to adequate policies, practices and procedures for evaluating the quality of assets and the adequacy of loan loss provisions and loan loss reserves. Source: Basle Committee on Banking Supervision (1997).

Sequencing Domestic Financial Reform

141

reducing inﬂationary pressure and currency stabilization is necessary for the success of reform in China. Moreover, its deduction concerning sequencing here is that it would be better not to phase in interest rate deregulation and capital convertibility at this stage. Legal and Accounting Reform According to Sheng (1996), legal and accounting reform should be implemented to ensure that ﬁnancial transaction and ﬁnancial institutions are as transparent as possible, with laws and court procedures that enforce ﬁnancial discipline. Domestic accounting standards should be brought in line with international standards so that domestic efﬁciency can be compared with international standards. Accounting and disclosure and audit standards should facilitate public and supervisory understanding of the performance trends of ﬁnancial institutions so that the market can appraise the true value of these institutions’ net worth. Financial Deepening Financial deepening involves two main areas: the creation of active money and capital markets and the enhancement of the operating efﬁciency of the payment system, such as check clearing and electronic fund transfer mechanism (Sheng, 1996).26 China now has a relatively large capital market.27 However, many deeprooted pitfall are facing this new market, including price manipulation, insufﬁcient ﬁnancial disclosure and ailing corporate governance, which need to be addressed in the next few years. The money market in China is still in a primitive stage, while interbank markets are relatively well developed and more efﬁcient compared with other segments of the money market (Chang et al., 2000). SOE Restructuring In most cases, restructuring in the ﬁnancial sector is the mirror image of that in the real sector. Combining bank restructuring with enterprise restructuring will be particularly important in TSEs because the enterprise will require signiﬁcant resources during the transition to a market economy. The adjustment may be unavoidable (Sheng, 26 By using more efﬁcient indirect instruments of monetary management, money and capital market can strengthen macroeconomic management and reduce the risk of ‘overbanking’. Better money and foreign exchange markets also improve the market skills of banks, particularly their short-term liquidity management. Improving timeliness and certainty of payments transmission, clearing, and settlement generally reduces the liquidity pressures in an economy (Summers, 1992). 27 By the end of 2001, the A-share market and B-share had 1020 and 113 companies, respectively. Of the 1227 securities listed on the Shanghai and Shenzhen exchanges, 93 percent are stocks, 35 are investment funds, 4 percent are Treasury bonds, and 1 percent consists of corporate bonds (Ma, Qing and Li, 2001).

142

The Chinese Economy after WTO Accession

1992). In China, there may be no quick solution to the state enterprise reform and any big-bang style reform may end in failure if it does not address the critical issue of corporate governance (Zhu, 2000). This chapter suggests a quick push forward for bankruptcy legislation or introduction of time-bound special legislation and a gradual approach of privatization. Steps 2 and 3: Sequencing Domestic Financial Liberalization In the 1970s and 1980s, a number of developing countries liberalized their ﬁnancial market. Theoretically, a more liberalized ﬁnancial system is more likely to achieve resource allocation efﬁciency compared to a system with ﬁnancial repression. Empirically, ﬁnancial reform in general was accompanied by increases in real interest rate and in the ratios of money, ﬁnancial assets, and credit to the private sector to GDP (Johnston and Sundararajan, 1999). However, the reforms initially produced chaos in a number of Latin American countries, as the much desired efﬁciency gains did not seem to materialize. The Asian countries that liberalized seemed to fare somewhat better temporally until 1997–98 ﬁnancial crises. Capirio, Atiyas and Hanson (1994) attributed the failure to three main factors: Short run of investment below the optimum level due to high interest rate, timing and inadequately designed ﬁnancial safety net—the government loosed the rein on the risky investment. In sequencing China’s ﬁnancial liberalization, these experiences and lessons merit reference. Financial reform usually entails a variety of steps to ease portfolio controls and directed credit, as well as to limit government intervention in the determination of interest. Reducing barriers to competition in the ﬁnancial sector, scaling back government ownership of ﬁnancial intermediaries, allowing new ﬁnancial products to appear, limiting excessive taxation of banks and other intermediaries, and reducing restrictions on ﬁnancial dealing of domestic households and business with counterparts abroad also may be a part of reform efforts (Capirio, Atiyas and Hanson, 1994). In this chapter, the study will focus on two major opponents of ﬁnancial reform: interest rate liberalization and its cooperation with entry opening. Country Experiences: What Should China Learn From? Country experiences (see Box 8.3) to liberalize the ﬁnancial sector varied in terms of the pace and the strategy. Chile and New Zealand took a big bang strategy, while Malaysia and Korea followed a gradualist approach. Different approaches were based on different initial conditions (see Table 8.7) and the complicated domestic situations in each country. In general, a big body of literature suggests a gradualist approach because the most needed institution building takes time to establish.28 28 According to Demirguc-Kunt and Dtragiache (1998), institutional development needs to be emphasized early in the liberalization process. Unfortunately, strong institutions cannot be created overnight, thus the path to ﬁnancial liberalization should be a gradual one, in

Sequencing Domestic Financial Reform

143

Box 8.3 Summary of Financial Reforms Chile Big Bang: complete interest rate deregulation in 1975, privatization of all banks by 1978, reserve requirements reduction 1974 onwards. Capital controls on nonbanks lifted within 2 years, on banks within 5 years; all ﬁnancial institutions given ‘universal’ banking powers. Simultaneous dramatic reform of real economy, with average tariff reduced from nearly 100% in 1973 to below 30% in 1976, and budget deﬁcit eliminated by 1975, with large surplus by 1979. Indonesia Eliminated bank-by-bank credit ceilings, reduced subsidized credit program, decontrolled most deposit and lending rates, and ended subsidies on deposit rates. Second stage (ﬁve years after start) saw lifting of ban on new entry onto banking, easing of branching restraints on domestic and foreign banks, sharp reduction in reserve requirements (15% to 2%). Later strengthened prudential regulation and abolished remaining subsidized credit lines from central bank. In 1990 partially reversed reforms by requiring that 20% of lending go to small ﬁrms. Concomitant with real sector reform program: 2 large devaluations, reduction of tariffs, shift away from commodities, especially oil. Korea Began relaxing regulation of branching and management in early 1980s, transferred ownership of banks to private sector by 1983. Most preferential interest rate on policy loans abolished by mid-82, restrictions on non-bank ﬁnancial companies eased. Selected interest rate control gradually abandoned 1988–91, but tacit intervention continued. NPLs remained large through most of the 1980s. Controls on international capital ﬂows maintained during 1980s. Malaysia Gradually lifted controls on long term deposits and opened capital account in early 1970s, full liberalization of deposit and lending rates in 1978, then re-introduced administering of rates through mid-1980s; complete deregulation again in 1991. Reduced scope of priority lending program from mid-1970s, with no bank credit below banks’ cost of funds. Active central bank role in developing money and securities markets throughout periods. NPLs rose in mid-1980s but declined by end of decade. Major budget deﬁcit reduction program in the early 1980s. New Zealand Big Bang in 1984–85: interest rate controls and bank-by-bank credit guidelines removed, currency devalued then ﬂoated, all capital controls lifted, portfolio requirements and reserve asset ratio dropped, tariff reduction announced. Allowed new entry into banking in 1987, began selling state-owned ﬁrms and commenced tariff reductions in manufacturing sector. Also eased restrictions segmenting various ﬁnancial institutions. Reserve Bank bill legislating price stability goal passed in 1989; tightening of monetary and ﬁscal policy 1987 onwards. Source: Capirio, Atiyas and Hanson, 1994.

144

The Chinese Economy after WTO Accession

However, it is also noteworthy that the gradualist approach should not take too long because the existing beneﬁciary may permanently capture the rent, and rent seeking could create more distortions (Liu, 2002). Precondition and Timing for Financial Liberalization There are several preconditions that need to be satisﬁed before embarking on a domestic ﬁnancial liberalization, especially for complete interest rate deregulation. Fiscal control should precede domestic ﬁnancial liberalization (McKinnon, 1991).29 Some other preconditions should be met, such as a minimal system of prudential regulation, including enforcement; recapitalization of the banking system; key monetary control reforms and support for changes in the money market (Sundararajan, 1992). Also reasonably stable macroeconomy; the sound ﬁnancial condition of banks and their borrowers; at least a minimal base of ﬁnancial skills; checks in place to limit collusive behavior among banks in the determination of interest rates (Capirio, Atiyas and Hanson, 1994). If the decision has been made to go ahead with ﬁnancial liberalization before the regulatory and supervisory framework has been upgraded, there is a second-best argument for limiting private capital inﬂows, or for imposing speed limits on the expansion in bank lending—at least until the quality of the supervisory regime has caught up with the pace of liberalization (Goldstein and Turner, 1996). Interest rate deregulation in general should proceed in stages, with complete deregulation awaiting later stages of reform. Timing is also essential for ﬁnancial liberalization. The general principle is to avoid a radical ﬁnancial reform when recession comes. The best way to beneﬁt from timing is to move aggressively in good times and more slowly when borrower net worth is being reduced by negative shocks such as recessions or losses due to terms of trade (Capirio, Atiyas and Hanson, 1994).

which the beneﬁts of each further step towards liberalization are carefully weighed against risks. Further more, the abruptness of ﬁnancial liberalization did not give private ﬁnancial institutions themselves the time to develop internal monitoring, credit appraisal, and risk management processes that would have been necessary safeguards in the more liberal ﬁnancial environment (Johnston and Sundararajan, 1999). Where a deregulation of interest rate and credit control occurred abruptly after a long period of control, countries have experienced signiﬁcant ﬁnancial deepening but also some problems of a loss of monetary control; a more gradual approach could avoid this loss of control (Johnston and Sundararajan, 1999). 29 Direct government spending is best limited to some small share of GNP, which could increase modestly as per capita income rises. Equally important, successful liberalizing governments must levy broadly based, but low-rate, taxes on both enterprise and households.

Sequencing Domestic Financial Reform

Table 8.7

Initial Condition of Financial Reform

Country Chile

Interest Rate Repression Substantially negative for decades Modestly negative in 1970s Modestly negatively

Indonesia Korea

Malaysia

Near zero or slightly positive New Zealand Modestly negative Country Chile

Ownership of Banks

Nationalized in 1971–73, privatized in 1975–76 Indonesia No private domestic banks until 1981 Korea Signiﬁcant public ownership until 1981–83 Malaysia Large banks publicly owned New Zealand Private

145

Portfolio Requirements Supervisory System Extensive Weak, rules only Modest

Mostly rules oriented

Substantial (Policy loans of 41-51% of bank assets in 70s and 39-48 in 90s) Relatively small or nonbinding Extensive until 1984, minimal after

Mostly rules oriented

Subsidized Credit Extensive

Moderate

Mild in 1970s, essentially absent in 1980s Minimal in 1970s and 1980s Extensive until 1984, minimal after Source: Capirio, Atiyas and Hanson, 1994.

Relatively well developed Informal, only 4 banks Openness to Capital Flows Closed

Extremely open since early 1970s Essentially closed in 1970s and 80s Relatively open, but with ad hoc intervention Closed until late 1984, open thereafter

Sequencing Interest Rate Liberalization and Entry Opening in China In 2000, PBOC started interest rate liberalization with the lifting of the control of foreign currency rates for loans and deposits larger than US$3mn. Further steps have been postponed due to concerns of over-competition among the SOCBs. However, late last year, the PBOC ofﬁcially announced that it would widen the ranges within which lending rates will be allowed to ﬂoat from January 2004. Speciﬁcally, commercial banks and city credit cooperatives will be allowed to set their lending rates from 90 percent to 170 percent of the legal rates stipulated by the PBOC; and the range for rural credit cooperatives will be from 90 percent to 200 percent. These new ranges are signiﬁcantly wider than the previous ones (e.g., 90 percent to 130 percent for bank lending to small and medium enterprises and 90 percent–110 percent for large enterprises).

146

The Chinese Economy after WTO Accession

This chapter does not stand by any hasty move in ﬁnancial liberalization due to the aforementioned precondition. On the basis of the country experiences,30 this chapter suggests a gradual approach, liberalizing loan rates ﬁrst before embarking on a gradual freeing of deposit rates; freeing long-term deposit and larger deposits before moving on to shorter rates and smaller deposits—to ensure the stabilization during the liberalization. However, a more harsh condition has been set by China’s commitment to the WTO entry. This sequencing, particularly the timing and timetable, have to be in line with the timetable of China’s ﬁnancial opening up. The timing and timetable of China’s interest rate liberalization appears to become a dilemma. Empirical evidence has shown that interest rate liberalization combined with entry requirement relaxation can be a deadly dose resulting in bank failures and possible systemic crisis (Liu, 2002). Nevertheless, the opening up towards foreign entrants has begun, currently, foreign banks are allowed to conduct foreign currency business with Chinese households, during 2005–2007, opening up will extend to RMB market. Despite the advantages of foreign entry,31 empirical studies also suggest that an increase in the share of foreign banks leads to a lower proﬁtability of domestic banks (Claessens, Demirguc-kunt and Huizinga, 1998), because foreign banks are able to take some of the highly proﬁtable banking business, if domestic banks that have positive but low net worth need to build up their capital, foreign new entry could become even more dangerous (Capirio, Atiyas and Hanson, 1994). There are two possible timings for China’s ﬁnancial liberalization. One is to initiate interest rate liberalization before a fully-ﬂedged opening up to foreign banks, i.e., free RMB loan interest rates fully before 2004 and free deposit interest rate before 2006, under the condition that bank recapitalization has been fully completed and minimal supervision norm has been established in the banking sector and PBOC. Under this very tight timetable, there is only a one-year adjustment period before foreign entry. It seems to be too harsh for the banking sector and is less likely to succeed, partly because the authority does not seem to have the political will to deal with the recapitalization issue in a radical manner. Any other timing will be only a ‘second best’ solution even if every step is well designed and managed, because 30 The Philippines preferred to liberalize long-term rates ﬁrst, before moving to shortterm rates. (Such a policy may not work best if the general level of interest rates is expected to fall during the reform period. This is because banks that carry long-term liabilities, previously contracted at high interest rates, may be reluctant to reduce loan rates following the fall in liability rates.) Korea proceeded gradually, liberalizing loan rates ﬁrst before embarking on a gradual freeing of deposit rates; long term deposit and larger deposits were freed, before moving on to shorter rates and smaller deposits, in the hope that this would prevent sudden portfolio shifts. Other countries (Indonesia and Malaysia) have preferred to liberalize both loan and deposit rate simultaneously (Johnston and Sundararajan, 1999). Thailand liberalized deposit rates ﬁrst because policy makers were more concerned with savings mobilization (Liu, 2002). 31 These advantages include: increased funding resources, improved quality of services, bank efﬁciency spillovers (Claessens, Demirguc-kunt and Huizinga, 1998), and greater stability of credit in time of ﬁnancial stress (Goldberg, Dages and Kinny, 2000).

Sequencing Domestic Financial Reform

147

the close-entry condition is never to be fully satisﬁed. Restrictions on domestic entry may be a remedy, but it is ridiculous that domestic capital cannot enter into a market that foreign capital is entitled to have a ﬁnger on. The feasible solution may be a gradualist foreign entry under some implementing rules that do not conﬂict with WTO spirit, with a combination of restriction on domestic entry, which will help to maintain the franchise value.32 But the restriction cannot last too long. It is very likely that China will continue its interest rate deregulation in a gradual manner and have a fully-ﬂedged loan rate deregulation in around 2006, when the banking sector has grown out from the NPLs and before full foreign entry. Other Important Issues in Financial Liberalization It is important to bear in mind that ﬁnancial liberalization would lead to increased bank fragility, associated with the erosion of rents, increased uncertainty, and erosion of the franchise value of banks (Capirio, Honohan and Stiglitz, 1999). For most developing countries, the level and dynamic rise of interest rates have been higher than those of the pre-liberalization period (Liu, 2002). Apart from the sequencing and timing of ﬁnancial liberalization, there are some other related important issues that need attention. First, the adjustment of interest rates to positive real levels and the adoption of interest rate ﬂexibility to contain inﬂationary pressures appear to be critical policy actions in the reform period to deal with the credit and monetary effects of reform.33 Secondly, government should avoid regulations that serve to concentrate the risk in their ﬁnancial sectors, limiting banks to ﬁnancing a single sector can easily produce problems when that sector turns down. Higher capital requirements—or risk-based capital requirement, which could be geared to rise with the exposure to individual sectors—are effective mechanisms for limiting exposure (Capirio, Atiyas and Hanson, 1994). Thirdly, most countries would like to establish a money market and an inter-bank market ﬁrst so that monetary authorities are able to use these markets to conduct indirect monetary policy operations (Mehran, Quintyn and Laurens, 1996). Regular issuing of bonds as well as allowing trading in the secondary market would allow the market to determine the yield curve, which 32 Another concern here is that as long as implicit or explicit deposit insurance is being provided, the basis for wide-open entry into banking will encourage risk taking with public funds, a dangerous combination. And if supervisory skills are in scarce supply, either raising capital requirements or creating a high franchise value will be especially important. Either action would limit competition and differs only by the means to distribute bank licenses (Capirio, Atiyas and Hanson, 1994). 33 The experiences of various countries suggest the tendency that credit grows more rapidly than deposits, but as long as the authorities can maintain the positive real interest rate, the tendency proved to be temporary. After an adjustment period, the growth of deposits and credit converge, allowing for balance with a higher level of overall resource mobilization, and the economy can tolerate a somewhat more rapid growth of money and credit without increasing inﬂationary pressures (Johnston and Sundararajan, 1999).

148

The Chinese Economy after WTO Accession

will be helpful to price bank loans and other ﬁnancial assets (Liu, 2002). Fourthly, ﬁnancial sector reforms may initially reduce net private savings and increase the pressure on resources. Maintaining macroeconomic balance during ﬁnancial sector reforms may, therefore, require a reduction in the ﬁscal deﬁcit, and/or an initial attraction of foreign resources to cover the balance of payments deﬁcit (Johnston and Sundararajan, 1999). Finally, there is evidence that foreign subsidiaries tend to provide all ranges of banking services to host countries compared with branch ofﬁces. So the optimal entry requirement could be designed to encourage subsidiaries (Liu, 2002). Sequencing Institution Building in Financial Liberalization To ensure a successful transition towards full interest rate ﬂexibility, a group of reforms of prudential regulation and supervision system should phase in to support the ﬁnancial liberalization (see Box 8.4). Some of these are also discussed earlier, see ‘Step 1(B): Institutional Building, 2003–2004’. Concluding Remarks Being the ‘weakest link’ in the economy, the ﬁnancial sector in China foresees an unprecedented reform plan and restructuring agenda to provide a solid footing for achieving efﬁciency and stability. With the timetable of China’s WTO entry commitment, the ﬁnancial system is receiving increasing pressure to step up its reform process dramatically. Indeed, sequencing the domestic ﬁnancial reform is of crucial importance to the success of China’s last step reform, this chapter seeks to explore an answer and draw a road map on this issue. Technically, China’s banking sector is insolvent, and this is by no means the only risk facing PRC’s ﬁnancial system. Recapitalizing the banks and liberating them from the huge burden of NPLs is the ﬁrst step. But the bad news is that the decline of the state’s ﬁscal capacity leaves limited room for immediate bailout. Allowing banks to work their way out of a serious problem of NPLs is politically attractive, however, this growing-out strategy has dangers. It is logically clear that China needs a radical approach to resolve all its NPLs and fully recapitalize its banking sector. This chapter illustrates several ways to solve China’s NPLs problem and suggests a combination of bond-bad-debt swap, foreign strategic investors and ABS. Institution building in China’s ﬁnancial sector should be strengthened pari passu. To an extent, it is partly a precondition for domestic ﬁnancial liberalization. In domestic ﬁnancial reform, this chapter suggests a gradualist model. However, the timing and timetable of China’s interest rate liberalization appears to be a dilemma. Empirical evidence has shown that interest rate liberalization combined with relaxing the entry requirement can result in bank failures and possibly a systemic crisis. It is very likely that China will take a ‘second best’ approach: a gradualist foreign entry under some implementing rules that do not conﬂict with the

Sequencing Domestic Financial Reform

149

Box 8.4 Sequencing Supervision and Financial Restructuring Stage 1: Preparatory Preparatory stage for interest rate liberalization; direct controls on credit and interest rate dominate; limited liberalization of interest rate – for example interbank rate – is initiated and the beginning stages of impediments to broader interest rate ﬂexibility are dealt with.

• • •

A minimal program of ﬁnancial restructuring policies is introduced to deal with ﬁxedrate loan, selected NPLs, capital adequacy, subsidized and selective credit. Legal and organizational arrangements for banking supervision are reviewed and adjusted. Licensing and entry regulations are strengthened. A framework for orderly intervention and liquidation of banks is put in place or streamlined.

Stage 2: Initiating market development Direct controls on credit begin to be phased out; simple market-based instruments of monetary and exchange policy – for example, open-market-type operations based on treasury bill and credit auctions – begin to be used. Discount window and Lombard credit provide liquidity to money market; efforts to are made so that policy interest rate will be sufﬁciently ﬂexible.

• • • • • •

Reform of commercial bank accounting and bank reporting systems are phased in to help to enforce prudential norms and facilitate monetary analysis. Prudential regulations, particularly loan classiﬁcation and provisioning, credit concentration limits, credit appraisal guideline, and foreign exchange exposure rules begin to phase in based on new accounting standards. Capital adequacy norms are strengthened and phased in, in line with bank restructuring strategy. A strategy to combine off-site, on-site, and external audits is introduced, the balance among the components initially dictated by the availability of resources and technical assistance. Institutional development of banks is pursued actively; portfolio audits of major banks initiated. A comprehensive program of bank restructuring, bank liquidations, and loan recovery and loan workout arrangements is formulated. As part of this program, simple ﬁnancial restructuring policies for banks—supported by enterprise ﬁnancial restructuring—are implemented.

Stage 3 Money market and secondary markets in government securities are strengthened through supporting reforms of institutional arrangements and payment system; open-market operations become more active; foreign exchange markets are fostered. Central bank manages money market liquidity at its own initiative, and interest rates are fully ﬂexible.

• •

Comprehensive reforms to foster bank and enterprise restructuring are continued systematically, in line with the program designed in Stage 2. Well-capitalized and well-supervised dealers in government securities (and money market instruments) are promoted as part of strengthening security market regulations and supervision.

The Chinese Economy after WTO Accession

150

• • • • •

Reforms of bank accounting and prudential standards are completed. Financial risk management in payment system is further strengthened. Reinforcement of supervision of asset-liability management (interest rate risks, liquidity management), internal controls, and management system of banks is continued. Appropriate balance between off-site supervision, on-site supervision, and external audit is achieved through technical assistance and training. Risk implications of ﬁnancial innovations and internationalization are monitored closely.

Note: Stages are used for analytical convenience and do not imply that the steps listed in different stages cannot be implemented simultaneously, if technically feasible. Source: Johnston and Sundararajan, 1999.

WTO spirit, with a combination of restrictions on domestic entry, which will help to maintain the franchise value. References Agénor, Pierre-Richard and Joshua Aizenman (1999), ‘Financial Sector Inefﬁciencies and Coordinate Failures: Implications for Crisis Management,’ NBER Working Paper No. 7446. Aizenman, Joshua (2002), ‘Financial Opening: Evidence and Policy Options,’ NBER Working Paper No. 8900. Barth, J., G. Capiro and R. Levine (2001), The Regulation and Supervision of Banks around the World: A New Database, World Bank. Basle Committee on Banking Supervision (1997), Core Principles for Effective Banking Supervision. Available on the Internet at www.bis.org. BIS (1999), Bank restructuring in practice. BIS Policy Papers No.6. Bordo, Michael and Barry Eichengreen (1998), ‘Implications of the Great Depression for the Development of the International Monetary System,’ in Michael Bordo, Claudia Goldin and Eugene White (eds), The Deﬁning Moment: The Great Depression and the American Economy in the Twentieth Century, University of Chicago Press, Chicago. Bottelier, Pieter (2002a), ‘China: Economic Reform Accomplishments and Challenges,’ Presentation for the Center for US-China Cooperation, Graduate School of International Studies, University of Denver. Bottelier, Pieter (2002b), ‘Managing China’s Transition Debt: Challenges for Sustained Development,’ working paper. Capirio, Gerard, Jr., Izak Atiyas and James A. Hanson (1994), Financial Reform: Theory and Experience, Cambridge University Press, New York. Capirio, Gerard, Jr., P. Honohan and J. E. Stiglitz (1999), ‘Financial liberalization: How Far and How Fast? Introduction and Overview,’ World Bank Manuscript.

Sequencing Domestic Financial Reform

151

CCER (China Center for Economic Research) (1998), ‘Seeking Effective Policy Combination under Multi Economic Objective: Analysis of China’s Macroeconomic situation in 1998 and some recommendations,’ Economic Research Journal (Jingji Yanjiu), Vol. 4. Chen, Baizhu, Dietrich J. Kimball and Yi Fang (2000), ‘An Overview of China’s Financial Market: Progress, Problems and prospects,’ in Baizu Chen (ed.), Financial Market Reform in China: Progress, Problems, and Prospects. Westview Press. Chow, Gregory C. (2002), China’s Economic Transformation, Blackwell Publishers. Claessens, Stijn, Asli Demirguc-kunt, and Harry Huizinga (1998), ‘How does Foreign Entry Affect the Domestic Banking Market?’ NBER Working Paper Series. CLSA (2002), ‘Banking in China: Problems and Virtues,’ CLSA sector outlook. Demirguc-Kunt, Ash and Enrica Dtragiache (1998), ‘Financial Liberalization and Financial Fragility,’ IMF Working Paper WP/98/83. Diamond, Douglas W. and Raghuram G. Rajan (2002), ‘Liquidity Shortages and Banking Crises,’ NBER Working Paper No. 8937. Dornbusch, Rudiger (1976), ‘Expectations and Exchange Rate Dynamics,’ Journal of Political Economy, Vol. 84. Dornbusch, Rudiger and F. Givazzi (1999), ‘Heading Off China’s Financial Crisis,’ unpublished paper, Department of Economics, MIT, Cambridge, Mass. Edwards, Sebastian (1984), ‘The Order of Liberalization of the External Sector in Developing Countries,’ Essays in International Finance. No. 156. Princeton University. Edwards, Sebastian and Michael D. Bordo (2002), ‘Crises Now and Then: What Lessons from the Last Era of Financial Globalization?’ NBER Working Paper No. 8716. Frenkel, Jacob (1982), ‘The order of Economic Liberalization: A Comment,’ in K. Brunner and A. Meltzer (eds), Economic Policy in a World of Change, Amsterdam, North-Holland. Furman, Jason and Joseph Stiglitz (1998), ‘Economic Crises: Evidence and Insights from East Asia,’ Brookings Papers on Economic Activity 2. Gelb, Alan and Patrick Honohan (1989), ‘Financial Sector Reforms in Adjustment Programs,’ Policy Research Working Paper 169, World Bank. Gibson, H. D. and E. Tsakalotos (1994), ‘The Scope and Limits of Financial Liberalization in Developing Countries: A Critical Survey,’ Journal of Development Studies, Vol. 30, No. 3. Goldstein, Morris and Philip Turner (1996), ‘Banking Crises in Emerging Economies: Origins and Policy Options, BIS Economic Papers No. 46. Goldberg, L., G. Dages and D. Kinney (2000), ‘Foreign and Domestic Bank Participation in Emerging Markets: Lessons from Mexico and Argentina,’ NBER Working Paper WP 7714. Hanson, James and Roberto Rocha (1986), ‘High Interest Rates, Spreads, and the Cost of Intermediation.’ Industry and Finance Series, Vol. 18, World Bank.

152

The Chinese Economy after WTO Accession

Hellman, Murdock and J. E. Stiglitz (2000), ‘Liberalization, Moral Hazard in Banking, and Prudential Regulation: Are Capital Requirements Enough?’ American Economic Review. Vol. 90, No.1. Honohan, Patrick (1997), ‘Banking System Failures in Developing and Transition Countries: Diagnosis and Prediction,’ BIS Working Paper No. 39. Huang Yiping (2002a), ‘Are Fiscal Conditions Getting Out of Hand?’ Citi Group Research Paper. Huang Yiping (2002b), ‘Is China Heading for a Banking Crisis?’ Citi Group Research Paper. Huang Yiping (2002c), ‘The Case for Breaking up the Large Banks,’ Citi Group Research Paper. Huang Yiping (2003), ‘Band Widening Inevitable, But Not Now,’ Citi Group Research Paper. Johnston, R. Barry and Sundararajan V. (1999), Sequencing Financial Sector Reforms: Country Experiences and Issues, International Monetary Fund, Washington DC. Kane, Edward (1992), ‘How Incentive – Incompatible Deposit Insurance Funds Fail.’ Research in Financial Services: Private and Public Policy, Vol. 4, JAI Press, Greenwich, Conn. Keidel, Albert (2001a), ‘Practical Issues in Using the Chinese Statistical System,’ China Perspectives, No. 33, Jan./Feb. Keidel, Albert (2001b), ‘China’s GDP Expenditure Accounts,’ China Economic Review, Vol. 12. Klingebiel, D. and G. Caprio (1996), ‘Bank insolvency: bad luck, bad policy, or bad banking?’ in M. Bruno and B. Pleskovic (eds), Annual World Bank Conference on Development Economies, World Bank. Krueger, Anne (1984), ‘Problems of Liberalization,’ in Arnold Harberger (ed.), World Economic Growth, ICS Press, San Francisco. Krueger, Anne (1986), ‘Problems of Liberalization,’ in A. Choksi and D. Papageorgiou (eds), Economic Liberalization in Developing Countries, Basil Blackwell, London. Lardy, Nicolas (1998), China’s Unﬁnished Economic Revolution, Brookings Institute Press, Washington DC. Liu, Li-Gang (2002), ‘Sequencing China’s Banking Sector Reform after the WTO,’ ADB annual conference paper. MacDougall, Ian (1968), ‘The Beneﬁts and Costs of Private Investment from Abroad: A Theoretical Approach,’ in R. Caves and H. Johnson (eds), Readings in International Economics, London. Ma, Guonan and Ben S. C. Fung (2002), ‘China’s asset management corporations,’ BIS Working Papers No. 115. Ma, Jun, Hai Qing and Hongsheng Li (2001), ‘China’s Post-WTO Industrial Restructuring,’ Deutsche Bank Global Markets Research Paper. Ma, Jun and David Folkerts Landau (2001), ‘China’s Financial liberalization Agenda,’ Hongkong Deutsche Bank Global Markets Research Paper.

Sequencing Domestic Financial Reform

153

Mathieson, D. J. and Richard Haas (1994), ‘Establishing Monetary Control in Financial System with Insolvent Institutions,’ IMF Paper on Policy Analysis and Assessment. McKinnon, Ronald I. (1973), Money and Capital in Economic Development, Brookings Institution, Washington DC. McKinnon, Ronald I. (1991), The Order of Economic Liberalization: Financial Control in the Transition to a Market Economy, The Johns Hopkins University Press. Mehran, H., M. Nordman, T. Quintyn and Bernard Laurens (1996), ‘Monetary and Exchange System Reform in China: An experiment in Gradualism,’ IMF Occasional Paper 141. Mundell, Robert (1968), ‘Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rate,’ in R. Mundell (ed.), International Economics, Macmillan, New York. Quinn, Dennis P. (1997), ‘The Correlates of Changes in International Financial Regulation,’ American Political Science Review, Vol. 91. Rawski, Thomas G. (2001), ‘China by the Number: How Reform has Affected China’s Economic Statistics,’ China Perspectives, No. 33, Jan./Feb. Roubini, N. and X. Sala-i-Martin (1992), ‘Financial Repression and Economic Growth,’ Journal of Development Economics, Vol. 39, No. 1. Sheng, Andrew (1992), ‘Bad Debts in Transitional Socialist Economies,’ World Bank, Financial Policy and System Division. Sheng, Andrew (1996), Bank Restructuring: Lessons from the 1980s, The World Bank Standard and Poor (2001), Global ﬁnancial system stress, 9 October. Stiglitz, J. (1998), ‘Sound Finance and Sustainable Development in Asia,’ Keynote speech to the Asian Development Forum, Manila, 12 March. Summers, Bruce J. (1992), ‘The Role of the Central Bank in Developing Payment System.’ Paper presented at the 19th SEANZA central banking course. Summers, Larry H. (2000), ‘International Financial Crises: Causes, Prevention and Cures,’ American Economic Association, May. Sundararajan, Vasudevan (1992), ‘Financial Sector Reform and Their Appropriate Sequencing,’ IMF, Monetary and Exchange Affairs Department. Tanzi, Vito and Mario Blejer (1982), ‘Inﬂation, interest rate policy, and currency substitution in Developing Countries: A Discussion of some Major Issues,’ World Development, Vol. 10, pp. 781–89. Villanueva, Delano and Abbas Mirakhor (1990), ‘Strategies for Financial Reform: Interest Rate Policies, Stabilization, and Bank Supervision in Developing Countries,’ Staff Papers, Vol. 37, IMF. Wolf, Holger C. (1999), ‘Transition Strategies: Choices and Outcomes,’ Princeton Studies in International Finance, No. 85. Woo, Wing Thye (1998), ‘Chinese Economic Growth: Sources and Prospects,’ in M. Fouquin and F. Lemoine (eds), The Chinese Economy, Ecomomica, London.

154

The Chinese Economy after WTO Accession

Yang, Yanqing (2001), ‘How will AMCs Deal with NPLs in China,’ Shanghai Comprehensive Economy, Vol. 1 Zhu, Tian (2000), ‘Restructuring China’s SOE: A Corporate Governance Perspective,’ in Financial Market Reform in China: Progress, Problems, and Prospects, Westview Press.

Chapter 9

Public Venture Capital and Its Private Strategies in China Changwen Zhao, Shuming Bao and Chunfa Chen

Introduction Sufﬁcient evidence reveals the signiﬁcant contribution of venture capital (VC) to the development of start-ups and economies of many countries, especially the United States in the last two decades. It has been embraced as a key tool for economic and technological development by policy-makers, academics and the business community, and the US venture capital industry is often held up as the model other nations should attempt to replicate. Venture capital is deﬁned as a professionally managed pool of money for the sole purpose of making actively managed direct equity investments in rapidly growing high-tech companies, and with a well-deﬁned exit strategy. Venture capital (VC) can be divided into two categories – Private Venture Capital and Public Venture Capital – by the source of funds and the performer. Private venture capital funds are professionally managed, and have risk-equity capital invested primarily in innovative and/or rapidly expanding enterprises (US National Venture Capital Association, 2000). The predominant form of venture capital funds is limited partnerships (Horvath, 1999). Along with providing funds by purchasing equity securities, these funds also add value by actively participating in the development of the company. Public venture capital is a subset of ‘public investment,’ and it is a variety of the mechanisms that have been created by the (federal, state and local) governments to encourage knowledge-based economic development. These programs have in common the commitment of public funds to support the entrepreneurial development of technology in situations where private venture capital ﬁnds it too risky to venture (Etzkowitz et al., 2002). A common understanding is that public VC involves central and local governments while private VC involves for-proﬁt VC corporations. Both private venture capital and public venture capital have been playing important roles in ﬁnancing the high-tech businesses with different strengths and limitations. However, the public venture capital has been neglected in the past while most academic attention has gone into private venture capital. And as a matter of fact, public venture capital has been a very important ﬁnancing alternative, especially

156

The Chinese Economy after WTO Accession

today when less private venture capital funds are invested across many countries and a large part of the funds is being reinvested in portfolio companies. Henry Etzkowitz et al. (2001) suggest that public venture capital is a hidden asset since many high-tech entrepreneurs often are not aware of the grants and matching funds available to them through government funding sources, i.e., a variety of government programs, which provide both initial and later-stage venture capital for their companies. The use of public resources to reduce risk in the development of new technology has long been accepted in the agricultural, military and health areas. In recent years, marked by controversy, public entrepreneurs have extended the role of government from the macro factors affecting innovation such as interest rates and money supply to the micro conditions of innovation (Etzkowitz, 2002). Public venture capital funds usually included in such projects generally expect lower rates of return than those sought by private venture capitalists and they often operate with a geographic or community focus. The funds may operate with a dual bottom line, considering both social and ﬁnancial returns on investments. A common characteristic of these nontraditional venture capital institutions is the attempt to ﬁll venture capital gaps that exist in their markets and to overcome impediments to making venture capital investments in their regions (Markley et al., 2001). Obviously different scholars approach the same concept in different ways, on different levels and from different perspectives. Meanwhile, some people argue that the deﬁnition for ‘Public Venture Capital’ contains many ambiguities, raising important questions and issues. For instance, what is public? What is private? (Canadian Venture Exchange, 2001.) Some people equate the word ‘venture’ to private opportunities. However, it is of great importance to separate the public venture capital from private venture capital. In our opinion, public venture capital is a typical government ﬁnancial program that supports the new technology R&D and commercialization of small businesses, and it is complex, consisting of a variety of grants, loans, subsidies, and even equity investment, at the central or local levels of the government. Public venture capital distinguishes itself from private venture capital primarily in the sources of capitalization, investment objectives, ﬁnancing stages, ﬁnancial instruments and organizational structures. 1. Different sources of capital: the central or local governments fund Public VC while private VC is funded by the private sector. 2. Different objectives: public VC programs are always targeted at gaining the best technology to meet government needs, usually for central government; encouraging the development and growth of high-tech ﬁrms for both central and local governments; helping to spread new technology throughout the economy for both central and local governments; and seeking to attain the growth of local economy and regional development. However, private venture capitalists are always willing to invest in those ﬁrms with higher expected returns.

Public Venture Capital and Its Private Strategies in China

157

3. Different ﬁnancing stages: Public VC usually ﬁnances the start-ups, because entrepreneurs have been least successful with raising appropriate ﬁnance for initial operations. In this stage most of them ﬁnanced their start-ups with their own savings; no ﬁnance is available either from banks, or from other providers of ﬁnance. During the expansion stage, funds for premises and equipment have been much easier to access. Basically, private VC is more attracted to the high-tech companies sitting on the expansion stage, so more stable, developed businesses are offered much more attractive and viable assistance by providers of private venture capital. 4. Different ﬁnancial instruments: Public VC ﬁnances small businesses with a variety of ﬁnancial instruments in the form of grants, loans, loan guarantee, direct subsidies and so on. In most cases, public venture capital programs do not require equity in exchange. However, the situation is changing in recent years, more and more public VC programs are seeking to balance the objectives and to combine the different ﬁnancial instruments. Meanwhile, private VC is involved in equity investment and therefore requires equity in exchange. 5. Different organization structures: In general, private VC funds are organized as limited partnerships with a predetermined life of ten years. However, the organization structure of public VC varies by the level, country and objectives. It has been recognized that the public VC is important for many new startups at their early stage of development, especially in those countries under a transition from centrally planned economy to market economy with a bankbased ﬁnancial system, although public VC and private VC may function in different ways. Focusing on China’s public VC, along with a comparison of public venture capital and private venture capital, this chapter will explore such issues as: Why and where should the public VC be invested and its relationship with private venture capital? How can the public VC be more efﬁcient for funding high-tech start-ups? And what are the strategies to privatize public VC? This chapter ﬁrst provides a literature review of public VC, and then it reviews the historical development of ﬁnancing systems and public VC institutions for small businesses in China, together with the current situation of VC and public VC in China. The third section focuses on some different models which combine public VC with private VC in China. Finally, we will present some discussion and suggestions on government policies. Literature Review The ﬁnance literature on venture capital has grown considerably both theoretically and empirically, research on public venture capital, however, has been limited. In accordance with modern economic theory, where market failures prevent the development of a vital industry, public policies must be adapted to correct or

158

The Chinese Economy after WTO Accession

diminish these market disincentives. In the case of venture capital, public programs that provide early-stage ﬁnancing to ﬁrms, particularly high-tech companies, have become commonplace in the United States, China and other countries, suggesting that the venture capital market itself might not be perfect both in developed and developing countries. Steven Casper (1999) ﬁnds that the venture capital and high-tech industry development in Germany represent an interesting challenge to prevailing institutional theory as applied to the study of advanced industrial economies, which tend to view the characteristics of organizations as strongly constrained by the orientation of a number of key national institutional frameworks. A growing body of empirical research suggests that new ﬁrms, especially technology-intensive ones, may receive insufﬁcient capital to fund all positive net present value projects due to the information problems. The literature on capital constraints (reviewed in Hubbard, 1998) documents that an inability to obtain external ﬁnancing limits many forms of business investment, particularly relevant are works by Hall (1992), Hao and Jaffe (1993), and Himmelberg and Petersen (1994). These show that capital constraints appear to limit R&D expenditures, especially in small high-tech ﬁrms. David L. Barkley et al. (1999) focus on the state-backed venture capital programs which serve one or more of the following goals: to encourage general economic development; to enhance economic opportunities for geographically isolated regions or economically disadvantaged populations; to enhance availability of early stage or seed capital investments; and to create venture capital infrastructure and management capacity within the state and so on. Barkley et al. (2001) also propose that the industrial focus and geographic concentration of venture capital activity contributed to the view that private venture capital ﬁrms underserve certain industries and regions of the country. The uneven allocation of venture capital investments among regions and industries potentially reﬂects the distribution of good investment opportunities. However, it also may reﬂect market failures that result from imperfect information and high transaction costs. Private venture capital is not willing to aggressively seek investment opportunities in small metropolitan areas and no metro communities because deal ﬂow is sparse, costs per investment are relatively high, exit opportunities are limited, and local business environments are less supportive. Economic historians document that the US federal government has played a key direct or catalytic role in developing many of the most important technologies. Additionally, federally chartered Small Business Investment Companies accounted for two-thirds of all the venture capital funding allocated to American business from the 1950s to 1969, and continued to play an important role even after private partnerships rose to preeminence. Lerner (1999) documents that one of the most important and visible federal programs, the Small Business Innovation Research (SBIR) grant program, which provides grants to small businesses for feasibility and development research in high-tech areas, designed to correct a perceived market gap in the equity ﬁnancing of young, technology-based companies, has proven very successful at fostering

Public Venture Capital and Its Private Strategies in China

159

technological innovation and corporate growth. On the other hand, Lerner’s ﬁnding that the positive effects of SBIR awards were conﬁned to ﬁrms based in zip codes with substantial private venture capital activity suggests that government programs are only effective when they supplement, rather than attempt to supplant or promote, private fund-raising. SBIR acted as ‘a complement to venture capital organizations and other private institutions that assist new ﬁrms, and the impact of the awards in regions without these private sector mechanisms was minimal’. However, we should not neglect the government inability when we emphasize the market failure for ﬁnancing small high-tech businesses. Indeed, some scholars regard government as a supplement, not the substitution for market. If the ﬁnancing difﬁculty for small high-tech business was due to moral hazards and information asymmetries, why would we encourage public ofﬁcials instead of private venture capital as a source of capital in this setting? Moreover, it is always an important issue for many government programs to avoid bureaucratic and ineffective decisions. Empirical research has increasingly suggested the inability of the new ﬁrms, especially technology-intensive ones, to receive sufﬁcient capital to fund all positive net present value projects due to the asymmetrical information problems. However, if public venture capital awards could certify that ﬁrms are of high quality, the information problems could be overcome and investors could conﬁdently invest in these ﬁrms (Lerner, 2002). Many SBICs made investments in ineffective or corrupt ﬁrms. It has been noted that SBIC manager’s incentives to screen or monitor portfolio ﬁrms were greatly reduced by the presence of government guarantees that limited their exposure to unsuccessful investments (Lerner, 2002). An extensive literature on political economy and public ﬁnance has emphasized the distortions that may result from government subsidies when particular interest groups or politicians seek to direct subsidies in a manner that beneﬁts them. As articulated by Olson (1965) and Stigler (1971), and formally modeled in works such as Peltzman (1976) and Becker (1983), the theory of regulatory capture suggests that direct and indirect subsidies will be captured by parties whose joint political activity such as lobbying is not too difﬁcult to arrange (i.e., when ‘free riding’ by coalition members is not too large a problem). These distortions may manifest themselves in several ways. One possibility (discussed, for instance, in Eisinger, 1988) is that ﬁrms may seek transfer payments that directly increase their proﬁts. Politicians may acquiesce in such transfers in the case of companies that are politically connected. A more subtle distortion discussed by Cohen and Noll (1991) and Wallsten (2000): ofﬁcials may seek to select ﬁrms based on their likely success, and fund them regardless of whether the government funds are needed. In this case, they can claim credit for the ﬁrms’ ultimate success even if the marginal contribution of the public funds was very low. In studying new control modes and emerging organizational forms of PrivatePublic Contracting in public administration, Willmott et al. (2001) examine the ‘reinvention’ of public service provision heralded in the present government’s plans for private-public partnership. Their conclusions question the currently fashionable

160

The Chinese Economy after WTO Accession

idea of ‘post-bureaucracy’ in public services, showing that public-private partnerships inevitably require the introduction of additional layers of bureaucracy to regulate the new self-managing organizations. While these discussions concentrate on what the public sector has to learn from private enterprise, minimal consideration is given to the question of how, in return, market disciplines can be bettered by the notions of civic virtue and democratic accountability which are deeply rooted in the public sector. Recently, several public-internet companies have been identiﬁed in the popular press as ‘public venture capital ﬁrms,’ but in the current legal environment these ﬁrms cannot operate as traditional private equity funds do. For a combination of legal and business reasons, publicly traded private equity funds have so far proven unworkable (Testa et al., 2000). Deborah Markley et al. (2001) conducted case studies of public VC programs to determine characteristics associated with successful and unsuccessful programs. The Research focused on the advantages and disadvantages of six general program types: publicly funded and publicly managed, publicly funded and privately managed, tax credits to encourage private VC investments, community level funds, community development funds, and SBICs. One of the reasons why debates on Public and Private Partnership (PPPs) tend to generate more heat than light is that there is little agreement as to what constitutes a partnership or the types of problem that they might help solve. Public managers do not have a clear account of how partnerships can help to improve the quality of services. We argue that PPPs are a risk-sharing relationship based upon an agreed aspiration between the public and private sectors to bring about a desired public policy outcome (Robinson, 2001). From the discussion above, it can be seen that both public and private VCs present both advantages and disadvantages. So a favorable way is to combine the advantages both of market and government, integrate the private and public VCs if possible. But the issue is how to establish an appropriate relationship between public venture capital and private venture capital. The Development of the Venture Capital Industry and Public Venture Capital in China Development of the VC Industry in China The current VC situation in China proves to be extremely difﬁcult to evaluate and measure. The initial development of the VC industry in China was essentially done through direct government participation in the market. In the mid-1980s, central and regional governments started to establish VC companies. Moreover, the governments have also made the establishment of venture capital communities a top priority. On the whole, China has enormous potential to develop the venture capital industry, though it faces many pitfalls and challenges.

Public Venture Capital and Its Private Strategies in China

161

Table 9.1

Venture Capital Pools and Disbursements in China by Year (in $US millions)

Year

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Numbers

N/A

Pools

N/A

8

15

28

36

38

66

128

202

236

246

269

878 1,422 2,384 3,458 3,612 3,500 3,112 3,735 5,647 7,899 8,200

Disbursements N/A 65 242 645 678 609 662 411 728 806 565 418 N/A—not available. Source: Asian Venture Capital Journal various years and Authors’ calculations based on this and the China Annual Report of venture capital in 2002 published by Zero2ipo Company.

Venture capital amount and structureThe venture capital industry in China is remarkably volatile. As can be seen in Table 9.1, the amount of funds raised increased dramatically until 1995, and then ﬂattened out. From 1999, it shot up again, reaching a peak in 2002. Similarly, the investments have oscillated violently, reaching between US$50–700 million per year from 1992 through 1997. But then investment plummeted in 1998, before increasing once again in 1999, and reaching the peak in 2000. In 2001 and 2002, there was ample or even excess venture capital in China, but a lack of promising investments. This is illustrated by the low-level of investment versus the total pool of funds. The Chinese VC industry is in a process of diversiﬁcation, in part driven by the increased competition. Many companies today direct their investments to a speciﬁc industry sector, region or a speciﬁc development stage of a company. There is no established pattern of investing in China and thus far much of the investing continues to be experimental and ranges from high-tech to infrastructure and consumer goods. Table 9.2 indicates that the greatest investments have been in infrastructure, consumer products, and transportation/distribution, areas that are not normally considered for VC investment. But given the growth of the domestic market these may be the signiﬁcant opportunities. From 1998 to 1999, there was a signiﬁcant increase in information technology investing, probably manifesting the Internet bubble in China. The amount of investment in the IT industry continued to grow from 2000 to 2002, but with a declining growth rate. Meanwhile, investment in Biomedicine, Telecommunications, and Utilities infrastructure keeps on increasing due to the increasingly open policy. An important aspect of the VC industry is in what stage of the company development process it invests. According to our study, around 5 percent of the number of investments in 2000 was made in the seed stage and 35 percent in the early growth stage. Given that many VC companies have invested only a small share of their total raised capital, the amount of capital aimed for investments in later stages may be higher. The Chinese venture capital industry is geographically clustered. 150 VC companies have ofﬁces in Beijing, Shanghai and Shenzhen, accounting for 60 percent of the total VC companies and 70 percent of total funds. The most obvious

162

Table 9.2

The Chinese Economy after WTO Accession

Venture Capital Investments in China by Sector (1998–2002, in $US millions*)

Sector Agriculture/Fisheries Computers Conglomerates Construction Consumer prod. Electronics Ecology Financial services Information technology Infrastructure Leisure/entertain. Manuf. heavy Manuf. light Media Medicine Mining & metals

1998 8 154 0 128 411 172 0 65 53 723 0 238 42 11 61 0

1999 3 227 0 168 398 203 0 70 200 759 4 221 71 16 128 0

2000 12 244 0 189 436 258 0 126 247 806 12 167 86 65 220 0

2001 15 251 0 191 422 273 12 144 305 944 18 188 120 87 368 2

2002 20 256 0 220 467 298 21 168 345 996 22 213 175 102 458 12

Retail/wholesale 28 31 68 140 216 Services- nonﬁnan. 112 127 126 187 232 Telecommunications 144 197 267 304 406 Textiles & clothing 5 11 25 30 53 Transport/distribution 297 257 268 388 405 Travel/hospitality 86 93 138 151 166 Utilities 259 224 376 402 456 TOTAL 2,997 3,408 4,136 4,942 5,707 * These represent the entire portfolio and not the annual investment. Source: 1998–1999 from Martin Kenney et al. (2002). 2000-2002 from Authors’ calculation based on this and the China Annual Report of venture capital in 2002 published by Zero2ipo Company.

cluster of research institutions and universities is the Zhongguancun, Beijing, which is located in close proximity to the top two Chinese universities and various national research institutes. However, undoubtedly this is an undercount as there has been a proliferation of venture capital funds. Investments are more dispersed as venture capitalists mentioned areas such as Guangzhou, Wuhan, Chengdu, Xi’an and other cities with good universities as possible sources of deals. Foreign venture capital companies in ChinaOn the whole, China is a capitalpoor country today so foreign VC could be very important in helping to create a good climate for VC investment as well as VC industry. Moreover, the government has expressed support for foreign VC investment. IDG, one of the biggest venture capital companies in the world, have exited from 30 projects with a 55 percent

Public Venture Capital and Its Private Strategies in China

163

average return in China while it only was 45 percent in the US and 35 percent in the EU over the same period. In total, IDG have invested over 120 projects with US$250 million in China from 1980 to 2002. Statistics show that among the 73 VC companies that invested in China with a total amount of US$418 million, there are 34 foreign companies accounting for 49.8 percent (China Venture Capital Forum in 2003, Beijing). However, the difﬁculties for foreign ﬁrms are probably even greater than for domestic ﬁrms. Up to now, there is only Venture Capital Company form in China, but there is no real Venture Capital Fund. So the alternative way for investing in a Chinese portfolio is through a Chinese-foreign Equity Joint Venture; however, this is cumbersome and requires approval from the Ministry of Commerce—an approval that may not come quickly. The other possibility for foreign venture capitalists would be to invest directly in a Chinese limited liability corporation (LLC). The difﬁculty here is that the law is vague on the legality of such an investment (Lo, 2000). Difﬁculties will be minimized if everything is operated according to plan. It is when difﬁculties arise that legal clarity or, at least, a transparent process for settling the difﬁculties is necessary. At this point, both methods are not so attractive to venture capitalists. Finally, the government controls the ability of Chinese ﬁrms to list on overseas markets and thus limits the range of potential acquirers. The exit strategyExit is presently a little difﬁcult, especially for international investors. Trade sales are possible, though it could be difﬁcult to remit the currency abroad if it is sold to another Chinese company. If the ﬁrm is sold to a foreign ﬁrm, then, with approval from the government, payment can be made in a foreign currency. In the case of an IPO in the Chinese stock market, the situation is far more complicated due to the regulations regarding liquidation of shares after listing because the stocks that can be traded are conﬁned to those held by the public. The other exit strategy is to list on overseas markets such as the Hong Kong GEM or the NASDAQ. However, there are difﬁculties, with both options. A second board was planned to open in Shenzhen in 2001, and it might create a viable domestic exit opportunity. However, the opening has been delayed repeatedly for fear that it would become a weakly traded, illiquid market. This would make it an unattractive exit channel. Summary There is consensus that China is the most important location for venture investing in the future, and there appears to be a large number of strategies for investing in China—a situation suggesting either many opportunities or uncertainty about what will work. The Chinese environment is not yet sufﬁciently mature or stable to present a ‘favorable’ climate for venture investing which is still a new practice with a few examples of successful exits with returns of ten times or greater. To improve the situation, the government must undertake to remove the barriers that are amenable to government action.

164

The Chinese Economy after WTO Accession

Public Venture Capital in China Within the past few years, public efforts to ﬁnance small high-tech ﬁrms have increased. Actually, from the beginning, Chinese VC has been supported and promoted by the government at the national, provincial, and municipal levels. Many government entities in China have been actively involved in VC investing. The situation is complex and difﬁcult to fully comprehend. Today most VC companies are those founded by the government. By one estimate, government capital makes up over 80 percent of the total VC investment (UltraChina.com 2000). Innovation Fund for Small Technology-based Firms (IFSTF) IFSTF is a special government fund set up with the approval of the State Council to support technological innovations of small businesses by providing ﬁnancing in the forms of appropriation, loan interest subsidy and equity investment. The fund was ofﬁcially launched on June 25, 1999 with an initial capital of US$120 million through budgetary allocations by the central government. As a policy-oriented fund, the Innovation Fund facilitates and encourages the innovation activities of STFs and the commercialization of research achievements by ways of ﬁnancing. In the meantime, it works to bring along and attract investment from the whole society so as to promote the establishment of a new investment mechanism conforming to the rules of market economy for technological innovations of STFs. From 2001, a total of US$250 million in policy-oriented loans work together with IFSTF. IFSTF takes three different forms depending on the speciﬁc characteristics of the projects. First of all, appropriation is applied primarily as start-up capital to small ﬁrms founded by research personnel bearing their own research achievements. It also provides partial subsidies to STFs for new product development and pilotproduction. The total volume of appropriation to each project will generally not exceed US$120,000 with a maximum of US$250,000 for key projects. Secondly, loan interest subsidy is adopted to provide interest for STFs requiring loans from commercial banks to expand the production scale of the innovation project. The total subsidy amount of each project is generally within US$120,000 and US$250,000 for key projects. The loan support is different from the loan of policy-oriented banks that generally can be used in many ﬁelds while the loan from IFSTF is limited to use in technology innovation. Thirdly, equity investment is targeted at a small number of projects with a high level of technology starting point, greater innovation capacity and market potential in emerging industries. Generally the investment will not exceed 20 percent of the registered capital of the invested company and will be redeemed within a time limit in accordance with law. As a special government fund with the objective of supporting technological innovation and facilitating commercialization of R&D results, IFSTF differs from other non-governmental funds or private VC in three main features. Firstly, it is macro policy-oriented, aiming to promote the development of new and high-tech industries by encouraging the technology innovation activities of STFs. Secondly, it serves as a

Public Venture Capital and Its Private Strategies in China

Table 9.3

165

Volume of IFSTF in China Time

Year 1999 Year 2000 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Total input of funds (in $US million) 12.5 17.2 26.0 43.8 17.3 21.0 18.7 The number of projects 125 205 277 481 188 230 198 Average supporting sum of one 9.97 8.37 9.40 9.10 9.17 9.14 9.43 project (US$104 ) Source: IFSTF Management Center, the Ministry of Science and Technology, PRC.

Total 156.5 1,703 9.18

‘priming-pump’, trying to attract more investment to STFs from local governments, corporations and ﬁnancial institutions. Finally, the innovation fund does not aim at proﬁt-making in its own right, but it takes revenue increase and job creations as the reward, therefore contributing to the national economic structure adjustment and the general health and growth of the economy. IFSTF is oriented towards small businesses registered in China with various ownership structures involved in R&D, production and services in new and hightech industrial technology. The successful applicant should be a business corporation with no more than 500 employees. Meanwhile, priorities will be given to those projects with innovative technology or independent intellectual property and high added value, to those started by the research personnel or returned overseas students to transfer their scientiﬁc achievements, to those innovation projects jointly initiated by the enterprises, universities and research institutions, and those that make use of new and high technology to revive the stock assets of the traditional industries and drive job creations. The statistics of 1999 show the different technology sectors, investment stages, business ownership structures, and company sizes of all the approved projects. IFSTF primarily supported the high-tech projects in the ﬁelds of IT, Automation Technology, Advanced Materials, Biotechnology and Medicine, Environmental Resources as well as New Energy and Energy Saving, accounting for 30.4 percent, 20.1 percent, 19.7 percent, 19.4 percent, 5.1 percent and 3.8 percent respectively. In terms of the stage of the company, 52 percent of IFSTF efforts went into pilot production with scale production and R&D taking up 26.9 percent and 20.3 percent. For the ownership structure, Limited Liability Company is the most popular form, accounting for almost a half, with the rest divided among limited companies by shares, private companies, joint ventures, state enterprises and collective enterprises. In terms of the company size, 35 percent of the companies were small, with the rest as medium-sized. Two forms of funding were provided, namely, appropriation and interest subsidy. Statistics by November 19, 2001 show that among the 10,338 applications 8,830 have been assessed and, after examination and evaluation by both technological and economic experts, 2,376 projects have been approved jointly by the Ministry of Science and Technology and the Ministry of Finance, with a total of US$219.6 million of funds being awarded, averaging approximately US$9,000 per project. The

The Chinese Economy after WTO Accession

166

Table 9.4

Ways to Use Innovation Funds for Small Technology-based Firms Time

Deductible Interest loan

Interestfree loan

Number of projects Sum (in $US million) Number of projects

Q1 54

Year 1999 Q2 Q3 71 84

5.84

6.50

71

134

Year 2000 Q2 Q3 32 42

Total

Q4 122

Q1 33

7.96

12.01

3.35

3.38

4.25

43.29

193

359

155

198

156

1266

438

Sum (in $US 6.63 10.67 18.08 31.78 13.88 17.63 14.41 113.08 million) Source: IFSTF Management Center, the Ministry of Science and Technology, PRC.

projects with appropriation account for 75.46 percent while the projects with loan interest subsidy take up 24.54 percent. To achieve the goals as designed, a new operational mode of government fund was explored and implemented with the principles of openness, fairness and transparency. First, a high-level expert committee, consisting of 17 well-known experts in the ﬁelds of technology, economy or business management, has been established to ensure the full implementation of the government policy. Then, an Expert Databank for IFSTF Project Appraisal has been established for the supervision and assessment of the awarded projects to ensure the quality of the IFSTF Projects. The qualiﬁed experts must possess not only excellent professional knowledge, but also a high professional morality and a profound understanding of the national industrial policies. After that, intermediary organizations are involved in carrying out the project appraisal. It has shown that this practice has not only upgraded the project quality but also provided a good example in forming a market-oriented project appraisal system. The Ministry of Science and Technology and the Ministry of Finance ratify the qualiﬁed intermediary organizations jointly. Still after this, measures are taken to strengthen the post-supervision of the awarded projects, including fund appropriation by stages, strict supervision and check before acceptance. The supervision system has been established with a view to facilitate a secured and effective operation of the fund. Successful projects will be encouraged to apply for further funding. Meanwhile, the Center will delegate interval supervision to the local Science and Technology Commission and High-tech Industry Parks, apart from conducting information collection and statistical analysis on Innovation Fund Projects through electronic reports. Finally, the Innovation Fund Information Management System was built up in the Center and put into ofﬁcial functioning in August 1999 in order to regulate its operation, to improve efﬁciency and increase transparency. With the interaction between an internal project-based management and a website, this system provides online inquiry and instructions on application, processing, supervision and acceptance check of the projects.

Public Venture Capital and Its Private Strategies in China

167

IFSTF has not only provided direct funding to the development of STFs, but also facilitated the establishment of various local funds. Many local governments have set up their own innovation funds, S&T achievement commercialization funds or STF development funds. By the end of October 1999, the matching fund allocated by 36 provincial or municipal governments to support the awarded Innovation Fund Projects had reached US$134 million. Progress has also been made in attracting the investment from the ﬁnancial institutions. Four major commercial banks signed cooperation agreements with Innovation Fund in 1999 to give preferred support by way of interest subsidies to the Innovation Fund projects. Assuming a most active role in innovation in the development of new and high-tech industries, STFs in China will become an essential force to enhance the nation’s economic competitive power and the foundation for a sustainable growth of the economy. The limited record of Innovation Fund has proved that the strategy of the central government to increase the ﬁnancial support for STFs through direct government funding could not have been more timely and wiser. It is well believed that, by supporting the development of STFs, IFSTF will produce a profound impact upon pulling the domestic demand, promoting investment, and upgrading and optimizing the industrial structure. However, there is a limit to the volume of total investment and a limited number of projects. Besides, appropriation and interest-free loan projects are a little too big; the return on the capital is low, which tends to push the ‘rent-seeking’ action of the government department involved. Provincial and Municipal Public Venture Capital Funds Many municipal and provincial governments see venture capital as an economic development strategy. Though recent, the scope of sub-national involvement in encouraging VC investing is broad. According to statistics by the Management Center of IFSTF of the Ministry of Science and Technology (excluding National High and New Technology Enterprise Development Zones, and a few provinces or cities), 28 provinces or cities had set up local innovation funds with a total amount of US$169 million by the ‘three-item expenditure’ on science and technology or ﬁnancial appropriation up to the end of 2002. The local innovation funds not only made joint efforts with IFSTF in supporting projects, they chose some high-tech projects themselves as well by way of appropriation, loan interest subsidy, guarantee, equity investment and so on. Jiangsu Provincial Venture Capital Fund (JPVCF) is a good example to show how the Provincial Venture Capital Fund operates in China. The operating system of JPVCF works on three levels. Firstly, as an owner of funds, the managerial committee, including the relevant government departments, is responsible for making decisions on investment and dividend policy, it is never involved in, or intervenes with, its operation. Meanwhile, JPVCF set up the Board of Supervisors. Secondly, Jiansu High-tech Venture Capital Investment Company operates the fund, but they could not use any extra money except for commission, in order to separate the Investment

168

Table 9.5

The Chinese Economy after WTO Accession

Provincial and Municipal Venture Capital Funds in China (in $US millions)

Management Department 1 Commission of S&T, Beijing City 2 Department of S&T, Hebei Province 3 Department of S&T, Inner Mongolia Municipality 4 Department of S&T, Liaoning Province 5 Bureau of S&T, Shengyang City 6 Department of S&T, Jiling Province 7 Commission of S&T, Shanghai City 8 Department of S&T, Jiangsu Province 9 Bureau of S&T, Nanjing City 10Department of S&T, Zhejiang Province 11Bureau of S&T, Ningbo City 12Department of S&T, Fujian Province 13Bureau of S&T, Xiamen City 14Department of S&T, Jiangxi Province 15Department of S&T, Shandong Province 16Department of S&T, Henan Province 17Department of S&T, Hubei Province 18Bureau of S&T, Wuhan City 19Bureau of S&T, Guanzhou City 20Department of S&T, Guangxi 21Department of S&T, Hainan Province 22Department of S&T, Sichuan Province 23Bureau of S&T, Chengdu City 24Department of S&T, Guizhou Province 25Department of S&T, Yunnan Province 26Department of S&T, Gansu Province 27Department of S&T, Xingjiang 28Department of S&T, Qinghai Province 29Commission of S&T, Tianjing City 30Department of S&T, Shanxi Province 31Bureau of S&T, Dalian City

Fund/Setting up Time Innovation Fund /2001

Amount 6.10/year

Source Three-item expenditure

Innovation Fund /2001

1.22/year

Financial Appropriation

Innovation Fund /2000

0.98/year

Financial Appropriation

Innovation Fund /2002

7.31

Financial Appropriation

Innovation Fund /2000 Innovation Fund /2000

0.61/year 3.05/year

Financial Appropriation Financial Appropriation

Innovation Fund /2002

12.2

Financial Appropriation

Innovation Fund /2002

24.4

Three-item expenditure

Innovation Fund /2000 Innovation Fund /2002

2.44/year 13.42

Financial Appropriation Three-item expenditure

Innovation Fund /2001 0.37-0.61/year Innovation Fund /2002 14.63

Financial Appropriation Financial Appropriation

Innovation Fund /2001 Innovation Fund /2000

3.17 0.61/year

Financial Appropriation Financial Appropriation

Innovation Fund /2001

1.83/year

Financial Appropriation

Innovation Fund /2002

1.83/year

Three-item expenditure

Innovation Fund /1999

12.2

Financial Appropriation

Innovation Fund /2002 12.20/5 years Innovation Fund /2002 7.32 Innovation Fund /2002 2.44/year Innovation Fund /2002 0.06

Financial Appropriation Three-item expenditure Financial Appropriation Three-item expenditure

Innovation Fund /2002

2.44

Jointly by Finance and S&C

Seed Fund/2001 Innovation Fund /2000

1.22 1.22/year

Three-item expenditure Financial Appropriation

Innovation Fund /2000

2.44/year

Financial Appropriation

Innovation Fund /2001

0.73/year

Innovation Fund /2002 Innovation Fund /2000

0.61/year 0.61/year

Commercializing Fund for research achievements Jointly by Finance and S&C Financial Appropriation

Special Complementary 0.61/year Fund /2001 Special Complementary 0.61/year Fund /2001 Special Complementary 0.37 Fund /2001 32Bureau of S&T, Changchun City Special Complementary 30% Fund /2001 complementary

Three-item expenditure Three-item expenditure Three-item expenditure Three-item expenditure

Public Venture Capital and Its Private Strategies in China 33Department of S&T, Heilongjiang Province 34Bureau of S&T, Haerbin City 35Department of S&T, Anhui Province 36Bureau of S&T, Qingdao City 37Department of S&T, Hunan Province 38Department of S&T, Guandong Province 39Commission of S&T, Chongqing City

Special Complementary 50% Fund /2001 complementary Special Complementary 1.22/year Fund /1999 Special Complementary 0.61 Fund /2001 Special Complementary 0.73 Fund /2001 Special Complementary Complementary Fund /2001 Special Complementary 50-150% Fund /2001 complementary Special Complementary 2.44 Fund /2001

169

Financial Appropriation Financial Appropriation Three-item expenditure Financial Appropriation Three-item expenditure Three-item expenditure Three-item expenditure

Notes: 1. The ‘three-item expenditure’ on science and technology is a kind of special fund that is limited for use on research and development of science and technology and managed jointly by Ministry of S&T and Ministry of Finance, including the expenditure on trialproducing new products, the expenditure on the middle stage tests, and the expenditure on subsidies for major scientiﬁc research projects. 2. Financial appropriation is the special ﬁnancial account that is in line with the local government budget and only used for the innovation of science and technology. Source: Management Center of IFSTF, the Ministry of Science and Technology, PRC.

Company from the fund. Moreover, the investment company does not operate speciﬁc projects, instead it only transfers money to start ups selected by the project Management Company so as to separate fund investment from fund management. The secrets of the dramatic growth of high-tech industry for nearly a decade lay in the systematic mechanism of venture capital. An analogy has been made: the government sows the seeds, the company plants the trees, and the bank waters, a process in which the government plays a leading role. Jiangsu High-tech Venture Capital Company was one of the public venture capital pioneers in China. It established the cooperative relationship with eight venture capital companies, attracted more than US$243 million in venture capital portfolio investment through a total amount US$79 million core government capital and became the ﬁrst partner of IDG in 2002. The local governmental efforts will be most likely to succeed in regions that already have successful entrepreneurial ﬁrms and traditions of informal support for new ventures. In 2001 we made a questionnaire survey with about 200 high-tech companies regarding driving forces, obstacles and things that need to be done by public authorities to promote innovation and local economic growth. Of the companies receiving the questionnaire 82 percent to 90 percent responded. The most interested issue is investment climate, such as high quality academic research, human capital and the legal system. There is also a strong need for more public seed ﬁnancing and support to start-ups and SMEs. Among the most frequently occurring issues brought up by the respondents is the lack of capital for different phases in a company’s development and the limited support offered to SMEs by public authorities. It is especially pointed out that there is an increasing need of public seed ﬁnancing since private venture capital companies rarely invest in early development phases. The support could be in the form of soft loans that are repaid when the company shows

170

The Chinese Economy after WTO Accession

a proﬁt, loans guaranteed by the government, favorable credits or more support to projects not only performed at universities but also in companies. Our questionnaire results show that most entrepreneurs appreciated public ﬁnancial support, especially those that characterized the milestones of the company, such as introducing new technologies, developing new products, attracting new markets. In most cases the support was provided as grants, either at the beginning of the investment or as reimbursement after the investment was accomplished. It was also indicated, however, that such plans should be widely available and easier to access. Some of the entrepreneurs have pointed out that the policy of ﬁnancing SMEs should be completely changed: small companies have little chance to attract appropriate funds regardless of their growth potential. Therefore, adequate forms of ﬁnancing should be more easily accessible, including seed capital, speciﬁc funds for developing new products, guarantees that would enable sharing of risks between the entrepreneur and the provider of capital. Public VC should greatly facilitate accomplishment of development tasks in early stages of growth. Some of the entrepreneurs would appreciate more standardized procedures. Business Models that Combine Public VC with Private VC Public-private partnerships are at the top of many issues in high-tech industry development. When the market fails to allocate ﬁnancial resources to small hightech enterprises and disadvantaged areas – especially in developing and transitional countries, partnerships between public and private organizations are often seen as offering an innovative method with a good chance of producing the desired results. What is a public-private partnership? A good working deﬁnition would include three points. First, these partnerships involve at least one private for-proﬁt organization and at least one non-proﬁt or public organization. Second, the partners have some shared objectives for the creation of social value, often for disadvantaged small businesses and local economic development. Finally, the core partners agree to share both efforts and beneﬁts. Partnership can involve a range of partners with different rights and responsibilities, including core partners, who assume key responsibilities for the joint enterprise, and in-country partners, whose participation is necessary for successful implementation. Some partnerships play prominent roles in the governance structures of the recipients while others do not. Many public VC companies or funds have established partnerships with private VC companies. Recent estimates suggest that close to 40 percent of venture or venture-like disbursements in the United States and more than half of early-stage investments came from ‘social’ sources in 2001: those whose primary goal was not a high economic return. This activity has not been limited to the United States. Governments in dozens of countries have established signiﬁcant public venture programs.

Public Venture Capital and Its Private Strategies in China

171

Partnerships can produce innovative strategies and positive consequences for public goals, and they can create powerful mechanisms for addressing difﬁcult problems by leveraging the ideas and resources of different partners. Yet knowledge about the conditions for successful partnerships is still limited. While the precise structures of these programs vary, the efforts have been based on two shared assumptions: (1) that the private sector provides insufﬁcient capital to new small ﬁrms, at least in certain regions or industries, and (2) that the government can identify ﬁrms where investments will ultimately yield high social and/or private returns. Varying with different cultures and social circumstances, public and private VC partnerships that provide a sustainable mechanism share some characteristics. Thus a brief examination of the public VC in the United States will be a good preparation for the analysis of the models employed in China. American Models In the US, the government has been playing a very important role in the development of venture capital, though more so in the early years of VC and indirect in most cases. The most important direct US government involvement was the passage of the Small Business Investment Act of 1958 authorizing the formation of small business investment corporations (SBICs). An SBIC is a privately owned and operated small business investment company that partners with the federal government to provide venture capital to small businesses. The New Markets Venture Capital (NMVC) Program has been designed to promote economic development of, intensive technology assistance to, and the creation of wealth and job opportunities in, low-income (LI) geographic areas and among individuals living in such areas, through public-private partnerships between SBA and newly formed NMVC Companies (NMVCCs) and existing Specialized Small Business Investment Companies (SSBICs). The Small Business Technology Transfer Program (STTR) expands funding opportunities in the federal innovation R&D arena. Central to the program is expansion of the public/private sector partnership to include the joint venture opportunities for small businesses and the nation’s premier nonproﬁt research institutions. It combines the strengths of both entities by introducing entrepreneurial skills to high-tech research efforts. The technologies and products are transferred from the laboratory to the marketplace by reserving a speciﬁc percentage of federal R&D funding for award to such partnership. During the 1990s, many states are sponsoring the formation of new VC ﬁrms to spur the creation and growth of new businesses. One key beneﬁt of these venture funds is the coordination of private and public sector resources and objectives that lead to a more focused approach to technology transfer and regional economic development. David L. Barkley suggested three types of public VC in terms of organization structure and a business model:

172

The Chinese Economy after WTO Accession

1. Publicly funded and publicly managed VC funds; 2. Publicly funded and privately managed VC funds; 3. Certiﬁed capital companies. Publicly funded and managed programs are usually capitalized with state money made available through appropriations or sale of bonds. Fund management is provided by a public or quasi-public agency. Institutional objectives generally are to promote economic development and/or improve state VC markets. Privately managed programs with public funding or incentives, state capital generally is placed in private VC institutions, and/or public incentives such as tax credits are used to encourage the placement of private capital in the funds. A private party manages the fund, though government representatives may serve on the institution’s board. Investment objectives generally are to maximize the fund’s internal rate of return (IRR) subject to state restrictions on the location and type of investments. Certiﬁed Capital Companies (CAPCOs) are privately funded, privately managed institutions created by state enabling legislation. Insurance companies, in exchange for 100 percent state tax credits, provide the capital over 10 years on premium taxes paid by insurance companies. The CAPCOs place approximately 40 percent of the raised capital in zero coupon bonds as collateral on the insurance companies’ investments, thus not all certiﬁed capital raised from insurance companies is available for investments in qualiﬁed businesses. CAPCOs must invest the certiﬁed capital in speciﬁc types of qualiﬁed in-state businesses according to a speciﬁc investment schedule, for example, 50 percent of certiﬁed capital invested in qualiﬁed businesses within 5 years. CAPCO investment goals are to maximize IRR from fund investments, while meeting the state’s regulatory requirements (Barkley et al., 2001). Each structure above has its advantages and disadvantages, so how to choose an effective organization structure still remains a vital problem. Chinese Models Government capital, government department managementThe overwhelming model of public VC in China is Government Capital, Government Department Management model of which the IFSTF is typical case in point. Firstly, all of its initial and followup capital comes from the Ministry of Finance and the State Policy-oriented Banks; Secondly, the Management Center of IFSTF is an institution under the Ministry of Science and Technology. Besides, municipal and provincial governments also provide VC programs. As an advantage, this model can be so designed as to meet the objectives of developing high technology and economy. Its performance and efﬁciency depend heavily on the operating and managerial mechanism, the staff’s dedication. As a matter of fact, IFSTF has been playing a very active role since its beginning. The government has made substantial efforts to designing a new operational mode for

Public Venture Capital and Its Private Strategies in China

173

the government fund, continuously improving the administration mechanism and strengthening the post-supervision of the awarded projects. However, political consideration and pressure can lead to investments in certain industries, provinces or ﬁrms, rather than areas that can bring more proﬁt. That also makes it very difﬁcult to attract most talented fund managers, or to co-invest with private VC, it can even result in bureaucracy and political corruption. A study shows that many SBICs made investments in ineffective or corrupt ﬁrms; SBIC managers’ incentives to screen or monitor portfolio ﬁrms were greatly reduced by the presence of government guarantees that limited their exposures to unsuccessful investments. So the government still faces the dilemma: stimulating development in certain areas, regions, industries, or encouraging economic proﬁts. For IFSTF, their hard work lies in how to balance the different objectives and keep on improving the managerial mechanism. Owned and operated by state-owned enterprisesA large proportion of Chinese VC companies are state-owned or state-holding companies. This is very important in terms of organization form and is quite different from the US. For instance, the main shareholders of Tsinghua Venture Capital Co., Ltd. are High-tech Group of Tsinghua University, Capital Steel Group, accounting for about 50 percent share. The rest of the shareholders are national ﬁnancial institutions, VC funds and some industrial or commercial enterprises. With another VC company, Tsinghua Unisplendour HiTech Venture Capita, most shareholders are state-owned enterprises or state holding listed companies, such as Tsinghua Unisplendour Corporation, Sichuan Investment Group, Beike Medical Industry Corporation, China Shipping Haisheng Co., Ltd. Beijing Yanjing Beer Co., Ltd, with average of 8 percent share, except for Tsinghua Unisplendour Corporation with 16 percent. It must be recognized that this is common VC organization form, since venture capital is still young and basic in China, following the state trust investment company model set up the VC investment companies. On the other hand, it was hard for stateowned VC institutions to take a favorable form, because ‘Investment Company’ is the only legal form for venture capital, especially limited liability companies. However, the nature of venture capital lies in the high return due to the high risk, and the investors expect to get the return as soon as possible. These two things keep the private investors reluctant to co-invest with state owned enterprises. Therefore, the ‘Investment Company’ form may not be a rule of thumb, though it has presently been adopted. Owned and operated by private companyThere are a quite limited number of private VC companies in China. Wanxiang Venture Capital Company is a representative of this type. It was established jointly by Wanxiang Group Corporation with shares of 70.28 percent, Zhejiang Province Department of Finance with shares of 10 percent, Shenzhen Wanxiang Investment Company with shares of 8.33 percent, Zhejian Oriental Communication Company with shares of 3.33 percent, Shanghai Linyu Automobile E-Business Ltd. with shares of 3.33 percent, Hangzhou City

174

The Chinese Economy after WTO Accession

Financial Development Company with shares of 1.67 percent, Xiaoshan State Assets Management company with shares of 1.67 percent, Venture Equities Management, Inc. (US) with shares of 0.84 percent and Power drive Europe Ltd. (UK) with shares of 0.84 percent in November 2000. The registered capital was 300 million RMB (US$36.58 million). So it is really a private holding VC company with government participation, the private partner plays the main role in decision-making. However, this kind of model is not prevailing because, except for some listed companies held by private groups, most individual investors always worry about the high risk of venture capital, and they also lack adequate knowledge about VC. Reﬁnancing modelThe young ﬁrms might need to resort to the capital market for the second round of funds for their expansion. They face the need and pressure to grow on the one hand, but there is also the constraint of capital resource on the other. The government then may provide a springboard for the ﬁrms to have more access to the needed capital resources by means of equity investment, loan guarantee and loan subsidies. Co-investment modelThe governments play their roles also by setting up matching funds with private VC to ﬁnance for the small businesses. In so doing, the governments share the risk with private VC and have reduced the risk of the projects. Meanwhile, the governments have also achieved the objective and strategy of using a certain amount of capital to encourage more activity of the private capital, and therefore guiding and orienting the private capital. This role of providing matching funds or co-investing has been quite popular in China. However, there are some limitations on the effects of state-funded VC. First, these funds often struggle to generate private funds for current or later rounds. Second, the number of publicly funded investments is still small and, therefore, has a limited impact. Finally, research has not been conclusive as to the effectiveness of publicly funded VC. In fact, some nontraditional VC institutions have been failures both in terms of ﬁnancial returns and economic development impacts. Discussion and Suggestions Chinese government should play a more important role in VC development It is very difﬁcult to deﬁne the role of the government in VC development in China, though from the historical perspective, government policy has produced an extraordinarily strong effect on the general health of VC business. However, government involvement always faces a dilemma. On the one hand, some entrepreneurs acclaim the government’s involvement, mostly in making VC more locally accessible. The most common model was a guarantee similar to SBA loans, where government mitigates some of the ﬁnancial risk associated with equity investment. Most other sources also believed that government has a ﬁnancial role

Public Venture Capital and Its Private Strategies in China

175

in entrepreneurship. Public VC programs can help change the local entrepreneurial culture and attract more investment talent to a region. A venture capitalist at St. Paul Venture Capital stated that the best role for the government is to put in place the right securities and regulatory incentives, like low capital gains taxes, that encourage risk taking by individuals. In fact, many claim the growth seeds of the VC industry were planted about 20 years ago – when there was a big capital gains cut and a federal clariﬁcation regarding VC investments by pensions. On the other hand, many entrepreneurs still favor a hands-off approach in principle: they prefer less government involvement in the VC development and less restrictions with as few roadblocks as possible. China is still in the process of transition from planned economy to market economy, and the distribution of resources depends on both government department and market mechanism. Since the market has not been fully established in China, the role of the government has become more prominent, and it should be a more important role in China than that in the United States. Meanwhile, China is a bank-based ﬁnancial system while the capital market is still in its primitive stage of development. The share of market capitalization in GDP is only 37.43 percent, still very low compared with developed countries. By contrast, the ratio of total loans to GDP in 2002 was 125.62 percent, suggesting the primary roles of banks and other ﬁnancial institutions in ﬁnancing for small hightech businesses. According to modern comparative ﬁnancial theory, capital market is more favorable than a bank system to VC development. Therefore, Chinese government faces the issue of how to improve the capital market. The relationship between ﬁnancial development and VC development, and the presence of signiﬁcant differences in the organization of ﬁnancial systems raise a series of important questions. For instance, how do the ﬁnancial institutions and markets play their roles in VC investment? What are the differences between the development patterns followed by China and the United States respectively? Is it possible to determine the types of institutions that are more conducive to VC development? How much can the investment boom be accounted for by the ability of efﬁcient and well-organized ﬁnancial markets to channel funds toward the most innovative and high returning projects in the United States? These issues have attracted great academic interest in recent years, since, ﬁrst of all, the ﬁnancial markets have played a critical role in supporting and stimulating the economic growth of the US in recent years, and secondly, animportant transformation has been taking place in Chinese ﬁnancial markets. Therefore, it is safe to conclude that Chinese government should play a more important role and assume a greater responsibility than that played by the government in the United States in the development of VC industry and in nurturing entrepreneurship. Indeed, the central and local governments in China have become intimately involved with economic development.

176

The Chinese Economy after WTO Accession

Public VC can ﬁll in the gap between the demands for and supplies of venture capital and help improve the investment climate for private VC, which may be much more important than the ﬁnancial transfer from government in China The ability to generate and sustain high-tech companies and employment is dependent on three general factors: equity investment capacity, human resources and a supply of technologies and business ideas. Investment capacity is clearly the key factor. Moreover, entrepreneurial region needs a culture of entrepreneurship and innovation, because an entrepreneurial climate is a cultural, social, and economic milieu that encourages and nurtures the creation of new business ventures (US National Commission on Entrepreneurship, 2002). In fact, any government committed to sustained economic progress must ensure that all aspects of its economic system are conducive to and supportive of increased levels of entrepreneurial activity. This includes minimizing taxation, ensuring access to labor, lowering non-wage labor costs, deregulating and making it easier to do business with the government. Although the government intends to encourage VC development, and an environment with a good innovative climate for the commercialization has been evolving in China, many of its laws discourage small ﬁrm formation. These include regulations on the minimum required capital of RMBҰ10 million to be considered a company with the exception of technology-based ﬁrms for which only RMBҰ100,000 is required. There are other legal provisions that by commission or omission retard entrepreneurship and thereby prohibit the formation of an active VC industry. One of the most signiﬁcant of these is the restriction on disposal of shares in the open market after an IPO by the investors and founders. In addition, many regulations negatively affect venture capital in China. For example, Company Law Article 24 provides that, when a company invests in other limited liability companies or limited companies, the aggregate amount of investment shall not exceed 50 percent of the net assets of the company, not including any capital increase of the other limited liability companies or limited companies arising from any conversion of proﬁts into capital following such investment, except for investment companies and holding companies speciﬁed by the State Council. Article 147 states that shares of a company held by a promoter of that company shall not be transferred within three years of the company’s establishment. Directors, supervisors and the managers shall report all the shares that they hold in the company, and shall not transfer them during their term of ofﬁce. Improving the environment for VC will require rethinking some of the central tenets of the way the central government controls the economic environment. For example, simplifying the environment for foreign venture capitalists would also be part and parcel of simplifying the environment for foreign investors in general. Allowing stock option plans for start-ups would likely be part of a more general creation of stock option plans. Similarly, removing restrictions on share sales by insiders would almost surely require an improved and more transparent stock regulatory system modeled after the US SEC. Thus, improving the environment

Public Venture Capital and Its Private Strategies in China

177

for VC also implies decisions about the future direction of the economy and the industrial structure. Besides the issues in government’s policies and regulations, there are many nongovernmental concerns. Some entrepreneurs have a very short-term outlook and tend to treat the venture capitalists’ investments as a ‘gift,’ thereby exposing the investor to a classic ‘principal-agent’ problem. This is not entirely surprising, as it will require some time before Chinese entrepreneurs are educated to the behavior patterns of a ‘bankable’ entrepreneur – but for investors this is an immediate problem. China is also in dire need of experienced professional managers, which leads to enormous difﬁculty in assembling a management team. How to make a strategy that integrates the public VC and private VC in China? There has always been a question of how to keep the appropriate balance between the free market and intervention by public authorities. The experiences of the United States before the 1970s show that, in contrast to the poor performance of private VC, public VC did well. It is clear that increasing the support system to start-up companies would promote growth of the high-tech industry. But since then, private VC has developed dramatically. So both private and public venture capitals have been playing important roles in ﬁnancing high-tech businesses with different strengths and limitations. The current public venture program is a mechanism to distribute the limited source of capital, but a more important issue is how to establish a fund-raising mechanism. Statistics show that non-government investment accounts for about 60 percent of total R&D expenditure of $14.16 million, suggesting that private capitalists are willing to invest more in high-tech R&D. It is our recommendation that there be different business models integrating public VC with private VC following a three-step process. First step, public VC plays the principal part of VC industry. As discussed, no individual or private companies would like to invest without a sound environment, which includes law and regulations, professional managers, exit channels and so forth. Therefore, public VC should play its roles in ﬁlling the capital gap, attracting foreign investment, training professional investor, promoting managerial system and improving legal environment. After that, an integration of public and private VC in different forms should be encouraged. The key issue is to design an efﬁcient governance structure, regardless of the model. Public VC may focus on the professional manager training while private VC may involve strategy and management. There is a long way for the Chinese VC industry to go, because government investments always are limited while the private companies and social capital have enough room to expand; there is an outstanding amount of savings deposited in urban and rural areas, close to US$900 billion. With the improvement of the investment environment and more professional managers with richer experiences and even greater achievements, an increasing number of individuals

178

The Chinese Economy after WTO Accession

and private companies will join the VC fund, and public-private relationship will be expected to deepen. Finally, private VC takes over and becomes the principal part of VC industry. Public VC functions as a tool for supplement and adjustment, and thus private VC plays a leading role in the capital market and development of the high-tech industry in China. References Abbott, Charles C. and Eugene M. Zuckert (1941), ‘Venture Capital and Taxation,’ Quarterly Journal of Economics, Vol. 55, No. 4, pp. 667–82. Backlund, Anna et al. (2000), The Swedish biotechnology innovation system, Working Paper, May. Barkley, David L., Deborah M. Markley and Julia Sass Rubin (1999), Public Involvement in Venture Capital Funds: Lessons from Three Program Alternatives, Rural Policy Research Institute, PB99-2, November. Barkley, David L., Deborah M. Markley, David Freshwater, Julia Sass Rubin and Ron Shaffer (2001), Establishing Nontraditional Venture Capital Institutions: Lessons Learned. Economic Development Quarterly. Becker, G.S. (1983), A Theory of Competition Among Pressure Groups for Political Inﬂuence,’ Quarterly Journal of Economics, Vol. 98, No. 3 (August), pp. 371– 400. Bergemann, Dirk and Ulrich Hege (1998), ‘Dynamic Venture Capital Financing, Learning and Moral Harzard,’ Journal of Banking and Finance, Vol. 22, pp. 703– 35. Bundesministerium der Finanzen (2001), ‘Einkommensteuerliche Behandlung von Venture Capital Funds und Private Equity Funds: Abgrenzung der private Vermögensverwaltung vom Gewerbebetrieb,’ Erlassentwurf – Geschäftszeichen IV A 6–S 2240–0/01_II, Stand: November. Canadian Venture Exchange Annual Report (2001). Casper, Steven (1999), High Technology Governance and Institutional Adaptiveness. Do technology policies usefully promote commercial innovation within the German biotechnology industry? June. Cohen, Linda and Roger Noll (1991), The Technology Park Barrel, The Brookings Institution, Washington, DC. Commission Staff Working Paper (2000), Trends in European innovation policy and the climate for innovation in the Union Brussels, xxx SEC 1564. Eisinger, Peter (1988), The Rise of the Entrepreneurial State. State and Local Economic Development Policy in the US, Wisconsin UP, Madison. Etzkowitz, Henry (2002), ‘The Triple Helix of University, Industry, Government: Implications for Policy and Evaluation,’ Science Policy Institute Working Paper 2002–11.

Public Venture Capital and Its Private Strategies in China

179

Etzkowitz, Henry, Magnus Gulbrandsen and Janet Levitt (2001), Public Venture Capital: Sources of Government Funding for Technology Entrepreneurs, 2nd edition, Aspen/Kluwer, New York. Feldman, Maryann P. and Maryellen R. Kelley, ‘How States Augment The Capabilities of Technology-Pioneering Firms,’ working paper. Galante, Steven P. (2001), ‘An Overview of the Venture Capital Industry & Emerging Changes,’ The Private Equity Analyst, newsletter, Wellesley, MA. Gentry, William M. and R. Glenn Hubbard (2000), ‘Tax Policy and Entrepreneurial Entry.’ American Economic Review, Vol. 90, No. 2, pp. 283–87. Gompers, Paul A. and Josh Lerner (1999), The Venture Capital Cycle, MIT Press, Cambridge. Hall, B.H. (1992), ‘Investment and Research and Development: Does the Source of Financing Matter?’ Working Paper No. 92-194, Department of Economics, University of California at Berkeley. Hall, Wendy, Lynn Foster and Rohit Shukla (2002), ‘Public venture, public gain?’ November 9, Larta University’s course on winning federal R&D funds. Hao, K.Y., and A.B. Jaffe (1993), ‘Effect of liquidity on ﬁrms’ R&D spending,’ Economics of Innovation and New Technology, Vol. 2, pp. 275–82. High-Tech Business Job Growth (2001), A White Paper prepared for the Wisconsin Economic Summit held November 29–December 1, Milwaukee, Wisconsin. Himmelberg, C.P., and B.C. Petersen (1994), ‘R&D and internal ﬁnance: A panel study of small ﬁrms in high-tech industries,’ Review of Economics and Statistics, Vol. 76, pp. 38–51. Horvath, Michael (1999), ‘U.S. Venture Capital Flows: Empirical Evidence and Implications,’ Stanford University, November. Hubbard, R.G. (1998), ‘Capital-market imperfections and investment,’ Journal of Economic Literature, Vol. 36, pp. 193–225. Kenney, Martin, Kyonghee Han and Shoko Tanaka (2002), ‘Scattering Geese: The Venture Capital Industries of East Asia - A Report to the World Bank,’ BRIE Working Paper 146. Keuschnigg, Christian and Søren Bo Nielsen (2000), ‘Tax Policy, Venture Capital, and Entrepreneurship.’ NBER Working Paper No. 7976, October (also: CEPR Discussion Paper No. 2626, November). Keuschnigg, Christian (2002), ‘Start-ups, venture capitalists and the capital gains tax.’ CEPR Discussion Paper No. 3263. Lerner, Josh (1999), ‘The Government as Venture Capitalist: The Long-run Effects of the SBIR Program,’ Journal of Business, Vol. 72, pp. 285–318. Lerner, Josh (2002), ‘When Bureaucrats Meet Entrepreneurs: The Design of Successful ‘Public Venture Capital’ Programs,’ The Economic Journal, Vol. 112, No. 477, F73–F84. Lo, Clarence (2000), ‘The Problem with Venture Capital in China,’ Virtual China Inc.

180

The Chinese Economy after WTO Accession

Markley, Deborah M., David L. Barkley, Julia Sass Rubin, David Freshwater and Ron Shaffer (2001), ‘RUPRI Rural Equity Capital Initiative: Case Studies of Nontraditional Venture Capital Institutions,’ P2001-11D, Part 4 of 4 of the Final Report. Megginson, William L. (2002), ‘Towards a Global Model of Venture capital?’ Journal of Applied Corporate Finance, Fall. Olson, M. (1965), The Logic of Collective Action: Public Goods and the Theory of Groups, Harvard University Press, Cambridge, MA. Peltzman, S. (1976), ‘Toward a More General Theory of Regulation,’ Journal of Law and Economics, Vol. 19, pp. 211–48. Poterba, James M. (1989), ‘Venture Capital and Capital Gains Taxation,’ NBER Working Paper No. 2832, July. Repillo, Rafael and Javier Suarez (1998), Venture Capital Finance: A Security Design Approach, CEMFT, Madrid, Mimeo. Robinson, Peter (2001), ‘Public Private Partnerships: The Evolving British Debate,’ Working Papers 7(e)/2001. Stigler, G. J. (1971), ‘The Theory of Economic Regulation,’ Bell Journal of Economic and Management Science, Vol. 2, pp. 3–21. Testa, Hurwitz & Thibeault (2000), ‘The Macro View: Can the Public Participate in Private Equity?’ http://www.tht.com/pubs/SearchMatchPub.asp?ArticleID=308. US National Commission on Entrepreneurship Annual Report (2002). US National Venture Capital Association (2000), National Venture Capital Association Yearbook 2000, Copyright 2000 Venture Economics. Wallsten, Scott (2000), ‘The Effects of Government-Industry R&D Programs on Private R&D: The Case of the Small Business Innovation Research Program,’ RAND Journal of Economics, Vol. 31, No. 1, pp. 82–100. Willmott, H., D. Grimshaw and S. Vincent (2001), ‘New Control Modes and Emergent Organizational Forms: Private–Public Contracting in Public Administration,’ Administrative Theory and Praxis, Vol. 23, No. 3: pp. 407–30.

Chapter 10

Financial Development and Urban-Rural Income Disparity in China Qi Zhang, Mingxing Liu, Yiu-Por Chen and Ran Tao

Introduction The relationship between ﬁnancial development and economic growth has received an increasing amount of attention and has induced a lively debate from economists. Most students of the subject believe in long run ﬁnancial system beneﬁts, economic growth through mobilizing and pooling savings, diversifying investment risks, screening investment projects, facilitating exchange, monitoring managers and exerting corporate control, and so on. Empirical evidences also support the captioned arguments. (Goldsmith, 1969; McKinnon,1973; Shaw, 1969; Stiglitz, 1985; Mayer, 1990; King and Levine, 1993a, 1993b; Beck and Levine, 2002.) However, relatively few studies focus on the relationship between ﬁnancial development and income distribution. Greenwood and Jovanovic’s (1990) paper is the ﬁrst to explore the association among economic growth, ﬁnancial development and income distribution, where income distribution is treated as exogenous. A recent study by Clarke, Xu and Zou (2003) explores how ﬁnancial development will inﬂuence income distribution by using cross-country data. They ﬁnd ﬁnancial development robustly reduces the level of income inequality. In this chapter we use a panel data of China’s 28 provinces for the period of 1978–98 to analyze the effect of ﬁnancial intermediation development on urban-rural income inequality. Compared to cross-country analysis, it has several advantages. First, the difference in institutions, cultures and legal systems among different countries is difﬁcult to control for in cross-country comparison. In addition, problems in data collecting and processing methodology make income distribution comparison among different countries less credible in an international setting (Wei and Wu, 2001; Atkinson and Brandolini, 2001). Second, China’s urban-rural income inequality constitutes the most signiﬁcant component of overall income inequality, according to some recent studies. Tsui (1993) decomposed regional disparity into ﬁve parts, namely within-province disparity, between-province disparity, within-countryside-area disparity, withinurban-area disparity and urban-rural disparity. He found urban-rural disparity has played a leading role in the enlargement of regional disparity. The World Bank

182

The Chinese Economy after WTO Accession

(1997), based on data in 1995, concluded that urban-rural income disparity accounts for no less than half of overall income disparity. By using the Theil Decomposition Method, Lin et al. (1998) also found that urban-rural per capita income disparity explained at least ﬁfty percent of overall regional disparity, while the within-rural disparity and within-urban-area disparity accounted for the other ﬁfty percent. Our chapter ﬁnds that the ﬁnancial development index (measured as the ratio of total loans extended by ﬁnancial intermediation to GDP) contributes signiﬁcantly to the enlargement of urban-rural income disparity since the late 1980s, after controlling for other factors such as provincial infrastructure, institutional transition in rural areas, and degree of international integration, and other related variables. The major driving force of the URID is the growing government reliance on the ﬁnancial system to exert control on its economy since late 1980s, whereas before the late 1980s ﬁscal policy was the principal policy instrument. In addition, we show the effect of provincial ﬁnancial intermediation is robust to the changes of sector structure in that province. The chapter is organized as follows: The next section describes the pattern of urban–rural disparity and ﬁnancial intermediation development for the period 1978– 99. Based on the pattern, we explore the relationship between Chinese ﬁnancial intermediation development and urban-rural income disparity; The third section sets up an econometric model and describes the data. The methods used in econometric estimation are introduced in the fourth section. The estimation results are presented and discussed in the ﬁfth section. The ﬁnal section provides a brief conclusion. How Financial Development Affects Urban-Rural Income Disparity Despite China’s impressive achievement in the overall economic growth rate over last two decades, averaging nearly 10 percent since the early 1980s, the income inequality, especially urban-rural income inequality, has been a pressing issue and has been more so in recent years. In Figure 10.1, we present the evolution of urbanrural income disparity (measured as the ratio of disposal income of urban residents to net income of rural residents) for the period of 1980–98. It is clear that urban-rural income disparity (URID) declined in the early 1980s, and has increased since the late 1980s. In fact, in many provinces, URID in the late 1990s even exceeds that in the late 1970s when market reform was initiated. For instance, Beijing’s URID rose from 1.63 in 1978 to 2.11 in 1998. In the central province of Anhui, the gap rose from 1.72 to 2.56 during the same period, while in the northeast province of Jinlin, it rose from 0.97 to 1.76. A cross province comparison shows that there is a negative correlation between disparity and per capita income (measured in 1978 price, Figure 10.2). These facts were also noted by other authors (Hu et al., 1995; Houkai Wei, 1997; Zhang, 2000). The primary goal of this chapter is to see what role ﬁnancial intermediation has played in the raising of URID, which so far has received little attention. Although many researchers (Shang-Jin Wei, 1997) point out there is signiﬁcant urban-bias of

Financial Development and Urban-Rural Income Disparity in China

183

Figure 10.1 Urban-Rural Income Inequality Across Provinces (1978–1998) Note: URID is measured as the ratio of urban disposable income per capita to the rural per capita net income per capita. Source: Compilation of Statistic Data of New China from 1978–1999.

the Chinese ﬁnancial system, the validity of such an argument needs to be explored further: First, the urban-bias perspective cannot fully explain why there is a turning point between the early-mid and late 1980s, since urban-bias policies pursued by the Chinese authority did not experience profound change over the last two decades. Second, the channels through which the urban-bias of the ﬁnancial system exerts its impact on URID needs to be further analyzed. In this chapter, it is argued that to fully understand the relationship between URID and ﬁnancial development in China, several things need to be considered, including the characters of ﬁnancial structure, the change of instruments by the government to regulate the ﬁnancial system and the role of the ﬁnancial system as a

Figure 10.2 Urban-Rural Income Inequality and Per Capita Income (Full Sample, 1978–1999, Average Values) Source: Compilation of Statistic Data of New China from 1978–99.

184

The Chinese Economy after WTO Accession

vehicle to support State-Owned Enterprises (SOEs), and the impacts of regulations of government on rural economy in China. Characteristics of Financial Structure in China China has a highly concentrated bank-based ﬁnancial structure, which goes against ﬁnancing conditions of small and medium scale farmer households and that of small and medium scale Township-and-Village enterprises (TVEs). First, the securities market in China is relatively small and underdeveloped. For example, although the growth of the securities market has exceeded the growth of ﬁnancial intermediation during the whole of the 1990s (Table 10.1), China still maintains a bank-based ﬁnancial structure in which four State-owned Commercial Banks (SCBs) together account for two-thirds of ﬁnancial assets and provide funds of more than 6 times the securities market. Second, the credit market in China is dominated by the four large SCBs (Table 10.1). Among the four large SCBs, only the Agricultural Bank of China (ABC) has the literal responsibility to provide loans to the agricultural sector, but in practice farmers cannot get ﬁnancial support directly from ABC. It is the Rural Credit Cooperatives (RCCs), other than ABC, that deal with farmers’ loans. We will go back to this issue. Given the monopolized banking market structure, it is easy for the government to direct credits to large SOEs (Zhang, 2002).1 In addition, we argue that largescale banks generally incur relatively higher transaction costs and information costs than small-and-medium banks in the case of extending loans to small and medium enterprises (SMEs) (Levonian and Soller, 1995; Berger and Udell, 1995; Peek and Rosengren, 1996; Strahan and Weston, 1996, 1998). Empirical researchers also ﬁnd that large banks have no comparative advantage in providing funds to SMEs (Lin, Zhang and Liu, 2003; Meyer, 1998). Given the nature of agricultural activities, the problem of asymmetry information between farmers and ﬁnancial intermediation is severe. In such cases, large banks are unwilling to provide loans to farmers or TVEs, which are both of normally medium or small scale and lack the necessary and traceable loan history. In contrast, large enterprises, which usually locate in urban areas, are more easily able to get ﬁnancial resources from large banks. It means that even if the regulation of government on ﬁnancial intermediation were removed, urban-bias of ﬁnancial intermediation in allocating credits would not disappear automatically.

1 China’s highly monopolized bank-based ﬁnancial structure has a long history. From 1949 to 1952, China nationalized the whole banking industry and canceled its ﬁnancial market. All banks were merged into the People’s bank of China, the only bank left in Mainland China, so that credit allocation was highly centralized and was easily assigned to priority sectors according to national development strategy. Reform of the banking industry in the early 1980s led to the establishment of the four state-owned commercials banks which have dominated the ﬁnancial intermediation market since then. A highly monopolized banking market structure makes it easy for the government to have control over the banking industry (Zhang, 2002).

Financial Development and Urban-Rural Income Disparity in China

Table 10.1 Variable

Financial Development Comparison Across Countries (%) Country

Ratio of M2 China to GDP Japan Korea U.S.A. High Income Countries Low & Middle Income Countries Ratio of Share Traded Value to GDP

Ratio of Credit to Private Sector to GDP

Ratio of Credit by Banks to GDP

Ratio of Capital Value of Shares to GDP

185

China Japan Korea U.S.A. High Income Countries Low & Middle Income Countries China Japan Korea U.S.A. High Income Countries Low & Middle Income Countries China Japan Korea U.S.A. High Income Countries Low & Middle Income Countries China Japan Korea U.S.A. High Income Countries Low & Middle Income Countries

1980

1990

1993

2000

70.29 126.35 35.61 63.98 -

Growth Rate (1980– 1990) 7.79 5.49 1.84 0.82 -

86.63 105.34 37.56 60.52 -

143.73 120.02 71.78 60.09 -

Growth Rate (1993– 2000) 7.50 1.88 9.69 -0.10 -

33.19 74.02 29.65 58.99 26.59

35.35

2.89

41.78

59.82

5.26

-

-

-

10.05 21.81 61.24 50.96 32.82

66.81 55.64 121.33 323.89 181.07

31.07 14.31 10.26 30.23 27.63

-

-

-

19.65

37.02

9.47

53.44 131.08 52.04 78.46 78.4

87.71 195.19 65.53 93.09 107.83

5.07 4.06 2.33 1.72 -

99.81 200.97 64.35 95.07 115.31

124.57 187.73 101.95 143.53 136.31

3.22 -0.96 6.79 6.06 2.41

31.67

41.65

2.77

48.6

55.34

1.87

53.62 188.11 56.47 94.44 -

89.98 259.66 65.65 110.91 -

5.31 3.27 1.51 1.62 -

103.23 272.42 65.31 118.45 147.5

132.73 310.45 104 161.72 175.15

3.65 1.88 6.87 4.54 2.48

55.27

-

-

65.88

68.65

0.59

-

-

-

9.4 68.56 40.33 78.02 63.85

53.8 65.21 32.51 153.54 120.61

28.30 -0.71 -3.03 10.15 9.51

-

-

-

34.96

38.72

1.47

Source: The World Bank (2001).

The Chinese Economy after WTO Accession

186

Table 10.2

Industrial and Commercial Bank of China Agricultural Bank of China Bank of China Construction Bank of China Total

How Do the Four State-Owned Commercial Banks Matter? Banking Market Structure in China (%) The proportion of asset value in total asset value of banks in China

The proportion of proﬁt value in total proﬁt value of banks in China 1997 11.04

The proportion of deposit in total deposit of banks in China 1996 1997 27.37 27.32

The proportion of loans in total loans of banks in China 1996 1997 28.03 26.63

1994 34.18

1996 34.59

1997 34.13

1994 19.4

1996 12.73

16.26

13.98

13.69

2.41

10.25

2.84

13.11

13.47

13.34

13.09

23.85

20.08

19.04

24.82

25.36

21.25

18.02

16.7

16.54

15.05

18.13

20.26

26.33

12.8

10.59

6.79

15.39

15.89

14.22

14.80

92.42

88.92

93.19

59.43

58.93

41.92

73.89

73.38

72.13

69.57

Source: Reforming Chinese Financial System: Prospects and Outlook, Working Paper Series, China Center for Economic Research, Peking University, No. 2000005.

Government Intervention: Budget Funds vs. Bank Credits Given a highly monopolized banking market structure, along with declining budgetary capacity since the 1980s, government relies more and more on its control of the ﬁnancial system, including maintaining an interest ceiling, providing policy loans, introducing quota systems in the securities market (abolished in 2001), etc, to realize its policy goals and more generally to support SOEs. At the beginning of the 1980s, the level of ﬁnancial development in China was relatively low. Bank deposit was the only ﬁnancial asset. Its level of ﬁnancial depth, in terms of the ratio of M2 to GDP, was a little more than that of low-middle income countries (Table 10.1). At that time, government had no other choice but to rely mainly on ﬁscal funds to promote reform, such as raising the prices of principal agricultural products in 1979, increasing salaries and subsidies to urban residents, and so on; and to provide low-cost or even no-cost resources to SOEs. However, the situation has undergone great changes since the late 1980s. After implementing a series of reform policies in ‘delegating more decision making power to and sharing proﬁts with enterprises’ in the 1980s, the government gradually lost its ﬁscal power in resource allocation. However, the ﬁnancial system is still under tight regulation and is controlled by the state as a means to provide ﬁnancial support

Financial Development and Urban-Rural Income Disparity in China

187

to large SOEs,2 most of them are capital-intensive and large-scale ﬁrms.3 In addition, with government ﬁscal capacity declining since the beginning of the reform, cheap banking loans (and also funds from the equity market in the late 1990s) became the principal tool of the cash-scrapped government to support state-owned enterprises. Figure 10.3 shows that during the 1980s there was a positive correlation between the loans by ﬁnancial intermediation and per capita GDP (FINDEV) at province level, which is consistent with the results demonstrated in literature in the context of cross country comparison. It is during the 1990s that the credit lending level in the lowest income provinces began to rise. It shows that in the relatively underdeveloped provinces, ﬁnancial intermediation was used as the primary tool by local governments to intervene in the economy during the 1990s. Because more funds were directed by government to meet the ﬁnancial demand of preferential enterprises, it is obvious that farmers and TVEs were more ﬁnancially constrained,4 which would be a detriment to URID. Regulations of Government on Rural Economy and Rural Financial System Relying on the four Stated-owned Commercial Banks, which dominate the banking market, the government successfully directs most credits to large SOEs. Although there are formal ﬁnancial agencies in rural areas that are responsible for providing ﬁnancial services to farmers (such as ABC and RCCs), however, the interventions of and the regulations imposed by government on the rural economy make those ﬁnancial agencies have no incentive to, or are inefﬁcient to, provide funds to farmers.5 2 According to IFC report (2000), in the period from 1991–97, the share of investment in the national total was the range of 15–27 per cent, with little resource to formal bank loans (less than 1 per cent of working capital loans went to private sectors). In addition, private ﬁrms’ access to the equity market is also very limited by size requirement and quota system. The IFC (2000) reports that, of the 976 companies listed on the Shanghai and Shenzhen Stock Exchanges, only 11 are non-state ﬁrms, while in 1998 and 1999 only 4 non-state IPOs took place. 3 For example, the economy and trade commission (ETCC) was set up in the early 1990s to cooperate with the National Planning Commission of China (NPCC) and the Ministry of Finance of China (MFC) to implement industrial policies. One most famous policy is called ‘guarantee big ﬁrms and let alone tiny ﬁrms’. Its key goal is to guarantee the development of 500 super-big ﬁrms, which were chosen by ETCC and NPCC. 4 One survey conducted by the Ministry of Agriculture about private enterprises ﬁnancing in Wujin, a city in Jiangsu Province, indicated that 45.1 percent of the enterprises with annual revenues of more than 5 million yuan asserted that it was difﬁcult for them to get loans, compared with 86.5 percent of the enterprises with annual revenues of less than 5 million yuan (China Economic Times, 2002). In their survey, Chen (2002b) and Liu (2002b) conclude that farmers get less ﬁnancial support from rural formal ﬁnancial agencies since the mid 1990s and are deadly ﬁnancially constrained. 5 Song (2000), for instance, pointed out the average growth rate of total agricultural loans as a whole was less than two percent per year. The ratio of newly added loans to total

The Chinese Economy after WTO Accession

Financial Intermediate Development

188

Per Capita Income at Province Level FINDEV

Fitted Values

Figure 10.3 Economic Development and Financial Intermediation Development Note: Financial Intermediation Development Index is measured as the ratio of loans of ﬁnancial intermediation to GDP in a Province. Source: Compilation of Statistic Data of New China from 1978–99.

First, although market reform in rural area has made great progress, many regulations imposed by the central government on rural society and economy are still maintained. Many of those regulations, such as compulsory procurement of the grain, family planning and nine-year compulsory education, etc, not only add to the loans was also less than 10 percent. In his survey, Chen (2002b) found that not only basic ﬁnancial services for rural residents dropped, but funds for productive and commercial activities were less than before. Moreover, after state-owned commercial banks retreated from the countryside in late 1990s, RCCs alone as the only formal rural ﬁnancial agency could not fully take the responsibility of providing ﬁnancial services for rural development due to its massive bad loan problem, accumulated over the last two decades. In Liu’s (2002b) survey on county ﬁnance, due to institutional problems rural ﬁnancial agencies were obsessed by huge bad loans and tremendous losses. He concluded because credit policies were tightened after stated-owned banks were commercialized, it would be more difﬁcult for local economic activities at county level to get ﬁnancial support from rural formal ﬁnancial agencies.

Financial Development and Urban-Rural Income Disparity in China

189

burdens of peasants, but keep peasants from exploiting their comparative advantage by adjusting production structure (Tao, Liu and Zhang, 2003). In consequence the low income from agricultural activities is inevitable. To be more speciﬁc, the policy trends prior to the mid and late 1980s in the countryside, such as abolishing Commune system, promoting Household Responsibility system (HRS), giving farmers more autonomy to manage its production, marketizing the prices of a lot of agricultural products (except for grain and cotton), opening product market and factor market, and even for a time trying to cancel compulsory procurement of grain, etc, can be viewed as market-oriented or deregulation-oriented. Accompanied with those trends was the rapid growth in rural income. But good times did not last long. Since the late 1980s those trends were gradually reversed. We can see it by tracking the change of grain policy in the 1990s: in 1990 the contract sale of grain was replaced with compulsory procurement by government so that in practice contracted sale of grain became a mandatory task for peasants; in 1994 the procurement and the wholesale of the grain were exclusively managed by state-owned grain ﬁrms; in 1995 the central government initiated the reform of the food circulation system and stressed the autarky of food within a province. All of those policies intensiﬁed the strength of regulations on rural economy. Second, since the late 1980s, to get promoted, local ofﬁcials have been enthusiastic over various ‘image projects’, set down economic and social development targets to lower level government, made falsiﬁed reports on farmers’ income level to upper level government, and even set up industrial ﬁrms regardless of local conditions and local comparative advantage. As a result, not only falsiﬁcation prevailed once again, but overproduction capacity was formed which led to a lot of bad debts. With respect to rural ﬁnance, two outcomes were caused: one is that bad loans were accumulated in rural ﬁnancial agencies and their asset quality was exacerbated;6 the other is the outﬂow of rural ﬁnancial resources for non-farm use. A lot of rural funds were not reinvested in agricultural sectors, but instead diverted to real estate and equity market in urban areas.7 The outﬂow of funds from the countryside could 6 Apart from those non-efﬁcient industrial ﬁrms constructed by local government, government’s grain procurement policy can also contribute to the growth of bad loans. Over the past few years, in order to ensure the grain supply center to procure agricultural products with a price that was higher than the market price, the Agricultural Bank of China (ADBC) took the responsibility of providing most of its loans to state-owned ﬁrms to purchase and sell grain as well as cotton. According to He’s (2002) estimation, ADBC provides about 200 billion Yuan in loans every year to be used for purchasing grain and cotton. However, those state-owned ﬁrms which were in charge of grain procurement had no incentives to repay the loans and would not be punished for their default. Therefore most loans to those ﬁrms became bad loans and had a negative effect on the central bank’s reclaim of loans (He, 2002). 7 He (2002) summed up the ways rural funds outﬂow from the countryside. One way is through the branch ofﬁces of state-owned commercial banks and post ofﬁce saving banks in rural areas to pool deposits and save those funds in upper level banks. There are also lots of funds ﬂowing out from rural areas via RCCs every year, by handing in reserve of deposit to the

190

The Chinese Economy after WTO Accession

be seen everywhere around 1994, the year when the Chinese economy was in its upsurge. But since the macroeconomic policy environment tightened after 1994, most of those funds that were invested in real estate and the equity market became bad loans accumulated within the banking system. Even for the loans for agricultural sectors, little was received by farmers. The Agricultural Bank of China (ABC), for instance, has no face-to-face business with farmers, but instead primarily does business with state-owned agricultural commercial ﬁrms and Township-and-Village enterprises (TVEs). The loans of ABC are mainly used for large infrastructure, purchasing public debts, and protecting ecological environment. But for those medium-and-small scale farm enterprises that are in urgent need of ﬁnancial support, the number of loans from ABC has been decreasing (Chen 2002a). Rural Credit Cooperatives of China (RCCs) are another group of formal ﬁnancial agencies to connect directly with farm households. But RCCs have incurred huge bad loans transferred from ABC, which have greatly hurt RCCs’ operative capacity.8 In addition, as referred to in the previous text, because proﬁts from agricultural activities have been reduced by regulations of the central government, RCCs are also reluctant to lend to farmers.9 Because it is difﬁcult for farmers to get loans from formal ﬁnancial agencies, they turn to informal channels as their primary source of ﬁnance. Based on Rural Fixed Point Survey Ofﬁce’s data which covers 20,294 farm households, Cao (2002) found that in 1999 loans borrowed from rural informal credit market accounted for 69.41 percent of total loans, averaging 1008.56 Yuan per household. Finally, in the name of preventing ﬁnancial risks, the Chinese authority is always cautious of, and even against, those non-ofﬁcial ﬁnancial agencies in civil society. Over the past two decades, non-ofﬁcial ﬁnancial agencies have been suppressed or reorganized by the authority in the name of preventing ﬁnancial risks. For example, the Rural Credit Foundations (RCFs), an active informal ﬁnancial agency in rural areas, was ordered by the authority to close in 1999, which led to 300 billion Yuan of the gap between fund supply and demand in the countryside (Chen, 2002). central bank, re-depositing to the central bank, buying public debts and ﬁnancial bonds, etc. He estimated in 2001 that funds ﬂowing out from the countryside via state-owned commercial banks and post banks were up to 30 billion Yuan and 59.11 billion Yuan respectively. Ma (2001, p. 133) also estimated that about 20 billion Yuan has ﬂowed out from the countryside per year since the 1990s. 8 Many researchers found that, in practice, RCCs have not manifested their cooperative nature during their operation. In contrast, the management and operation of RCCs suffered from the intervention of local government. This is another reason why RCCs were loaded with so many low quality assets (Cao, 2002). 9 According to IFAD (2001), RCCs only covered 20 percent of the poorest farmers. Many case studies on RCCs also showed whether in highly developed regions or in inner land provinces as well as some eastern provinces in which agriculture is their comparative advantage, RCCs’ operation showed more and more urban bias, including more loans allocated to urban areas, more residence of its employees in the city, and more of its subsidiaries located in towns (Shen, 2001).

Financial Development and Urban-Rural Income Disparity in China

191

All in all, in the early 1980s, regulations (such as the promotion of HRS, the increase of procurement price of some principal agricultural products, and the extension of rural product and factor market, etc) were eased off, leading agricultural Total Factor Productivity (TFP), output of grain production and net income per capita of peasants to increase signiﬁcantly and rapidly.10 As a result, URID dropped during this period (Figure 10.1). However, since the mid 1980s the effect of reform of the cultivation system was exhausted, while at the same time other institutional arrangements which were detriments to rural economic development did not undergo major change at all: On the one hand, both the highly monopolized banking market structure and the regulations of government on rural economy have a negative effect on farmers to get funds from formal ﬁnancial institutions. And in fact these formal ﬁnancial agencies have become the channel of funds to ﬂow out from rural areas; on the other hand, even those nonofﬁcial ﬁnancial agencies are under strict suppression so that they are unable to provide sufﬁcient ﬁnancial services to farmers. In contrast, most credit was allocated to large SOEs through the state-owned banking system. Thus ﬁnancial development in China may probably mean that the gap between rural ﬁnancial development and urban ﬁnancial development is widened and, as a result, URID enlarged. Model Speciﬁcation and Data We want to empirically test the relationship between the gap URID and ﬁnancial development in China. Our basic econometric model is as follows: INEit = C + α1 · RPGDPit + α2 · RPGDP2it + α3 · FINDEVit + α4 · AFINDEVit + ∑jβj · D + εit

(3.1)

where subscript i and t denote ith province and tth year respectively. ε is the error term with standard normal distribution of N (0,δ2). Dependent variable INE is the ratio of urban disposable income per capita to the rural per capita net income. We use it to gauge the gap of URID. RPGDP and RPGDP2 are real value of income per capita (in 1978 price) at province level and its square respectively. We add square of RPGDP to see whether there is Kuznets effect, namely an inverted ushaped relationship between income inequality and income level. FINDEV is the ratio of loans extended by ﬁnancial intermediation in one province to its GDP. We use this variable to measure ﬁnancial intermediation development at province level. AFINDEV is the ratio of loans to agriculture sectors to total loans. Because we argue that formal ﬁnancial intermediation plays a negative role in the matter of URID, we expect the coefﬁcient of FINDEV, namely α3, to be positive and statistically signiﬁcant. But the sign and signiﬁcance of α4, the coefﬁcient of AFINDEV, cannot be predetermined since there is the possibility that the data on AFINDEV does not 10 According to authors’ calculation, the growth rate of agriculture and net income per capita (in 1950 prices) of peasants averaged 7.7 percent and 14.4 percent per year respectively for the period of 1978–1984.

The Chinese Economy after WTO Accession

192

Table 10.3

Mean MAX MIN Std. Dev Observations

Descriptive Statistics (Average Values, 1978–1998) INE 2.212 3.044 1.457 0.436 27

RJGDP FINDEV 1193.668 0.735 5652.509 1.080 400.309 0.448 1055.826 0.159 27 27

AFINDEV 0.092 1.036 0.006 0.19 27

FSASP 0.015 0.045 0.002 0.011 27

HRS 0.807 0.855 0.696 0.037 27

OPEN 0.105 0.447 0.025 0.103 27

ROAD 0.216 0.565 0.017 0.126 27

accurately reﬂect the true number of loans going to agricultural sectors. We will turn to this problem in the robustness test using IV method. X is a vector of control variables (or the conditioning information set), which includes: ROAD, the length of road per capita. We use this variable to measure the infrastructure condition in one province; OPEN, the ratio of total value of export to GDP at province level; FDI, the ratio of foreign direct investment to GDP at province level. OPEN and FDI represent the international integration level; HRS, the proportion of farm households that adopt the Household Responsibility System in a province. It measures the progress of rural household responsibility reform in a province. It needs to be noted that in 1987 all rural households had adopted the Household Responsibility System, therefore HRS is 1 after this date; FSASP is the share of ﬁscal expenditure used to support agricultural production in a province.11 Data from 28 Chinese provinces for the period of 1978–98 are used for empirical analysis.12 All data are collected from Liu (2002b) and The Compilation of Statistic Data of New China for 1978–1999 (NBS 2000). Variables and data resources are listed in the appendix. All variables are in logarithmic forms. Table 10.3 lists descriptive statistic results of some principal variables for the period of 1978–99. From it we can clearly see that on average INE is the variable with most variance. Gansu Province has the largest value of INE, 3.04. Shanghai’s value of INE is the smallest. It is obvious that the provinces with higher per capita income are also provinces with lower value of INE. In addition, Tianjin has the highest level of FINDEV, 1.08, in contrast with Zhe jiang province which has the lowest value of FINDEV, 0.448. Estimation Results Table 10.4 reports several tests relating to the basic regression equation 3.1. The results of the Hausman test, shown in the bottom of column (1) and column (2) in Table 10.4, reject the null hypotheses of random-effect models in favour of the alternative ﬁxed-

11 Due to the potential problem of endogeneity, the estimation does not include urbanization index in a province. However if urbanization is measured as the ratio of urban population to total population in a province, including this variable will not change the estimation results much. 12 Due to data availability two provinces, namely Hainan and Tibet, are not included.

Financial Development and Urban-Rural Income Disparity in China

Table 10.4

193

Financial Intermediation and URID: Empirical Result I -

(1)

(2)

(3)

(4)

RJGDP

0.64***

0.57***

0.22*

0.24*

-

0.14

0.15

0.11

0.12

RJGDP2

-0.04***

-0.04***

-0.03***

-0.03***

-

0.01

0.01

0.01

0.01

FINDEV

0.13***

0.12***

0.10***

0.10***

-

0.04

0.04

0.03

0.03

AFINDEV

0.00

0.00

-

-

-

0.01

0.01

-

-

PAFINDEV

-

-

-7.15***

-7.09***

-

-

-

0.85

0.87

HRS

-

-0.15*

-

-0.05

-

-

0.08

-

0.07

ROAD

-

0.04

-

0.06

-

-

0.06

-

0.05

OPEN

-

0.02**

-

0.01

-

-

0.01

-

0.01

FDI

-

0.01

-

0.01

-

-

0.01

-

0.01

FSASP

-

0.02

-

0.02*** 0.01

-

-

0.01

-

-

-

-

-

-

P value of Hausman Test

0.00

0.00

0.00

0.00

Adj-R2

0.87

0.88

0.89

0.89

F Value

72.00

66.3

86.98

78.83

Observations

506

504

542

538

Notes:

1. Numbers below estimated coefﬁcients are robust standard errors. 2. *, **, *** indicate signiﬁcance at 10%, 5%, 1% signiﬁcant level.

effect models. Therefore in column (1) and (2) the estimation results are the results from the two-way ﬁxed-effect model. The F-values in all regressions are signiﬁcant and the adjusted-R2s are very high in both results. The focus of our analysis is on the assessment of the signiﬁcance of FINDEV, the proxy for ﬁnancial intermediation development. The coefﬁcient is consistently signiﬁcant at the 1 percent level in columns (1) and (2). These results are consistent with our expectation that ﬁnancial development as a whole contributes to the enlargement of URID. The estimated coefﬁcient (α4) for the proportion of agricultural loans is not signiﬁcantly different from zero, indicating that such loans did not signiﬁcantly affect URID. These results again conﬁrm our argument concerning the

194

The Chinese Economy after WTO Accession

inefﬁciency of Chinese formal ﬁnancial intermediation in providing ﬁnancial services to agricultural sectors. In columns (1) and (2), the sign of the estimated coefﬁcients of RPGDP and its square term are positive and negative respectively, and both coefﬁcients are signiﬁcant at the 1 percent level, suggesting there is an inverted-U shaped curve relationship between per capita income and URID. Not surprisingly, the coefﬁcient of HRS is signiﬁcantly negative in column (2), indicating deregulation did facilitate the fall of URID. In addition, while OPEN and FDI both have positive signs, only the coefﬁcient on OPEN is signiﬁcant. ROAD has an insigniﬁcantly positive coefﬁcient. This result may indicate that infrastructure constructions are mainly concentrated in urban areas and are used mainly to serve urban residents. Finally, it is somewhat surprising that FSASP has a positive but insigniﬁcant coefﬁcient, indicating that the more ﬁscal funds were put into agricultural sectors, the larger URID would be, while this effect is tiny. One possible explanation is the reversed causality between FSASP and URID. In other words, in the province where the gap of URID was widened, the local authority would be politically forced to increase ﬁscal funds to support agricultural sectors. Robustness Test It is widely believed that there is large deviation between ofﬁcial data, namely AFINDEV, and the actual number of loans to agricultural sectors, as pointed out in the previous text. To solve this problem, we follow Aziz and Duenwald’s (2002) methodology in which they try to decompose total credit into two parts, namely the credit to state-owned sector and those to non-state-owned sector, by using the share of credit to SOEs predicted by a province’s share of SOEs value added in GDP. Here we also use PAFINDEV, the ﬁtted value by regressing FINDEV on the proportion of the primary industry’s GDP to total GDP, to replace AFINDEV as the credit ratio to the agricultural sector and estimate equation 3.1 again. New results are reported in columns (3) and (4) of Table 10.4.13 According to the Hausman test value, both results are based on the two-way ﬁxed model effect. The results in columns (3) and (4) are similar to the previous results in columns (1) and (2) except the estimated coefﬁcients of PAFINDEV. In both columns they become statistically signiﬁcant and signs are negative, consistent with our expectation. Those results conﬁrm again our argument that formal ﬁnancial intermediation is unable to provide credit to agricultural sectors efﬁciently. So far we have explored an overall effect of ﬁnancial intermediation development on URID for the period of 1978–98. However, we want further to see whether this effect is different before and after the late 1980s, since we argue that only since the late 1980s did government use banks as principal policy instruments to intervene 13 Surely we cannot directly view PAFINDEV as the actual amount of credit to the agricultural sector. We only use it as a alternative way to test our empirical results obtained in a previous regression.

Financial Development and Urban-Rural Income Disparity in China

Table 10.5

195

Financial Intermediation and URID: Empirical Result II

RJGDP

(1) 1.72***

(2) -0.98***

(3) 1.11***

(4) -0.86***

RJGDP2 FINDEV

0.34 -0.15*** 0.03 0.00

0.24 0.07*** 0.01 0.37***

0.30 -0.11*** 0.02 0.03

0.20 0.04*** 0.01 0.28***

AFINDEV -

0.06 0.01 0.01

0.07 0.03 0.03

0.04 -

0.05 -

PAFINDEV -

-

-

-3.31*** 1.33

-10.27*** 1.20

HRS -

-0.12* 0.07

-

-0.11 0.07

-

ROAD

-0.06

0.14*

-0.05

0.11

-

0.07

0.08

0.07

0.07

OPEN FDI FSASP P value of Hausman Test Adj-R2 F Value Observations

0.01 0.02 0.06*** 0.02 0.01 0.02 0.000 0.89 43.46 240

-0.05** 0.02 0.003 0.01 0.05* 0.03 0.04 0.92 73.98 264

0.02 0.01 0.05*** 0.02 0.01 0.02 0.13 0.88 49.75 265

-0.06*** 0.02 0.003 0.01 0.04* 0.02 0.000 0.94 104.82 275

Notes:

1. Numbers below estimated coefﬁcients are robust standard errors. 2. *, **, *** indicate signiﬁcance at 10%, 5%, 1% signiﬁcant level.

into the economy. Therefore we split the whole sample period into two parts, i.e., 1978–88 and 1989–98. It must be noted that because the value of HRS will be timeinvariant after 1987, this variable will not be included in the regression for 1989–98 due to the collinearity problem. The results are listed in Table 10.5 in which we follow the same procedures and sequence as in Table 10.4. All four results in Table 10.5 suggest that the effect of FINDEV is only signiﬁcant in the period for 1989–98 (column 2, column 4), while in the period for 1978–88 such effect is not signiﬁcant (column 1, column 3). This is because the period around 1989 is when rural reform stagnated and the policy emphasis was turned to urban sector reform. Regulations on rural economy and the rural ﬁnancial system were tightened again so that the gap of URID would inevitably be enlarged. Another point is that an inverted-U curve relationship between URID and the level of economic development only existed in the ﬁrst period, but was reversed in the second period (Table 10.5). In addition, OPEN has a signiﬁcantly negative

196

The Chinese Economy after WTO Accession

coefﬁcient in the second period, while FDI has a signiﬁcantly positive coefﬁcient in the ﬁrst period. In columns (2) and (4) the marginal effects of OPEN are far bigger than those of FDI, which suggests that exploiting comparative advantage in terms of foreign trade signiﬁcantly reduced the gap of URID during 1990s. Moreover, in section 2 we argue that the government use ﬁscal policy and the ﬁnancial system alternatively in different periods to implement its policy goals. If our argument is true, combined with the results we have in Table 10.5, we expect that in the ﬁrst period of 1978–88 ﬁscal policy should be considered in the regression and its estimated effect on URID should be signiﬁcant and be bigger that that of ﬁnancial intermediation development, whereas in the second period of 1989–98 the effect of ﬁscal policy should be smaller than that of ﬁnancial intermediation development and is likely to be insigniﬁcant. In order to test those effects we add FSGDP, the ratio of ﬁscal expenditure to GDP in a province, to equation (3.1) and do the regression by employing the same procedures and sequence as we do in Table 10.5. The results are reported in Table 10.6. The results in column (1) of Table 10.6 shows that the overall effect of ﬁscal policy on URID in full sample periods is highly signiﬁcant and is larger than that of ﬁnancial intermediation development. However, if we split the whole periods into two sub-periods, a different picture emerges. For the period of 1978–88, the estimated coefﬁcient of FSGDP is signiﬁcant and is bigger than that of FINDEV (column 2 and column 5). But for the period of 1989–98, the estimated coefﬁcient of FINDEV is signiﬁcant and larger than that of FSGDP (column 3 and column 6). We conclude that both ﬁnancial intermediation development and ﬁscal policy contribute to the enlargement of URID, but the former plays a leading role for period of 1989–98, whereas the second is the most prominent factor for the period of 1978– 88. Those results are consistent with our theoretical argument. To be more speciﬁc, in the early 1980s, although procurement prices of many agricultural products were considerably raised, the urban residents received a great deal of living allowance from government. On the other hand, the proportion of ﬁscal funds in total ﬁscal expenditure for supporting agricultural development (FSASP) is very small (3 percent on average). Thus, even in the early 1980s, ﬁscal policy contributed signiﬁcantly to the enlargement of URID. Since the late 1980s, government only increased its support for agricultural development slightly (9 per cent on average in terms of FSASP), and at the same time the principal policy instrument used by government changed from ﬁscal policy to ﬁnancial policy. Therefore we have the results of Table 10.6. Extended Model Speciﬁcation and Results The approach we outlined so far only enables us to identify an overall and direct effect of FINDEV on URID. However, such analysis does not allow us to simultaneously identify both the effect of ﬁnancial intermediation and the effect of structural features of the economy. For example, in a province with a higher share of

Financial Development and Urban-Rural Income Disparity in China

Table 10.6

197

Financial Intermediation and URID: Empirical Result III

RJGDP RJGDP2 FINDEV AFINDEV PAFINDEV HRS ROAD OPEN FDI FSASP FSGDP P value of Hausman Test Adj-R2

(1) 0.77*** 0.15 -0.05*** 0.01 0.11** 0.04 0.00 0.01 -0.15** 0.08 -0.05 0.06 0.02* 0.01 0.00 0.01 -0.01 0.01 0.23*** 0.03 0.000 0.89

(2) 1.84*** 0.33 -0.15*** 0.03 0.03 0.06 0.00 0.01 -0.13* 0.07 -0.09 0.07 0.02 0.01 0.05*** 0.02 0.003 0.02 0.20*** 0.06 0.07 0.90

(3) -0.86*** 0.25 0.06*** 0.02 0.36*** 0.07 0.03 0.03 0.12 0.09 -0.05* 0.02 0.002 0.01 0.04 0.03 0.06 0.04 0.006 0.93

(4) 0.47*** 0.13 -0.05*** 0.01 0.09*** 0.03 -6.05*** 0.86 -0.07 0.07 0.00 0.05 0.01 0.01 0.00 0.01 0.002 0.01 0.19*** 0.03 0.000 0.90

(5) 1.39*** 0.27 -0.12*** 0.02 0.05 0.03 -2.32** 0.99 -0.11** 0.06 -0.02 0.03 0.03** 0.01 0.05*** 0.01 0.01 0.01 0.22*** 0.05 0.19 0.89

(6) -0.83*** 0.21 0.04*** 0.01 0.28*** 0.05 -10.21*** 1.21 0.10 0.07 -0.06 0.02 0.00 0.01 0.04* 0.02 0.01 0.03 0.001 0.94

F Value Observations

73.04 504

45.37 240

75.82 264

83.73 538

47.61 263

98.17 275

Notes:

1. Numbers below estimated coefﬁcients are robust standard errors. 2. *, **, *** indicate signiﬁcance at 10%, 5%, 1% signiﬁcant level.

agriculture output in GDP, will the gap of URID be larger or smaller when its banks lend more? Or what’s the effect of the growth of bank lending on URID when nonstate enterprises in a province become stronger? In order to work out such effects, we modify our basic model speciﬁcation of equation (3.1) by adding the interaction between FINDEV and variables that are regarded to reﬂect the structural features of the economy. The ﬁrst extended model speciﬁcation is as follows: INEit = C + α1 · RPGDPit + α2 · RPGDP2it + α3 · FINDEVit + α4 · AFINDEVit + α5 · (FINDEV · AGR) + α6 · AGR + ∑jβj · D + εit

(4.1)

198

The Chinese Economy after WTO Accession

where AGR is the share of primary industry’s GDP in total GDP. Kuznets (1955) argues that as people move from the low-income, but more egalitarian, agricultural sector to the high-income, but less egalitarian, industrial sector, income inequality would increase. In other words, income inequality depends on the sectoral structure of an economy. In order to control for such an effect, we add AGR in equation (3.1). We also add an interaction term FINDEV • AGR in equation (3.1). As Clarke, Xu and Zou (2003, hereafter CXZ) argue, sector structure will affect how ﬁnancial depth impacts inequality. In their paper, CXZ hold that if entry into modern sector (industry and service) is made easier when it is easier to gain access to ﬁnance, inequality will be greater in economies with larger modern sectors. Consequently, inequality will be higher in countries with large modern sectors and greater ﬁnancial depth than in countries with only one (or neither) of these characteristics. As far as the relationship between URID, ﬁnancial development and sector structure is concerned, if CXZ’s augmented Kuznets hypothesis is right, the estimated coefﬁcient of (FINDEV • AGR) should be negative and statistically signiﬁcant; on the other hand, we believe this interaction term also can be used as a vehicle to test the urban-bias hypothesis pertaining to the relationship between URID and ﬁnancial development. If urban-bias does exist, the coefﬁcient on the interaction term should be insigniﬁcant (ﬁnancial intermediation will be insensitive to the variation of sector structure). The estimation results of equation (4.1) are reported in columns (1)-(3) of Table 10.7. We can see from the result in column (1) that α 6 is signiﬁcantly negative, consistent with the ﬁnding in CXZ (2003) that supports the Kuzents’ hypothesis that URID increases during the transition from agriculture to modern industry. Moreover, our analysis focuses on the signs and signiﬁcance level of α 3 and α 5 . While α 3 and α 5 are both insigniﬁcant in column (1), α 3 is signiﬁcantly positive and α 5 is insigniﬁcant in column (3) of which the results are based on the sub-period of 1989–98. Those results seem to support the urban-bias hypothesis as opposed to Kuzents’ hypothesis since the late 1980s, for credit allocation is insensitive to sector structural change. Although results in columns (1)-(3) show that urban-bias does exist, they cannot tell us whether such effect covers all ﬁrms in urban areas indiscriminately or only those SOEs. In order to examine which effect is operative, we replace AGR in equation (4.1) by NRSOE, the share of non-state gross industrial output value in total gross industrial output value, and also add an interaction between NRSOE and FINDEV. NRSOE to a large extent measures the ownership structure of a province and was generally used as the proxy for the progress of market reform in that province. So we estimate the following equation: INEit = C + α1 · RPGDPit + α2 · RPGDP2it + α3 · FINDEVit + α4 · AFINDEVit + α5 · (FINDEV · NRSOE) + α6 · NRSOE + ∑jβj · D + εit

(11.2)

Financial Development and Urban-Rural Income Disparity in China

Table 10.7

199

Financial Intermediation and URID: Empirical Result IV

RJGDP RJGDP2 FINDEV AFINDEV AGR*FINDEV AGR NRSOE*FINDEV NRSOE HRS -

(1) 0.419** 0.176 -0.041*** 0.011 0.201 0.079 0.010 0.010 0.040 0.040 -0.287*** 0.047 -0.068 0.075

(2) 1.324*** 0.347 -0.121*** 0.025 0.062 0.115 0.003 0.010 0.021 0.059 -0.144** 0.059 -0.066 0.062

(3) -0.742*** 0.233 0.031** 0.014 0.402*** 0.104 0.039* 0.023 0.027 0.054 -0.486*** 0.058 -

(4) 0.944*** 0.159 -0.060*** 0.011 0.318*** 0.077 -0.005 0.010 0.154*** 0.045 -0.089 0.094 -0.155** 0.075

(5) 1.553*** 0.299 -0.130*** 0.022 0.175 0.124 -0.005 0.010 0.106 0.073 -0.307 0.191 -0.137** 0.058

(6) -0.967*** 0.279 0.067*** 0.017 0.297*** 0.099 0.029 0.027 -0.057 0.069 0.079 0.097 -

ROAD -

-0.009 0.056

-0.024 0.035

0.070 0.076

-0.050 0.055

-0.014 0.035

0.116 0.085

0.013 0.015 0.059*** 0.017 0.005 0.015 0.165*** 0.049 0.434 0.90 43.31 229

-0.069*** 0.020 0.003 0.010 0.029 0.025 0.014 0.037 0.000 0.95 96.71 254

0.020* 0.011 -0.007 0.009 -0.009 0.013 0.241*** 0.032 0.000 0.89 72.41 504

0.024 0.014 0.042** 0.017 0.004 0.015 0.203*** 0.048 0.13 0.90 43.9 240

-0.051** 0.023 -0.001 0.012 0.038 0.028 0.050 0.042 0.000 0.93 72.27 264

OPEN 0.010 0.011 FDI 0.002 0.009 FSASP 0.002 0.013 FSGDP 0.189*** 0.032 P value of Hausman Test 0.000 Adj-R2 0.90 F Value 76.67 Observations 483

Notes:

1. Numbers below estimated coefﬁcients are robust standard errors. 2. *, **, *** indicate signiﬁcance at 10%, 5%, 1% signiﬁcant level.

If the effect of ﬁnancial intermediation on URID can be explained by the government’s intervention into banks to support SOEs, α 5 should be insigniﬁcant and α 3 should be signiﬁcant. The underlying reason is because no matter how ownership structure changes, as long as government ofﬁcials determine to ensure SOEs have the priority over other ﬁrms to get loans, formal credits would ﬂow into state-owned sectors other than non-state sectors. The results are reported in columns (4)-(6) of Table 10.7.

200

The Chinese Economy after WTO Accession

However, while in column (4) the coefﬁcient on FINDEV ( α 3 ) is signiﬁcantly positive, the coefﬁcient on interaction term ( α 5 ) is also signiﬁcantly positive which suggests that in those provinces with a higher share of non-state sectors, output will have a larger URID when credit increases. The three municipalities of Beijing, Tianjin and Shanghai have the lowest level of NRSOE and highest level of FINDEV but the lowest level of URID. We believe for political considerations the government mobilizes all possible resources to ensure the overall development of the three municipalities, including keeping a low level of URID, whereas in other provinces governments have only limited resources to maintain a high level of share of SOEs at the expense of the development of other sectors. So the relationship among NRSOE, FINDEV and INE that emerges in other provinces does not exist in the three municipalities due to political reasons. Allowing for this possibility, we treat Beijing, Tianjin and Shanghai as outliers and drop them off from the sample and carry out regressions of equation (4.1) again. The estimation results are reported in Table 10.8. The results of columns (1) to (3) in Table 10.8 are similar to those in the corresponding columns in Table 10.7 except the interaction term of column (1) in Table 10.8 between AGR and FINDEV becomes signiﬁcantly positive. However, we argue this result is still consistent with an Urban-bias hypothesis. On the one hand, the direct effect of FINDEV is still signiﬁcantly positive; on the other hand, if government directs credits to SOEs, due to limited ﬁnancial resources the pace of provincial industrialization would be slower because those non-state industrial ﬁrms whose operations are more efﬁcient would be ﬁnancially constrained. Since larger modern sectors mean larger gaps of URID ( α 6 is signiﬁcantly negative), the marginal effect of FINDEV through industrialization on URID would be negative. However, when we split the full sample into two sub-periods, this effect disappears. In addition, as is expected, the results of columns (2) and (3) demonstrate again that the effect of ﬁnancial intermediation is signiﬁcant since the late 1980s and ﬁscal policies contributed more to a larger URID in the 1980s. More interestingly, after three municipalities were dropped off from the sample, the coefﬁcients on the interaction term between NRSOE and FINDEV are no more signiﬁcant (columns 4–6), which means the impact of ﬁnancial development on URID does not depend on the ownership structure change within a province. Those results are also consistent with an urban-bias hypothesis. Based on the results of columns (1)-(6), we ﬁnd although a larger modern sector will increase the gap of URID, a higher share of non-state production will reduce URID. As is often pointed out in the literature, a higher value of NRSOE may possibly mean more job opportunities for migrants in cities, more development of TVEs, and more favorable market conditions for farmers to sell their products, etc. We should remember that under current regulation, one farmer that ﬁnds a job in the city cannot be identiﬁed as an urban resident so that most of his non-farm income will be sent back to his village and is still calculated as rural income. So it is not surprising to see a signiﬁcantly negative coefﬁcient on NRSOE. In addition, this signiﬁcantly negative

Financial Development and Urban-Rural Income Disparity in China

Table 10.8

201

Financial Intermediation and URID: Empirical Result V

RJGDP RJGDP2 FINDEV

(1) 0.649*** 0.239 -0.064*** 0.016 0.324***

(2) 1.471*** 0.553 -0.133*** 0.043 0.112

(3) -0.166 0.332 -0.014 0.022 0.519***

(4) 1.078*** 0.229 -0.074*** 0.015 0.147**

(5) 1.645*** 0.467 -0.145*** 0.037 0.116

(6) -0.227 0.378 0.013 0.025 0.213**

AFINDEV AGR*FINDEV AGR NRSOE*FINDEV NRSOE

0.100 0.017 0.010 0.150* 0.069 -0.334** 0.063 -

0.194 0.009 0.012 0.069 0.149 -0.124 0.106 -

0.148 0.034 0.027 0.144 0.099 -0.474*** 0.065 -

0.074 0.010 0.011 0.063 0.045 -0.042*

0.173 0.006 0.012 0.100 0.103 0.058

0.091 0.026 0.031 -0.110 0.067 -0.044**

-

-

-

-

0.022 -

0.070 -

0.019 -

HRS ROAD OPEN FDI FSASP

-0.156 0.097 0.047 0.057 -0.003 0.011 -0.002 0.010 -0.025

-0.068 0.078 0.005 0.041 0.019 0.017 0.076*** 0.022 0.014

0.182** 0.078 -0.086*** 0.021 -0.007 0.011 0.021

-0.240** 0.101 -0.024 0.058 0.008 0.011 0.000 0.010 -0.017

-0.096 0.076 0.009 0.034 0.025* 0.015 0.068*** 0.023 0.027

0.206** 0.089 -0.068*** 0.023 -0.009 0.012 0.023

FSGDP P value of Hausman Test Adj-R2 F Value Observations

0.018 0.193*** 0.036 0.000 0.89 63.75 422

0.032 0.216*** 0.066 0.714 0.864 29.54 198

0.024 -0.019 0.038 0.01 0.94 76.77 224

0.018 0.202*** 0.037 0.000 0.88 57.7 443

0.030 0.198*** 0.062 0.113 0.86 30.43 209

0.028 0.009 0.043 0.083 0.91 56.04 234

Notes:

1. Numbers below estimated coefﬁcients are robust standard errors. 2. *, **, *** indicate signiﬁcance at 10%, 5%, 1% signiﬁcant level.

coefﬁcient may also reﬂect the fact that industrialization rose ﬁrst in the rural areas (in the form of TVEs) in the early 1980s because urban non-state industrial ﬁrms were not allowed at that time. Until recently, many non-state industrial ﬁrms took the form of TVEs since non-state ﬁrms are tightly regulated in urban areas. Indeed

202

The Chinese Economy after WTO Accession

in some cases the reducing of URID is the result of underdevelopment of non-state ﬁrms in urban areas.14 Conclusion In this chapter we use a panel data of 28 provinces for the period of 1978–98 to empirically test the relationship between ﬁnancial intermediation development and URID in China. We ﬁnd there is urban-bias of credit allocation in formal ﬁnancial agencies, for ﬁnancial intermediation development contributes signiﬁcantly to the enlargement of URID. We also divide the whole sample period into two sub-periods and ﬁnd that the effect of ﬁnancial intermediation development on URID is mainly apparent during 1989–98 because from the late 1980s the government relied more on banks than other ﬁscal powers to intervene into its economy. Moreover, (after dropping off three municipalities) we ﬁnd a modest negative interaction of ﬁnance with the size of the modern sector, whereas the interaction between ﬁnance and the size of non-state industrial production is not signiﬁcant. Those results are consistent with the character of the Chinese ﬁnancial and economic system, which we explored earlier. There are some limitations in our chapter. For example, we have not empirically explored the impact that the banking market structure had on ﬁnancial intermediation’s urban-bias behavior and on URID. This was due to a data availability problem. We also did not directly control for government’s own interventions on its economy. These are issues to be considered in future research. References English Atkinson, Anthony B., and Andrea Brandolini (2001), ‘Promise and Pitfalls in the Use of ‘Secondary’ Data-Sets: Income Inequality in OECD Countries as a Case Study,’ Journal of Economic Literature, Vol. 39, No. 3, September. Aziz, Jahangir and Christoph Duenwald (2002), ‘Growth-Financial Intermediation Nexus in China,’ IMF Working Paper Series, WP/02/194, Nov 2002. Beck, Thorsten and Ross Levine (2002), ‘Stock Markets, Banks, and Growth: Panel Evidence,’ NBER Working Papers 9082, National Bureau of Economic Research. Beck, Thorsten, Aslı Demirgüç-Kunt, Ross Levine and Vojislav Maksimovic (2000), ‘Financial Structure and Economic Development: Firm, Industry, and Country Evidence,’ mimeo, June 14. 14 Another explanation is that massive labor force supply from the countryside lowers the growth rate of average wage levels of non-state ﬁrms in urban areas, which will also reduce the gap of URID. However, this is another hypothesis to be tested.

Financial Development and Urban-Rural Income Disparity in China

203

Berger, A.N. and G.F. Udell, , (1995), ‘Relationship lending and lines of credit in small ﬁrm ﬁnance,’ Journal of Business, Vol. 68, pp. 351–82. Chen, Jianbo (2002a), ‘County economy has lost its energy for ﬁnancial repression,’ China Economic Times, December 20. Clarke, George, Lixin Colin Xu and Heng-fu Zou (2003), ‘Finance and Income Inequality: Test of Alternative Theories,’ World Bank Policy Research Working Paper 2984, March. Demirgüç-Kunt, Asli and Ross Levine (1999), ‘Financial Structures Across Countries: Stylized Facts,’ Washington, D.C.: World Bank Policy Research Working Paper 2143,. Goldsmith, R. 1969, Financial Structure and Economic Development, New Haven: Yale University Press. Greenwood, Jeremy and Boyan Jovanovic (1990), ‘Financial Development, Growth, and the Distribution of Income,’ Journal of Political Economy, Vol. 98, No. 5, pp. 1076–107. IFAD (2001), ‘Rural Financial Services in China. Thematic Study,’ Volume I – Main Report. Report No. 1147-CN Rev. International Finance Corporation (2000), China’s Emerging Private Enterprises— Prospects for the New Century, Washington D.C. King, Robert G. and Ross Levine (1993a), ‘Finance and Growth: Schumpeter Might Be Right,’ Quarterly Journal of Economics, Vol. 108, pp. 717–38. King, Robert G. and Ross Levine (1993b), ‘Finance, Entrepreneurship, and Growth: Theory and Evidence,’ Journal of Monetary Economics, Vol. 32, pp. 513–42. Kuznets, S. (1955), ‘Economic Groth and Income Inequality,’ American Economic Review, Vol. 61, pp. 1–28. Levine, Ross (1997), ‘Financial Development and Economic Growth: Views and Agenda,’ Journal of Economic Literature, June, pp. 688–726. Levine, Ross; Norman Loayza and Thorsten Beck (2000a), ‘Financial Intermediation and Growth: Causality and Causes,’ Journal of Monetary Economics, Vol. 35, pp. 688–762. Levine, Ross; Norman Loayza and Thorsten Beck (2000b), ‘Finance and the Sources of Growth,’ Journal of Financial Economics, Vol. 58, pp. 261–300. Levonian, M.E. and J. Soller (1995), ‘Small banks, small loans, small business,’ mimeo, Federal Reserve Bank of San Francisco, San Francisco, CA (December). Lin, J.Y. (2003), ‘Development Strategy, Viability, and Economic Convergence,’ Economic Development and Cultural Change, Vol. 51, No. 2. Liu, Shouying (2002), ‘Survey on County Economy,’ Report on Economy for 21st Century, August 10. McKinnon, Ronald (1973), Money and Capital in Economic Development, Brookings Institution, Washington DC. Mayer, C. (1990), ‘Financial Systems, Corporate Finance and Economic Development,’ in G. Hubbard (ed.), Asymmetric Information, Corporate Finance and Investment, The University of Chicago Press, Chicago.

204

The Chinese Economy after WTO Accession

Meyer, Laurence H. (1998), ‘The present and future roles of banks in small business ﬁnance,’ Journal of Banking and Finance, Vol. 22, pp. 1109–16. NBS (2000), The Compilation of Statistic Data of New China for 1978–1999, The National Bureau of Statistics, China Statistics Press. Park, Albert, Loren Brandt and John Giles (2002), ‘Competition Under Credit Rationing: Theory and Evidence from Rural China,’ mimeo of the World Bank. Peek, J. and E.S. Rosengren (1996), ‘Small business credit availability: How important is size of lender?’ in A. Saunders and I. Walter (eds), Universal Banking: Financial system design reconsidered. Irwin, Chicago, IL, pp. 628–55. Shaw, Edward (1969), Financial Deeping in Economic Development, Oxford University Press, Oxford. Stiglitz, Joseph E. (1985), ‘Credit Markets and the Control of Capital,’ Journal of Money, Credit and Banking, Vol. 17, No. 2 (May), pp. 133–52. Strahan, P.E. and J.P. Weston (1996), ‘Small business lending and bank consolidation: Is there cause for concern?’ Working Paper, Federal Reserve Bank of New York, New York (March). Strahan, P. E. and James P. Weston (1998) ‘Small business lending and changing structure of the banking industry,’ Journal of Banking and Finance, Vol. 22, pp. 821–45. Tsui, Kai Yuen (1993), ‘Decomposition of China’s Regional Inequalities,’ Journal of Comparative Economics, Vol. 17, pp. 600–627. Wei, Shang-Jin (1997), ‘Siamese Twins—Is There State-Owned Bias,’ China Economic Review, Vol. 8, No. 1. Wei, Shang-Jin and Yi Wu (2001), ‘Globalization and Inequality: Evidence from Within China,’ NBER Working Paper Series, No. 8611, Nov 2001. World Bank (1997), Sharing Rising Incomes—Disparities in China, World Bank, Washington DC. World Bank (2001), World Development Indicators, World Bank, Washington DC. Chinese Cao, Liqun (2002), ‘Research on rural ﬁnancial reform and credit for farmers,’ available at http://www.rcre.org.cn/ztyj/rcre-a3-4.htm. Chen, Jianbo (2002b) ‘Financial Repression Hurts Deadly County Economy,’ China Economic Times, December 20. Gao, Peiyong and Laichen Wen (2001), ‘How Chinese Fiscal system Developed Since Market, Renmin University of China Press, Beijing, pp. 192–93. He, Guangwen (2002), ‘Rural Credit Policy in China: Measures, Effects and Adjustment,’ in Xiaojian Fan (ed.), The Changes of Rural Economic Policies in China After Its Accession to WTO, Chinese Agriculture Press, pp. 249–69. He, Mengbi, Xingyuan Feng and Guangwen He (2002), ‘On the Diversiﬁcation of Financial Agencies I Rural China,’ (Unpublished), December 28. Hu, Angang, Shaogang Wang and Xiaoguag Kang (1995), ‘Report on Regional Disparity in China,’ Liaoning People Press, pp. 7–8.

Financial Development and Urban-Rural Income Disparity in China

205

Lin, Justin Yifu, Fang Cai and Zhou Li (1998), ‘An analysis of Regional Disparity in China during a Period of Economic Transition,’ Journal of Economic Research, Vol. 6. Lin, Justin Yifu, Qi Zhang and Mingxing Liu (2003), ‘Financial Structure and Economic Growth: the case of Manufacturing Industry,’ Journal of World Economy, Vol. 1. Lin, Justin Yifu, Feder, Zhunyi Liu and Xiaopeng Luo (2000), ‘Rural Credit and Farm Performance in China’, in Justin Yifu, Lin (ed.), On Institution, Technology and Rural Development in China, Peking University Press, pp. 188–207. Liu, Shouying (2002), ‘Survey on County Finance’, Economic Report in 21st Century, August 10. Ma, Jun (2001), ‘On China’s Rural Credit Cooperatives,’ Journal of Finance Research, Vol. 10. Research Group for Economic Development (2002), ‘The construction of Technology Choice Index’ (unpublished), China Center for Economic Development, Peking University. Shen, Yongwei (2001), ‘Serving for the City or for the Countryside: The Operation of Rural Credit Cooperatives and Rational Choices of Reforming Them,’ Journal of Finance Research, Vol. 5. Song, Hongyuan (ed.), (2000), The Evolution of Chinese Rural Economic Policies since Market Reform, China Economy Press, pp. 149–73. Tao, Ran, Mingxing Liu and Qi Zhang (2003), ‘The Taxation and Fee Burdens on Peasants in Rural China: a Political Economy Analysis,’ China Social Science Review, Vol. 3. Wei, Houkai (ed.) (1997), Regional Economic Development in China: Economic Growth, Institutional change and Regional Disparity, China Finance Press, pp. 41–52. Zhang, Qi (2000), ‘An analysis on Regional Development Disparity,’ Journal of Management World, Vol. 6. Zhang, Qi (2002), ‘Factor Endowment, Development Strategy and Financial Structure,’ Ph.D. dissertation (unpublished).

This page intentionally left blank

PART III International Trade and Industrial Development

This page intentionally left blank

Chapter 11

China in the Antidumping War Against China Jason Z. Yin

Introduction The WTO Cancun ministerial conference of September 2003, failed after major developing countries and the US deadlocked in reaching an agreement on antidumping and removing other market protection measures. The failure represents a major setback for the WTO to move forward in establishing a new order for international trade and to prevent the abusive use of antidumping measures. It also represents a major setback for China in its effort to reverse the course of having been the major target in the antidumping war. China became the 143rd WTO member in December 2001. While celebrating its membership for the trade club, one of the very ﬁrst problems China had to deal with was antidumping litigation. Chinese enterprises were shut off from many industrial sectors as part of antidumping actions against China and suffered major losses in international markets. After the failure of the last two WTO ministerial conferences, it is unrealistic for China to wait for the WTO to pass new rules to correct the abusive antidumping practices and to reverse its defensive positions. It needs to formulate its own strategy to defend its own interests. However, it seems that China has not yet formulated clearly deﬁned strategies to cope with the trade nightmares. These developments call for serious research and policy attention. In the last decade or so, ‘dumping’ and antidumping have become a major source of trade disputes among WTO member countries. The number of antidumping cases pending before various legal forums remains at a very high level, as shown in Table 11.1. China remains the target of antidumping investigations, from both before and after its WTO accession (Jiang and Ellinger, 2003; Messerlin, 2004). Looking at the antidumping data for the period January 1, 1995 to December 31, 2003, there were 41 member countries initiating 2,416 anti-dumping investigations against exports from a total of 97 countries or customs territories. The three major initiators were the US (329), the European Union (274) and India (389), as shown in Table 11.2. China was the number one target of antidumping litigation (356 cases), followed by South Korea (182 cases). Among the total 356 investigations against China, 35 percent came from four developed economies: the US (51 cases), the European Union

The Chinese Economy after WTO Accession

210

Table 11.1

Antidumping Cases in which China is Involved (1/1/1995– 12/31/2003)

Total AD AD Cases Filed AD Measures Taken AD Cases Initiated Cases Filed against China against China by China 1995 157 20 26* NA 1996 224 43 16 NA 1997 243 33 33 NA 1998 256 28 24 NA 1999 356 41 20 0 2000 294 43 30 6 2001 366 53 30 14 2002 311 51 37 30 2003 210 45 39 22 Total 2416 356 254 72 * Includes cases processed under GATT. Source: WTO, Statistics on Anti-Dumping, September, 2004. http://www.wto.org/english/ tratop_e/adp_e/adp_e.htm

(43 cases), Australia (16) and Canada (14); and 39 percent from four developing countries: India (69 cases), Argentina (38), South Africa (18) and Brazil (14), as shown in Table 11.2. This chapter will take a close examination of the WTO’s antidumping provision and cases to see what actual and potential problems lurk in its shadows. It will also analyze the structure and characteristics of all the antidumping litigations against China. Based on the analysis, strategic and policy recommendations are made for the Chinese government and enterprises to be better prepared in the antidumping trade disputes and to better protect its economic and trade interests. What is Antidumping? Dumping is often described as when the sale of products for export is at a price less than normal value and that there is ‘material injury’ or the threat of that to the domestic industry that is caused by dumped imports. Normal value means roughly the price for which those same products are sold in the home country or a third country. Today it is still the core international rule regarding dumping (Jackson, 1989: 226). Antidumping actions are intended to counter price discrimination. A country can impose antidumping duties on imports if it can show that a ﬁrm is selling a product abroad at less than its ‘normal value’ and that there is ‘material injury’ or the threat of that to the domestic industry that is caused by dumped imports (Lardy, 2002: 86).

China in the Antidumping War Against China

Table 11.2

211

Antidumping Cases for Major Reporting Parties and Affected Parties (1/1/1995–12/31/2003)

Reporting Party

Affected Party China South US Taiwan Japan Indonesia Thailand India Korea

Total cases initiated Argentina 38 9 10 10 2 2 2 3 180 Australia 16 15 7 9 5 16 13 4 163 Brazil 14 3 17 2 2 1 2 3 109 Canada 14 7 12 7 2 3 2 5 122 EU 43 21 7 14 8 11 13 25 274 India 69 27 20 26 20 14 15 -379 South Africa 18 13 7 5 1 6 4 16 166 US 51 22 16 29 13 6 16 329 Subtotal of above 8 263 117 80 89 69 66 59 72 1722* Total cases investigated 356 182 135 123 106 99 91 98 2416 Source: WTO, Statistics on Anti-Dumping, September, 2004. http://www.wto.org/english/ tratop_e/adp_e/adp_e.htm

WTO rules deﬁne normal value as the price at which the good is sold in the home country or in a third country. The dumping margin is deﬁned as: Dumping Margin = Home Market Sales Price – Export Sales Price

When that margin is greater than zero, there is ‘dumping’ in the sense used in international trade policy. The concepts of ‘underselling’ or ‘dumping’ of foreign competitors has a long history in international trade (Jackson, 1989). A special provision, Article VI, was made for dealing with dumping-related trade disputes in GATT in 1947. The GATT provision allows GATT member countries to use antidumping duties to offset the margin of dumping of dumped goods as long as such dumping is causing or threatens to cause ‘material injury’ to a domestic industry producing similar or competing goods. The basic guideline of the WTO antidumping agreement is to prevent the use of dumping as non-tariff trade barriers and to promote free trade between its member countries. However, dumping and antidumping can be a double-edge sword. While dumping can construct trade barriers, antidumping measures can also be used as trade barriers. To prevent the abusive use of antidumping as trade barriers, the WTO Antidumping Agreement sets forth procedural requirements (investigation for determination of dumping) and substantive requirements (determination of material injury to domestic industry) and constraining the use of such antidumping measures. If, and only if, these two conditions are met, then the importing country may apply antidumping measures – imposing antidumping duties on the imported products. A failure to respect either the substantive or procedural requirements can be taken

212

The Chinese Economy after WTO Accession

to the WTO Dispute Settlement Body and can be the basis for invalidation of the measure. It is important to note that antidumping measures (mostly tariff duties) are not the only regulatory trade barriers. The WTO safeguard provisions also allow tariff increases and other measures to deter imports of products in high quantities and under other conditions that cause or threaten to cause injury to a domestic industry producing the same or similar products. A recent example was that South Korea and the US imposed safeguard measures on garlic imports from China and other agricultural products by excising more than 300 percent import duties, this resulted in a major decline in China’s garlic exports in the following years (Ning and Lester, 2001). It is also important to note that, unlike the Agreement on Subsidies and Countervailing Measures, the Antidumping Agreement does not establish any disciplines on dumping itself. It is primarily because dumping is a pricing practice engaged in by business enterprises, and this is not within the direct reach of multilateral disciplines (www.wto.org: The Antidumping Agreement). In brief, the WTO Agreement deﬁnes antidumping as an ‘escape clause’ allowing import-restraining responses regardless of whether imports have beneﬁted from unfair trade practice. The basic idea is to let the importing country offset the effects of the unfair actions. The problem with this provision is that antidumping, on one hand, can be used as an ‘escape clause’ for unfair competition, but, on the other hand, it sometimes can be used as non-tariff trade barriers – disserving the purpose of this provision. In the past half-century, the tariff trade barriers for developed countries have been reduced from 19 percent on average in 1950 to less than 4 percent in 2000. The major reduction in tariffs has promoted free trade worldwide. However, more nontariff measures have been used to protect home markets and domestic industries. Developed countries have used antidumping policies as a safeguard mechanism to offset imported goods to protect domestic industries from ‘unfair’ foreign trade competition for decades. And the US has been the leading user of large antidumping duties (Mah, 2000: 721). Strategic trade theories for developed economies suggest that free trade might not be optimal once government intervention to foreclose markets and shift the rules of global competition in favor of domestic ﬁrms (Grossman, 1986). It provides more rationale for the use of antidumping as a competitive strategy for a nation in international trade. Many developing countries voiced their dismay that the developed world has abused this provision against the interests of developing countries in the WTO Seattle meeting of 1999. Meanwhile, many developing countries, including China, have caught on to the effectiveness of this protective tactic, and have started their own antidumping investigations (White, Gelb & Jones, 2000).

China in the Antidumping War Against China

213

Characteristics of Antidumping Actions Analyzing the 2416 antidumping cases ﬁled with WTO since 1995, we have found the following ﬁve characteristics are of great importance for China to formulate its strategy in the antidumping war. Who China Has to Deal With Four developed countries, the US, EU, Australia and Canada initiated about 90 percent of antidumping investigations in the 1980s and about 37 percent for 19952003. They are referred to as the ‘big four’. Since 1995, four developing countries, India, South Africa, Argentina, and Brazil, became active in initiating antidumping litigations and they account for about 34 percent of the antidumping investigations. The four developed countries together with the four developing countries composed the ‘big eight’ in using antidumping duties measures to protect their markets from foreign competition. The eight members dominated the antidumping litigation and contributed 71 percent of the total 2416 investigations. In contrast, there are eight major affected parties with 90 cases or above, as shown in Table 11.2. Those eight members accounted for 49 percent of total cases. Among the eight affected parties, six were in developing economies and two (US and Japan) were in developed economies. The US, the largest economy and trading partner, was the only member country that played as both a leading initiator and a major defender in the antidumping war. In contrast, China is the single member that was subject to the largest number of investigations with no antidumping investigation system against its trading partners until 2000. South Korea was the second country subject to antidumping litigations. What Areas of Antidumping Disputes China Is Involved In The WTO Council for Trade in Goods puts all trading goods in 21 sections. As of December 31, 2003, the antidumping disputes for China’s goods reported in 17 sections, as shown in Table 11.3. However, the investigations were concentrated in two major areas: products of chemical or allied industries (Section VI) with 69 cases and base metals and articles of base metals (Section XV) with 62 cases. The disputes in these two sections represent the heavy market competition in chemical and metal related industries. The other areas of antidumping disputes include mechanical and electrical appliances (22 cases), miscellaneous manufactured articles (23), plastics and rubber products (15 cases), and small daily use items, such as footwear, umbrellas, artiﬁcial ﬂowers, etc. (12 cases). What Are the Bases of a Dumping Determination The methodologies normally used to determine whether an exporting country is dumping are home market (HM), constructed value of factors from a speciﬁed

214

Table 11.3

The Chinese Economy after WTO Accession

AD Measures Taken against China, by Sector (1/1/95–12/31/03)

AD Description Cases Sectors I Live animals; animal products 2 II Vegetable products 6 IV Prepared foodstuffs; beverages, spirits and vinegar; tobacco and 2 manufactured tobacco substitutes V Mineral products 5 VI Products of the chemical or allied industries 69 VII Plastics and articles thereof; rubber and articles thereof 15 1 VIII Raw hides and skins, leather, furskins and articles thereof; saddlery and harness; travel goods, handbags and similar containers; articles of animal gut (other than silk-worm gut) IX Wood and articles of wood; wood charcoal; cork and articles of cork; 1 manufactures of straw, of esparto or of other plaiting materials; basketware and wickerwork X Pulp of wood or of other ﬁbrous cellulosic material; recovered (waste and 3 scrap) paper or paperboard; paper and paperboard and articles thereof XI Textiles and textile articles 11 XII Footwear, headgear, umbrellas, sun umbrellas, walking-sticks, seat-sticks, 12 whips, riding-crops and parts thereof; prepared feathers and articles made therewith; artiﬁcial ﬂowers; articles of human hair XIII Articles of stone, plaster, cement, asbestos, mica or similar materials; 10 ceramic products; glass and glassware XV Base metals and articles of base metals 61 XVI Machinery and mechanical appliances; electrical equipment; parts 22 thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles XVII Vehicles, aircraft, vessels and associated transport equipment 4 XVIII Optical, photographic, cinematographic, measuring, checking, precision, 4 medical or surgical instruments and apparatus; clocks and watches; musical instruments; parts and accessories thereof XX Miscellaneous manufactured articles 23 Total 254 Source: WTO, Statistics on Anti-Dumping, September, 2004. http://www.wto.org/english/ tratop_e/adp_e/adp_e.htm

country (CV) or cost of production analysis (COP). In contrast, the methodologies used against China are more discriminatory in nature. Most of the antidumping investigations against China initiated by the US were based on non-market economy (NME). Using non-market economy as a base to ﬁle antidumping itself is discriminatory against nations with transitional economy. The presumption is that the Chinese economy is in transition from a planned economy to a market economy, not all of its domestic prices fully reﬂect the supply and demand. Thus the comparisons of the price of Chinese goods at home with the prices of the

China in the Antidumping War Against China

215

same goods sold abroad would not necessarily indicate whether the good is being sold at less than ‘normal value.’ Thus the American producers might be forced to compete unfairly with the Chinese producer that had subsidized inputs. This non-market economy methodology discriminates against China in three ways: (a) Labor costs used are much higher than the actual labor costs prevailing in China. These approaches discount Chinese producers’ comparative advantage over their worldwide competitors; (b) Firms in market economies sometimes sell their products for less than average total cost, but at a price to cover the marginal cost of production. The constructed value method or the US Department of Commerce always includes proﬁt when it calculates normal value. Thus, they can establish that the Chinese goods are sold at ‘dumping’ prices; and (c) WTO rules do not deﬁne the market economy conditions. Each member has broad discretion in setting or even changing the conditions under which it applies non-market economy provisions in antidumping cases against Chinese ﬁrms. The antidumping cases ﬁled by the EU were mostly determined by speciﬁed third country market (TM). The presumption is that factor value of inputs to establish the normal value for a speciﬁc good is not available or the information provided by the Chinese producers is not reliable. Instead, they use costs of production in a third ‘surrogate’ country to calculate the ‘normal value’ of Chinese exports. The problem is that the third country selected, such as the US or Mexico, whose prices are used in the constructed value approach sometimes have labor or material costs that are much higher than those prevailing in China. Thus with this method, the normal value for a Chinese good in question is often above the selling price and thus the Chinese enterprise becomes the subject of an antidumping penalty. The major problem in non-market economy and third country approaches used to determine Chinese enterprises’ dumping behavior is they totally refuse to recognize China’s domestic costs of production in comparison to its price of the good at the market of the importing country. These methodologies disadvantage China in various ways in antidumping investigations. Longevity of Antidumping Measures against China As mentioned earlier, antidumping action is designed to be temporary to offset price discrimination against domestic producers. Article 11.3 of the Anti-Dumping Agreement states that ‘any deﬁnitive antidumping duty should be terminated on a date not later than ﬁve years from its imposition…’ unless an (obligatory) review ﬁnds that the expiry of the duty would be likely to lead to continued or recurred dumping. However, it was actually used by some of China’s trading partners as a long-term non-tariff barrier to prevent domestic producers from trade competition. There was a total of 42 antidumping duty orders imposed by the US against China in effect as of December 31, 2001. Among the 42 cases, 10 started to be in effect in the 1980s. Two cases related to cotton cloth and towels were in effect as early as 1983. The longevity of the ﬁnal antidumping measures imposed constitutes a severe problem. In the US the average duration of ﬁnal duties was more than 9 years. The

216

The Chinese Economy after WTO Accession

oldest duty has been in force for more than 32 years as of 2001. Over 90 percent of all US measures lasted more than 5 years (Neufeld, 2001). EU duties in force have an average of 3.5 years and some were as long as 8 years. The existence of these long-term orders indicates that the obligation to conduct 5-year reviews of antidumping duties did not show the anticipated results. In other words, the ‘sunset provisions’ of the Anti-Dumping Agreement have failed to signiﬁcantly reduce the number of duties stemming from the pre-Uruguay era (Neufeld, 2001). Antidumping Duty Level for China Based on the data available for deﬁnitive duties imposed, according to the WTO Committee’s report on antidumping practices (March 2002), the average antidumping duty that the US imposed on China for the second semester of 2001 was 47.6 percent. The antidumping duty that the EU imposed on China was about 28.5 percent on average. It was 9.6 percent and 6.8 percent average for the antidumping duties that the US imposed on South Korea and Taiwan respectively for the same time period (second semester of 2001). Overall the average antidumping duty level imposed was much higher on Chinese ﬁrms than ﬁrms from other countries or customs territories. Final Measures against China Research has shown that most of the investigations would not necessarily lead to ﬁnal measures due to either a lack of evidence of dumping or injury (Neufeld, 2001). In contrast to the increased volume of investigations, the success ratio, deﬁned as the composition of ﬁnal measures, is relatively low. According to Neufeld (2001), out of all antidumping investigations initiated in 1998, only 11.6 percent resulted in the imposition of ﬁnal measures. The 1999 data shows an even lower ‘success level’ – 5.4 percent. For the US cases, 80 percent of all investigations ended without ﬁnal measures were because of lack of injury, and 6.6 percent of investigations found no dumping. For EU cases, among all investigations without ﬁnal measures, 22 percent for no injury found, 26 percent for withdrawing, and 22 percent were terminated for expiry of the deadline to impose a deﬁnitive measure. For the period 1995–2003, the US and EU were the leading parties frequently imposing ﬁnal measures on their trading partners, especially partners in developing economies. The very low success ratio demonstrates that the antidumping actions appear to be a method of harassment with an expectation that the mere opening of an investigation has a signiﬁcant impact on the affected countries’ imports. The petitioners themselves may assume that the investigations will not lead to the imposition of a ﬁnal measure. However, China has been most often affected by ﬁnal measures, being the target of more than 17 percent of all cases for the period of 1995–2001. As shown in the third column of Table 11.1, 72 percent of the total cases ﬁled against China had

China in the Antidumping War Against China

217

been imposed on the ﬁnal measures, which is an extremely high ‘success rate’ in comparison with any other country. For instance, among the 42 cases initiated by the US in the second semester of 2001, there were 19 impositions of ﬁnal measures. A major reason is that the US treated China as a non-market economy. The US ﬁrms did not have to provide the evidence of ‘dumping’ and ‘material injury’ to establish the cases. The high success level may also be a result of the fact that the Chinese producers and exporters seldom chose to respond to the charges and to defend their interests before China joined WTO. The above six attributes of all antidumping cases indicate that China was discriminated against by its major trading partners and by the US particularly. The main reasons for such discrimination include its non-WTO membership, non-market economic system, and inability to defend itself in antidumping investigations. It is widely expected that the Chinese disadvantageous position will be reversed after its entry into WTO. But, unfortunately, China accepted discriminatory treatments in its WTO entry protocol, which further complicates the trading environment and restricts China’s ability to protect itself. China Accepted Discriminatory Treatments In its WTO accession negotiation, China agreed to fully comply with market access provisions. It has made a full commitment to reduce or remove tariff and non-tariff barrier and to open its markets to foreign competition. In addition to its market access commitment, China accepted discriminatory treatment in two important area: safeguards and antidumping, restricting its export policy and behavior (Lardy, 2002). Under certain conditions set forth in the WTO Agreement on safeguards, a country may impose quantitative restrictions on imports. Since this is a major departure from the most basic WTO principle of eliminating all-quantitative trade restrictions and promoting free trade, the country imposing the restriction must demonstrate that increased imports have caused or threaten to cause serious injury to domestic ﬁrms producing similar or competing products. Except under special circumstances, restrictions on imports imposed under a safeguard measure must be applied on a most-favored-nation basis, that is, proportionately on all suppliers. Thus an import restriction imposed under the WTO general safeguard provision cannot be imposed only on goods originated in a single country (Lardy, 2002: 81). However, in its entry negotiation, in exchange for its membership, China made a concession by allowing WTO members to apply the terms of this transitional product-speciﬁc safeguard for a period of 12 years from the time of accession, which is a far bigger burden than those that have been imposed on any other country as a condition for WTO membership. Only after 2012, will China be subject to the less onerous provision of the WTO Safeguard Agreement. Clearly, under the terms of the transitional product-speciﬁc safeguard clause in China’s protocol of accession to the WTO, it will be fairly easy for the US and other countries to impose restrictions on

218

The Chinese Economy after WTO Accession

goods imported from China. During this period, China is disarmed from responding to such restrictions including the unilateral imposition by the importing country of increased tariffs or even quotas. (Lardy, 2002: 82). Lardy argued that China’s ability to retaliate when this clause is invoked is more restricted than under the WTO safeguard. Once a restrictive quota has been imposed against Chinese imports, there is no requirement for progressive liberalization for the quota over time. Furthermore, China agreed to accept discriminatory terms for antidumping in its protocol of WTO accession, another important area in protecting its trading rights. China agreed in the protocol of WTO accession to allow the US and other WTO members to use the non-market economy (NME) methodology in dumping cases for 15 years from the time of its accession. Thus the US and other WTO members have broad discretion in determining the conditions under which they apply non-market economic provisions in antidumping cases against Chinese ﬁrms. In fact, the US has treated China as non-market economy in most of its antidumping cases for many years. With the discriminatory provision, the US is going to continue its practice to 2016 at least. Overall it was a tradeoff for China. On balance, there might be gains for China to be a WTO member. The membership may speed up the integration of the Chinese economy into a world market economy and it would create the potential for gains in economic efﬁciency through fair competition in global markets. In addition to the economic gains, China will be able to use the WTO dispute settlement procedure to protect its national interests. Antidumping is an important area where China can use the system to protect itself. Before joining WTO, it was difﬁcult to challenge the dumping charges because it was not a WTO member. Now China can use the WTO antidumping procedure and dispute settlement platform to deal with its trade partners more fairly. The Impact of the Antidumping War As analyzed above, to a large degree, antidumping has been abused and has a variety of negative implications. It creates serious distortion on international trade and fair competition with a damaging effect to exporting ﬁrms, especially the exporting ﬁrms in developing economies. The AD duty measures or threatening to take retaliatory measures may lead the exporting ﬁrms to change production and/or seek for new markets. On the other hand, instead of being an ‘escape clause’, antidumping has been frequently used as a non-tariff trade barrier and protects the domestic market from competition of goods and services imported. It encourages rent-seeking behavior by import-competing ﬁrms through trade defense measures. Consequently, the lack of international competition results in a higher price that the consumers have to pay for their consumption of goods and services. China, as the major AD target, is especially hurt by the AD actions. The Chinese producers are particularly vulnerable to the adverse effect of the AD actions because

China in the Antidumping War Against China

219

they lack the international trade experience and ﬁnancial capabilities that the wellestablished producers and exporters in developed economies use to defend themselves in the international arena. A good example is the AD actions taken by the US, South Korea and Japan against China’s garlic export. China is the largest garlic producer in the world with an annual output of 59 million metric tons, which was about 66 percent of the total world production. China’s fresh garlic export has created major trade disputes with its trade partners, including the US, South Korea and Japan. These countries have ﬁled several antidumping investigations against China. Starting in 1994, the US government decided to excise 376.67 percent antidumping duties on garlic from China for 5 years. In the 5-year review in 1999, no Chinese garlic ﬁrms attended the hearing and so the high import duty has been automatically retained to date. Even worse, in 2002 the US government suspected that the large volume of garlic imported from Thailand in the previous two years might have originated from China and started an investigation on this. The Japanese and Korean governments have taken or threatened to take similar actions. The sharp increase in duty and the sanction have resulted in a sharp decrease of China’s fresh garlic export. Consequently, Chinese garlic producers and garlic exporters have become the immediate victims of the antidumping actions. China’s Antidumping Actions After years of being charged for dumping, China promulgated Anti-Dumping and Anti-Subsidy Provision on March 23, 1997 with immediate effect. The aim of the regulation was to hit back with its own antidumping and countervailing duty laws. The Minister of Foreign Trade and Economic Cooperation (MOFTEC) was in charge of initiating dumping investigations and reporting the ﬁndings and the State Economic and Trade Commission (SETC) was in charge of the investigation of industrial injury. However, this regulation did not specify the legal procedure. According to the requirements for WTO membership and the promise China made in its WTO accession agreement, China promulgated a revised antidumping and antisubsidy provision on November 26, 2001, in which it speciﬁed the legal procedures for investigation and administrative ruling. Its Supreme Court also issued its related regulations for court appeal on December 3, 2002.1 Chinese industries, learning from their counterparts in other countries, have increasingly invoked the antidumping regulatory measures to curb the impact of imported products. China has become active in initiating AD investigations against its trading partners after China became a WTO member in November 2001. To protect its chemical producers against cheap imports, the Chinese government has taken several antidumping measures against its trading partners. In November 2000, MOFTEC imposed antidumping duties of 32 percent on acrylic ester imports from Japan and the US for a ﬁve-year period. Acrylic ester imported from Japan

1

China Daily (Overseas edition), December 14, 2002, p. 3.

220

The Chinese Economy after WTO Accession

and the US exceeded 29 percent of the Chinese market in 1999 (Mollet, 2001: 20). MOFTEC imposed provisional antidumping duties on imports of methylene chlorides from six countries and the duties range from 28 percent to 75 percent (Chemical Week, August 22, 2001, p. 15). China has a major trade conﬂict with Japan over imports of PVC (polyvinyl chloride) to the Chinese market. Japanese chemical industry exports about onethird of total Japanese PVC products to China. As part of a punitive penalty, China turned away a Japanese cargo of PVC in May 2000 from ports in southern China in retaliation against the Japanese restriction imposed on leeks, mushrooms, and towels from China (Mollet, 2001). In the past, Chinese companies seldom responded to foreign antidumping investigation. In 2002, the response rate on foreign antidumping investigations has been improved to 70 percent. The response rate for dumping charges from the US and EU was raised to 100 percent (Xie Yuandong, 2002). Recently, Chinese enterprises won four out of ﬁve cases in their responding to antidumping investigations ﬁled in the US (Xie Yuandong, 2002). It indicates that China is exhibiting an increasing willingness and forceful efforts to use its antidumping laws to protect its industries and enterprises. Strategic and Policy Options for China and Chinese Enterprises Based on the above analysis, the following strategic and policy suggestions are made for Chinese government and Chinese enterprises involved in international trade. The Strategic Use of Antidumping It has become clear that the WTO antidumping agreement has not been implemented as it was intended, to temporarily offset unfair competition arising from price discrimination and to provide a remedy to the related injury. In practice, it has been used as a safeguard measure to protect domestic producers from open competition. When China’s exports to the US increased substantially in many sectors in the 1980s and 1990s, the related US manufacturers suffered from the sharp increase of imports of low-price goods. The US also suffers from the increasing trade deﬁcits. In response, the US government increased its AD measures against China more than any other country. It seems that this antidumping behavior was justiﬁable. However, many of the dumping allegations made by the US International Trade Commission (ITC) against China had little linkage to the alleged ‘injury’ and the AD duties were reinforced for extensive time periods. It indicates that those AD actions were taken as harassment tactics to deter or block the imports from China. The Chinese enterprises should realize the intrinsic problem embodied in WTO AD agreements and practice and prepare themselves for both eventualities (dual tactics). On the one hand, they should acknowledge that it is common practice for governments to subsidize exports in some sectors and/or some products were in fact

China in the Antidumping War Against China

221

sold in foreign markets at a price lower than their ‘normal value.’ Such ‘dumping’ creates market distortion with damaging effects on trade and competition. On the other hand, they should be aware that antidumping can be strategically used as trade protection measures by importing countries which encourages rent-seeking behavior by domestic ﬁrms producing competing products. Assistance in Antidumping Actions Quite often the Chinese producers and the exporters are the direct victims of AD actions. They usually do not have experience of the international marketing and legal framework; they have little expertise to predict the international trade relations and limited ﬁnancial capabilities to defend themselves in AD investigations. Very often an on-going AD case or a mere threat to open an AD investigation can induce the producers to withdraw their exports although they may win the case if they choose to ﬁght on. And the importers are scared off by the possible prolonged battle and seek alternative sources of supply. Additionally, Chinese producers and exporters have much less capability to absorb the negative economic consequences caused by AD duties and other retaliations than well-established exporters in developed economies. In many cases, the exports from China sharply declined or completely stopped after AD investigations or AD duties were imposed. The Chinese government should provide AD-related legal and administrative services to the domestic producers and exporters. The government should help them to be better organized and better trained for taking AD actions, and assist them to respond to alleged charges and to apply for investigation against foreign dumping and other unfair competition. Being Active in the Antidumping Club As discussed earlier, antidumping is no longer a weapon solely equipped for developed economies. More and more developing countries have become the signatories of WTO AD agreements and have aggressively armed themselves with the AD code and used it as a major instrument for regulating imports. Argentina, India, and South Africa were the pioneers, each having initiated more than a hundred AD cases since the WTO was formed. They have become active members in the international AD club and are quite effective in protecting their trade interests. Many other developing countries, such as Brazil, Mexico and Chile, are increasingly playing active roles in the AD war to protect their domestic markets from international competition. China should learn from those developing countries. Chinese government and Chinese enterprises should closely monitor their own markets for possible dumping behavior of foreign companies and related industrial injury. Instead of being passive AD targets, they should be armed with antidumping regulations and be more active in initiating AD investigations against unfair trade practice from their trading partners, while encouraging and assisting Chinese ﬁrms

222

The Chinese Economy after WTO Accession

to actively defend themselves in litigations against them, especially the litigations from the ‘big four’ mentioned above. Reform of the AD Agreement A growing tendency can be found where developed economies use AD measures against imports of specialized commodities or products from China, such as when the US’s import of Chloropicrin from China was shut down with its duty enforcement. AD measures are also used against some major Asian companies, which are often major suppliers to a region or even the world markets. As Prusa (1999) pointed out, ‘ …it often seems that just when developing countries begin to efﬁciently operate and become more competitive in a particular markets, industrialized countries shut down those precise markets…’. In its current format antidumping turns out to be a policy of anti-competition rather than pro-competition. It in fact discriminates against developing economies. WTO AD Agreement needs reform to undo its tremendous negative effects on exports from developing economies. The Chinese government should actively promote and participate in the reform of WTO AD Agreement and its ill-fated applications. First of all, the requirements for opening an AD investigation should be tightened by raising the threshold for the determination of ‘dumping.’ Given the signiﬁcantly lower labor costs and other factor endowments, there should be different dumping margins set for developed and developing economies. Furthermore, as Neufeld (2001) proposed, unjustiﬁed AD investigations should be eliminated. A stricter test of injury should be conducted before the opening of an investigation and the injury determination should be more on a ﬁrm-speciﬁc basis. It should be mandatory to consider the inﬂuence of factors such as price range in normal business conduct, quality difference, and exchange rate ﬂuctuation. In addition, the life expectancy of AD orders should be reduced to 2-3 years and a new re-imposition of a duty should be subject to a complete new investigation with examination of all relevant factors from the beginning. Taking a leading role in reforming WTO antidumping could both minimize China’s exposure to foreign AD actions and improve its trade outcomes. The Doha round of trade negotiation has offered China great opportunities to negotiate stricter disciplines both on WTO contingent protection and on the use of the special provisions included in China’s accession protocol (Messerlin, 2004). Conclusions It is less debatable that WTO accession will lead China to greater potential for economic efﬁciency and will beneﬁt Chinese consumers. China’s trading partners have and will lift most of their quantitative import restrictions on a range of products from China. China can resort to the WTO dispute settlement mechanism to protect its trade interests and is able to participate in multilateral negotiation on trade rules and trade liberalization. China’s long-term gains from accessing foreign markets, however,

China in the Antidumping War Against China

223

could be eroded by its acceptance of discriminatory provisions in safeguarding and antidumping areas and its embattlement of the escalating antidumping war. A trading partner can invoke a product-speciﬁc safeguard in the next 12 years to impose restriction on Chinese imports based only on ‘market disruption or threat to market disruption’ without proof of ‘injury.’ And the other trading partners can take actions to prevent diversion of China’s exports to their countries even without establishing evidence of market disruption. Any trading partners in the next 15 years can use non-market economy methodology to bring dumping charges against China. These discriminatory provisions can be serious threats for exports of Chinese enterprises. China had already been the target of antidumping actions before its WTO accession. It remains the major target after its accession so far. Coping with antidumping actions of its trading partners presents a greater challenge for China as a new WTO member with a commitment to fully comply with market access provisions. It would be particularly difﬁcult for China in the short and medium term to implement its commitment to open up its market to foreign competition. But China is now in a more favorable position to protect trade interests than ever before. Before WTO accession, China was helpless to challenge antidumping litigation. Now as a WTO member China can use the WTO trade dispute settlement mechanism to challenge its trading partners that do not apply fair procedures or the decision-making process is not transparent. China can also challenge its trading partners if they do not use the non-market economy methodology in a non-discriminatory manner to all non-market economies. Besides defending itself, more importantly, WTO membership enables China to actively initiate antidumping investigations against foreign companies to protect its vulnerable industries and to bring the disputes to WTO for settlement. China and Chinese enterprises need to formulate their antidumping strategy and be fully prepared for the antidumping war. With a clearly deﬁned strategy and proper organizational resources, China will be able to maximize its beneﬁts as a WTO member and to minimize possible damages in the antidumping war. References Grossman, G.M. (1986), ‘Strategic Export Promotion,’ in Krugman, P.R. (ed.) Strategic Trade Policies and the New International Economics, MIT Press, Cambridge, MA, pp. 47–68. Jackson, J.H. (1989), The World Trading System: Law and Policy of International Economic Relations, MIT Press, Cambridge, MA, pp. 217–48. Jiang, Bin and Alexander Ellinger (2003), ‘Challenges for China—the world largest antidumping target,’ Business Horizon, Vol. 46, No. 3, p. 25. Lardy, Nicholas R. (2002), Integrating China into the Global Economy, Brookings Institution Press, Washington, DC. Mah, Jai S. (2000), ‘The United States’ Antidumping Decisions against the Northeast Asian Dynamic Economies,’ The World Economy, Blackwell Publishers Ltd. MA, pp. 721–32.

224

The Chinese Economy after WTO Accession

Messerlin, Patrick A. (2004), ‘China in the world trade organization: antidumping and safeguards,’ The World Bank Economic Review, Vol. 18, No. 1, p. 105. Mollet, Andrew (2001), ‘Making Greater Use of Antidumping Duties,’ Chemical Week, New York (August 29-September 5), p. 20. Neufeld, Inge Nora (2001), ‘Antidumping and Countervailing Procedures—Use or Abuse? Implications for Developing Countries,’ United Nations Conference on Trade and Development: Geneva, Switzerland, Policy Issues In International Trade and Commodities Study Series No. 9. Ning, Susan and Lester Ross (2001), ‘Perfecting Protectionist Procedures: An update on China’s Antidumping Regulations,’ The China Business Review, Washington DC (May/June), pp. 42–43. Prusa, T.J. (1999), ‘On the Spread and Impact of Antidumping,’ NBER Working Paper Series, Working Paper 7404. White, Z. Ying, Catherine Gelb and Katherine M. Jones (2000), ‘The Sun Sets on US Antidumping Orders,’ The China Business Review, Vol. 27, Washington DC (May/June), pp. 34–39. Xie Yuandong (2002), ‘China calmly ﬁghts antidumping battles,’ People’s Daily (Overseas Edition), December 14, p. 3.

Chapter 12

Price Competition in the Chinese Pharmaceutical Market Y. Richard Wang

Introduction Previous economic studies on pharmaceutical price competition found wide variations across major developed countries. In the mostly free-pricing US, the price of originator products remains unchanged or even rises slightly in response to generic competition, while the price of generic products decreases with more generic entries and presumably moves toward the marginal cost of production (for example, see Caves, Whinston and Hurwitz, 1991, Grabowski and Vernon, 1992, and Frank and Salkever, 1997). Consequently, generics typically account for the bulk of molecule sales shortly after patent expiration in the US. In an extensive cross-country comparison, Danzon and Chao (2000) found that generic competition leads to lower prices in unregulated or less regulated countries such as the US, the UK, Canada, and Germany but is ineffective in countries with strict price regulation such as France, Italy, and Japan. However, their evidence on therapeutic competition, i.e., between products of similar molecules in the same class, is less conclusive due to selection biases associated with entry decisions (Danzon and Chao, 2000). The pharmaceutical market in developing countries differs from that in developed countries in many ways (Kremer, 2002). Beyond low income and lack of drug coverage, many developing countries offer inadequate intellectual property right (IPR) protection for pharmaceuticals and their markets are ﬂooded with low-quality generics or even counterfeits. Although per capita pharmaceutical expenditure is small, pharmaceuticals typically account for over one half of the total healthcare dollars in developing countries. There is little empirical evidence on pharmaceutical price competition in developing countries. The pharmaceutical market in China shares some problems of the developing world (IMS Health, 2004). Over 75 percent of the population is uninsured, therefore having no drug coverage. Personal or out-of-pocket expenditures accounted for 60.5 percent of national healthcare expenditures in 2001. In 2002, 55.4 percent of outpatient expenditures and 44.4 percent of inpatient expenditures were spent on pharmaceuticals. Although the total pharmaceutical market reached US$ 7.5

226

The Chinese Economy after WTO Accession

billion at the ex-manufacturer level in 2003 and is the second largest in Asia, GDP per capita is only US$1,045 and pharmaceutical expenditure per capita is only $5.8 in 2003. However, there are large regional variations, with higher income and more pharmaceutical spending in the east-coast provinces and big cities. Patent protection for pharmaceuticals became law in China in 1993. Also since 1993, China has been granting administrative protection up to seven and one half years for foreign patents awarded between 1986 and 1992 (Wang, Ji and Lin, 2003). Attracted to its booming economy, large market size, and improving IPR protection, all major global, research-based pharmaceutical companies sell their products in China and many also have local manufacture facilities. In addition, there are over 2,000 local generic companies in China. Many are small-scale operations producing low-quality generics of older molecules, a legacy of China’s policy of self-sufﬁciency. As local generics are not required to be bioequivalent for approval, they are considered to be of lower quality than global products. In contrast, global products in the same class are deemed similar in quality and better substitutes. In China, most outpatient visits take place in the hospital outpatient setting and patients typically ﬁll prescriptions in the hospital’s pharmacy. Even though retail pharmacies have been growing in recent years, hospital pharmacies account for about 85 percent of all dispensed drugs. Pharmaceutical sales are a hospital’s main revenue source, typically accounting for over one half of its total revenue. Pharmacists do not routinely substitute originator products with local generic products, in part due to quality difference. The Chinese government regulates pharmaceutical pricing through maximum retail price, which is mainly based on originator status, quality, and manufacture cost. However, companies can apply for independent pricing if they can present evidence of better quality, presumably easier for global or originator companies. For those with public health insurance, there is a national reimbursement bulletin for prescription drugs, but almost all products are listed. Regional governments and hospitals only use the national reimbursement bulletin as a guide in developing their own formularies. Because global products are considered to be of higher quality than local products, they often fall under a separate category in these formulary decisions. Theoretical Framework China provides an interesting market to study price competition between highquality global products and low-quality local products, a unique feature of the pharmaceutical market in developing countries. For a global company maximizing local proﬁts, the price of a global product B is a function of product characteristics X such as molecule age and strength, number of local generic competitors NG, and number of global therapeutic competitions in the same class NTB, i.e., PB = f (X, NG, NTB)

(1)

Price Competition in the Chinese Pharmaceutical Market

227

Similarly, for a local company, the price of a local product G is a function of molecule characteristics X, number of local generic competitors NG and number of local therapeutic competitors in the same class NTG, i.e., PG = g (X, NG, NTG)

(2)

We hypothesize that local generic competition drives down local product price but not global product price, i.e., | d PG / d NG | >> |d PB / d NG | 0

(3)

As local generics are not bioequivalent and are deemed of lower quality, global companies are shielded from perfect generic competition. In addition, global companies tend to have a higher cost structure than local companies. It may be more proﬁtable for global companies to focus on less price-sensitive segments of the population, e.g., those with higher-income or better drug coverage, than to compete solely on price. On the other hand, local generic products are similar in quality and compete solely on price, so their prices decrease with more local competitors and move towards the marginal cost of local production. In addition, we hypothesize that global and local products react differently to therapeutic competition from similar products in the same class. If local generic competition already drives price towards the marginal cost of production, local therapeutic competitors have no additional effect. On the other hand, for a global product, similar global products in the same class pose real competition, as they are considered similar in quality. In other words, we hypothesize that | d PB / d NTB | >> |d PG / d NTG | 0

(4)

Data and Methods Our study sample is based on IMS Health’s MIDAS database, speciﬁcally its Chinese Hospital Pharmaceutical Audit (hereafter called MIDAS China). IMS Health is an international pharmaceutical market research company headquartered in Plymouth Meeting, Pennsylvania, USA. Started in 1999, MIDAS China is a quarterly survey on hospital purchase of pharmaceuticals. It stratiﬁes hospitals based on total number of beds and covers approximately 11 per cent of the total beds in the 6,862 hospitals with 100 or more beds in China. The price reported in MIDAS is the weighted average hospital purchase price. Main variables include product name, company, molecule, therapeutic class, quarterly sales in RMB (local currency, with 1 US$ roughly equal to 8.20 RMB), and quarterly sales in standard units (one capsule, one tablet, etc). Our study period covers the 16 quarters between 1999 and 2002. We deﬁne a product as a unique combination of molecule, therapeutic class, company, and

228

The Chinese Economy after WTO Accession

product name. Therapeutic class is deﬁned as a unique 3-digit code according to the IMS Health anatomical therapeutic classiﬁcation (ATC) system, for example, C10A for cholesterol-lowering drugs. The unit of observation is a product in a speciﬁc quarter, with a maximum of 16 observations for the same product. We deﬁne companies with positive sales in one of the top 5 markets worldwide, i.e., the US, Japan, the UK, Germany or France, as global companies and their products in China as global products. The other companies and products in China are deﬁned as local companies and local products respectively. We limit the study sample to singlemolecule products whose molecules are also sold in the US or the UK. The two countries have higher regulatory standards for safety and efﬁcacy, so their approved molecules are likely globally important ones (Danzon, Wang and Wang, 2003). We also limit the study sample to molecules with one and only global product in a given quarter, so as to focus on the originator product and its local generic competitors. It should be noted that products of non-US-UK molecules, molecules without a global product, and molecules with more than one global product are included in variables of therapeutic competition. Our main interest variable, product price, is deﬁned as the weighted average price per standard unit, aggregated over all formulations (excluding injectables) and packs. In MIDAS China, sales are reported in thousands in both local currencies and standard units (SUs). To reduce measurement errors, we only calculate price for products with sales over 10,000 RMBs and over 10,000 SUs in a quarter. The distribution of price is approximately lognormal, so we use the log transformations of price and all linear explanatory variables (plus 1 for measures of local generic competition and therapeutic competition, to be deﬁned) in regression analysis. Our analytical framework is similar to that of Danzon and Chao (2000). The explanatory variables for product price include molecule age, average strength, average pack size, number of local generic competitors, and number of global or local therapeutic competitors. As the launch date is missing for most products in MIDAS China, we use the earlier launch date in the US or the UK as a proxy for the ﬁrst launch worldwide date and deﬁne molecule age as the difference in months between this US/UK date and December 2002, the end of our study period. Average strength is deﬁned as the mean milligrams of molecule per standard unit. Average pack size is deﬁned as the average number of standard units per pack. Local generic competition is deﬁned as the number of local generic competitors for the same molecule. For a global product, therapeutic competition is deﬁned as the number of other global products for different molecules in the same class. For a local product, therapeutic competition is deﬁned similarly as the number of other local products for different molecules in the same class. Other explanatory variables include 13 main therapeutic class (1-digit ATC) indicators, e.g., C for cardiovascular drugs, and 16 quarter/year indicators. The above analytical framework differs from that of Danzon and Chao (2000) in three aspects. First, MIDAS China does not contain a launch date for most products, so we cannot deﬁne other measures of competition, i.e., generic entry lag and therapeutic substitute molecule entry lag. In addition, we run global and local

Price Competition in the Chinese Pharmaceutical Market

229

product price regressions separately because local products are not bioequivalent. Danzon and Chao (2000) ran pooled analyses of all products in a country and their results for the US mainly reﬂect generic competition. We will compare results of local generic competition in China with theirs for the US. Finally, we use all 16 quarters of data available while Danzon and Chao used the full-year data of 1992 in a cross-sectional analysis. To adjust for repeated quarterly observations from the same product, we employ the generalized linear model estimated by the generalized estimating equations (GEE) method. Speciﬁcally, we use PROC GENMOD with its REPEATED SUBJECT statement in SAS version 8.01. As GEE is based on maximum likelihood estimation, we report a pseudo R-Square based on an OLS estimation with the same speciﬁcation. Similar to Danzon and Chao (2000), our measures of generic and therapeutic competition are potentially endogenous or subject to company entry decisions. These estimates are upwardly biased, if the extent of generic and therapeutic competition is positively correlated with price. China’s introduction of stronger IPR protection in 1993 may serve as an instrument to estimate the unbiased effect of local generic competition on global product price. A new drug typically takes 8-12 years of clinical development before market authorization. Therefore, marketing exclusivity for global products in China during our study period is likely administrative protection (for foreign patents awarded between 1986 and 1992). Administrative protection prevents new generic entries but does not shut down existing local competitors. Some global companies might apply for and receive administrative protection in time to preempt local generic competition, while others might not act until it is too late. For global products eligible for marketing exclusivity in China, the extent of local generic competition is exogenous, if it depends on company instead of molecule characteristics. We do not have access to individual product’s patent data and assume the middle point of 10 years for clinical development. Therefore, global products ﬁrst launched in or after 1996 in the US or the UK are assumed eligible for administrative protection in China. For this subgroup of newer global products, the estimated effect of local generic competition on global product price is unlikely biased. We do not have any instrument to test whether our measure of therapeutic competition is endogenous. Results Our study sample includes 229 unique global products and 983 unique local products falling in 13 main therapeutic classes (Figure 12.1). Over one half of the products (global and local combined) are cardiovascular (C) or alimentary tract/metabolism (A) products, while systemic hormones (H) and antiparasitics (P) jointly account for less than 1 percent of the full sample. Figure 12.2 describes the distribution of global and local product observations by quarter. The number of global product observations increases steadily over time, while the number of local product observations ﬂuctuates but remains mostly stable.

230

The Chinese Economy after WTO Accession

Figure 12.1 Distribution of Global (N=229) and Local (N=983) Products by Main Therapeutic Class (1-digit ATC code)

Figure 12.2 Distribution of Global and Local Product: Observations by Quarter (1999–2002) Note that a product is excluded in Figure 12.2 if its sales are less than 10,000 RMBs or 10,000 SUs in a quarter. Table 12.1 describes characteristics of global and local product observations respectively. Compared with local products, global products have a higher mean price, smaller mean molecule age, higher mean strength, and smaller mean pack size. In addition, there is much less local generic competition and less therapeutic competition for global products than for local products. All the above differences are statistically signiﬁcant (p<0.001) in Student t-test with unequal variances.

Price Competition in the Chinese Pharmaceutical Market

Table 12.1

231

Characteristics of Global and Local Product Observations

Variable Global Products (N=2447) Local Products (N=7020) Mean STD Mean STD Price per Unit (RMB) 6.12 8.82 1.29 3.73 Molecule Age (month) 225 137 290 135 Strength (MG) 221 1195 120 336 Pack Size (SU) 24.8 29.9 54.5 69.7 Generic Competition Local Generic Competitors 3.6 7.0 17.2 11.0 Therapeutic Competition Global Therapeutic Competitors 5.0 4.3 6.1 4.9 Local Therapeutic Competitors 42.3 42.5 54.6 47.6 Notes: STD = standard deviation; RMB = Renminbi (local currency); SU = standard unit; MG = milligram.

Table 12.2 reports the generalized linear model results on the determinants of global and local product price respectively. The signiﬁcant determinants of global product price include molecule age (p<0.001), pack size (p<0.001), and number of global therapeutic competitors (p<0.05). As we hypothesized, therapeutic competition, measured by the number of global therapeutic competitors, lowers global product price. In addition, older molecules are cheaper and there are signiﬁcant volume discounts for large packs. The effect of strength is insigniﬁcant, reﬂecting the mix of within-molecule effect (expected to be positive if the therapeutic effect increases with strength) and between-molecule effect (expected to be negative if more potent molecules have weak strength). As we hypothesized, the effect of number of local generic competitors is small and insigniﬁcant, indicating that the price of a global product is not responsive to local generic competition. Therefore, global products are more responsive to therapeutic competition than local generic competition in China. In contrast, the signiﬁcant determinants of local product price include molecule age, strength, pack size, and number of local generic competitors. It is interesting to note that, except for molecule age, our estimates of local generic competition in China are similar to those in the US (see Table 4 in Danzon and Chao, 2000). The similar parameters include strength (0.080 vs. 0.103), pack size (-1.167 vs. -0.946) and number of generic competitors (-0.567 vs. –0.503). It seems that the local generic market in China, with over 2,000 domestic companies, is highly competitive. The only difference is that molecule age is signiﬁcant in China but not in the US (-0.467 vs. –0.027). For off-patent molecules, generic product price reﬂects the marginal cost of production, which may not vary by molecule age. One possible explanation is that local products vary in quality – those for newer molecules are of higher quality and incur higher marginal cost of production. As we hypothesized, therapeutic competition has no additional effect on local product price.

232

Table 12.2

The Chinese Economy after WTO Accession

Generalized Linear Model Results on the Determinants of Global and Local Product Price

Variable Global Products (N=2447) Local Products (N=7020) Coefﬁcient SE P Value Coefﬁcient SE P Value Molecule Age -0.626 0.095 <.0001 -0.467 0.069 <.0001 Strength 0.031 0.028 0.276 0.080 0.022 <.001 Pack Size -0.564 0.080 <.0001 -1.167 0.050 <.0001 Generic Competition Local Generic Competitors -0.013 0.054 0.808 -0.567 0.054 <.0001 Therapeutic Competition Global Therapeutic -0.170 0.076 0.026 Competitors Local Therapeutic Competitors -0.034 0.045 0.455 Therapeutic Class Indicators Yes Yes Quarter/Year Indicators Yes Yes Pseudo R-Square 0.654 0.791 Notes: SE = standard error. The dependent variable and all explanatory variables (plus 1 for variables of local generic competition and therapeutic competition) are log-transformed.

Table 12.3 describes characteristics of the subgroup of newer global products ﬁrst launched in the US or the UK in or after 1996. We assume that these global products are eligible for marketing exclusivity in China. There are 41 unique global products and 339 global product observations in our study sample. Eighteen percent of these global product observations face local generic competition, with a maximum of 4

Table 12.3 Variable

Characteristics of 1996 and after Global Product Observations

Global Products >= 1996 (N=339) Mean STD Price per Unit (RMB) 9.46 12.16 Molecule Age (month) 62.0 15.7 Strength (MG) 62.0 100.3 Pack Size (SU) 15.1 14.8 Generic Competition Local Generic Available 18% 38% Local Generic Competitors 0.38 0.95 Therapeutic Competition Global Therapeutic Competitors 6.3 5.1 Local Therapeutic Competitors 55.6 55.5 Notes: STD = standard deviation; RMB = Renminbi (local currency); SU = standard unit; MG = milligram.

Price Competition in the Chinese Pharmaceutical Market

Table 12.4

233

Generalized Linear Model Results on the Determinants of Global Product Price for 1996 and after Global Products

Variable Global Products >=1996 (N=339) Model 1 Model 2 Coefﬁcient SE P Value Coefﬁcient SE P Value Molecule Age -0.148 0.174 0.394 -0.164 0.163 0.316 Strength 0.068 0.052 0.191 0.065 0.052 0.207 Pack Size -0.713 0.131 <0.001 -0.716 0.132 <.0001 Generic Competition Local Generic Available 0.165 0.239 0.492 Local Generic Competitors 0.105 0.225 0.640 Therapeutic Competition Global Therapeutic -0.267 0.142 0.061 -0.277 0.143 0.052 Competitors Therapeutic Class Indicators Yes Yes Quarter/Year Indicators Yes Yes Pseudo R-Square 0.743 0.745 Notes: Global products >=1996 are those ﬁrst launched in the US or the UK in or after 1996. SE = standard error. The dependent variable and all linear explanatory variables (plus 1 for variables of local generic competition and therapeutic competition) are log-transformed.

local generic competitors. This is consistent with the assumption that these global products are eligible for marketing exclusivity in China. Compared with the other global product observations (not shown in Table 12.3), this subgroup has a higher mean price, smaller mean molecule age, smaller strength, smaller pack size, fewer local generic competitors, and more global therapeutic competitors (all p<0.01 in Student t-test with unequal variances). Table 12.4 reports the generalized linear model results for this subgroup of newer global products. In addition to the speciﬁcation in Table 12.2 (Model 1), we substitute the number of local generic competitors with an indicator variable for local generic availability (1 for Yes and 0 for No) in an alternative speciﬁcation (Model 2). Results from both models are qualitatively similar to those of the full sample in Table 12.2. The effect of molecule age is negative but insigniﬁcant, owing to small age differences amongst these newer products. The effect of pack size is negative, indicating volume discounts for large packs. Neither the number of local generic competitors (Model 1) nor the indicator variable for local generic availability (Model 2) is signiﬁcant; in fact, both are positive. Both results verify that global product price does not decrease in response to local generic competition. The effect of number of global therapeutic competitors is negative and marginally signiﬁcant (p<0.10), indicating that global product price decreases with therapeutic competition.

234

The Chinese Economy after WTO Accession

Conclusions Nearly all previous studies on pharmaceutical price competition focused on the developed countries, in which generic products are bioequivalent. We provide empirical evidence on price competition between high-quality global products and low-quality local products in a developing country, i.e., China. We ﬁnd that local generic competition drives down local product price but has no effect on global product price, in part due to local generic products’ lack of bioequivalence. In addition, we ﬁnd that therapeutic competition lowers global product price but has no additional effect on local product price, as other global products in the same class are considered better substitutes than local generic products. Our ﬁndings shed light onto the nature of pharmaceutical price competition in developing countries. This study has two main limitations. First, our measures of generic competition and therapeutic competition may be endogenous or subject to company entry decisions. Similar results from the subgroup of newer global products eligible for marketing exclusivity verify our ﬁnding on how global product price responds to local generic competition, but we do not have any instrument for therapeutic competition. Our estimate of therapeutic competition therefore serves as the low boundary or underestimates the true competition effect, if entries are positively correlated with price. Secondly, we do not know how pricing and reimbursement regulations in China affect competition between global and local companies. It is generally believed that the current regulation schemes leave room for price competition. However, too strict price regulation may discourage generic and therapeutic entries or drive out competition. Both of these questions are important topics for future research. Policies that encourage bioequivalent local generics and accelerate global product approvals in China will enhance price competition and lead to signiﬁcant cost savings. The local generic market is highly competitive but fails to inﬂuence global product pricing, in part due to local products’ lack of bioequivalence. Without bioequivalent local generics, physicians and patients cannot tell whether it is low generic quality or the molecule itself that is responsible for treatment failure. With bioequivalent local generics, patients can simply shop for the lowest price without any quality concerns. Although local generic price may rise in absolute terms to reﬂect higher manufacture costs, generic competition should drive price towards the marginal cost of production and deliver the best value for off-patent molecules. The opportunity for cost savings is tremendous, considering that over 90 percent of the WHO-designated essential drugs are off-patent. Secondly, China’s current policy of requiring local clinical trials for global product approval undermines therapeutic competition. Both the cost of local clinical trials and the associated proﬁt loss (due to fewer years of marketing exclusivity) reduce the expected return, lead to launch delays and possibly fewer launches, and result in higher prices for existing global products. China should follow international guidelines and rationalize the requirement for local clinical trials, in order to reap the full beneﬁt of therapeutic competition.

Price Competition in the Chinese Pharmaceutical Market

235

Acknowledgments I wish to thank Patricia Danzon and Lizheng Shi for comments on an earlier version of this chapter, Zhihua Qiao for research assistance, Sha Ling for correcting the language, and Lisa Croll for administrative support. References Caves, Richard E., Michael D. Whinston and Mark A. Hurwitz (1991). ‘Patent expiration, entry, and competition in the U.S. pharmaceutical industry,’ Brookings Papers on Economic Activity: Microeconomics, pp. 1–66. Danzon, Patricia M. and Li-wei Chao (2000), ‘Does regulation drive out competition in pharmaceutical markets?’ Journal of Law and Economics, Vol. XLIII, pp. 311– 57. Danzon, Patricia M., Y. Richard Wang and Liang Wang (2003) ‘The impact of price regulation on the launch delay of new drugs: Evidence from twenty-ﬁve major markets in the 1990s,’ NBER Working Paper No. w9874, July 2003. Available at http://www.nber.org. Frank, Richard G. and David S. Salkever (1997), ‘Generic entry and the pricing of pharmaceuticals,’ Journal of Economics and Management Strategy, Vol. 6, No. 1, pp. 75–90. Grabowski, Henry G. and John M. Vernon (1992), ‘Brand loyalty, entry, and price competition in pharmaceuticals after the 1984 drug act,’ Journal of Law and Economics, Vol. XXXV, pp. 331–50. IMS Health (2004), IMS Market Prognosis, Asia 2004–2008, May, London, UK. Kremer, Michael (2002), ‘Pharmaceuticals and the developing world,’ Journal of Economic Perspectives, Vol. 16, No. 4, pp. 67–90. Wang, Y. Richard, Lei Ji and Y. Aileen Lin (2003), ‘Intellectual property right protection for pharmaceuticals in developing countries: A case study of administrative protection in China,’ Pharmaceutical Development and Regulation, Vol. 1, No. 4, pp. 277–82.

This page intentionally left blank

Chapter 13

The Performance of Commodity Futures Markets: A Comparison of China and US Wheat Futures Wen Du and H. Holly Wang

Introduction The prosperity of wheat trading in the world’s major commodity futures markets has provided an effective channel for market participants to hedge price risks and ensure proﬁts (Yang and Leatham, 1999). If the futures market is efﬁcient, price will reﬂect the equilibrium level in the spot market, and will help the formation of a rational market expectation of price in both the short run and the long run. China is the biggest wheat production and consumption country in the world (USDA, 2004). Wheat production, consumption and trade account for a major share of China’s food system. In 2002, the share of wheat in all grains is 27.0 percent, 28.9 percent, and 32.9 percent for production, consumption, and imports, respectively. Agricultural commodity futures markets emerged in China in the early 1990s, when China was stepping into an advanced phase of its market-oriented economic reform. China’s ﬁrst exchange market, the China Zhengzhou Commodity Exchange (CZCE) was founded in 1990.Wheat futures trading started in May 1993. CZCE is the only exchange trading wheat futures contracts in the country today. In 1999, CZCE accounted for 50 percent in total trading value and 49 percent in total trading volume of all commodity exchanges in China. Since 1997, the trading of wheat futures has experienced a stable growth except for 1999 (Figure 13.1).1 In 2002, the total trading value amounted to 225.25 billion Yuan2 and total trading volume was 18.27 million contracts. After about 10 years of development, the wheat futures price in CZCE is on the way to becoming an important indicator of China’s wheat price. The correlation between the spot price and futures price is as high as 0.96 and a strong association is identiﬁed between the wheat futures price of CZCE and that of Chicago Board 1 The low trading in 1999 was mostly affected by the regulatory change in CZCE, which was designed to discourage the mung bean trading. As a result, mung bean trading declined sharply and disappeared in the following years. 2 One US dollar equals about 8.3 Chinese Yuan.

238

The Chinese Economy after WTO Accession

Trading volume (Million Contracts)

Trading value (Billion Yuan)

Figure 13.1 Annual Wheat Futures Trading in CZCE (1993–2002) Source: CZCE 2002 Annals (http://www.czce.com.cn/home/annals.asp).

of Trade (CBOT) in the United States (CZCE Report, 2001). China’s wheat futures price became more important to the world after November 2001 when China obtained full membership in the WTO. Given the enhanced interrelation between China’s markets and the world market, China’s integration to the world is on a fast lane and has shown impacts on both sides. Facing challenges from major wheat exporters such as the US and Canada, China’s previously over-valued domestic wheat price is expected to undergo a downward shift. Meanwhile, the futures price in CZCE may become more volatile due to the stronger linkage to the world commodity markets and the unpredictable factors in the world economy, or less volatile because some irrational behavior of domestic traders that had a fairly strong inﬂuence on prices in the past (Durham and Si, 1999) will not be able to affect an integrated world market price. Founded in 1848 and by far the largest and most developed agricultural commodity market in the world, the CBOT in the US has been playing a leading role in the world commodity market. The wheat futures price in CBOT is highly volatile and directly reﬂects the supply and demand in both US and world markets. It has been one of the most important wheat price indicators in the world market. In this chapter, the CBOT wheat futures market is chosen to represent the world market and the integration of

The Performance of Commodity Futures Markets

239

CZCE to the CBOT will be analyzed as an approximation of CZCE’s integration to the world market. This research is a quantitative assessment of China’s wheat futures price performance and the integration of China’s wheat futures market to the world market. The objective is to identify the best time series models to characterize the price behavior in both CZCE and CBOT, and the interrelationship between them. We then use the identiﬁed models to compare the price patterns in both markets and investigate the outlook of China’s market integration to the world market. Speciﬁcally, this analysis will: 1) estimate and identify an appropriate ARCH/GARCH model for China’s and US’s wheat futures prices; 2) investigate the interrelations between the two price series, including cointegration in the ﬁrst moment and autoregressive heteroskedasticity in the second moment, in a multivariate framework; and 3) compare the estimations between the two markets and assess the role of China’s wheat futures market in the world market. Previous Studies Since the 1980s, China’s successful economic reform has drawn the whole world’s attention for more than two decades, and the membership to the World Trade Organization (WTO) enhanced this attention to a new level. However, studies on China’s agricultural commodity futures markets are quite limited, particularly with regard to wheat futures. Moreover, most of the existing studies are descriptive analyses on regulatory and market development issues rather than quantitative investigations of futures prices. Such studies include Tao and Lei (1998); Fan, Ding and Wang (1999); and Zhu and Zhu (2000). A historical perspective on the development of China’s futures market is shown in Yao (1998), which includes a detailed structural analysis of the commodity futures markets and the government’s legislative and regulatory attempts. Some quantitative analyses have been attempted in recent years. Williams, et al. (1998) investigated mung bean trading in CZCE to test for market efﬁciency. Durham and Si (1999) examined the relationship between the China Dalian Commodity Exchange (CDCE), another commodity futures market in China, and the CBOT soybean futures prices through a regression model. Wang and Ke (2005) investigated the information efﬁciency of the CZCE wheat and CDCE soybeans futures in a framework of cointegration between cash and futures markets. Despite that, quantitative studies dealing with the time series properties of price on China’s wheat futures market, especially on the issue of world market integration, have not been found. Modeling time series data usually starts from the moving average (MA) model, autoregressive (AR) model, or more generally, autoregressive integrated moving average (ARIMA) model for the ﬁrst moment of the data. However, stochastic trend or unit root is discovered as a characteristic property of many high frequency commodity price series (Ardeni, 1989; Baillie and Myers, 1991). More complete

240

The Chinese Economy after WTO Accession

but complicated price models focusing on the second- or higher-order moment variability were introduced in the early 1980s and have enjoyed great attention in the last two decades. The autoregressive conditional heteroskedasticity (ARCH) model, developed by Engle (1982), allows the shocks in nearby earlier periods to affect the current volatility. The generalized ARCH, (GARCH) model (Bollerslev, 1986) allows, in addition, previous volatilities to affect current volatility, so that the volatility behaves like an AR process. ARCH and GARCH models have been widely applied in ﬁnancial time series analysis (Bollerslev, Cho, and Kroner, 1992) as well as in agricultural commodity prices (Baillie and Myers, 1991; Yang and Brorsen, 1992; Tomek and Myers, 1993; Myers, 1994). Excess kurtosis, namely heavier tails compared to normal distribution, is also found in commodity prices (Gordon, 1985; Deaton and Laroque, 1992; Myers, 1994). Although ARCH and GARCH models can partially alleviate the excess kurtosis problem (Engle, 1982; Myers, 1994), empirical studies have shown that they cannot capture all of it if the normal distribution is assumed on the price innovations (Bollerslev, 1987; Baillie and Myers, 1991; Yang and Brorsen, 1992). One possible solution to this problem is to use t-distribution instead of normal distribution to describe the price innovations in the ARCH/GARCH model (Myers, 1994). In the context of multivariate analysis, time-series of variables are often interrelated. Based on the theoretical framework derived by Engle and Granger (1987), and the empirical test methods by Johansen and Juselius (1990, 1992), studies on international futures markets have started to focus on using cointegration as an indication of market integration. Cointegration is a phenomenon that multiple nonstationary variables are driven by some common stochastic trends. Yang, Zhang, and Leatham (2003) examined the price and volatility transmission in a threevariable system for the US, Canadian, and EU markets. They found no cointegration in the system. Bessler, Yang, and Wongcharupan (2003) examined the wheat futures markets in the US, Canada, Australia, EU, and Argentina, and found cointegration. The present chapter contributes to the existing literature on China’s wheat futures prices in two ways. First, it incorporates both China’s and US’s wheat futures markets into a multivariate time series model so that the price interactions in the two markets can be studied simultaneously. Second, besides the interaction at the mean level as investigated in the cointegration studies, the interaction at the variance level is also carefully examined. Third, the assumed conditional error distribution of price changes is extended from normal distribution to t-distribution in a multivariate situation; therefore improvement of excess kurtosis can be examined and compared.

The Performance of Commodity Futures Markets

241

Models Univariate Conditional Heteroskedastistic Models We start with the univariate ARCH and GARCH models, which allow the volatility of error terms to change over time. An ARCH(q) model is commonly deﬁned to include a mean equation: Yt = Xt'β + εt, where εt | Ωt–1 ~ (0,ht)

(1)

and a variance equation: q

ht = ω + ∑αi ε2t–i

(2)

i=1

where Y t denotes the dependent variable; X t denotes the vector of explanatory variables which can include a constant, a time trend, lagged dependent variables, and/or any (lagged) exogenous variables; t denotes the time period; εt is the error component in the ARCH model whose conditional distribution has a zero mean and time-varying variance t h ; Ω t-1 is the information set available at t -1; βis the parameter vector for the exogenous variables; ω(ω > 0) is the parameter for intercept in the variance q

equation; and αi (αi ≥ 0 and ∑αi < 1) for i = 1,2,…,q is the parameter for ARCH i=1

effect. εt’s are serially uncorrelated, however, their dependency lies on the second moment evolution. A GARCH (p, q) model is deﬁned in the same way except that: q

p

ht = ω + ∑αi ε2t–i + ∑γjht–j i=1

(2’)

j=1

with γ for j =1, 2, …, p as additional parameters for past volatilities; ω > 0,αi,γj ≥ 0 q

p

and ∑αi + ∑γj < 1. i=1

j=1

The basic ARCH (q) model is a short memory process in that only the most recent q shocks have an impact on the current volatility. The GARCH(p, q) model is a long memory process, in which all the past shocks can affect the current volatility indirectly through the p lagged variance terms. Multivariate Conditional Heteroskedastic Models Multivariate ARCH and GARCH models allow more than one series to be modeled together so that the interrelation between different series can be examined and tested through cross equation parameter constraints.

The Chinese Economy after WTO Accession

242

An m-variate GARCH (P, Q) model can be deﬁned as: Yt = BXt + εt, where εt | Ωt–1 ~ (0,Ht) Q

P

i=1

j=1

Ht = W + ∑Ai(εt–iε't–i)A'i + ∑ГjHt–jГ'j

(3) (4)

where Y t is now an m×1 dependent variable vector; Β is the coefﬁcient matrix corresponding to the explanatory variable vector Xt ; εt, the error vector, is conditionally distributed with a mean of an m×1 null vector and an m×m variance-covariance matrix Ht; W is an m×m parameter matrix for the constant terms, and Αi and Γj are m×m parameter matrices for GARCH coefﬁcients. By deﬁnition, the ARCH(Q) model is a special case of the GARCH(P,Q), when the coefﬁcient matrices for the past variancecovariance matrices, Γj ’s, are set at zeroes. Alternative deﬁnitions of the variance equations and restrictions on matrices Αi and Γj exist, which lead to different versions of multivariate ARCH/GARCH models. The above deﬁned model allows each element of the current variance-covariance matrix Ht to be affected by all elements of the past variance-covariance matrices and/ or the squared error matrices. This is called BEKK model (Engle and Kroner, 1995). Other commonly used models, like constant conditional correlation (CCC) model (Bollerslev, 1990) and the diagonal-vec (DVEC) model (Bollerslev, Engel, and Wooldridge, 1988), have simpler forms with different assumptions on parameters. The variance equation in a CCC model has the following form: q

p

hmm,t = ωm + ∑αmi ε2m,t–i + ∑γmjhmm,t–j i=1

(4’)

j=1

hmn,t = ρmn√hmm,t hnn,t, m ≠ n where h mn, t is the mnth element in Ht, and ρmn is the constant correlation between hmm,t and hnn,t .With an imposed constant correlation coefﬁcient to the independent univariate ARCH/GARCH models, the CCC model largely reduces the number of parameters to be estimated. However, this model only allows the conditional variances to evolve based on their own past levels and past shocks, and the relationship between one another is constrained to null and cannot be revealed. The DVEC model deﬁnes the variance equation as: Q

P

i=1

j=1

Ht = W + ∑Ai Ĵ(εt–iε't–i) + ∑Гj ĴHt–j

(4’’)

where Ĵ is the Hadamard product operator, i.e. element-by-element multiplication. Αi and Γj are restricted to be symmetric matrices. This model is called the diagonalvec (DVEC) model, which allows each element of the current variance-covariance

The Performance of Commodity Futures Markets

243

matrix H t to be affected only by its own past values and/or corresponding element in the past squared error matrices. Similar to the CCC model, information about the relationships between variances of different series is not available in DVEC form. Furthermore, when the parameter matrices are set to be diagonal, the model will degenerate into separate univariate models. Data Description The CZCE price data in the study are collected from CZCE’s online database at http://www.czce.com.cn. The CBOT data are collected from http://www.turtletrader.com. The daily settlement price series for the September wheat contracts from January 4, 2000 to September 13, 2002 are used for both CZCE and CBOT series.3 The prices are taken as continuous for each trading day. A switching contract dummy variable, SD t , is introduced in the explanatory variable vector X in addition to the constant term for both series to indicate when the price series switches from an old contract to a new one. The switching points are set at the last trading day of the old contract following the method as in Myers and Hanson (1993).4 SD t equals 1 at the switching points and 0 otherwise. The time-series plots of CZCE and CBOT prices are given in Figure 13.2. Both series show strong nonstationarity and stochastic trend, while CZCE prices look more chaotic than CBOT prices. For both price series, the sample autocorrelation functions show very slow exponential decay, and the sample partial autocorrelation functions show a large spike in the ﬁrst lag. The augmented Dickey-Fuller (ADF) unit root test yields a P value of 0.60 for CZCE prices and 0.90 for CBOT prices, conﬁrming the existence of unit root. Therefore, the ﬁrst difference is taken. From the time series plots (Figure 13.3) of the squared ﬁrst difference data, evidence of a time-varying volatility pattern is visible in CBOT series. That is, big changes are often followed by other big ones and small changes followed by small ones. This pattern is consistent with the ARCH/GARCH processes. Furthermore, the P values of Portmanteau Q statistic and the LaGrange Multiplier statistic, for testing H0: no ARCH effect, are all less than or equal to 0.0001 for ﬁrst twelve lags for CBOT prices, indicating strong ARCH effect. For our selected CZCE prices, however, the two statistics are not statistically signiﬁcant.

3 The original data are in Yuan per metric ton for CZCE prices and in US dollar per bushel for CBOT prices. To make the two data series directly comparable in our study, we converted CBOT data into Yuan per metric ton using a constant factor, one metric ton equaling 36.74 bushels of wheat, and the exchange rate. The exchange rate stays constant in the study period. 4 For CBOT data, trading of the September 2000 contract started in July 2000. These early prices were not included until September 2001 when the September 2000 contract expired. CZCE data are arranged similarly. The old contract trading prices are chosen for this overlapping period because the old contracts are traded more actively than the new one during the period.

The Chinese Economy after WTO Accession

244

CZCE Daily Settlement Prices Unit: Yuan/Ton

CBOT Daily Settlement Prices Unit: Yuan/Ton

Figure 13.2 CZCE and CBOT Wheat Futures Price for September Contract Normality check of CZCE and CBOT price changes implies strong evidence against normality. For CBOT price changes, the kurtosis coefﬁcient is 1.80, indicating the distribution has fatter tails than the normal distribution.5 The skewness coefﬁcient is 0.24; close to zero, meaning it is quite symmetric. The distribution of CZCE price changes is also quite symmetric with the skewness coefﬁcient of -0.57

5

A normal distribution has both skewness and kurtosis at 0 as a benchmark.

The Performance of Commodity Futures Markets

245

CZCE September Contract 2

2

2

2

Unit: Yuan /Ton

CBOT September Contract Unit: Yuan /Ton

Figure 13.3 Squared First Difference of CZCE and CBOT Wheat Futures Prices Note: Since some squared ﬁrst difference prices are so much higher than the rest, such as those at switching points, we have cut off a large part of the spikes to ﬁt them into the plot.

but even fatter tails. The kurtosis coefﬁcient is 6.67. These coefﬁcients indicate the non-normality is mostly caused by excess kurtosis rather than the skewness.

246

The Chinese Economy after WTO Accession

Results Univariate Analysis In this section, we study the price performances of CZCE and CBOT separately. Due to the existing unit root, ﬁrst differences of the data are ﬁt into alternative time series models. The mean equation of the model is deﬁned as price change dependent on the constant term and the contract-switching dummy, SD. A univariate framework is applied to ﬁnd the best speciﬁcations of ARCH/GARCH process. The GARCH procedure in GARCH module of S-PLUS is used to estimate these models. For the visible heavy tails of the price distributions, we estimate price models under both normality and t distributions. Results based on the two distributions are examined and compared. Selection of model speciﬁcationAlternative speciﬁcations in terms of the lags of the ARCH/GARCH models are ﬁtted. Two goodness-of-ﬁt criteria, Akaike Information Criterion (AIC) and Schwarz’s Bayesian Information Criterion (BIC), along with signiﬁcance criterion, are used in coordination to select the best model. The model ﬁtting results show that ARCH (1) has the best ﬁt for both CZCE and CBOT price changes under the normality assumption. In the estimation of above models, however, the Shapiro-Wilk and Jarque-Bera statistics for normality test reject the normality assumption in all cases. It conﬁrms what we observe in the earlier normality check in the data section and implies change of distribution is relevant and necessary. Following the existing empirical literature, we assume the error terms in the mean equation is t distributed, i.e. εt | Ωt–1 ~ t(v), where v denotes the degree of freedom, and with mean 0 and variance ht. With obvious gain in the goodness-of-ﬁt, the best models under t distribution are ARCH (1) for CBOT and GARCH (1,1) for CZCE. The results are not reported in this chapter, but are available from the authors upon request. EstimationTable 13.1 gives the estimation results of the choice models for CZCE and CBOT under both normal and t assumptions. In general, the signiﬁcance and sign of each parameter are consistent between the two sets of results. The main difference lies in the magnitude, or weight, of the estimated coefﬁcients in the mean and variance equations. The models capture more contract switching and ARCH/GARCH effects when the conditional distribution is t. From the results, we see both CZCE and CBOT price changes contain no drift. The contract switching has insigniﬁcant and negative effects on CBOT price changes, but a signiﬁcant and positive contribution to CZCE price changes, indicating a jumpup of the price from the mean at the switching point in CZCE.6 In all the ARCH (1) 6 At switching points, the estimated mean equation becomes ∆Pt = αˆ + δˆ + εˆt, where the right-hand-side of the equation represents the level of price changes when the contract switches.

The Performance of Commodity Futures Markets

Table 13.1

247

Estimates of Selected Univariate ARCH/GARCH Models εt | Ωt–1 ~ normal

Model

εt | Ωt–1 ~ student t

CZCE - ARCH(1)

CBOT - ARCH(1)

CZCE - GARCH(1,1)

CBOT - ARCH(1)

β0

-0.77 (0.60)

0.54 (0.62)

-0.22(0.20)

-0.13 (0.40)

δ

298.6* (0.76)

-18.26 (70.27)

322.43* (18.08)

-19.11 (37.80)

ω

110.96* (0.89)

168.45* (2.41)

2.54* (1.11)

130.33* (12.93)

α1

0.05* (0.02)

0.10* (0.02)

0.04* (0.02)

0.12* (0.06)

γ1

-

-

0.94*(0.02)

-

AIC

4974.8

5449.5

4332.9

4359.7

BIC

4992.7

5467.6

5215.0

5237.6

Notes:

1. “*” denotes signiﬁcance at 5% level. 2. Standard errors are listed in parentheses. 3. The estimated GARCH(1,1) model is deﬁned as: ∆Pt = β0 + δSDt + εt, εt | Ωt–1 ~ N(0,ht),

where ht = ω + α1ε2t–1 + γ1ht–1 with Pt denoting price at time t. The ARCH(1) is obtained when γ1 is set to zero in the above speciﬁcation.

models for CBOT and CZCE, the ARCH coefﬁcient has a signiﬁcant but relatively weak impact on the variance, compared with the intercept term. In the GARCH (1,1) for CZCE under t, however, the inﬂuence from the intercept reduces enormously. Both GARCH and ARCH coefﬁcients are signiﬁcant, and the GARCH coefﬁcient is a lot more inﬂuential, implying that a large part of the current volatility in ZCE is due to the last period of volatility. Multivariate Analysis Multivariate analysis allows both price series to be estimated simultaneously. As a result, cross-market relations that are unable to be detected in univariate analysis can now be captured. In our multivariate version of the models, the mean equation still follows the same structure as in univariate case, but the variance equation becomes a system of equations. Cointegration testIn order to identify the possible interrelations that exist in the comovement of the price levels, we ﬁrst conduct the cointegration test on the original prices of CZCE and CBOT wheat futures, before moving to examine the second moment (volatility) relation. According to the ADF unit root test, both price series are integrated to order one, satisfying the conditions for cointegration test. Proceeding with the Johansen’s cointegration test, however, we fail to reject the null hypothesis of no cointegration. Results indicate that our data on CZCE and

248

The Chinese Economy after WTO Accession

CBOT wheat futures prices have no cointegrating relation on the ﬁrst moment, and a vector regression model is appropriate for the following time series analysis. Here the dependent variable vector is the ﬁrst difference of prices, and the independent variable vector includes constant and two contract switching dummy variables. Again, the results can be made available upon request. Selection of model speciﬁcationTo fully disclose the interrelations between CZCE and CBOT wheat prices, we apply three types of multivariate GARCH models, BEKK, CCC, and DVEC, to estimate the CBOT-CZCE bivariate series under both normal and t distributions.7 By deﬁnition, the BEKK model contains information about the cross-market ARCH/GARCH effects. The CCC and DVEC models are simpliﬁed multivariate GARCH models with different model structures. Since these two models have different ranks in restrictiveness and robustness, we include both in the estimation for a comparison purpose. The MGARCH procedure in the GARCH module of S-PLUS is used for analysis in this section. The results show that GARCH (1,1) in DVEC form and CCC form has better ﬁt under both normal and t distributions, while in BEKK form ARCH (1) performs better (Table 13.2). EstimationInformation about the interactions between elements in the conditional variance matrix and relationships between the price changes of China’s and the US’s wheat futures are now reﬂected in the estimates (Table 13.2). When BEKK model is ﬁtted with normal distribution, CZCE price changes appear to have a small drift while CBOT data do not. The coefﬁcient matrix Λ for the switching dummy vector provides full information about within and cross equation relationships of contract switching in the two markets. For own market effect, the CZCE switching dummy has a signiﬁcant positive impact on the mean of price changes, similar to the univariate case. The CBOT dummy has a signiﬁcant positive impact on its price changes, quite different from the univariate case. The cross effects, however, are both statistically insigniﬁcant, implying the interactions of contract switching between the two markets are weak. The variance equation estimates, noteworthily, do not directly reﬂect withinand cross-market effects on volatilities. Certain combinations of these estimates can only show those effects. In the Appendix we provide a detailed derivation of such combinations. The calculated estimated effects are reported in Table 13.3. When normal distribution is assumed, current volatility of CBOT price changes is positively correlated with the last period shocks in its own market and that in CZCE. The own market effect of 0.1132 dominates the cross-market effect by a ratio of 15:1. Therefore the volatility in CBOT is mostly affected by the previous shock in its own market. The previous shock in CZCE has very limited inﬂuence on the 7 Although usually the joint t distribution is not well-deﬁned, unlike its normal counterpart, it is deﬁned in a certain way in s-plus GARCH module (S+ GARCH User’s Manual, pp. 107–108, Mathsoft, inc., March 2000). This deﬁnition is followed in our analysis.

The Performance of Commodity Futures Markets

Table 13.2

Estimates of Selected Multivariate ARCH/GARCH Models BEKK Modeling

Model

ARCH(1) - N

ARCH(1) - t

μB

0.009 (0.49)

-0.002 (0.48)

μ

-0.97* (0.63)

Z

Λ BB

249

CCC Modeling

DVEC Modeling

GARCH(1,1) - N GARCH(1,1) - t GARCH(1,1) - N GARCH(1,1) - t

-0.48* (0.31)

-0.16 (0.48)

-0.21 (0.47)

-0.26 (0.48)

-0.20 (0.47)

-0.15 (0.43)

-0.15 (0.26)

-0.13 (0.40)

-0.16 (0.27)

141.80* (11.49) 138.20* (6.76) 135.30* (4.69) 133.98* (3.44) 129.60* (6.92) 133.77* (3.49)

Λ ZB

6.05 (56.13)

3.48 (13.75)

48.28* (1.96)

-1.09 (10.50) -39.21* (2.10) -0.81 (10.65)

Λ BZ

3.21 (150.47)

1.74 (548.57)

3.98 (169.18)

7.28 (119.49)

Λ ZZ

*

3.58 (4.37)

8.53 (63.28)

*

280.89* (5.66) 303.09* (22.73) 321.02 (20.25) 437.80 (7.95) 360.83* (2.97) 437.92* (3.65)

ωu

11.34* (0.28)

11.40* (0.33)

ωuv

0.07 (0.64)

-0.17 (0.35)

ωv

10.83* (0.08)

2.13* (1.64)

1.99 (1.81)

1.98 (1.59)

1.95 (1.83)

-

0.02 (0.08)

-2.29 (6.88)

5.13 (0.90)

5.08* (1.19)

5.15* (0.91)

*

6.72* (0.20)

5.64 (1.34)

*

Αuu

0.34* (0.05)

0.21* (0.06)

0.03 (0.009)

0.03 (0.010)

0.03* (0.009)

0.03* (0.010)

Α vu

-0.04 (0.14)

-0.01 (0.05)

-

-

-0.02* (0.01)

-0.007 (0.03)

Αuv

0.09 (0.10)

0.002 (0.08)

-

-

-

0.24 (0.04)

0.24* (0.04)

0.24* (0.04)

Α vv

0.20* (0.05)

*

*

*

0.46* (0.05)

0.22 (0.04)

*

Γuu

-

-

0.96 (0.02)

0.96 (0.02)

0.96* (0.02)

0.96* (0.02)

Γvu

-

-

-

-

0.99* (0.009)

-0.92 (0.25)

Γvv

-

-

0.79* (0.04)

0.69* (0.03)

0.78* (0.03)

0.69* (0.03)

ρuv

-

-

-0.04 (0.05)

-0.02 (0.05)

-

-

AIC BIC

9687.1 9744.8

9398.4 9460.5

9438.0 9491.2

9293.4 9351.0

9406.1 9472.6

9299.2 9370.2

*

*

Notes: 1. * denotes signiﬁcance at 5% level, and standard errors are listed in parentheses. 2. The estimated ARCH(1) model in BEKK form is deﬁned as: ΔPtB μB SDtB ut = +Λ + , where ΔPtZ μZ SDtZ vt

( ) () ( ) () ( )| (( ) ( )) ( ) ( ) ( ut

vt

Ωt–1 ~

0

0

,

σ2u,t

σuv,t σ2v,t

,

σ2u,t

σuv,t σ2v,t

=

ωu

+A

ωuv ωv

u2t–1 ut–1 vt–1 v2t–1

)

A',

the GARCH(1,1) model in CCC form has the variance equation as:

σi2, t = ωi + Αii it2−1 + Γii σi2, t −1 , i = u , v , σij , t = ρij σi2, t σ 2j , t , i, j = u , v, and i ≠ j , and the GARCH(1,1) model in DVEC form has the variance equation as: ωu u2t–1 σ2u,t–1 σ2u,t = + A☼ + Г☼ . σuv,t σ2v,t ωuv ωv ut–1 vt–1 v2t–1 σuv,t–1 σ2v,t–1

( )( ) (

) (

The superscript B denotes CBOT and the superscript Z denotes CZCE.

)

The Chinese Economy after WTO Accession

250

Table 13.3

Within and Cross Market Effects of Bivariate BEKK–ARCH(1)

σ u2, t

σ uv , t

σ v2, t

2 ut−1

0.1132

-0.0126

0.0014

ut −1vt −1

0.0579

0.0639

-0.0149

0.0074

0.0172

0.0398

2 ut−1

0.0453

-0.0022

0.0001

ut −1vt −1

0.0007

0.0973

-0.0096

H t | Ωt −1 ~ normal

2 t−1

v

H t | Ωt −1 ~ student t

2 t−1

v

0.29E-5 0.0008 Notes: 1. The reported values are calculated as in the Appendix. 2. The estimated ARCH(1) model in BEKK form is deﬁned as: ΔPtB μB SDtB ut = +Λ + , where ΔPtZ μZ SDtZ vt

( ) () ( ) () ( )| ( ) ( )) ( )( ) ( ut

Ωt–1 ~

vt

σ2u,t

σuv,t σ2v,t

0

0

=

,

σ2u,t

0.2091

,

σuv,t σ2v,t

ωu

ωuv ωv

+A

u2t–1

ut–1vt–1 v2t–1

)

A'.

volatility in CBOT. The volatility in CZCE is even more dominated by own market effect rather than cross-market effect. The ratio increases to 28:1. The insigniﬁcancy of both Auv and Avu indicates that the cross impact in the variance equation may not exist, and the BEKK is not superior to DVEC or CCC. The smaller AIC’s and BIC’s of DVEC and CCC also indicate they actually have a better ﬁt. From Table 13.3, in the bivariate DVEC form of GARCH (1,1), when the underlying conditional distribution is normal, the intercepts in the mean equation are consistent with the univariate cases, i.e. neither of the series shows signiﬁcant drift in prices. The own market contract switching dummies have a similar pattern as in the BEKK model. The cross equation terms, more informatively, show that the contract switching in CBOT has signiﬁcant negative inﬂuence on the price changes in CZCE, while switching in CZCE does not affect CBOT price signiﬁcantly. The contract switching dates are different for CBOT and CZCE, around September 15 for CBOT, and one week later for CZCE. This implies that when the old contract expires at CBOT, the switching to a new contract in CBOT enhances the decreasing trend of old contract prices in CZCE. But the contract switching in CZCE doesn’t have a comparable effect on CBOT new contract prices when it is one week after.

The Performance of Commodity Futures Markets

251

In the variance equation, only CZCE volatility has a signiﬁcant drift. For CBOT price changes, the last period volatility has much more inﬂuence on current volatility than last period shock, which indicates CBOT price has a long memory. The former has an estimated coefﬁcient of 0.96 while the latter only has 0.03. The own market effect of volatility for CZCE price changes shows a similar pattern. The gap between the inﬂuences of last period volatility and last period shock seems smaller in CZCE. This indicates that CZCE price has a shorter memory than CBOT, so that a shock tends to have a larger impact on price of next period in China. That in a certain way indicates prices in CZCE are apt to be more volatile, even chaotic. Estimation results from CCC model are mostly very close to those from DVEC. In terms of model ﬁtting, the DVEC under normal distribution outperforms both CCC and BEKK with smallest AIC and BIC. When the underlying distribution is t, results are different. Especially all the parameters for interaction terms between the two markets, including the switching dummy and covariance, are insigniﬁcant. This may indicate the two markets do not demonstrate signiﬁcant interactions. Market integration of CZCEThe integration of CZCE to the world market, especially the CBOT, has been an interesting issue to many researchers as well as government ofﬁcials since the establishment of CZCE. Actually, there exists general belief, or good will, that CZCE prices have developed or are developing a close relationship with CBOT prices based on 1) the fact that CZCE was established a decade ago with help from CBOT so that many institutional features of the two markets are the same, and 2) some preliminary statistical calculation reveals the prices from the two markets have a strong association (CZCE Report, 2001). Although China becomes more integrated to the world economy and its trade policy turns more liberalized after its WTO accession, the relationship between CZCE and CBOT has not shown a clear pattern yet. Although there are a priori reasons to expect wheat futures prices in the two markets to move together more closely, there are also reasons to expect otherwise. Such reasons include the facts that the physical wheat trading volume between the two countries is still a small proportion compared to the domestic production and consumption levels, regulatory and institutional barriers still interfere with China’s market development, and futures traders involved in the two markets are not largely observed. Based on our data on wheat futures prices around the WTO accession, the cross-market effects are not yet evident in terms of cointegration in the mean level. It implies the long-run equilibrium relationship that binds the price movements in CZCE and CBOT are not existent. However, we still ﬁnd transmission in the contract switching effect and volatility, depending on model speciﬁcation, between the two markets. Although the linkage is not strong, it implies an asymmetric pattern. That is, CBOT has a stronger inﬂuence on CZCE than CZCE on CBOT, which implies CBOT’s leading role in the market interaction while CZCE is more like a follower. Such an interaction discloses the existence of a weak connection between China’s and America’s wheat futures market on one hand, but on the other hand, the asymmetric property of the

252

The Chinese Economy after WTO Accession

relationship indicates China’s wheat futures market is not strong enough to inﬂuence world market but is inﬂuenced by it. Conclusion After more than ten years of development, the CZCE has built up to be the biggest commodity futures market in China and its wheat futures trading has important effects on wheat prices in China’s agricultural price system. This chapter makes an effort to investigate the wheat futures price behavior in CZCE and more importantly the integration of CZCE to the world market, using CBOT as a representative. Previous studies on the market integration of China’s wheat futures market such as CZCE report (2001) focused on the pairwise correlation analysis of prices and assumed constant price volatility. In this chapter, we consider the cointegration relation and model price behavior of two wheat futures markets simultaneously based on time variant conditional variances, i.e. ARCH/GARCH. Model ﬁtting shows that both CZCE and CBOT price can be best modeled by ARCH (1)/GARCH (1,1) process. These results are consistent with the empirical studies of high frequency commodity prices (Myers, 1994; Poon and Granger, 2003). Bivariate analysis of CZCE and CBOT prices shows the two series are not cointegrated. The existing cross-equation effects, i.e. the interrelations, between the two markets are signiﬁcant but weak, and asymmetric under normal distribution. CBOT plays a leading role in the interactions and CZCE is more like a follower. This result reveals that the two prices evolve in a similar way and coincide to one another through the season, but there is not strong evidence for information ﬂow from one market to the other. However, under t distribution, no signiﬁcant evidence can be found for any interaction between the two markets. This means the relationship between the two markets has not shown a clear pattern. The results indicate that the price in China’s wheat futures behaves in a similar way to the price in the represented world market, which is a good sign showing that the Chinese agricultural commodity market is performing in line with world markets. On the other hand, the short memory feature of CZCE compared to CBOT indicates that the CZCE is more volatile and chaotic, a sign showing that either the Chinese traders are less mature or the Chinese food market environment is less stable. The one-way impact from CBOT to CZCE and the weak relation between the two markets indicate that China’s wheat market is not fully integrated with the world market yet. References Ardeni, P.G. (1989), ‘Does the Law of One Price Really Hold?’ American Journal of Agricultural Economics, Vol. 71, No. 3, pp. 661–69. Baillie, R.T. and R.J. Myers (1991), ‘Bivariate GARCH Estimation of the Optimal Commodity Futures Hedge,’ Journal of Applied Econometrics, Vol. 6, No. 2), pp. 109–24.

The Performance of Commodity Futures Markets

253

Bessler, D.A., J. Yang and M. Wongcharupan (2003), ‘Price Dynamics in the International Wheat Market: Modeling with Error Correction and Directed Acyclic Graphs,’ Journal of Regional Science, Vol. 43, No. 1, pp. 1–33. Bollerslev, T. (1986), ‘Generalized Autoregressive Conditional Heteroskedasticity,’ Journal of Econometrics, Vol. 31, No. 3, pp. 307–27. Bollerslev, T. (1987), ‘A Conditional Heteroskedastic Time Series Model for Speculative Prices and Rates of Return,’ Review of Economics and Statistics, Vol. 69, No. 3, pp. 542–47. Bollerslev, T. (1990), ‘Modeling the Coherence in Short-Run Nominal Exchange Rates: A Multivariate Generalized ARCH Model,’ Review of Economics and Statistics, Vol. 72, No. 3, pp. 498–505. Bollerslev, T., R.Y. Cho and K.F. Kroner (1992), ‘ARCH Modeling in Finance: A Review of the Theory and Empirical Evidence,’ Journal of Econometrics, Vol. 52, Nos 1–2, pp. 5–59. Bollerslev, T., R.F. Engel, and J.M. Wooldridge (1988), ‘A Capital Asset Pricing Model with Time-varying Covariance,’ Journal of Political Economy, Vol. 96, No. 1, pp. 116–31. CZCE Report (2001), http://www.czce.com.cn/Publish/research/200108/498.asp. Deaton, D.A. and G. Laroque (1992), ‘On the Behavior of Commodity Prices,’ Review of Economic Studies, Vol. 59, No. 1, pp. 1–24. Durham, C. and W. Si (1999), ‘The Dalian Commodity Exchange’s Soybean Futures Contract: China’s Integration with World Commodity Markets,’ Chinese Agriculture and the WTO Proceedings, Western Coordinating Committee, December. Engle, R.F. (1982), ‘Autoregressive Conditional Heteroskedasticity with Estimation of the Variance of U.K. Inﬂation,’ Econometrica, Vol. 50, No. 4, pp. 987–1008. Engle, R.F. and C.W.J. Granger (1987), ‘Co-Integration and Error Correction: Representation, Estimation, and Testing,’ Econometrica, Vol. 55, No. 2, pp. 251– 76. Engle, R.F. and K.F. Kroner (1995), ‘Multivariate Simultaneous Generalized ARCH, Economic Theory,’ Vol. 11, No. 1, pp. 122–50. Fan, Y., X. Ding, and H. Wang (1999), ‘Factors Affecting the Development of Agricultural Commodity Futures Markets in China,’ Farm Economic Management, No. 1, pp. 35–36. Gordon, J.D. (1985), ‘The Distribution of Daily Changes in Commodity Futures Prices,’ Technical Bulletin No. 1536, Economic Research Service, USDA, Washington DC. Haigh, M.S. and M.T. Holt (2000), ‘Hedging Multiple Price Uncertainty in International Grain Trade,’ American Journal of Agricultural Economics, Vol. 82, No. 4, pp. 881–96. Johansen, S. and K. Juselius (1990), ‘Maximum Likelihood Estimation and Inference on Cointegration: With Applications to the Demand for Money,’ Oxford Bulletin of Economics and Statistics, Vol. 52, No. 2, pp. 169–210.

254

The Chinese Economy after WTO Accession

Johansen, S. and K. Juselius (1992), ‘Testing Structural Hypotheses in a Multivariate Cointegration Analysis of the PPP and the UIP for UK,’ Journal of Econometrics, Vol. 53, pp. 211–44. Myers, R.J. (1994), ‘Time Series Econometrics and Commodity Price Analysis: A Review,’ Review of Marketing and Agricultural Economics, Vol. 62, No. 2, pp. 167–82. Myers, R.J. and S.D. Hanson (1993), ‘Pricing Commodity Options when the Underlying Futures Price Exhibits Time-Varying Volatility,’ American Journal of Agricultural Economics, Vol. 75, No. 1, pp. 121–30. Poon, S. and C.W.J. Granger (2003), ‘Forecasting Volatility in Financial Markets: A Review,’ Journal of Economic Literature, Vol. 41, No. 2, pp. 478–39. Tao J. and Lei H. (1998), ‘Futures Market and the Grain Circulation Reform,’ Economic Problems, No. 9, pp. 29–31. Tomek, W.G. and R.J. Myers (1993), ‘Empirical Analysis of Agricultural Commodity Prices: A Viewpoint,’ Review of Agricultural Economics, Vol. 15, No. 1, pp. 180– 202. USDA Foreign Agricultural Service (2004), World Wheat Production, Consumption and Stocks, USDA FAS Production, Supply and Distribution Online, available at: http://www.fas.usda.gov/psd/complete_tables/GF-table9–80.htm. Wang, H.H. and B. Ke (2005), ‘Efﬁciency Tests of Agricultural Commodity Futures Markets in China,’ The Australian Journal of Agricultural and Resource Economics, Vol. 49, No. 2, pp. 125–41. Williams, J., A. Peck, A. Park and S. Rozelle (1998), ‘The Emergence of a Futures Market: Mung Beans on the China Zhengzhou Commodity Exchange,’ Journal of Futures Markets, Vol. 18, No. 4, pp. 427–48. Yang, J. and D.J. Leatham (1999), ‘Price discovery in Wheat Futures Markets,’ Journal of Agricultural and Applied Economics, Vol. 31, No. 2, pp. 359–70. Yang, J., J. Zhang, and D.J. Leatham (2003), ‘Price and Volatility Transmission in International Wheat Futures Markets,’ Annals of Economics and Finance, Vol. 4, No. 1, pp. 37–50. Yang, S.R. and B.W. Brorsen (1992), ‘Nonlinear Dynamics of Daily Cash Prices,’ American Journal of Agricultural Economics, Vol. 74, No. 3, pp. 706–15. Yao C. (1998), Stock Market and Futures Market in the People’s Republic of China, Oxford Express, Oxford, New York. Zhu L. and Zhu J. (2000), ‘Futures Market and Contract Agriculture,’ Journal of Zhengzhou Grain College, Vol. 21, pp. 29–34.

The Performance of Commodity Futures Markets

255

Appendix Derivation of Own- and Cross-market Effects of Conditional Variances in a Multivariate ARCH Model (BEKK form) By the deﬁnition of BEKK-ARCH (1), the conditional error variance equation can be speciﬁed as:

( ) ( ) ( )( ( )( σ2u,t

ωu

Auu

=

Auv

u2t–1

+

σuv,t σ2v,t

ωuv ωv

Avu Avv

Auv '

)( )

ut–1vt–1 v2t–1

Auu

Avu Avv

A2uuu2t–1 + 2AuvAuuvt–1ut–1 + A2uvv2t–1

ωu

=

+

)

.

AuuAvuu2t–1 + (AuuAvv + AuvAvu)ut–1vt–1 + AuvAvvv2t–1 A2vuu2t–1 + 2AvuAvvvt–1ut–1 + A2vvv2t–1

ωuv ωv

To obtain the own and cross market effects from this speciﬁcation, we take partial derivative of each dependent variable with respect to the respective explanatory variables to get the needed effects. Own market effects: ∂σu2t ∂ut2−1

2 , = Αuu

∂σv2t ∂vt2−1

= Α vv2 .

Cross market effects: ∂σu2t ∂vt2−1

2 , = Αuv

∂σv2t ∂ut2−1

2 . = Α vu

Other effects: ∂σu2t ∂ut −1vt −1

= 2Αuv Αuu ,

∂σv2t ∂ut −1vt −1

= 2Α vu Α vv .

Covariance effects: ∂σuvt ∂u

2 t −1

= Αuu Α vu ,

∂σuvt ∂v

2 t −1

= Αuv Α vv ,

∂σuvt ∂ut −1vt −1

= Αuu Α vv + Αuv Α vu .

This page intentionally left blank

Chapter 14

Home Market Effect and Its Impact on Production and Trade: An Empirical Study of China and the US Fan Zhang and Zuohong Pan

Introduction Two principal theories have been developed to explain why countries or regions trade. Comparative advantage believes that trade arises to take advantage of differences in resource endowments, while increasing returns maintains that trade arises to take advantage of specialization and scale economy. The recent increasingreturns models, labeled ‘economic geography’ by Krugman, allow researchers to distinguish comparative advantage from increasing returns by testing the impact of home market demand on a country’s production and trade. As Davis and Weinstein (1999) put it, in a world of comparative advantage, a country with strong demand for a good will import that good. In a world with increasing returns, when trade costs exist, a country with strong demand for a good makes that country the site of production and the exporter of the good. In other words, the extra demand in the home market leads to large-scale production and high efﬁciency, and creates exports. This core concept of new economic geography is labeled by Krugman (1980) as ‘home-market effects’. China has become a world major production site and exporter for many manufactured products. The prevailing explanation of China’s role in world trade is its comparative advantage in labor-intensive industries because of its relative endowment with labor. If comparative advantage is the sole reason for China’s role in world trade, China will lose its current position as soon as its labor costs rise to a certain level, as has happened or will happen in some small-sized East Asian countries. If other reasons, such as increasing returns exist, China may keep its current position as a world major manufacturer for a longer time, due to the demand from its own market. Moreover, cheap labor can only explain the development of China’s labor-intensive sectors; it cannot explain the development of its high-tech sectors. In this chapter, we try to ﬁnd the role of increasing returns in explaining the reasons for international trade in China and the United States.

258

The Chinese Economy after WTO Accession

The increasing-returns theory has been used in different ﬁelds of economics since Dixit and Stiglitz (1977) developed a formalized CES model of Chamberlinian monopolistic competition. In the late 1970s and early 1980s, a number of economists applied the tools to international trade (e.g., Helpman and Krugman, 1985) and economic growth (e.g., Grossman & Helpman 1991). The theory of ‘new economic geography’ is considered as a new wave of the increasing-returns revolution. Krugman’s (1980) model of the ‘home market effect’ argues that countries will tend to export those kinds of product for which they have relatively large domestic demand. Krugman shows that if high demand leads the good to be exported, production must rise by more than demand in two equal-sized economies. Weder (1995) extends this result to the unequal-sized case. Krugman’s theory was extended in a series of papers and a book by Fujita, Krugman, and Venables (2000). While theoretical economic geography has made signiﬁcant progress, only a few empirical works have been done to test the home market effects. Using Japanese data, Davis and Weinstein (1999) ﬁnd support for the existence of economic geography effects in eight of nineteen manufacturing sectors in Japan. This contrasts with the result of the same authors, Davis and Weinstein (1996), which found scant economic signiﬁcance of economic geography for OECD countries using international data. Feenstra et al. (1998) ﬁnd home market effect by using a free entry, imperfect competition, homogeneous good model. More recently, Head and Ries (2001) develop two alternative models of trade in differentiated products between Canada and the US, increasing returns and national product differentiation models, and ﬁnds the preponderance of the evidence supports national product differentiation. Brülhart and Trionfetti (2001) develop a discriminating criterion reliant upon the assumption that demand is home biased. They test the hypothesis in 29 industries, covering 22 OECD countries for 1970–85 and ﬁnd that 17 industries can be associated with the increasing returns to scale paradigm. There is an increasing literature on China’s industrial growth from a geographic perspective. Bao, Chang, Sachs and Woo (2002) examine the geographic effects on regional economic growth in China, using a regional growth model characterized by foreign direct investment and mobilization of rural surplus labor. Brun, Combes, and Renard (2002), and Batisse (2002) study the spillover effects on economic development from a geographic and from a sectoral perspective, respectively. Theoretical Framework The empirical work of this chapter is based on theory developed by Krugman (1980) and Fujita, Krugman, and Venables (FKV, 2000). With minor revisions, we derived the model used in this chapter directly from FKV. With that, we identiﬁed the relation between FKV’s theory and the empirical testing model in Davis and Weinstein (1996 and 1999). FKV (2000) gives a formal version of Krugman’s model, and derives the home market effect in the following steps.

Home Market Effect and Its Impact on Production and Trade

259

The consumer’s problem is to maximize representative consumer’s utility function for manufactured goods M and agricultural good A: μ

1−μ

U =M A

⎤ ⎡n = ⎢⎢ ∫ m(i )ρ di ⎥⎥ ⎥⎦ ⎢⎣ 0

μ/ρ

A1−μ

subject to budget constraint. Where m is the consumption of each variety, μ is the expenditure share of manufactured goods, ρ is the intensity of the preference for variety in manufactured goods, and n is the number of varieties. The elasticity of substitution between varieties is σ=1/ (1- ρ). The consumer problem is solved in two steps. First, m is chosen to minimize the cost of attaining M to solve the demand for m. The second step on the upper-level of consumer’s problem is to choose A and M to maximize utility. FKV then assumes an ‘iceberg’ form of transport cost, which means if a good is shipped from location r to s, only a fraction 1/Trs arrives. Then the price index becomes a function of price and transport cost in all locations. Given a ﬁxed input of F, marginal input requirement for manufactured goods cM, price of manufactured goods at location r, prM , and quantity of manufactured goods at r, qrM , a ﬁrm producing a speciﬁc variety at location r chooses its price to maximize proﬁt, taking price indices G and wage rate wrM as given: πr = prM qrM − wrM ( F + c M qrM ) . Assuming there is free entry and exit, the zero-proﬁt condition implies the equilibrium output of any ﬁrm q*, equilibrium labor input of any ﬁrm l*, and the number of the varieties produced at r, nr. FKV then derived the price and wage equations for a ﬁrm located at r. After making some normalizations, FKV writes the price and wage equations in a more convenient form. The two-location version of these equations is: G11−σ =

1 ⎡ 1−σ 1−σ L1 w1 + L2 ( w2T ) ⎤⎥ ⎦ μ ⎢⎣

G21−σ =

1⎡ 1−σ L1 ( w1T ) + L2 w12−σ ⎤⎥ , ⎢ ⎦ μ⎣

w1σ = Y1G1σ −1 + Y2 G2σ −1T 1−σ , w2σ = Y1G1σ −1T 1−σ + Y2 G2σ −1 where Y is income or demand, L is the number of manufacturing workers; the subscripts indicate countries 1 or 2, the superscript M is dropped because the focus is

The Chinese Economy after WTO Accession

260

on manufacturing only. Differentiating the price indices and wage equations around 1− T 1−σ the symmetric equilibrium and deﬁne a new variable Z ≡ , an index of 1 + T 1−σ trade cost values between 0 and 1, FKV ﬁnds the home market effect as: ⎡σ ⎤ dw dL dY ⎢ + Z (1− σ )⎥ +Z = ⎢⎣ Z ⎥⎦ w L Y

(1)

FKV points out that suppose labor supply to manufacturing is perfectly elastic (DW=0), then ‘a 1 percent change in demand for manufactures (dY/Y) causes a 1/Z (>1) percent change in the employment, and hence production of, manufactures, dL/L.’1 Following some of the normalization assumptions made by FKV, the authors of this chapter transfer equation (1) into a relationship between the change in output and the change in demand (see Appendix for details): ⎡σ ⎤ dw dX dY ⎢ + Z (1− σ )⎥ +Z = ⎥⎦ w X Y ⎣⎢ Z

(2)

where X=nq* is the total output of manufacturing goods at a location. Multiplying by output X and then adding average output X on both sides of the equation, assuming dw/w equals zero, gives: X + dX = X +

1 dY X Z Y

(3)

Equation (3) says that the average output plus the deviation in output is a function of the average output plus the deviation in idiosyncratic components of demand. This relationship is similar to the relationship identiﬁed in equation (4) in Davis and Wienstein (1999), except they use dY instead of dY/Y. The Model The hypothesis tested by this research is whether demand in a country can improve our understanding of production and trade relative to the hypothesis that all production and trade are determined by endowments. The design of analysis is based on the framework developed by Helpman (1981), where endowments served to determine the broad sector structure of a country while monopolistic competition led to intrasector specialization at the industry level. A typical empirical model of the classical comparative advantage theory is the square Heckscher-Ohlin model, which uses the endowments of a country to explain 1

FKV (2000), pp. 45–58.

Home Market Effect and Its Impact on Production and Trade

261

its output. Assume that there are R regions, N sectors with Gn industries in sector n and the number of industries equals the number of endowments. Assume that the inverse of the technology matrix mapping output into factors is Ω , the output of industry g is determined by endowments: X gnr = Ωng E r

(4)

where Xgnr is industry g’s output for region r and sector n, Er is the vector of factor endowments in region r, and Ωng is the corresponding row of matrix Ω . This chapter will choose an alternative research approach from the above classical Heckscher-Ohlin model. The basic testing strategy in this chapter is taken from Davis and Weinstein (1996 and 1999), while the testing models are derived from FKV (2000) by the authors of this chapter. On the sector level, we test the above hypotheses by estimating the following Heckscher-Ohlin model, in which endowments determine the structure of production by sector: Gn

X nr = ∑ X gnr = Ωn E r

(5)

g =1

where Xnr is the vector of output for region r and sector n, Er is the vector of factor endowments in region r, and Ωn is an NxN matrix which maps endowments into output for sector n. Note that the difference between equation (4) and (5) is that equation (4) is on industry level while equation (5) is on sector level. On the industry level, we will test the hypotheses that the production is determined by economic geography. We will test the empirical versions of FKV’s theoretical model, equation (2) or Model I in this chapter. In Model I, we use X gnr to measure industry g’s production in sector n for region r, and Ygnr to measure the demand for industry g in sector n for region r. Then ( X gnr − X gnr * ) / X gnr * and (Ygnr − Ygnr * ) / Ygnr * present the percentage deviation of production of and demand for industry g of region r from the average of the rest of the countries in sector n, respectively. We then deﬁne the empirical version of Model I as the following: X gnr − X gnr * X gnr *

= a + b1

Ygnr − Ygnr * Ygnr *

+ b2

Lr − Lr Lr

+ b3

Kr −Kr Kr

+ εgnr

(6)

where L represents labor endowment, K represents capital endowment, and bars above variables represent the average across countries. The term on the left is the relative deviation of output from the average across different countries and industries, which is a function of the relative deviation of demand from the sector average across different countries and industries, the second term on the right. To avoid the omitted variables bias, we also include the changes in labor and capital, which can be viewed

262

The Chinese Economy after WTO Accession

as a Rybczynski effect of factor endowments (if a country accumulates a factor more rapidly than does the rest of the world, that country’s production and exports will sift toward industries that more intensively use that factor). Coefﬁcient b1 in Model I (Equation 6) plays a key role in hypothesis testing, since the economic geography framework predicts that the positive responsiveness of production to the changes in demand will be more than one-to-one. When b1 equals zero, we are in a world of frictionless comparative advantage in which transport costs do not matter. If b1 is between zero and one, we are in a world of comparative advantage in which transport costs affect the production location, but there are no scales economies. If b1 is greater than one, there are home market effects that affect the location of production. Data Using 1987–2001 industrial outputs, exports and imports data of China and the US, we ﬁnd domestic demands for each of the eleven industries in China and the US by subtracting net exports from GDP. Data of eleven manufacturing industries are used in the regression analysis (Table 14.A-1). These industries are organized into three sectors, non-durable goods, material-related goods, and durable goods. The capital endowments are ﬁxed investment in GDP, and the labor endowments are non-farm employment in the US and employment in the second and third industries in China, respectively. Econometrics Issues A potential problem is serial correlation or spatial autocorrelation, arising from the time-series nature of the data and the fact that regression is based on a relationship at industries level, many of which are from same sectors or same areas. It is highly probable that some of the common factors affecting different industries’ output, such as interest rate, technology, local regulation and policies, are omitted from the individual equation. Their inﬂuences, along with the state of the economy, are all collected by the error terms. The resultant correlation between cross-sectional error terms violates the Gaussian assumption of no serial correlation or spatial autocorrelation. Under this situation, although the standard OLS estimators remain unbiased and consistent, they are not efﬁcient (even asymptotically) any more. The Seemingly Unrelated Regression (SUR) method is then used to address the issue. In doing so, we assumed two scenarios: (1) The output relations all have one common coefﬁcient vector at industries level; (2) There is no coefﬁcient restriction at all, each industry has its own coefﬁcient. Another problem is heteroskedasticity, stemming from the cross-sectional nature of our data. It is possible that the industries’ output relationship as speciﬁed in equation (6) varies across sectors, as well as across geographical areas. One possible source of the varying effects is that in a larger geographic area, sectors tend to be

Home Market Effect and Its Impact on Production and Trade

263

larger, the output of one industry is more likely to be affected by other factors such as scope of economy, urbanization level, infrastructure and facilities, and it might be less related to endowments and geographic economy. In other words, the variance may be larger in larger areas and smaller in smaller areas. We are then faced with the problem of heteroskedasticity. The normal procedure to ﬁx this problem is to use weighted least squares method (WLS), the same method used by previous studies as in Davis and Weinstein (1996, 1999). However, the problem with WLS in this context is, after specifying a variance function and estimating the transformed equation, it cannot address the serial correlation problem. Since the two problems are related to the structure of the variance-covariance matrix, we can address both of them by the method of Generalized Least Squares (GLS), as employed in SUR model. If we really knew the exact form of the variance function, using WLS may be better than GLS. Under the situation of unknown variance function and the joint presence of serial correlation, using GLS should be the better choice. Results Based on the theoretical framework, two levels of industrial division, industries level and sector level, are examined separately. With the assumption that endowments determine the broader sector structure of a country while monopolistic competition determines the specialization on the industries level, we conducted two groups of tests. In one version of their paper, Davis and Weinstein (1996) considered sectors to be at 3-digit level and goods (equivalent to industries in this chapter) at 4-digit level. In another version (Davis and Weinstein, 1999), when analyzing Japanese data, they chose 19 sectors (equivalent to industries in this chapter), which is at 2-digit level. In this chapter, we deﬁne industries at 2-digit level, because it is the most detailed data of both production and trade by industry available. We group industries into three sectors, non-durable, material-related, and durable, based on traditional deﬁnition of industries. The industrial composition in each of the three sectors is listed in Appendix Table 14.A-1. Comparative Advantage in Heckscher-Ohlin Speciﬁcation We ﬁrst test the above hypothesis on the sector level, by estimating the Heckscher-Ohlin model in which endowments are assumed to determine the structure of production by sector. Since we use relative change in variables in our economic geography models, we also estimate Rybcznski model, in which relative change in endowments are assumed to determine the structure of production by sector. The results reported in Table 14.1 seem to support, to certain degree, the hypothesis that factor endowments determine the sector structure. The effect of the capital factor is signiﬁcant at the 1 percent level in all three sectors by both models and the factor of labor has negative signs in all of the six cases under investigation. The unlimited labor supply from the huge rural surplus labor and the inefﬁcient use

The Chinese Economy after WTO Accession

264

Table 14.1

Sector Level Production on Factor Endowments

Heckscher-Ohlin Model Sector I Non-durable Sector II Material-related Sector III Durable

Labor -9.09E-05** (1.78E-05) -0.0001** (1.69E-05) -0.0003** (7.95E-05)

Capital 0.045** (0.004) 0.064** (0.004) 0.074** (0.016)

Rybcznski Model Sector I Non-durable

-0.650** 0.695** (0.213) (0.100) Sector II Material-related -0.426** 0.320** (0.093) (0.044) Sector III Durable -1.047** 0.543** (0.182) (0.086) Notes: SUR regression. In Rybcznski models, labor and capital are percentage deviations to the average for the rest of the countries. Standard deviations are below coefﬁcients. * indicates signiﬁcance at 5%, ** 1%.

of labor in state owned enterprises in China may explain the negative signs of the labor factor. The heteroskedasticity and contemporaneous correlation in the errors is addressed by the Generalized Least Squares method implied in Seemingly Unrelated Regression. Home Market Effects We then test for the home market effects based on the hypotheses that the production is determined by economic geography on the industries level. According to the theoretical formulation in the previous sections, home market effect is indicated by a more than one-to-one relationship from the change in idiosyncratic demand to the change in total production relative to other areas. We ﬁrst test the hypothesis with a pure economic geography model, then incorporate labor and capital endowments in the economic geography model to see if the result is robust. We use model I (Eq 6) and its variations to compare the results. The results of aggregate estimation using data of three sectors are summarized in Table 14.2. Home market effects are indicated by the coefﬁcients of idiosyncratic demand that are both signiﬁcant and greater than one in sectors I (nondurable goods) and II (material-related goods) in both Model I and Model I*. In Model I (pure economic geography model), the percentage deviation in production is positively affected by the percentage deviation in demand, with a coefﬁcient of 1.270 and 1.265, respectively in sectors I and II. In Model I* (endowments incorporated in the economic geography model), in addition to the similar effects of its own demand as in Model I, production

Home Market Effect and Its Impact on Production and Trade

Table 14.2

Sector Level Production on Economic Geography

(Ygnr − Ygnr * ) / Ygnr * dL L dK K

265

Sector I Model I Model I* 1.270** 1.127** (0.021) (0.088)

Sector II Model I Model I* 1.265** 1.247** (0.004) (0.008)

Sector III Model I Model I* 0.003 0.0003 (0.008) (0.0038)

-

0.314** (0.119)

-

-0.013 (0.009)

-

-1.007** (0.198)

-

0.250** (0.052)

-

0.008* (0.004)

-

0.560** (0.093)

Observations 90 90 150 150 90 90 Notes: SUR regression. Dependent variable is the percentage deviation in the share of industrial production. Standard deviations are below the coefﬁcient.

of a sector is also affected by the labor and capital endowments in sector I and capital endowments in sector II). There are no home market effects shown in sector III (durables) in both models. Under the situation of increasing returns, when the demand is stronger than the rest of the countries, it causes its production to increase by a larger than proportionate amount, leading to aggregation of production in one country and thus export of that industry from the country. This is how the home market effects help shape the trade pattern. Nesting Rybczynski speciﬁcation in the economic geography models as in Model I* in Table 14.2 did not change the above results much. As a matter of fact we got almost the same coefﬁcients for the idiosyncratic demand variable in models with and without endowments.2 On the industry level, we break down the home market effects further and did a disaggregated estimation for each of the industry types. The results are reported in Table 14.3. Again, Model I is a pure economic geography model, while endowments are incorporated in the economic geography model in Model I*. Looking for coefﬁcients of idiosyncratic demand that are signiﬁcantly greater than unity, Model I in Table 14.3 shows that the home market effects concentrate on 5 out of the 11 industries under examination. They are Food, Beverage and Tobacco, Chemicals, Nonmetal Minerals, and Metals. In the speciﬁcation of Model I* we ﬁnd 7 industries with the home market effect: Food, Beverage and Tobacco, Paper, Chemicals, Plastic and Rubber, Nonmetal Minerals, and Metals. Overall, we can see 5 of the 11 industries showed strong home market effects estimated by both models.

2 The coefﬁcients for the factor endowments in the nested economic geography model convey mixed messages. Capital is highly signiﬁcant but labor has a wrong sign in two of the three cases. This is consistent with our analyses of the factor endowments (Table 14.1). The theoretical implication could be interpreted as some evidence for the comparative advantage.

266

Table 14.3 Industry Name Food Beverage and Tobacco Textile Paper

The Chinese Economy after WTO Accession

Disaggregated Estimation on Industry Level Model I Econ. Geo. (Ygnr − Ygnr * ) / Ygnr * ** 2.361** (0.139) ** 2.294** (0.013) 0.89** (0.112) -

Chemicals

**

Plastic and Rubber Nonmetal Minerals Metals

-

Machinery and Electric Equipment Transport Equipment Instruments

** ** -

0.836** (0.038) 1.308** (0.058) 0.815** (0.033) 1.481** (0.042) 1.221** (0.043) 0.021** (0.008)

Model I* Econ. Geo. ** ** ** ** ** ** ** -

(Ygnr − Ygnr * ) / Ygnr *

2.395** (0.019) 2.028** (0.046) 0.994** (0.033) 1.228** (0.006) 1.548** (0.014) 1.277** (0.015) 1.063** (0.015) 1.134** (0.012) 0.630** (0.114)

Obs. 30 30 30 30 30 30 30 30 30

-

0.0004* 0.004* 30 (0.0002) (0.003) -0.004 0.016 30 (0.005) (0.013) Notes: SUR regression. Dependent variables are percentage deviation. Standard deviations are below coefﬁcients. The changes in labor and capital endowments are included in Model I as independent variables.

Conclusions In this chapter, we derived a model for empirical testing of the home market effect directly from Fujita, Krugman, and Venables (2000) and linked it with the empirical testing model in Davis and Weinstein (1996 and 1999). We use the models to test if production and trade patterns in China and the US are shaped by home market effects. Our study on the regional production and trade patterns uses production, demand and endowment data of China and the US for eleven industries from 1987 to 2001. The results show the existence and importance of home market effects in determining production and trade structure in China and the US. We found that on the sector level, the production and trade patterns are determined mainly by factor endowment, speciﬁcally, the capital endowment. The home market effect is found in two of the three manufacturing sectors, non-durables and material-related, but no such effect

Home Market Effect and Its Impact on Production and Trade

267

is found in the durable sector. On the industry level, estimated by both the pure economic geography model and the economic geography model with endowments incorporated, the home market effect tends to focus on at least ﬁve of the eleven industries, they are: Food, Beverage and Tobacco, Chemicals, Nonmetal Minerals, and Metals. To some degree, the ﬁnding is comparable to Davis & Weinstein (1999), where they identiﬁed strong empirical support for home market effects on regional trade patterns in Japan. References Bao, S., G.H. Chang, J D.Sachs, and W.T.Woo (2002), ‘Geographic factors and China’s regional development under market reforms, 1978–1998,’ China Economic Review, Vol. 13, pp. 89–111. Batisse, C. (2002), ‘Dynamic externalities and local growth: a panel dataanalysis applied to Chinese provinces,’ China Economic Review, Vol.13, pp. 231–51. Brülhart, M. and F. Trionfetti (2001), ‘A test of trade theories when expenditure is home biased,’ The Management Centre Research Papers No. 005, Kings College, University of London. Brun, J. F., J.L. Combes and M.F. Renard (2002), ‘Are there spillover effects between coastal and noncoastal regions in China?’ China Economic Review, Vol. 13, pp. 161–69. China State Statistical Bureau (1998), China Statistical Yearbook 1998, Statistical Press, Beijing. China State Statistical Bureau and Euromonitor/Soken (2000), China Marketing Data and Statistics, Euromonitor International, London, Chicago. Davis, Donald R. and D.E. Weinstein (1996), ‘Does economic geography matter for international specialization,’ Working paper No. 5706. NBER, Cambridge, MA. Davis, Donald R., and D.E. Weinstein (1999), ‘Economic geography and regional production structure: an empirical investigation,’ European Economic Review, Vol. 43, pp. 379–407. Dixit, K. Avinash and J.E. Stiglitz (1977), ‘Monopolistic competition and optimum product diversity,’ American Economic Review, Vol. 67, No. 3, pp. 297–308. Feenstra, Robert C., James A.Markusen, and Andrew K. Rose (1998), ‘Understanding the home market effect and the gravity equation: the role of differentiated goods,’ National Bureau of Economic Research Working Paper No. 6804. Fujita, M., P. Krugman and Anthony J. Venables (2000), The Spatial Economy: Cities, Regions, and International Trade, MIT Press, Cambridge. Grossman, G. and E. Helpman (1991), Innovation and Growth in the World Economy, MIT Press, Cambridge. Head, K. and J. Ries (2001), ‘Increasing returns versus national product differentiation as an explanation for the pattern of US-Canada trade,’ American Economic Review, Vol. 91, No. 4, pp. 858–76.

268

The Chinese Economy after WTO Accession

Helpman, E. (1981), ‘Imperfect competition and international trade: evidence from fourteen industrial countries,’ in A.M. Spence and H. Hazard (eds), International Competitiveness, Ballinger Publishing, New York. Helpman, E. and Paul. R. Krugman (1985), Market Structure and Foreign Trade, MIT Press, Cambridge. Justman, M. (1994), ‘The effect of local demand on sector location,’ Review of Economics and Statistics, Vol. 76, No. 4, pp. 742–53. Krugman, Paul R. (1980), ‘Scale economics, product differentiation, and the pattern of trade,’ American Economic Review vol. 70(5), pp. 950–59. Krugman, Paul R. (1991), ‘Increasing returns and economic geography,’ Journal of Political Economy, Vol. 99, No. 3, pp. 483–99. Linder, S.B. (1961), An Essay on Trade and Transformation, Wiley, New York. Trionfetti, F. (2001), ‘Using home-biased demand to test for trade theories,’ Weltwirtschaftliches Archiv, Vol. 137, pp. 404–26. Weder, R. (1995), ‘Linking absolute and comparative advantage to intra-sector trade theory,’ Review of International Economics, Vol. 3, No. 3.

Home Market Effect and Its Impact on Production and Trade

269

Appendix Table14.A-1 Sector Classiﬁcation Sector Industry I. Non-durable Food Beverage and Tobacco Textile II. MaterialPaper Related Chemicals

III. Durable

Description Food processing and manufacturing Beverage manufacturing and tobacco processing Textile, garments, leather products and footwear Timber processing, paper making, and paper products Raw chemical materials, chemical products, and medical products Plastic and Rubber Plastic and rubber products Nonmetal Mineral Nonmetal mineral products Metals Metals melting and pressing and metal products Machinery and Electric Machinery and electric equipment Equipment Transport Equipment Transport equipment Instruments Instruments, meters, cultural and ofﬁce machinery

Derivation of Equation (2) from the Theory in FKV (2000) We start from the price index (Equation 4.34 in FKV 2000): 1

1

⎤ 1−σ ⎤ 1−σ ⎡ 1 R ⎡ R Gr = ⎢ ∑ ns ( psM TsrM )(1−σ ) ⎥ = ⎢ ∑ LMs ( wsM TsrM )(1−σ ) ⎥ ⎢ μ s =1 ⎥ ⎥⎦ ⎢⎣ s =1 ⎣ ⎦ where ns is the number of varieties (ﬁrms) at location s, psM is the price of variety produced at s, TsrM is the transport cost from location s to r, LMs is the number of manufacturing workers at location s, μ is the expenditure share of manufactured goods, M stands for manufacturing, and σ is the elasticity of substitution between varieties. Deﬁne output of all manufacturing ﬁrms at s as X sM = ns q * , where q* is equilibrium output of any active ﬁrm, then the price index becomes: 1

⎡1 R ⎤ 1−σ Gr = ⎢ ∑ X sM ( psM TsrM )(1−σ ) ⎥ ⎢ q * s =1 ⎥ ⎣ ⎦ Because prM = c M wrM / ρ (FKV 4.20):

270

The Chinese Economy after WTO Accession 1

⎡1 R c M wsM M (1−σ ) ⎤⎥ 1−σ Gr = ⎢ ∑ X sM ( Tsr ) ⎢ q * s =1 ⎥ ρ ⎣ ⎦ where c M is the marginal input requirement, wsM is the wage rate for manufacturing workers at location s, ρ is the intensity of the preference for variety. Using FKV’s normalizations assumptions c M = ρ , q* = μ (FKV 4.29 and 4.33), the price index becomes: 1

⎡1 R ⎤ 1−σ Gr = ⎢ ∑ X sM ( wsM TsrM )(1−σ ) ⎥ ⎢ μ s =1 ⎥ ⎣ ⎦ The only difference between this equation and FKV’s equation (4.34) (shown at the beginning of this section) is that FKV’s LMs is replaced by X sM . Following FKV’s steps, we can build a two location symmetric model, and derive the home market effect like FKV’s equation 4.42 or equation 1 in this chapter with dX/X (instead of dL/L) on the left-hand side: ⎡σ ⎤ dw dX dY ⎢ + Z (1− σ )⎥ +Z = ⎢⎣ Z ⎥⎦ w X Y which is equation (2) in this chapter. (For details of the steps of FKV’s derivation see Fujita, Krugman, and Venables (2000) pp. 45–58 or the section entitled ‘Theoretical Framework’ in this chapter.)

PART IV Economic Performance and Labor Market

This page intentionally left blank

Chapter 15

Comparing Selected Aspects of Economic Performance: China versus India Peter E. Koveos and Yimin Zhang

Introduction and Background India entered the 1990s facing severe macroeconomic and balance of payments crises. On the ﬁscal side, deﬁcits in 1990–91 were running close to 10 percent of GDP. The rate of inﬂation was more than 10 percent. The current account deﬁcit was about 3.1 percent of GDP, and foreign exchange reserves were dwindling. Output growth came to a halt. As the country battled internal challenges, the international environment was also changing. One of the most signiﬁcant international developments entailed the abandonment of the ‘central planning’ model in favor of a market-based framework. The Soviet Union itself was both disintegrating and undergoing radical economic reforms. In its search for change and meaningful economic progress, Narasimha Rao’s government had no other alternative but to initiate radical reforms aimed at changing India’s own economic landscape and its relationship to the fast moving outside world. The reforms undertaken covered foreign trade and investment, as well as exchange rate and industrial policies. The objective was to make the transition to a market based economy while integrating India within the global economy.1 China began its departure from the planned economy model in 1978. The accompanying reforms have had a purely Chinese ﬂavor. Observers of these reforms have noted that the country’s experiences may be distinguished from reform experiences in other countries through its use of a gradual as well as pragmatic approach and its deployment of intermediate adjustment mechanisms. Furthermore, the objective of changing the economy while maintaining the basic elements of the political environment was characterized as ‘market socialism.’ The impact of the reform has varied between the internal and external sectors, among geographical areas, industry groups, and income levels. The challenges faced have been immense and multifaceted, involving the country’s economic, political, and social frameworks.

1

See Srinivasan and Tendulkar, 2003, pp. 1–2.

The Chinese Economy after WTO Accession

274

According to at least one source, China’s recent reform period may be distinguished into four sub-periods (Bell, 1993). The ﬁrst sub-period (1978–84) introduced reform practices by allowing for a larger role by the market and the establishment of material incentives. Agriculture was identiﬁed as the emphasis of reform. A ‘household responsibility’ system was put in place to improve on the results of collectivization. A ‘dual price’ system was instituted, allowing farmers to sell any level of output exceeding the ofﬁcial quota at market determined prices. The ‘dual price’ system was accompanied by the ‘dual land-use’ system. Special economic zones were also introduced during this sub-period. The second sub-period (1984–88) featured implementation of several sector reforms, including wage, ﬁnance and currency reforms. Enterprise reforms were undertaken, based on the ‘managerial responsibility’ concept. A two-tier system was implemented in foreign exchange, allowing some enterprises to retain a portion of foreign currency receipts. To attract needed capital and technological expertise, 14 coastal cities were designated as targets of foreign investment. The third subperiod (1988–91) included a retrenchment of the reform effort, as inﬂation became a signiﬁcant obstacle to stabilization in 1988. The most recent period (1992-present) featured renewed emphasis on reforms and further opening (Adams, 1994; Koveos and Tang, 2002). China’s entry into the WTO may signal the beginning of a new period in the reform process, one characterized by an accelerated phase of opening to the rest of the world. By most accounts, Chinese reforms have shown uneven success. While incentives proved effective in agriculture, they did not work as well in reforming the State Owned Enterprises (SOEs). Managerial responsibility did not generate the same positive results as its household counterpart did. Instead, SOEs experienced a fall in their share of national output and in their ability to compete. To this day, they remain one of China’s most serious challenges. It is doubtful whether Chinese reforms can ultimately succeed without satisfactory resolution of the SOE challenge. In addition to the development experience of each of these countries individually, the relationship between them has also been fascinating. This relationship has had political, military, and economic dimensions and has developed into a modern day rivalry. Each country is trying to present itself as more successful in its economic reforms than the other, a more attractive destination for foreign investment, and a more important global economic power in the years to come. At the beginning of the current decade, India’s GNP per capita is about half of that of China’s.2 India is growing at a 6 percent rate, compared to 10 percent for China. On the international front, China appears to be attracting more investment, as well as more intense overall interest, than that directed towards India. The Economist reports that China received $52.7bn in foreign investment in 2002, compared to India’s $2.3bn.3 On a less tangible basis, the eyes of the world seem to be ﬁxed on China’s economic progress, while India receives relatively less attention. And, to 2 3

McKinsey, 2001, MGI reports, p. 1. Economist, June 21, 2003, p. 22.

Comparing Selected Aspects of Economic Performance

275

many outside observers, India seems to be increasingly concerned with catching up China. As David Wessel says: India-despite its impressive call centers, software ﬁrms and pharmaceuticals factoriesis viewed as a less formidable competitor. The Chinese have order, discipline, modern telephones and roads, less poverty and faster economic growth. The Indians have democracy, chaos, lousy phones and roads, more poverty and slower growth (The Wall Street Journal, Thursday, July 24, 2003, p. A2).

This study presents a descriptive view of the economic performance of these two giants using both traditional, GNP-based comparisons, as well as presenting alternative measures of performance. The theoretical and empirical concerns encountered in making cross-country comparisons are well known and acknowledged. This study cannot possibly address all relevant concerns. It should then be viewed as another building block in our efforts to understand the various paths countries take in pursuing internal and external economic objectives. The next section provides a comparison using GDP measures. The third section presents a comparison between the two economies using additional measures. The fourth section analyzes the time lags observed, while the ﬁfth section presents additional comparisons. The ﬁnal section offers concluding comments. Analyzing Economic Performance Using GDP India, which has the second largest population in the world, has made marked advances in its economy, especially in its computer software sector. However, when we examine its economic performance in terms of per capita GDP in US$, we ﬁnd that India still has not caught up with China. According to Figure 15.1 and Figure 15.2, during the period from 1980 to 1992, per capita GDP in US$ was about the same in the two countries. Since then, India has fallen behind China in both per capita GDP

Figure 15.1 Per Capita GDP in US$

276

The Chinese Economy after WTO Accession

Figure 15.2 Per Capita GDP Annual Growth Rate in US$ and per capita GDP growth. The gap between the two countries widened after 1994. By 2000, China’s per capita GDP in US$ was twice that of India’s. Analyzing Economic Performance Using Additional Measures The use of GDP measures to compare economic performance across countries has been an important ﬁrst step in analyzing national economic performance. As Barro et al. indicate, even small differentials in GDP growth rates, compounded over a long period of time, can result in vast differences in per capita GDP.4 However, GDP based measures have also been criticized in the literature. According to one such criticism, assessing economic development through an index based on per capita GNP or GDP in US$ may be misleading if the overall index is inﬂuenced by its individual components, such as exchange rate and price movements.5 To address such criticisms, and therefore arrive at a more representative measure of the differential in the economic performance between the two countries, we use two composite indices: the per capita production index and the per capita living standard (or consumption) index.6 The appropriate equations and measures are as follows: 1

⎡ k ⎤k PCPDI t = ⎢∏ ( PRDPCi )t ⎥ ⎢⎣ i =1 ⎥⎦

(1)

1

⎡ l ⎤l PCLSI t = ⎢⎢∏ ( LSIPC j )t ⎥⎥ ⎣ j =1 ⎦

(2)

4 Barro and Sala-I-Martin, 1999, p. 5. 5 See Appendix. See also Dowrick and Quiggin, 1997, for an approach of constructing multilateral true quantity indices; also, see Dowrick, Dunlop, and Quiggin, 2003, for a recent treatment of indicators using revealed preference analysis. 6 See Appendix for rationale for use of these indices.

Comparing Selected Aspects of Economic Performance

Table 15.1

277

Indices Selected for PCPDI and PCLSI Classiﬁed by Group

13 per capita indexes related to production 1. Heavy industry production per capita cement output (kg) per capita petroleum output (kg) per capita motor vehicles output (unit/10000 persons) per capita fertilizer output 2. Light industry production per capita cloth output (m) per capita newsprint output (kg) per capita chemical ﬁber output (kg) per capita refrigerator output (set/10000 persons) 3. Agriculture production per capita grain output (kg) per capita ﬁsheries output (kg) per capita hog output (head) per capita milk output (kg) 4. Export per capita export (US$)

11 per capita indexes related to living standard (consumption) 1. Raw material per capita electric power consumption (kwh) per capita steel consumption (kg)

2. Infrastructure telephone sets per 10000 persons per capita freight trafﬁc (ton-kms) per capita passenger trafﬁc (passenger kms)

3. Culture university students per 10000 persons physicians per 10000 persons

4. Living quality life expectancy (years) per capita deposit (US$) spending on health care/GDP

where PCPDIt – per capita production index at year t; (PRDPCi)t – selected index related to per capita output of principle products at year t; PCLSIt – Per capita living standard/consumption index at year t; (LSIPCj)t – selected index related to per capita living standard/consumption index at year t. Following Equation (1), Equation (2) and Table 15.1, we generate Figures 15.3 and 15.4. Accordingly, China outperformed India in both PCPDI and PCLSI. In particular, since 1991, both China indices began to accelerate, thus widening the gap between the two countries. This behavior precedes by three years the behavior of the change in US$ per capita GDP. One explanation for this differential may lie in the relative movements of the values of each country’s currency versus the US$. Thus, the exchange rate between the US$ and the RMB moved from 1:5.76 in 1993 to 1:8.62 in 1994, which dropped China’s per capita GDP in US$ from $511 in 1993 to $457 in 1994, even though per capita GDP in local currency increased from 2939 yuan to 3923 yuan. During the same period, the exchange rate between the US$ and India’s rupee remained around 1:31.38. So, the drop in China’s per capita GDP in US$ in 1994 shown in Figure 15.1 is not reﬂected in Figure 15.3, since PCPDI and PCLSI are not affected by exchange rates.

278

The Chinese Economy after WTO Accession

Figure 15.3 PCPDI of China and India

Figure 15.4 Annual Growth Rate of PCPDI

Figure 15.5 PCLSI of China and India

Comparing Selected Aspects of Economic Performance

279

Figure 15.6 Annual Growth Rate of PCLSI

Figure 15.7 Comparison between China and India Alternatively, when we divide China’s PCPDI, PCLSI and per capita GDP in US$ by India’s PCPDI, PCLSI and per capita GDP in US$ respectively, we ﬁnd that the ﬁrst two ratios are higher than the third, see Figure 15.7. This implies that China advanced to a greater relative extent than India in terms of PCPDI and PCLSI than it did in terms of per capita GDP in US$. There may be three reasons for this ﬁnding.7 The ﬁrst may be the inﬂuence of the exchange rate. In 1985, the exchange rate between the US$ and the Indian rupee was 1:12.166. By 2000, it reached 1:43.6. Over a period of 15 years, the rupee declined in value by 3.38 times while, over the same period, the RMB depreciated 2.59 times against the US$. According to the equation below, we conclude that the exchange rate change should be excluded from consideration.8 %GNPPC$=%GNPcon + %PRIGNP - %POP - %FER 7 8

See Appendix. For details, please see Zhang, Chen and Koveos, 2003.

(3)

280

The Chinese Economy after WTO Accession

where %X denotes the growth rate of variable X; GNPPC$ – per capita GNP in US$; GNPcon – per capita GNP in local constant price; PRIGNP – price index for GNP; POP – population; FER – foreign exchange rate. The second reason may be the inﬂuence of the price index. From 1980 to 1998, the price index for GDP in India increased by 449%, whereas China’s increased by only 321%. According to Equation (3), India’s per capita GDP was disproportionately inﬂuenced by the price index. So, the gap between the two countries in terms of per capita GDP in US$ was narrower than that expressed in terms of PCPDI and PCLSI. The third reason may lie in the nature of the components included in the PCPDI and PCLSI. The items selected, that is, may show China under a relatively better light. Due to data limitation, however, it is difﬁcult to select all appropriate items. Leads and Lags in Economic Performance In this section, we examine the manner in which China’s performance led India’s, or vice versa. According to Table 15.1, we categorize production into four groups: heavy industry, light industry, agriculture and export. Consumption (living standard) is also categorized into four groups: raw material, infrastructure, culture and living quality. We consider geometric means of ratios (index of China over that of India) over items within each group. 1

china ⎡ ai ( PRDPCkChina )t ⎤⎥ ai RPCPDI i = ⎢∏ India Inida ⎢ k =1 ( PRDPC ) ⎥ k t ⎦ ⎣

(4)

1

China ⎡ aj ( LSIPCkchina )t ⎤⎥ aj RPCLSI j = ⎢∏ India Inida ⎢ k =1 ( LSIPC ) ⎥ k t ⎦ ⎣

(5)

where RPCPDIi – relative per capita production index between China and India; RPCLSIj – relative per capita living standard/consumption index between China and China –selected index of per capita product output of China in group India; ( PRDPCk )tIndia i at year t; ( PRDPCk )t –China selected index of per capita product output of India in group i at year t; ( LSIPCk )t –Indiaselected index of per capita living standard of China in group j at year t; ( LSIPCk )t – selected index of per capita living standard of India in group j at year t. The results of the calculations are shown in Figure 15.8 and Figure 15.9. China held a lead over India in exports and raw materials. Both ratios between China and India were about 5 in 2000, implying that China’s exports and raw materials in 2000 were ﬁve times as much as India’s. China also has an advantage over India with respect to heavy industry, light industry, agriculture and infrastructure. The gap narrows only with respect to the culture and living quality items.

Comparing Selected Aspects of Economic Performance

Figure 15.8 Relative Per Capita Production Index

Figure 15.9 Relative Per Capita Living Standard/Consumption Index

Figure 15.10 Time Lag of PCPDI and PCLSI Index, India and China

281

282

The Chinese Economy after WTO Accession

Finally, we use PCPDI and PCLSI to uncover time lags in the stages of the economic performance between China and India. Thus, we assume that, if India’s PCPDI in year t is identical to China’s in year t-n, then we conclude that India’s PCPDI lags China’s by n years. The calculation of the time lag for PCLSI follows the same principle. According to Figure 15.8, India has fallen further and further behind China since 1987 in both PCPDI and PCLSI. In relative terms, the time lag between China and India was narrower in terms of PCLSI. By 2000, India was 14 years behind China in PCPDI and 10 years in PCLSI. Selected Additional Comparisons There are indeed a number of approaches researchers can use in comparing economic performance of countries. Some approaches may be presented as substitutes and enhancements of already existing approaches, while yet others may be presented as complements. For example, the approaches presented above apply to comparisons that deal with the past performance of the two countries, primarily at the macroeconomic and sector-wide levels. More relevant, but even more difﬁcult to assess, are comparisons involving future performance. If the past ‘belongs’ to China, according to one recent study, ‘the future belongs to India.’9 Huang and Kanna state that India’s economy is more dynamic at the microeconomic level. Thus, India’s approach of internal reliance and development of the knowledge-based sector will in the future prove more effective than China’s export- and FDI-driven reliance. Following another approach,10 Khanna and Palepu focus on the environment with which business groups operate in India and other countries. They indicate that India has underdeveloped and illiquid equity markets, nationalized banks, inadequate management training, and limited enforcement of liability laws, high corruption, and unpredictable enforcement mechanisms. Indeed, assessing a country’s ability to sustain its pace of economic development and compete effectively entails examination of its entire economic infrastructure. This is an important task, but also an extremely difﬁcult one to be undertaken in its entirety. Most studies review selected aspects of countries’ economic infrastructure. For example, in its Global Competitiveness Report, the World Economic Forum has been offering a series of country rankings involving each country’s ability to compete. The rankings according to the Growth Competitiveness Index, or GCI, represent ‘a best estimate of 102 economies’ underlying prospects for growth over the coming ﬁve years.’11 Selected aspects of these rankings are presented in Table 15.2. In addition to the rankings for China and India, rankings for S. Korea and the Russian Federation are also presented. S. Korea is included because of its rise to economic prominence and proximity to China and India. Russia was included to offer another perspective of another major economy in transition. Both China and India are ranked behind S. 9 Huang and Khanna, 2003. 10 Khanna and Palepu, 2000. 11 Porter, et al., 2004.

Comparing Selected Aspects of Economic Performance

Table 15.2

283

Growth Competitiveness and Business Competitiveness

KEY FACTS

GDP Per Capita (at PPP 2002) GDP Per Capita Rank Change in GDP Per Capita Relative to the USA (1995 - 2002) - Rank Change in GDP Per Capita Relative to the USA (1995 - 2002) - Hard Data GROWTH COMPETITIVENESS RANK Technology Index Rank Information and Communication Technology Subindex Rank Innovation Subindex Rank Public Institution Index Rank Corruption Subindex Rank Contracts and Law Subindex rank Macroeconomic Environment Index Rank Macroeconomic Stability Subindex Rank Country Credit Rating Rank Government Waste Subindex Rank BUSINESS COMPETITIVENESS INDEX Source: Porter et al., 2004.

CHINA INDIA S.Korea

Russian Federation

2003– 2004

2003– 2004

2003– 2004

2003–2004

$4,475 66 7

$2,571 78 17

$16,465 30 56

$7,926 49 79

4.24

1.47

-0.6

-1.79

44 65 62

56 64 75

18 6 11

70 69 56

70 52 50 60 25 4 34 35 45

66 55 80 35 52 43 48 72 37

7 36 38 34 23 6 27 30 23

27 81 75 91 61 61 55 76 63

Korea, but ahead of Russia in most categories. China and India are very close in terms of their respective Technology and Public Institution ranks. China’s macroeconomic environment, however, is ranked decidedly ahead of India’s. However, India’s Business Competitiveness Index (BCI) is more favorable than China’s. The BCI emphasizes the micro-economic foundations of a country’s productivity, relying on measures that address sophistication of company operations and the quality of the national business environment. Comments and Conclusions The study presented a series of comparative analyses between the economies of China and India. Comparing economic performance between countries has always been a controversial undertaking (see, for example, Usher, 1980, for an early work on the subject). Comparisons become even more problematic when the countries involved are clearly in a state of macro and micro disequilibria, as countries in transition are by deﬁnition. Even when we assume that there is indeed a sense of credibility in the comparison, the natural next question is: how was the differential in economic

284

The Chinese Economy after WTO Accession

performance generated and what are its implications? The study does not purport to offer a full analysis of relative economic performance and the underlying reasons for the existence of such performance differences. It should therefore be viewed as a mere contribution to a very timely and important debate conducted around the world among economists and public policy makers. The analysis conducted leads to two conclusions. First, whether we use per capita GDP in US$ or PCPDI and PCLSI, India is found to lag behind China considerably. The gap between China and India is wider when we measure economic performance through PCPDI and PCLSI. Second, due to the inﬂuence of the exchange rate and the price index, using per capita GDP in US$ to measure economic growth can be misleading. By comparison, using PCPDI and PCLSI circumvents such inﬂuences. However, there also exist certain shortcomings when substituting PCPDI and PCLSI for per capita GDP/GNP in US$. To be sure, in order to reﬂect a country’s true economic development, enough items should be collected to appropriately calculate PCPDI and PCLSI. But this effort will be surely limited by data availability. It is also realized that measuring economic growth through PCPDI and PCLSI tends to ignore the development of the service sector. Therefore, to more accurately measure a country’s economic development, an eclectic approach, using a variety of indices, is recommended. Certainly, the work of the World Economic Forum group has been offering such multifaceted comparisons. Even at that, however, conclusions may be limited by the context within which they apply. Has China’s recent economic performance been superior to India’s? If so, how can it be explained? There have been many attempts to answer this question, most from inside India. One type of answer refers to the lack of credibility of the data provided by China. Simply put, Chinese data are unreliable. Any kind of analysis conducted using ofﬁcial Chinese data is, therefore, equally suspect in its conclusions. Another type of answer, however, examines the nature and true objective of each country’s reforms. According to Kaushal (2003), for example, Chinese reforms came from within. India’s reforms were forced from outside. China’s reforms were, therefore, successful because: Chinese economic reforms had a human face. We continue to lack that, as if our reforms are not for Indians but for the World Bank and the IMF. Maybe that’s why China is emerging as an economic superpower, but India is not. (Kaushal, 2003)

What does the future hold for these two countries? Will they face it as antagonists or collaborators? According to Bhattacharjee (2003), ‘the two sides should think of a Hindi-Chini Saath Saath scenario, without trying to resurrect the earlier Bhai Bhai syndrome.’ There is, indeed, a great deal of potential beneﬁts associated with cooperation between these two countries. Bhattacharjee concludes: ‘Singing about camaraderie may not work, but the two economies can deﬁnitely march in step with each other.’

Comparing Selected Aspects of Economic Performance

285

References Adams, G.F. (1994), ‘Economic Transition in China: What makes China Different?’ in C.H. Lee and H. Reisen (eds), From Reform to Growth, China and Other Countries in Transition in Asia and Central and Eastern Europe, Organization for Economic Cooperation and Development, Paris, France. Bell, M.W. (1993), China at the threshold of a market economy, International Monetary Fund, Washington, D.C. Bhattacharjee, Jay (2003), ‘India, China can march in tandem economically,’ The Times of India, August 31. Dowrick, S. and J. Quiggin (1997), ‘True Measures of GDP and Convergence,’ American Economic Review, Vol. 87, No. 1, pp. 501–29. Dowrick, S., Y. Dunlop, and J. Quiggin (2003), ‘Social Indicators and Comparisons of Living Standards,’ Journal of Development Economics, Vol. 70, pp. 501–29. Economist (2003), ‘India and China: Two systems, one grand rivalry,’ Economist: Special Report, June 21, pp. 21–23. Huang, Yasheng and Tarun Khanna (2003), ‘Can India Overtake China?,’ Foreign Policy, July/August, pp. 74–81. Kaushal, Neeraj (2003), ‘Why China, why not India?’ The Economic Times, November 26, Dow Jones Interactive Publications Library. Khanna, Tarun and Khrishna Palepu (2000), ‘Is Group Afﬁliation Proﬁtable in Emerging Markets? An Analysis of Diversiﬁed Indian Business Groups,’ The Journal of Finance, Vol. LV, No. 2, pp. 867–91. Koveos, Peter and Linghui Tang (2002), ‘Chinese Economic and Financial Reforms,’ in A. Young, I. Teodorovic, and P. Koveos (eds), Economies in Transition: Conception, Status, and Prospects, World Scientiﬁc Publishing. McKinsey (2001), Global Institute Report, India, www.mckinsey.com/knowledge/ mgi/, pp. 1–9. Porter, Michael E., Klaus Schwab, Xavier Sala-I-Martin, and Augusto Lopez-Claros (2004), The Global Competitiveness Report, 2003-2004, World Economic Forum, Geneva. Srinivasan T.N. and Suresh D. Tendulkar (2003), Reintegrating India With the World Economy, Institute for International Economics, Washington DC, March. Srivastava, Sadhana (2003), ‘What is the True Level of FDI Flows to India?’ Economic Political Weekly Commentary, February 15. Usher, Dan (1980), The Measurement of Economic Growth, New York, Columbia University Press. Wessel, David (2003), ‘India Could Narrow Its Economic Gap With China,’ The Wall Street Journal Thursday, July 24. Zhang Yimin, Chung Chen, Peter Koveos (2003), ‘Further Evidence of Economic Development of Shanghai and Taiwan’, Unpublished manuscript.

286

The Chinese Economy after WTO Accession

Data Sources United Nations, Statistical Yearbook for Asia and the Paciﬁc. China National Statistical Bureau (2001), China Statistical Yearbook, 2001, China Statistical Publishing House.

Comparing Selected Aspects of Economic Performance

287

Appendix One of the most important areas of economics is the study of the relationship among investment, consumption and growth. The basic macroeconomic identity includes consumption, investment and net export. The indices used in this study basically reﬂect the GDP identity, incorporating consumption (net export belongs to consumption) and production activities. Expressing in terms of per capita units, we employ the per capita consumption index – PCLSI (referred to in this study as per capita living standard index, or per capita consumption index) and per capita production index – PCPDI. The table accompanying this Appendix illustrates the importance of Consumption/Standard of Living and Investment/Production to GDP growth rate derivations. According to the above concept we select four subcategories in PCPDI, namely: Heavy industry production index – supply investment equipments, Light industry production index – provide consumer goods, Agriculture production index – supply food products and Exports. Also four subcategories are included in PCLSI, namely: Raw material, Infrastructure, Culture and Quality of Life. The selection of the individual data series within each subcategory is based on common sense, the authors’ experiences, and data availability. The individual series within each index belong to the same subcategory, so they are to some extent related to each other. Lack of appropriate and reliable time series information, however, prohibits us from ascertaining the exact relationship among series. The process of sub-index and individual series selection and construction suffers from lack of complete objectivity and from possible obsolescence in the data used. There are, also, speciﬁcation problems. The per capita production index does not account for new products and sectors, such as the tertiary and high tech sectors. Similarly, the per capita standard of living index does not account for such aspects as environmental quality. GDP Growth Rates, Consumption, and Investment GDP=Consumption + Investment + Net Export Then the growth rate of GDP can be determined by: ΔGDP ΔConsum + ΔInvest + ΔNExpo = GDP GDP Using the data from China Statistical Yearbook we obtain following table.

288

The Chinese Economy after WTO Accession

Table 15.A-1 The GDP Growth Rate Contributed by Investment, Consumption and Net Export GDP Growth rate 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

7.60% 7.80% 5.20% 9.10% 10.90% 15.20% 13.50% 8.80% 11.60% 11.30% 4.10% 3.80% 9.20% 14.20% 13.50% 12.60% 10.50% 9.60% 8.80% 7.80% 7.10% 8.00% 7.50% 8.00%

Contributed by consumption 7.08% 6.20% 5.63% 4.79% 7.11% 9.32% 9.00% 4.58% 6.00% 7.62% 3.08% 0.29% 5.42% 8.66% 4.72% 6.12% 6.13% 6.68% 4.76% 5.13% 5.67% 5.86% 3.14% 2.88%

Contributed by investment 0.72% 1.47% -1.00% 2.73% 4.53% 6.70% 9.25% 2.78% 2.97% 4.92% 1.16% -0.50% 3.40% 7.22% 12.08% 2.98% 3.84% 2.29% 2.02% 2.32% 2.37% 2.15% 4.47% 4.55%

Contributed by net export -0.20% 0.13% 0.57% 1.58% -0.73% -0.82% -4.75% 1.43% 2.63% -1.24% -0.15% 4.02% 0.38% -1.68% -3.30% 3.50% 0.53% 0.63% 2.02% 0.35% -0.95% -0.01% -0.10% 0.57%

From the above table we can see GDP growth in China was mainly impacted by growth in consumption and investment. Therefore, PCPDI and PCLSI are selected as major category indices.

Comparing Selected Aspects of Economic Performance

289

GNP and Its Components Let GNPPC$, GNPPCcur, GNPPCcon, and FER denote per capita GNP in US$, per capita GNP in local current price, per capita GNP in local constant price and foreign exchange rate respectively. We have: GNPPC $ =

GNPPC cur GNPPC conx PRT GNP GNP conxx PRI GNP = = OPxFER FER x FER PO

(1)

where PRIGNP is price index for GNP and POP is population. Taking the logarithm of both sides of Equation (1), we observe that: log e( GNPPC $ ) = log e( GNP con )+ log e( PRI GNP ) - log e(POP) - log e(FER)

(2)

Since the growth rate of a variable may be approximated by the change in logged observations, we have: Δlog e( GNPPC $ ) = log e( GNPPC $(t)) - log e( GNPPC $(t - 1)) Δlog e( GNPPC $ ) = Δlog e( GNP con )+ Δlog e( PRI GNP ) - Δlog e(POP) - Δloog e(FER)

(3) (4)

Hence: %GNPPC $ = %GNP con +% PRI GNP - %POP - %FER

(5)

where %X denotes the growth rate of variable X. From Equation (5) it is obvious that devaluation of the local currency, population growth, decrease in price index for GNP or decrease in constant price GNP will have a negative effect on per capita GNP in US$.

This page intentionally left blank

Chapter 16

Low Wage and Low Labor Standards in China: A Substitute Explanation of ‘The Race-to-the-Bottom’ Sheng Li

Introduction Globalization has accelerated in recent decades with the support and encouragement of industrialized countries. Nonetheless, globalization has proved to be a twoedged sword for these selfsame industrialized countries. Globalization has meant an increase in trade and markets for exports, but globalization has also meant increased competitive pressure for both the companies in industrialized countries and for the labor markets and workers in industrialized countries. Consequently, some in developed countries are having second thoughts about globalization due to the negative results of free trade and capital mobility in an age of invigorated globalization. One view, commonly held by many politicians in industrialized countries, is that nations who face these competitive pressures unprotected have and must lower wages and labor standards as far as bearable and even further in order to increase exports relative to imports and stimulate local investment. Such desperate measures are commonly referred to by critics of free trade as an inevitable and unfortunate ‘Race-to-the-Bottom.’ The corollary of this mournful dynamic asserts that when countries adopt or ratify one or more International Labor Organization (ILO) standards, labor costs will rise effectively truncating any hope at being competitive in a globalized economy. In contrast, countries that do not adopt international labor standards are presumed to gain an advantage in trade and investment at the expense of those who do. Critics of the Race-to-the-Bottom dynamic assert there is only one way out of this negative spiral, and that is world-wide labor standards: Many observers believe that the liberalization of international trade will lower the level of workers’ wage because only companies with the lowest prices and the lowest cost will survive and all ﬁrms that pay social beneﬁts or offer their workers better working

The Chinese Economy after WTO Accession

292

conditions or higher pay will simply disappear. Only one obstacle is said to stand in the way of this outcome – a move to raise labour standards in every country in the world.1

Race-to-the-Bottom critics argue that all countries should be compelled to adopt uniform labor standards such as the currently non-compulsory ILO standards by denying the lower tariffs established by the World Trade Organization (WTO) to any country that fails to adopt these labor standards. Such a program is designed particularly to protect the industrial countries’ workers, especially unskilled labor, from the shocks of globalized competition. The essence of this theory is trade protectionism. China, both populous and of growing economic importance, is a central target of censure from Race-to-the-Bottom critics of globalization, especially since the ongoing Chinese export boom began in the early 1990s. Ofﬁcial Chinese data show that exports from China increased dramatically from 62 billion (USD) in 1990 to 266 billion (USD) in 2001. Accompanied by a continuous favorable balance of trade for nine years, China is seen by most in the industrialized world as a powerful and dangerous competitor in the world economy. Especially in the ﬁeld of laborintensive industries such as footwear, leather, textiles and clothing industries, China with a huge and growing industrial labor force is seen as the dominant power in these industries. Recently, two criticisms of China have been widely raised in the industrialized world both attempting to limit the advantage of Chinese exports. On the monetary side, the renminbi (RMB) is said to be undervalued especially relative to the US dollar to which the RMB is pegged at a ﬁxed rate. Thus, according to this criticism, the Chinese should revalue their currency. People also cast their eyes on China’s labor costs, which are among the lowest in all export-oriented countries. Chinese wages and labor standards are seen as unfairly low and lax respectively, consequently crowding out potential smaller competitors in the Third World and destroying competitors in industrialized countries. However, over the last decade or so, China has taken several measures to raise local wages and labor standards. Since 1993, prevailing wage rates have been legislated in China to ensure a minimum income for employees. According to internationally accepted core labor standards proposed by the ILO, basic labor standards consist of (i) elimination of exploitative use of child labor; (ii) prohibition of forced labor; (iii) elimination of discrimination in employment; (iv) freedom of association; and (v) provision of the right to organize and bargain collectively. Currently China has ratiﬁed 22 fundamental ILO conventions. In 1994 China passed the benchmark China Labor Law which forms the backbone of Chinese labor market regulations. From an economics perspective, labor legislation is an exogenous shock to the economic system. Although possibly inspired by generous and altruistic motivations aimed at protecting workers, labor standard effects need to be tested and can fall well short of their goals. An example of falling short is laws aimed at improving worker safety. Labor standards may be viewed as ‘the acquisition and ﬁnancing of public 1

Cited from André Raynauld (1998, p. 22).

Low Wage and Low Labor Standards in China

293

goods that are linked to employers in some way because of where they are consumed or how they are ﬁnanced’ (Raynauld, 1998). In Brown’s paper (Brown, Deardorff and Stern, 1996) analyzing labor market failure to provide safe working conditions, a model built on the assumption of endogenous working conditions and an imposed safety labor standards is proposed, the authors conclude that ‘establishing minimum labor standards rarely corrects, or even partly corrects, labor-market failure associated with worker safety’. This example shows that the presence of legislated labor standards do not necessarily result in improved actual labor standards. To complicate matters further, in developing countries such as China, the development of labor standards as well as the growth of wages are interactive and endogenous to the development process. Labor standards are less exogenous shocks to the labor market and more outgrowths of the development of these labor markets. So the traditional view of labor market regulations devised in and applied to developed countries needs modiﬁcation to be applied to China’s case. Generally, labor markets in developing countries are basically different from mature labor markets in industrialized countries. Analysis of Chinese labor markets requires the use of theories from development economics applied to labor markets. However, China’s characteristics are not fully addressed in the original development theories and models that had in mind smaller countries. This calls for new theoretical elucidation and improvement. In recent years, many scholars researching the Chinese economy have focused on the different aspects of Chinese labor markets, especially the key issue of rural-to-urban immigration. Sicular examines the microeconomic determinants of rural employment and incomes in China, especially the phenomenon of multiple wages determining the rural labor supply (Sicular and Zhao, 2002). Wang, Wu and Cai focus their study on analyzing the reason for low unemployment rates of rural migrants in urban areas compared to urban residents in those areas (Wang, Wu and Cai, 2002). This chapter utilizes the results of this previous theoretical and empirical research to provide a more comprehensive view of the Chinese labor market. Rural markets, urban markets and the link between these two are studied. A plausible explanation of endogenous wage and labor standard determination based on analyzing Chinese labor markets from both the theoretical and practical aspects is proposed. The rest of the chapter is organized as follows. Following this introductory section, there is in the next section a general review of the present Chinese labor market. The third section then formulates a simple model used to analyze endogenous wage and labor standard determination in urban/rural segmented Chinese labor markets. The fourth section provides some data evidences and empirical applications. Finally the ﬁfth section provides a summary along with a discussion of policy implications and a forecast of the future development of Chinese labor markets.

294

The Chinese Economy after WTO Accession

A Brief Introduction to China’s Labor Market The study is built on the theoretical foundation of Lewis’ dualistic labor market theory and rural-urban migration mechanism on developing countries. The former views economic development as the progressively structural transformation of a traditional sector into a modern sector. Dualism is the coexistence of the traditional and modern sectors. The traditional sector is often equated to the agricultural sector; in contrast, the modern sector is the industrial sector. Ongoing capital accumulation in the modern sector provides the fuel for sustained transfers. Lewis argued that the traditional sector is characterized by surplus labor. In principle, this permits industrial development with unlimited supplies of labor, at least until the surpluslabor phase comes to an end. Harris and Todaro inherited Lewis’ two-sector analysis (Harris and Todaro, 1970). Their migration mechanism gives a reasonable explanation for rural-urban migration ﬂow in the transition stage of developing countries. Their basic model posits that rural-urban migration depends on expected rural-urban income differentials. With the existence of formal and informal sectors in cities, rural migrants enter the informal sector at ﬁrst and, while working in the sector, seek opportunities to enter the formal sector. Therefore, the expected income of migrating to urban areas is the weighed average of wages in the formal and informal sectors with their employment probabilities. Fields further introduced the job-searching behavior (Fields, 1975), revealing the impacts of urban rigid wages and relative probabilities of employment on the process of migration. He argues that a signiﬁcant underemployment in the urban informal sector would ensure a lower equilibrium unemployment rate than predicted by the Harris-Todaro model. Both the dualism theory and the immigration theory are the starting point of this study with some modiﬁcations. Before the 1980s, the registered permanent residence system (or Hukou) limited the formation of the Chinese labor market. People were limited to living and working within their Hukou, generally the place where they were born. It is a ‘land bondage’ system that creates great attachments to the land for many rural Chinese. In brief, it is impossible for rural residents to migrate to the city permanently and thus it factitiously divides the unique labor supply pool into two sides, the urban side and the rural side. By 2001, the whole economically active population in China was 744.32 million, with 239.4 million employment in the urban sector and 490.85 in the rural sector. In 1984, a reform allowed rural residents to temporarily immigrate to the urban region. These rural migrants could be hired in construction, manufacture, service industries, and so on, which have no preliminary requirement of urban Hukou. To a large degree, the complexity of the whole problem is ascribed to the ruralurban migration activated by economic reform. The ﬁrst step of China’s economic reform is to put household responsibility systems into practice, which is considered as a main motivation of rural-urban migration. Under the old work point system, in which individual effort was not rewarded, people had little incentive to work hard and so it brought the classical free-rider problem and inefﬁciency. Instead, the new

Low Wage and Low Labor Standards in China

295

system of household responsibility entirely solves the free-rider problem, dramatically increases the rural productivity, and releases agricultural surplus labor from the land. Leaving the assigned contract land to their parents, spouses, or relatives, those surplus laborers, usually youths, could work in the rural informal labor market, and then migrate temporarily to the urban area and seek employment in the urban informal labor market. The formation and maturity of the Chinese labor market depended on the gradual reform of the Hukou system. At the end of the 1980s, the framework of the segmented labor market could be observed clearly as shown in Figure 16.1. By then, the urban labor force pool and rural labor force pool had different accesses to the labor markets. The four segmented labor markets in the whole system are follows: urban non-competitive labor market (UN), urban competitive labor market (UC), rural informal labor market (RI) and rural agriculture labor market (RA). UN covers both state-owned enterprise (SOE) and urban collective-owned units, which are almost monopolized by urban residents.2 Being employed in the SOE, lifetime employees may have access to subsidized housing, subsidized medical care, and schooling for their children. Lacking such beneﬁts, the other sectors are generally inferior. Rural residents are excluded from this market although competition among urban residents does exist. The organizational form of the UC is diversity that includes cooperative units, joint ownership units, limited liability corporations, share holding corporations Ltd., private enterprises, private units with funds from Hong Kong, Macao and Taiwan, foreign funded units and self-employed individuals. The diversity of the market is also manifest in the different skill requirements of the laborers. It covers the highest paid careers in China, such as ﬁnancial agents, as well as the low paid part-time positions, such as mechanical operators or nursery maids. The UC may absorb laborers from both the urban and the rural side and thus can be looked at as an opened competitive market. The organizational form of the RI is comparatively simple and includes township and village enterprises, private enterprises and self-employed individuals. Although no barrier is set against urban residents to enter the RI market, only rarely are cases found. The RI is unattractive to urban residents due to its low wages and beneﬁts. Besides agriculture, the UC and the RI are the only two choices for rural residents although there does exist an entrance barrier to UC. After understanding the structure of the Chinese labor market and the signiﬁcant rural-urban migration, I now turn to investigate the reasons for the low labor costs of export industries. Firstly, it is necessary to understand that the labor market of export industries is a sub-market of the whole. The whole labor market decides the labor cost in export industries. Secondly, to get the signiﬁcant factors in deciding the low labor cost, a model for the segmented market based on the theory mentioned above should be built.

2 Beginning in the 1990s, SOE and collective-owned units began to hire small quantities of rural labor for the short term.

The Chinese Economy after WTO Accession

296

Figure 16.1 Structure of China’s Segmented Labor Market A Development of Immigration Model The model is built to explain that both wage level and labor standards in China are endogenous and are basically determined by the income level of the agriculture sector. To break the equilibrium by ﬁctitiously promoting the wage level and labor standards in the other sectors except the agriculture sector would trigger adjustment of the labor market structure. Wage Determination I start from the migration equation improved from the Fields’ model (Fields, 1975).3 His model is built on the dual sector structure of formal and informal sectors. To meet the reality of China, I develop it into a four-sector model that is composed by UN, UC, RI and RA. Instead of the unique choice between formal and informal sectors considered in Fields’ model, China’s rural residents are faced with three choices, to stay in the agricultural sector, to work in the rural informal sector, or to move out to the nearby town or city. To stay in the rural area or to move to the city is the key decision for laborers considering immigration. The individual migration equation is shown as formula (1): PUN WUN +(1–PUN)WUC –C=(1–PRA)WRI +PRA IRA

(1)

where PUN and PRA are the share of employment in the urban non-competitive and the agricultural sector respectively. Two points should be clariﬁed here. On the urban side, people are generally employed by only one employer in either the UN or the UC, so PUN would be equal to 0 or 1. Yet, laborers in the countryside may make the choice to work in both the informal and the agricultural sectors, that is, to divide their 3 To simplify, Fields’ model can be written as: PFWF+(1-PF) WI =WA where PF is share of employment in urban formal sector, WF is minimum wage in urban formal sector, WI and WA are wages in the urban informal sector and agricultural sector, respectively.

Low Wage and Low Labor Standards in China

297

time working on the land or for the RI, which means 0< PRA <1. In rural China, WUN, WUC , WRI are wages of the UN, UC and RI sectors respectively; IRA is the income from agriculture production that is comparable to wage; and C is the migration cost from a rural to urban area. The travel cost and the living expenditure gap between rural and urban are counted as the two main costs of immigration. While low quality dwellings are usually offered by employers freely, food is weighted highly in the whole living expenditure cost. The travel cost depends on both travel distance and frequency. Basically, temporary migrants go back home during Spring Festival and at the busy harvest season in the fall. Besides these travels, the shorter the distance, the more frequently migrants go home. In the present stage, migrants cannot break the barrier to enter the urban formal sector, i.e., PUN =0, equation (1) is turned to (2): WUC –C=(1–PRA)WRI +PRA IRA

(2)

From (2), we can see that the wage level of temporary migrants relies on the level of rural wage or income, both income from agriculture and the rural informal sector. Labor Standards Determination In the labor standard associated study, a different methodology is used. Labor standards are looked at as an integrative value (Brown, Deardorff and Stern, 1996) that is appended to labor’s marginal value product. Martin (Martin and Maskus, 2001) shows that low labor standards impair the competitive power of a country in exports, based on a detailed study of ﬁve core standards in the WTO studied separately using welfare analysis. In this chapter, the focus is on whether we can promote labor standards with external power. Other arguments about competitive advantage in international trade brought about by the low labor standards or effects of not adopting core labor standards are the next questions to be asked, and are beyond the discussion of this chapter. To simplify the study, improved labor standards are looked at as an increase in the welfare of laborers, which equals an increase in the wage. Improved labor standards triggered by government generally takes on the form of legislation or regulation. Labor standards are commonly promoted in all sectors except for self-employed industries such as agriculture. Assume the quantity of increased labor standards is S, ﬁrst only consider the countryside, a laborer tries to maximize his individual income, that is: (1–PRA)(WRI +S)+ PRAIRA

(3)

Assume that WRI= IRA, the hypothesis is reasonable, which is illuminated in the empirical study of the next section, thus (3) is changed to max (1-PRA)(IRA + S) + PRA I RA or in the form of IRA +(1-PRA )S. The maximum result is PRA= 0, i.e. no labor input in the agriculture industry.

298

The Chinese Economy after WTO Accession

Secondly, we turn to consider the urban side, that is, labor standards are improved in both the urban competitive sector and the rural informal sector. A new immigration equation (4) is derived from (2): WUC + S–C =(1–PRA)( I RA+ S)+ P RA I RA

(4)

In assuming that WRI =IRA, we also get WUC= IRA +C from (2). To make the right side equal the left side of equation (4), we should have PRA S=0, or PRA=0 when S is set to be positive. Both results of PRA =0 deducted from above equations conﬂict with the truth. The number of laborers working in the agricultural sector cannot decrease to zero, which means the equilibrium will be broken by increasing the labor standards. Theoretically, two solutions are available. The ﬁrst is that old equilibrium is broken while the improved labor standards in the other sector attract laborers from agriculture. The outﬂow of labor from the agricultural sector will induce a rise of the agricultural product price and thus increase the income of peasants, a new equilibrium is realized once IRA' = IRA +S, where IRA' is the new income of the agricultural sector. The other solution is to maintain the stable equilibrium that the wage of the rural informal sector decreases to a lower level as WRI = IRA–S; the wage of the urban competitive sector decreases to a lower level as WUC=IRA–S–C. The labor standards of the UC and RI sectors are determined by the wage level and labor standards of the agricultural sector. Ofﬁcial Data Evidence To test the immigration equation in the last section, the wages of migrants in different sectors and immigration costs are required. However, individual data and immigrant costs are not available in the general census. The available accumulative annual data of wages and number of laborers are from three resources: the Chinese Statistical Yearbook, the China Labor Statistical Yearbook and the Sannong.4 Although economists usually impugn ofﬁcial data for their accuracy, it is still the most accessible resource and can be used in the approximate estimation. When analyses of the framework of China’s labor market are taken, the market size and share are the ﬁrst items to be measured. As shown in Figure 16.A-1 in the appendix, with the natural increase of labor, the UN shrank in the past decade by policy adjustment and the UC expanded dramatically by absorbing laborers from both the urban and rural sectors. The RI experienced a continuous but slow increase. The RA, with the largest share of the whole labor market, experienced a slow decrease. The incomes of different sectors are shown in Figure 16.A-2. To be observed from the graph, the income of the RI is close to that of the agricultural sector, which means there is little to no entrance barrier between the agricultural sector and the 4 The branch of National Bureau of Statistics of China is in charge of the Sannong rural data survey.

Low Wage and Low Labor Standards in China

299

rural informal sector. Rural residents may transfer freely between those two sectors. As expected in the SOE and collective-owned units, employees of those two urbanmonopoly labor markets enjoy higher wages. A surprising phenomenon exists in that the UC has the highest wage of all the categories. One possible explanation is that the UC is a market mixture of corporations hiring both the highest paid employees and lowest paid laborers. To distinguish the wage of rural migrants from that of all UC, I dig into data of three low technology but labor-intensity industries. It is common knowledge that rural migrants are frequently employed in manufacturing, construction, and catering services in the urban sector. In Figures 16.A-3, 16.A-4 and 16.A-5, rural labor forces are weighted more than half in these three industries. Wages in these industries can represent the general wage level of rural migrants to a large degree. I further compare the income of these three industries with that of rural sectors. The result is shown in Figure 16.A-6. A large gap still exists between the two. Can we count it as the immigration cost? The average per capita annual living expenditure of urban households in 2001 is 5309.01 Yuan, with that of rural households being 1364.08 Yuan of the same year, which means the expenditure gap is 3944.93 Yuan. Even adding in an approximate estimate of 1000-Yuan for immigration costs, the move into the urban area is still proﬁtable. Thus the immigration cost does not contribute alone to the large gap. An immigration barrier must exist and does matter, such as the institutional regulation of ‘temporary stay credentials In Urban (Zhanzhuzheng)’. This rough analysis shows that there still exists the drive power and space for rural-urban migration and immigration will not stop until no proﬁt can be made. Once the institutional barriers are removed, the price of the labor force will keep dropping because of the extremely large labor force pool in the countryside. The ultimate purpose of Chinese economic reform is to transfer to a market economy. This result is a dynamic process changing from a multi-labor market to a single labor market. Although a small sample of annual data that generally covers 17 years from 1985 to 2001, means the econometrics test is of little stringency, an endeavor is made to test the model that the wage of the RI and the UC depend on the income of agricultural sector. WRI =u1+α1 IRA+β1 Cos(ω1 T+θ1)

(5)

WUC= u2+α2IRA+β2 Cos(ω2 T+θ2)

(6)

where T is a time unitary function that is equal to YEAR minus 1985, the initial year of the data. A cosine time trend function is added into the basic model that shows time affects wages in cycles. Regression results of models (5) and (6) are reported in Table 16.1. Dependent variables in (5) are statistically signiﬁcant at the 1 percent level, and 5 percent in (6). The sign and slope estimate of (5) means that, if the income of the RA increases by one percentage point, the wage in the RI increases around two-thirds of a percentage point; likewise, a remarkable increase of 8.485

The Chinese Economy after WTO Accession

300

Table 16.1

Wage Regression Results

Independent variable

Dependent variable = wage of Dependent variable = wage of UC RI sector between 1985 and 2001 sector between 1985 and 2001 Constant -112.716 -571.992 (11.331) (184.95) Wage of RA 0.722 8.485 (0.036) (0.597) Time Cosine Trend 21.684 268.149 (3.971) (63.229) Adj. R2 0.961 0.931 16 16 n: Notes: Standard errors are in parentheses. ω1=0. 48, ω2=0.69 and θ1= −2, θ2=1.14 in (5) and (6).

percent appears on the wage of UC sector as shown in (6), which implies a much faster rate of change. The slope estimate of the time cosine variable for the wage of the RI is quite small, which implies that wage of RI basically depends on the income of the RA with a small inﬂuence by time. Policy adjustments concerning the labor standards were frequent in the past twenty-ﬁve years. Two main concerns are the prevailing wage law ﬁrstly approved in 1993 and the benchmark labor law approved in 1995. The inﬂuence of 1993 and 1995’s laws would be a good test to the two deductions in the last section. Figure 16.A-6 indicates that no decrease in the wage after 1993 or 1995 can be observed as expected in the results of the last section. Another deduction is that laborers in the agricultural sector ﬂow to the RI or the UC after the shock of improving labor standards. The deduction might be observed in Figure 16.2, where the number of RA laborers is decreasing while the number of laborers in the RI and UC sectors is increasing. To test the effect of prevailing wage law on wages in different sectors, dummy variables ‘LAW’ are added into both model (5) and (6), which is equal to 1 if the year includes or is after 1995. The results show that LAW is not statistically signiﬁcant at the 5 percent level. Thus, the variable LAW is practically, as well as statistically, insigniﬁcant. Concluding Remarks The 744.32 million economically active people in China are foundation of China’s low labor costs, and form the comparative advantage in the international trade ﬁeld. When comparing the Chinese labor supply and the international labor supply, it is easy to conclude that the low labor costs of China have their roots in the enormous labor supply. Since mid-2001, China has expanded the geographic scope of reforms to its household registration system that formerly conﬁned to roughly 200 pilot towns

Low Wage and Low Labor Standards in China

301

and small cities. The reforms enable qualiﬁed rural migrants to register as urban residents. This is signiﬁcant in that it allows labor mobility and helps in forming a unique labor market. Accompanying this process, temporary rural migrants may change status so as to become permanent urban laborers, the immigration cost will disappear and the average wage is expected to decrease. China, as a developing country, experiences low wage levels and labor standards as a result of its natural economic status and historical inequities in the international economic order. To improve the endogenous labor standards, which are based on the rise of agricultural income, is a long-term process. References Benjamin, Dwayne and Loren Brandt (2002), ‘Property Rights, Labor Markets, And Efﬁciency In a Transition Economy: The Case of Rural China,’ Canadian Journal of Economics, November, Vol. 35, No. 4, pp. 689–716. Brown, Drusilla K., Alan V. Deardorff and Robert M. Stern (1996), ‘International Labor Standards and Trade: A Theoretical Analysis,’ in Jagdish N. Bhagwati and Robert E. Hudec (eds), ‘Fair Trade and Harmonization: Prereq sites for Free Trade?’ MIT Press, Vol. 1, pp. 227–80. Busse, Matthias (2003), ‘Do Transnational Corporations Care About Labor Standards?’ Journal of Developing Areas, Spring, Vol. 36, No. 2, pp. 39–57. Cai, Fang and Dewen Wang (2003), ‘Migration As Marketization: What Can We Learn From China’s 2000 Census Data?’ The China Review, Fall, Vol. 3, No. 2, pp. 73–93. Fields, Gary S. (1975), ‘Rural-Urban Migration, Urban Unemployment and Underemployment, and Job-Search Activity in LDCs,’ Journal of Development Economics, June, Vol. 2, No. 2, pp. 165–87 Flanagan, Robert J. and William B. Gould IV (2003), International Labor Standards: Globalization,Trade, and Public Policy, Stanford University Press, November. Harris, John R. and Michael P. Todaro (1970), ‘Migration, Unemployment and Development: A Two Sector Analysis,’ American Economic Review, March, Vol. 60, No. 1, pp. 126–42. Hatton, Timothy J. and Jeffrey G. Williamson (1991), ‘Integrated and Segmented Labor Markets: Thinking in Two Sectors,’ Journal of Economic History, June, Vol. 51, No. 2, pp. 413–25. Li, Peilin (1996), ‘Social Network of Rural-Urban Labor Migration in China’, Se Hui Xue Yan Jiu (Chinese), No. 4, p. 45. Martin, Will and Keith E. Maskus (2001), ‘Core Labor Standards and Competitiveness: Implications for Global Trade Policy,’ Review of International Economics, May, Vol. 9, No. 2, pp. 317–28. Mehmet, Ozay and Akbar Tavakoli (2003), ‘Does Foreign Direct Investment Cause A Race To The Bottom? Evidence From Four Asian Countries,’ Journal of the Asia Paciﬁc Economy, Vol. 8, No. 2, pp. 133–56.

302

The Chinese Economy after WTO Accession

Meng, Xin (2000), Labour Market Reform in China, Cambridge University Press, Cambridge, UK and New York. Organisation for Economic Co-operation and Development (2000), International Trade and Core Labour Standards, Organisation for Economic Co-operation and Development, Paris. Raynauld, André and Jean-Pierre Vidal (1998), Labour standards and International Competitiveness: a comparative analysis of developing and industrialized countries, Edward Elgar, Cheltenham, UK and Northampton, MA. Sicular, Terry and Yaohui Zhao (2002), ‘Earnings and Labor Mobility in Rural China: Implications for China’s WTO Entry,’ RBC Financial Group Economic Policy Research Institute (EPRI) Working Paper Series No. 2002-8, December. State Statistical Bureau (various years), China Statistical Yearbooks, People’s Republic of China, Beijing: Zhongguo Tongji Chuban She,. Wang, Dewen, Yaowu Wu and Fang Cai (2003), ‘Immigration, unemployment and Segmented Urban Labor Market: Why the unemployment rate of rural migrant is low?’ Institute of Population and Labor Economics, Chinese Academy of Social Sciences. Working Paper No. 26.

Low Wage and Low Labor Standards in China

303

Appendix

Figure 16.A-1

Scales of China’s Segmented Labor Markets (unit: million)

Figure 16.A-2

Wages of Different Sectors (unit: yuan)

304

The Chinese Economy after WTO Accession

Figure 16.A-3

Labor Market Scale in Industry (unit: million)

Figure 16.A-4

Labor Market Scale in Construction (unit: million)

Low Wage and Low Labor Standards in China

305

Figure 16.A-5

Labor Market Scale in Wholesale, Retail Trade and Catering Services (unit: million)

Figure 16.A-6

Comparison of Urban Low-end Labor Market and Rural Market (unit: yuan)

This page intentionally left blank

Chapter 17

Migration and Regional Development in China Shuming Bao, Anqing Shi and Jack W. Hou

Introduction Migration has played an important role in regional development in Chinese history. As early as the Yuan dynasty, the government made efforts to move the people from the east or central regions to develop the wild lands of the west.1 This strategy continued through the Cultural Revolution years of the PRC, and has only recently reversed itself from this historical trend. Due to climate, altitude, and other natural limitations, the Western region has always had limited arable land, despite its vast range. After nearly eight centuries of development, population has become a tremendous pressure straining the cultivated land. Since 1978, however, with the success of the CER (Comprehensive Economic Reform) and the relaxation of the tight state-controlled household registration system, more and more rural labors have been moving out from their farmlands to nearby cities or other provinces for new job opportunities. This is due to the unbalanced nature of the CER, where the development has concentrated (especially since 1985) in the urban areas and favored the Coastal or eastern region. Consequently, most provinces in the West have experienced a net loss of population due to outward migration, which has been accelerating since 1982. Between 1982–87, there was a net loss of 0.44 million, but for the same ﬁve year period between 1995–2000, the net loss swelled to 6.63 million.2 The objective of this chapter is to document and assess the changes in the migration pattern over the last three decades. In the past, a lack of data had hampered the documentation and assessment of Chinese migration. Questions on migration were 1 The three regions are deﬁned in accordance with the 7th Five Year Plan of the State Commission of Planning and Development. The east region includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Guangxi, and Hainan. The central region includes Shanxi, Inner Mongolia, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, and Hunan. The west region includes Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang. 2 The data is collected based on a ﬁve-year period. A person is counted as a migrant if they have moved within ﬁve years. If a person moved to the place more than ﬁve years ago, they are not counted as a migrant.

308

The Chinese Economy after WTO Accession

not included in the ﬁrst three population censuses, 1952, 1964, and 1982 (Goldstein and Goldstein, 1990, Liang and White, 1996). Since 1987, China has conducted several national surveys and censuses on population that included information on migration (Liang, 2001; Lovely, 2001). By combining these data, we can examine the changes in migration patterns since the initiation of the market reform in 1978. This chapter has signiﬁcant policy implications for the regional development of China. To relieve the poverty in the West and to improve overall development across the regions, past studies have proposed many different strategies on migration.3 Some suggest that special policies be adopted to attract laborers to the west to meet the demand for labor, especially the well educated, in the west (Hu, 2001, Zhen, 2003). Others suggest that institutional barriers (such as the traditional hukou system) should be abandoned or redesigned to encourage more migrations from the West to other developed regions to relieve the population pressure so that the relative per capita income can be improved in the West. Could a spatial equilibrium of population distribution be reached within a short period by self-motivated inter-region migration driven by the market force itself? Or should the government play an active role in such regional population migration even under the marketing economy? If so, what kind of policies should be adopted? Before such questions can be answered, it is imperative that we have a clearer understanding of the changing patterns of migration, the motivation of migration, and the future trend of migration. In the next section, we will ﬁrst give an overview of the institutional development of the hukou system and the CER since 1978. In the third section, we will characterize the migration ﬂows between urban cities, towns and rural areas, and across different regions. We will also compare the changes in the spatial patterns of migration from the three different regions over the last twenty years. In section four, we will apply Roberto Bachi’s Migration Preference Index of inter-provincial migration to examine the changes in the regional migration patterns. This will be followed by a detailed examination of the determinants of regional migration. We will conclude with a discussion of the migration policy for regional development in China. Background: Institutional Changes in the Hukou System and Market Reforms Between 1949 and 1978, the regional distribution of population was strictly controlled by the government’s hukou system (household registration system), a system that is 3 The vital contribution made by migrants to the rapid growth of China’s economy is well recognized (Ping and Pieke, 2003). On the one hand, migrant laborers have made a positive contribution to the rapid growth of the coastal areas by building much of China’s new urban infrastructure and by making the labor-intensive industries in the coastal areas internationally competitive. On the other hand, migrants have a stake in the development of their area of origin: migrants are more likely to transfer resources (remittances, investment, human capital and information) back to their home base, and returned migrants, in particular, are vital sources of investment, entrepreneurship and experience. Thus migration has indeed played a positive role in reducing some regional inequality.

Migration and Regional Development in China

309

designed not only to provide population statistics and to identify personal status, but also to directly regulate population redistribution (Chan, 1999).4 In order to control population migration, a household registration system was practiced in conjunction with the job assignment and rationing of living necessities in urban areas. Migration to other places needed to be approved by local governments. As a result, intra-provincial and inter-provincial migration was not common except for government “planned” migration from the eastern regions to the sparse western region during the Cultural Revolution period of the 1960s and 1970s. The hukou system has undergone many changes since the initiation of the CER in 1978. The ﬁrst one was the introduction of identity cards in the late 1980s, which enabled individuals, for the ﬁrst time since the late 1950s, to travel around the country without showing any ofﬁcial “permission” letter from their local authorities. The second one was the abolition of the grain rationing coupons in the early 1990s, allowing migrants access to food necessities. In the past, these coupons were issued to residents of a town or city and could only be used in that particular location, thus effectively tying people to the place of residence. The third was the reforms in the hukou system itself, in 2001. Under this umbrella, residency in small towns and townships was open to all rural workers who legally had a job and a place to live; medium-sized cities and some provincial capitals have abolished the cap on the number of rural laborers who can apply for permanent residence status; large metropolitans such as Shanghai and Beijing have also adopted more lenient policy towards migrants. These institutional changes in the hukou system have made migration much easier, and have consequently impacted the regional redistribution of population. The change in the hukou system is only a necessary condition for migration, the sufﬁcient condition is in the changes in the economic institutions. After the two food crisis (1968 and 1971), China abandoned the “self-reliant” food policy and began experimentations with agriculture production in the rural areas (Hou, Mead, and Nagahashi, 2004). These experiments led the Central Committee to formally declare the beginning of China’s Comprehensive Economic Reform in 1979. The reform was isolated to the rural agricultural sector and was not extended to the urban industrial sector until 1984 (Central Committee, 1984). Though China was interested in attracting FDI, it was not the centerpiece until the early 1990s (a major revision of the Joint Venture Law was done in 1990).5 To serve as a timeline, we will follow the convention of Fei and Hou (1994) and designate China’s reform into three

4 The hukou system was established in cities in 1951 and extended to the rural areas in 1955. It was formalized as a permanent system in 1958. For more elaboration on the Chinese hukou system and its role in blocking rural-urban migration and in social control, please see papers by Christiansen (1990), Chan (1994), Cheng and Selden (1994) and Mallee (1996). For a different interpretation of its role, please see recent paper by Chan and Zhang (2003). 5 For a more detailed discussion of the sequential expansion of China’s reform, refer to Fei and Hou, 1994; Hou, 2004a and 2004b).

310

The Chinese Economy after WTO Accession

stages: rural reform (1978–84), urban reform (1984–89), and the external reform (after 1990). There are two economic reform initiatives that are particularly relevant in explaining the rapid growth of migration. First was the institution of household responsibility system (de-collectivization of agriculture) in rural areas, which ushered in the economic reform in 1978. This decentralized the labor allocation decision to the household. The household responsibility system greatly enhanced agricultural productivity, and released large amounts of labor from the agricultural sector. In the mid-1980s, farmers and the rural governments embarked on a movement to develop Township and Village Enterprises (TVEs), which helped to absorb a sizable portion of the surplus rural laborers. The share of township and village enterprises (TVEs) in China’s total exports increased ﬁvefold between 1985 and 1990, and their share was estimated to be 25 percent by 1992. The bulk (90 percent) of the TVE exports was manufactured products, with arts and craft, textile, and garments as the three major categories. Most of these TVE export earnings (88 percent in 1990) were generated in Eastern China (World Bank report 1994). However, since there were still very limited job opportunities in urban cities during this early stage of the market reform, they were mainly attracting and absorbing migrants from nearby rural areas. However, as the market reform deepened, a large regional income disparity was created, which pushed migrants from the inland regions to seek cash opportunities in the well-developed Eastern/costal regions. This is the second reform initiative that was conducive to large-scale migrations. Encouraged by the success in the household responsibility system (which was speciﬁc to the rural areas), the government embarked on market-oriented reforms in the urban areas in the late 1980s. To attract foreign direct investment (FDI), many economic incentives were introduced (Hou, 2004c). As part of the overall strategy, favorable provisions (such as special tax concessions and liberalized land leasing) were made available to many coastal cities to establish economic development areas and high technology development zones. In an even more progressive move, the policy ideology shifted to “letting some get rich ﬁrst.” In the 1990s, a policy package offered special treatment for designated area, to the furthered development of the private sectors and foreign invested enterprises (joint ventures). The most famous of the designated areas are the special economic zones of Shenzhen, Xiamen, Zhuhai and Shantou. These became the gateway to the Chinese economy and, with their special provisions, succeeded in attracting large quantities of foreign direct investment. The success of these policies cumulated in the fact that China is now the second largest recipient of FDI (second to the US), and the total delivered amount reached US$45.46 billion in 1998. Despite the success of these reforms, as evidenced by 8 percent annual growth of real GDP, some not so unexpected side effects emerged: regional inequalities. Some of these inequalities have long existed and were further strengthened by these reform initiatives, while others are direct results of the reform. The inequalities are between the coastal versus the inland areas, and between urban and rural areas. In the western and central parts of the country, the average income for rural people is much lower than that of the national level. This poverty is a key “push” factor of outward

Migration and Regional Development in China

311

migration. It is, therefore, not surprising that most of the cross-province migrants were from these western and central areas and their destinations were mostly in the eastern coastal region. Compared to the earlier periods, the later migrations are increasingly of much longer distance. These massive “blind” migrations resulted in well-documented problems for the coastal urban cities. Guangdong province was the ﬁrst to announce a coordinated effort with eight neighboring provinces to manage and regulate the mobile population. By the end of 1994, the central government had to implement a series of policies and regulations to restrict the “blind migration.” The central government has, obviously, been aware of this problem. After 20 years pro-East/Coastal development (with the Central as a secondary priority and the West as lowest in the rank), the government has shifted the national strategy in the late 1990s to “xibu da kaifa” (go west) policy by increasing government investment in the West, which has attracted some private investment and offered new job opportunities in some western regions. This has attracted some migrants from other regions to the West, as well as some early migrants to return back from the coast to their home of origin in the West to take advantage of these new opportunities. As a result, some western provinces such as Xinjiang, Qinghai, Tibet, Yunnan and Ningxia have actually experienced a positive net migration inﬂow in recent years. This inﬂow was reinforced by the economic restructuring of state-owned enterprises (SOE) in urban cities, which led to many redundancies in urban cities and a rise in the unemployment rate, making it more difﬁcult for new migrants in the urban cities. Changes in Spatial Patterns of Migration (1982–2000) In this section, we attempt to paint the overall picture of the migration pattern in China based on the data from 2000 population Census and some previous population Census and surveys. To fully capture the nature, direction, magnitude, and composition of the migrating population, we will discuss the issue in several different but interrelated layers. Changes in the Migration Flows between Urban Cities, Towns and Rural Areas Based on the one percent population sample survey of 1987, the total migrating population is estimated to be 30.44 million between 1982–87 (Yan, 1998). The numbers increased only slightly between the periods 1985–90 and 1990–95. However, the number exploded for the period of 1995 to 2000 (Bao and Woo, 2004). The total migrating population rose to 144.39 million, or 11.62 percent of the whole population. This sharp rise in migration is the direct result of the aforementioned reform initiatives and changes in the hukou system. In the early 1980s, the long-standing government regulations made it difﬁcult for migration from the rural villages to urban cities. Consequently, small towns with

The Chinese Economy after WTO Accession

312

Table 17.1

To \ From Total City Town Rural

Migration Flows between City, Town and Rural Areas (1995–2000) Total 100.00 59.40 19.16 21.43

City 100.00 89.58 3.83 6.58

Town 100.00 35.11 32.80 32.09

Rural 100.00 29.89 39.30 30.81

Inter-county 100.00 65.01 15.48 19.52

Inter-province 100.00 54.50 20.02 25.48

Total 100.00 21.26 14.42 9.77 25.17 29.38 City 100.00 32.06 8.52 4.92 27.54 26.95 Town 100.00 4.25 24.68 20.05 20.33 30.69 Rural 100.00 6.53 21.59 14.05 22.91 34.92 Source: 1a-1c are from Shanping Yan (1998). The migration ﬂows in 1a–1c are based on the total migration, including both intra-province and inter-province migration 1d–1e are derived from the 9.5 percent 2000 Census sample data. Note: Migration numbers are in 10,000 persons.

fewer restrictions became the primary attraction to rural laborers released from their farmland in the wake of household responsibility. As the economic reform extended to the urban cities in mid/late 1980s, the overall picture changed. As the hukou system became more relaxed, job opportunities in cities began to attract the lion’s share of migrant workers. Based on the 1990 Census and 1995 one percent population sample survey (Yan, 1998), more than 60 percent of migrants moved to cities while the share attracted by towns dropped considerably. Though the magnitude of the migration increased dramatically in this latest period (1995–2000), the share attracted by cities has been stable around 60 percent (Table 17.1). A regional difference in migration ﬂows is presented in Table 17.2. As the megacities (the metropolis) of Beijing, Tianjin, and Shanghai have garnered national attention, it is natural to start with them. As can be seen in the table, these metropolises attract a disproportional amount of migrants from other cites, with town and rural areas only contributing less than 19 percent of the total inward migrants. This may not be surprising as these metropolis areas are hubs for national and international corporations, as well as centers of government. As such, the metropolises are much more service oriented than ordinary cities. Hence, the job opportunities they offer and the work force they demand are more educated and better skilled than the national average. Across regions, cities are also the primary attraction for the migrants, especially for the east and central regions of China. For the inland regions (West), the rural area and towns seem to have a stronger lure. This may reﬂect the aforementioned shift in national strategy of “xibu da kaifa” (go west) in the late 1990s. The rural areas in the northwest appear relatively more attractive (attracted 31.18 percent of total immigration in the northwest while 48.52 percent of migration were attracted by cities), especially for the inter-province migration (attracted 38.97 percent of total inter-province immigration). This pattern is also seen in the Plateau region (Tibet and

Migration and Regional Development in China

Table 17.2 Region National

Total

The Intra-province and Inter-province Migration Flows City Town Rural

Intra- City Town Rural Inter- City Town Rural Province Province 100.00 59.40 19.16 21.43 100.00 61.44 18.81 19.75 100.00 54.50 20.02 25.48

Metropolis 100.00 81.57 Coast

313

9.73

8.69 100.00 85.29

9.24

5.46 100.00 78.13 10.19 11.68

100.00 56.75 21.66 21.59 100.00 62.35 20.09 17.56 100.00 48.84 23.87 27.29

Northeast 100.00 67.23 11.60 21.17 100.00 67.91 11.58 20.51 100.00 62.79 11.73 25.48 Central

100.00 60.31 17.42 22.27 100.00 60.98 17.86 21.15 100.00 54.21 13.44 32.35

Northwest 100.00 48.52 20.30 31.18 100.00 49.42 21.99 28.59 100.00 45.82 15.21 38.97 Southwest 100.00 53.80 23.97 22.23 100.00 53.24 24.29 22.46 100.00 56.77 22.23 21.00 Plateau

100.00 53.04 21.56 25.39 100.00 51.89 19.32 28.79 100.00 55.54 26.39 18.07

Source: 2000 China population census data. Note: The Metropolis includes Beijing, Tianjin, and Shanghai. Coast region includes Hebei, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Hainan. Northeast includes Jilin, Heilongjiang, and Liaoning. Central includes Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan. Northwest includes Inner Mongolia, Shaanxi, Gansu, Ningxia, and Xinjiang. Southwest includes Chongqing, Sichuan, Guizhou, Yunnan, and Guangxi. Plateau includes Tibet and Qinghai.

Qinghai), though with a slight variation. The rural area still attracts more migrants than the towns, but the gap is smaller. Furthermore, the relative attractiveness of the rural area is more prevalent among intra-province migration (attracted about 28.79 percent of intra-province migration), while the town appeared more attractive to inter-province migration (attracted about 26.39 percent of total inter-province migration). For the southwest region, the town and rural areas seem to be of equal attraction. All in all, in contrast to the central and east regions, cities in the 12 western provinces appear to be relatively weak in attraction while the town and rural area attracted about 46 to 51 percent of migration. Changes in the Regional Pattern of the Inter-region Migration Flows We now turn our attention to the migration dynamics across regions. As expected, the East has consistently been the primary destination for inter-region migration and experienced continuous increasing of net gain of migration, while the West and Central China have experienced persistent net loss to migration. Tables 17.3 through 17.5 summarize these patterns and the changes that have occurred over time. In the early phases (1982–87) of the reform, the East absorbed 52.95 percent of the migrant workers (Table 17.3). This absorption continued to increase over time and reached 79.18 percent in 1995–2000. In contrast, the inter-region migration towards the Central and the West have experienced a shrinking share: from 27.32 percent and 19.72 percent in 1982–87 to 9.03 percent and 11.79 percent in 1995 and 2000 respectively.

Table 17.3

Changes in the Inter-province Immigration Flows: By Destination Region

Region To \ From Total Total 100.00 East (12) 52.95 Central (9) 27.32 West (10) 19.72

Table 17.4 Region To\ From Total East (12) Central (9) West (10)

1982–87 East Central West 100.00 100.00 100.00 53.35 61.94 39.64 33.72 25.13 22.03 12.93 12.93 38.34

Total 100.00 55.87 26.29 17.84

1985–90 East Central West 100.00 100.00 100.00 61.65 60.13 42.50 27.81 27.36 22.86 10.55 12.51 34.65

Total 100.00 66.17 17.29 16.54

1990–95 East Central West 100.00 100.00 100.00 69.74 72.26 52.94 19.74 16.72 15.35 10.52 11.03 31.71

Total 100.00 79.18 9.03 11.79

1995–2000 East Central West 100.00 100.00 100.00 77.93 86.21 67.70 13.23 7.83 7.55 8.83 5.96 24.75

Changes in the Inter-province Emigration Flows: By Originating Region

Total 100.00 100.00 100.00 100.00

1982–87 East Central 35.16 38.09 35.42 44.56 43.40 35.04 23.05 24.97

West 26.75 20.02 21.56 51.98

Total 100.00 100.00 100.00 100.01

1985–90 East Central 38.40 34.13 42.37 36.73 40.61 35.51 22.71 23.93

West 27.47 20.90 23.88 53.37

Total 100.00 100.00 100.00 100.00

1990–95 East Central 31.50 41.06 33.20 44.84 35.95 39.70 20.03 27.37

West 27.44 21.96 24.35 52.60

Total 100.00 100.00 100.00 100.00

1995–2000 East Central 23.58 48.99 23.21 53.34 34.58 42.47 17.66 24.76

West 27.43 23.45 22.95 57.58

Migration and Regional Development in China

315

In terms of where the migrants are coming from, there have been signiﬁcant changes over time also. The percentage of emigration from the East rose in the 1980s, but declined to 23.58 percent in 1995–2000, while the percentage of emigration from the Central region has increased from 38.09 percent in 1982–87 to 48.99 percent in 1995–2000 (Table 17.4). This indicates that as the economic reform deepened, the center of the mobile population has shifted to the Central region. Given that the destination is still the East, this translates into an increase in the distance of migration. The West exhibits a different but stable pattern. It consistently accounts for about 27 percent of the emigrants (though the absolute size of the migration has increased strongly over time). Furthermore, invariably, more than 50 percent of the emigrants from the West migrate to another province within the West. This may be due to the much longer distance of migration (to the east), the lack of networks, or the migrants’ inadequate work experience or skills. Throughout our sample period, the East has been the only region with a positive net migration. Both the Central and the West lost population consistently. Though we have already portrayed the dramatic increase in the migration ﬂows over time, it is still imposing when we look at the numbers. As shown in Table 17.5, the net inmigration to the East increased 21 times over our sample period; from 1.11 million in 1982–97 to 23.58 million in 1995–2000. In contrast, the net outward-migration from the Central region grew from 0.67 million to 16.95 million in the same time periods, and the loss for the West was from 0.44 million to 6.63 million. The relative size of the West versus the Central reﬂects two facts. First, the Central is more densely populated than the West, and (as we saw earlier) the migration in the West tends to be inter-province movement within the region. To have a better look at migration ﬂows both within and across different areas, we further classiﬁed all provinces into eight areas (a more detailed grouping than the traditional three regions of East, Central, and West): North-coast (Beijing, Tianjin,

Table 17.5

Net Migration by Regions (in 10,000 persons)

Region

(1982–87) (1985–90) (1990–95) (1995–00) East Central West East Central West East Central West East Central West East (12) 0.00 73.50 37.92 0.00 108.94 84.37 0.00 249.93 119.48 0.00 1659.21 699.29 Central (9) 0.00 6.05 0.00 22.26 0.00 -3.38 0.00 -36.00 West (10) 0.00 0.00 0.00 0.00 111.42 -67.45 -43.97 193.31 -86.68 -106.63 369.41 -253.31-116.102358.50 -1695.2 -663.30 Net migration

Note: The three regions were deﬁned by the State Commission of Planning and Development for the 7th Five-year Plan. The east includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Guangxi, and Hainan; the central, Shanxi, Inner Mongolia, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, and Hunan; the west, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang. Sources: 1980 and 2000 China population Census. 1987 and 1995 1 percent population sample survey.

The Chinese Economy after WTO Accession

316

Table 17.6

To \ From Northcoast East-coast Southcoast Northeast Central Northwest Southwest Plateau Note: Source:

Inter-province Net Migration Flows between Regions (1995–2000)

Total North-coastEast-coastSouth-coast Northeast Central NorthwestSouthwest Plateau 266.41 0.00 24.07 -5.27 54.13 130.62 19.06 43.16 0.63 601.99 1623.15

-24.07 0.00

0.00 0.00

-40.43 0.00 0.00 -1590.79 0.00 0.00 66.35 0.00 0.00 -938.49 0.00 0.00 11.81 0.00 0.00 Units in 10,000 persons 2000 China Population Census.

-15.91 0.00

-2.62 10.90

0.00 0.00 0.00 0.00 0.00

0.00 0.00 0.00 0.00 0.00

490.73 887.82

-7.01 30.53

161.49 672.34

-0.61 0.36

20.43 -4.24 0.00 -53.65 0.00 0.00 0.00 0.00 0.00 0.00

5.88 -5.10 52.89 0.00 0.00

-0.09 -2.44 -1.84 -7.83 0.00

Hebei, and Shandong), East-coast (Shanghai, Jiangsu, and Zhejiang), South-coast (Guangdong, Fujian and Hainan), Northeast (Jilin, Heilongjiang and Liaoning), Central (Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan), Northwest (Inner Mongolia, Shaanxi, Gansu, Qinghai, Ningxia, and Xinjiang), Southwest (Chongqing, Sichuan, Guizhou, Yunnan and Guangxi), and Plateau (Tibet and Qinghai). We use these classiﬁcations to present the migration data in the four panels in Table 17.6. As previous tables have pointed out, the coastal area (the East region) is the main destination for migrants. What this new classiﬁcation shows in more detail is the Coast/East region, it is really the South-coast area that is the primary destination. All three coastal areas have strong within-area migration, especially in the North-coast and the East-coast. For the migrants originating form the Central and Southwest areas, their destination is mainly the South-coast, with the East-coast a relatively distance second choice. The migration originating from the Northeast and Northwest shows strong within area movement. In terms of across area migration, migrants coming from the Northeast tend to favor the North-coast, while those from the Northwest almost equally favor the north- and south-coast area, with the Central as another attractive alternative. These are consistent with our economic logic underyling the cost of migration. The Plateau is different, which is not surprising either. The within area migration is low (second lowest, behind Central), and their primary destination is the Northwest area. This is due to both economic cost (of migration) and cultural ties. In terms of the origination area of the inward migrants, the Central area is the largest source for all three coastal areas. This is especially true for the Eastand South-coast area. Migrants from the southeast region are the primary source of immigration for the northwest and plateau regions. The above describes the migration pattern in percentages of both origination and destination areas. As we have already made clear, the Central and Southwest areas are the main sources of migration. Within this period, the Central area lost 15.91 million and the Southwest lost another 9.38 million. It is not surprising that the three coastal

Migration and Regional Development in China

Table 17.7 Region Total Beijing Tianjin Hebei Shanxi Inner Mongolia Liaoning Jilin Heilongjiang Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong Henan Hubei Hunan Guangdong Guangxi Hainan Chongqing Sichuan Guizhou Yunnan Tibet Shaanxi Gansu Qinghai Ningxia Xinjiang

317

Inter-province Migrants as Percentage of Total Migrants 1985–90 32.08 51.82 45.56 37.57 33.14 29.53 27.19 19.42 17.40 43.02 31.89 38.21 21.72 34.58 20.67 27.16 26.21 25.29 17.73 47.61 12.97 44.82

1990–95 31.58 47.50 38.03 34.90 31.65 35.34 26.70 20.93 20.93 54.24 30.09 31.60 15.97 32.56 17.45 23.52 20.13 23.28 13.17 44.99 13.25 40.61

16.68 26.96 24.72 51.22 23.51 25.51 27.35 41.53 66.97

11.17 33.31 24.96 62.15 29.95 26.11 29.82 41.94 72.68

1995–2000 29.38 53.11 33.69 19.06 17.94 14.31 16.12 10.46 10.26 58.22 27.88 42.90 6.47 36.29 7.52 13.84 9.16 10.69 7.94 59.53 13.24 39.03 15.36 8.04 16.91 30.08 50.83 18.01 14.64 23.81 28.53 49.87

Change + + + + + NA + -

Source: 2000 China Population Census.

areas have the largest positive net gains (especially the South-coast), but it is worth noting that what is generally regarded as the West (the Northwest area and the Plateau) has had a slight net gain. This might be partially reﬂective of the new “Go West” strategy. Another important change in the migration picture has been the increased importance of intra-province migration. Although the total migrant population has been increasing in the last decade, the relative importance of intra-province migration has been rising (from 68 percent in 1990 to 70 percent in 2000). As shown in Table 17.7, inter-province migration has accounted for a lower percentage of total migrants

318

The Chinese Economy after WTO Accession

Figure 17.1 The Percentage of Net Migration, by Province (1995–2000) for most provinces (with the exception of Beijing, Shanghai, Guangdong, Fujian, Zhejiang, and Yunnan.). Linking this (Table 17.7) back with Table 17.5, we see that the coastal areas of the East (Guangdong, Fujian, Hainan, Zhejiang, Shanghai, Jiangsu, Beijing, Hebei, and Tianjin) have been the primary reception of the inter-province migration. Between 1982 and 2000, the coastal East has accounted for 73.27 percent (or 31.08 millions) of the total inter-province immigration. The Western region absorbed 14.09 percent (or 5.97 millions) of inter-province migrations, while the Central region only took in 12.64 percent inter-province migration (or 5.36 million). This last number is signiﬁcant as the Central region (made up of ten provinces: Liaoning, Jilin, Heilongjiang, Shanxi, Anhui, Jiangxi, Shandong, Henan, Hubei, and Hunan) account for 43.52 percent (or 540.72 million) of national population. Figure 17.1 shows a regional distribution of net-migration by provinces. For the most part, this conﬁrms our analysis earlier that the three coastal areas (the East region) have net gains, while the central and southwest regions have net losses of migration. In terms of the three costal areas, the south-coast region absorbed about 41.47 percent of total migration, the east-coast absorbed 22.07 percent, and the north-coast absorbed 12.17 percent. Combined, the coastal regions have a net gain of 24.9 million, or 5.63 percent of the total population of the region. The central region has a net loss of 15.9 million migration, which is about 4.6 percent of the region’s population. The West region is less homogenous. While the southwest has a net loss, the northwest has experienced a net gain of migration. The most noticeable gainer is Xinjiang province, which has become a new center of attraction for migration with a net migration of about 1.25 million from 1995 to 2000. Regional Differences in the Composition of the Migrating Population Up to now, our analysis has focused on painting a spatial picture of the migration issue in China. We will now turn our attention to the characteristics and composition of the mobile population, with the objective of understanding the underlying socioeconomic reason for migration that is not visible in the above numerical data.

Migration and Regional Development in China

319

Figure 17.2 The Percentage of Profession Migration by Regions Within the last ten years, the fast-growth experience has blanketed every corner of China. However, this growth is not uniform across regions (Bao et al., 2002). Early FDI were almost exclusively concentrated in the coastal area for both economical and policy reasons (Hou 2004c). Even though FDI has penetrated deeper inland, very little FDI has ﬂowed to the West region. Accordingly, the rapid economic growth in the East coast regions have generated disproportionately more job opportunities than the west. In addition, these jobs are better paying, and in the preferred sectors: in manufacturing or service sectors, rather than in agriculture. According to 2000 population Census, migrants working in the production and service sector (1995– 2000) account for 66 percent of the total migrant population, while professional or administrative occupations account for less than 15 percent. Table 17.8 shows a noticeable difference across areas. The migration ﬂows to the East or coastal areas were mostly in the production and service sectors, while the migration ﬂows to the West region were mostly attracted by agriculture and service sectors. Figure 17.2 demonstrates a spatial tendency of migration in production from east (high) to west (low) and a vice versa direction for migration in agriculture related ﬁelds. This regional difference is partially due to the geographic characteristics, but is also undeniably due to the unbalanced economic development. Table 17.8 also shows us the gender breakdown of the migrant population. Nationwide, males account for almost 51 percent of all migrants, with the highest male ratio in the metropolises (Beijing, Shanghai and Tianjin) of 58.66 percent, followed by 54.24 percent in the Northwest area. Statistics also reveal that there is a relatively large percentage of single male migration in those regions, which may bring some unstable factors and potential social problems in those regions, as the gender imbalance (which is a well documented result of the one-child policy) is further ampliﬁed.

Table 17.8 National

Total Metropolis Coast Central Northeast Northwest Southwest Plateau

Distribution of Migrants by Occupation and Region Total

Male

Female

100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

50.89 58.66 50.11 53.38 48.22 54.24 48.90 52.22

49.11 41.34 49.89 46.62 51.78 45.76 51.10 47.78

Administrator Professional & Manager & Technical 3.16 4.68 2.47 5.06 3.66 3.59 2.89 3.80

11.76 15.80 8.99 15.60 14.35 13.60 12.92 16.22

Clerical & Related Workers 7.16 11.04 6.17 8.32 7.81 7.46 6.77 7.55

Sales & Service Workers 22.29 28.64 19.45 23.11 24.50 22.77 25.10 26.89

Agriculture & Related Workers 15.93 4.04 9.59 18.54 24.93 24.98 28.24 25.61

Production & Related Workers 39.58 35.78 53.26 29.23 24.51 27.45 23.88 19.60

Not Stated

0.12 0.02 0.06 0.15 0.24 0.15 0.21 0.32

Migration and Regional Development in China

321

Compared to the 1990 population Census, there has been an increase in the percentage of female migration from 43.31 percent in 1985–90 to 52.29 percent in 1995 to 2000 (Tables 17.9a and 17.9b). The motivation or “cause” of the migration decision differs greatly between the genders, and has also changed signiﬁcantly over time. Categorically, in the earlier period (1985–2000), men’s migration was mainly motivated by work/business or job related, while females tend to be because of marriage or family related. By 1995–2000, the motivation for migration had converged considerably between the genders. Using “work or business” motivation as an example, female migration for this reason was less than 30 percent in 1985–90, but rose to more than 43 percent in 1995–2000. This change may be due to the increase in the education level of the female population. In addition, the deepening of the reform most likely played a signiﬁcant role, as it led to the growth of the Table 17.9a Migration (both inter- & intra-province) Motivation by Gender (1985–90) Motivation Total Work and business Job transferring Job assignment Study/training Retirement Marriage Moving with family Live with relatives or friends Other

Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Male 56.69 70.70 75.18 75.05 67.90 82.87 8.69 39.46 44.98 67.40

Female 43.31 29.30 24.82 24.95 32.10 17.13 91.31 60.54 55.02 32.60

Source: Compiled from Zhaoliang Hu (1994).

Table 17.9b Migration (both inter- & intra-province) Motivation by Gender (1995–2000) Motivation Total Working and Business Job transferring Job assignment Study/training Moving housing Marriage Moving with family Live with relatives or friends Other

Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00

Source: 2000 China population Census (L7-5).

Male 47.71 56.96 67.74 58.60 52.66 52.19 11.11 39.63 45.11 52.95

Female 52.29 43.04 32.26 41.40 47.34 47.81 88.89 60.37 54.89 47.05

322

The Chinese Economy after WTO Accession

service sector, which may be disproportionately attractive to female workers (Fuess and Hou, 2004). Further analysis is obviously warranted, and will be the focus of the remainder of the chapter. The Regional Preference of Migration The statistics that we have presented clearly show the direction of immigration and emigration. We now attempt to develop an empirical model to explain the motion of migration and the major determinants. We ﬁrst need to create an objective measure to gauge the attractiveness of regions for immigration, and on the ﬂip side, the pressure for emigration. To this end, we computed the immigration and emigration preference indices (Shryock, 1976). For an outline on the technical aspect of the computation, please refer to the Appendix. By construction, the “average” of the IPI and EPI is set at 100. Hence, generally speaking, the higher the immigration preference index (IPI), the stronger the attraction of the region to immigration. In contrast, the higher the emigration preference index (EPI), the greater the pressure on emigration in the region or the higher the population mobility in the region. Table 17.10 presents the migration indices (both IPI and EPI) for all the provinces, and sorted by the three ofﬁcial regions, covering the entire sample period of 1982– 2000. This will allow us to identify the areas (both by region and by province) that are attractive to migration. Furthermore, we can gauge the relative intensity of the attraction, and track the changes in this attraction. For example, the attraction of the metropolis is well established. The high IPI of Beijing and Shanghai has persisted (and grown) over time, yet the third metropolis—Tianjin—was not an attractive destination until the 1990s. Given that we are more concerned with the recent developments, we will conﬁne our discussion to the most recent period. Between 1995 and 2000, provinces with an IPI that is above the threshold of 100 include Beijing (532), Tianjin (219), Shanghai (560), Jiangsu (102), Zhejiang (235), Fujian (184), Guangdong (518), Hainan (148), Tibet (122), Ningxia (102), and Xinjiang (224). If we look at these provinces in a spatial cluster perspective, we can categorically divide them into Beijing-Tianjin, Shanghai-Jiangsu-Zhejiang, Guangdong-Fujian-Hainan, and Xinjiang-Tibet. With the exception of the last cluster (Xinjiang-Tibet), all clusters are in the East coast area. In terms of the West region, with the exception of Xinjiang, Tibet and Ningxia, all other provinces show a below average preference level (IPI). Those provinces with an emigration preference index higher than 100 include Anhui (215), Jiangxi (267), Hubei (138), Hunan (199), Guangxi (163), Sichuan (237), and Guizhou (133). These form two major clusters that are the primary supply of migrants: the lower central region (Anhui-JiangxiHubei-Hunan) and the southwest regions (Sichuan-Guizhou-Guangxi). Both the IPI and EPI analysis and the clustering that results from them conﬁrm the observed statistical data presented earlier. We next turn our attention to the gradient or changes of these indices over the past decade. With this, we hope to identify possible trends of migration ﬂows. The

Migration and Regional Development in China

323

Table 17.10 The Migration Preference Index by Province Region

Province National

East

Population 1982–87 (in 10,000) 124261 IPI EPI

1990–95 1995–2000 Changes (90–2000)

IPI

IPI

IPI

EPI

ΔIPI

ΔEPI

306 227 647 127 681 115 532

20

-149

-95

75 219

25

-51

-50

87

72

41

54

-46

-18

76 117

53

73

25

-44

-28

99 509 101 573

97 560

25

-13

-72

91 120

71 102

69

-51

-2

87

81 152 119 132 235

95

116

-37

69

71

87

82 122

78 184

70

62

-8

91

82

74

65

48

34

36

-32

-12

37 518

15

189

-22

237 167 168 164 148

46

-20

-118

Beijing

1357

Tianjin

985

73

37 285

Hebei

6668

95

82

Liaoning

4182

118 111 139

Shanghai

1641

179

Jiangsu

7304

99

Zhejiang

4593

77

Fujian

3410

Shandong

8997

Guangdong

8523

Hainan

756

Central Shanxi

West

1985–90 EPI

84 270

88 109

140 132 208

94 153

66

41 329

3247

123 125 110

52

60

28

2

-24

Jilin

2680

159 170

97 145

64 127

34

67

-30

-60

Heilongjiang

3624

107 133 104 173

68 185

31

95

-37

-90

Anhui

5900

68

73

62

97

29 141

11 215

-18

74

Jiangxi

4040

63

67

61

80

35 144

18 267

-17

123

Henan

9124

51

54

58

71

34

92

15

99

-19

7

Hubei

5951

130 127

82

66

53

74

30 138

-23

64

Hunan

6327

46

89

37 123

Inner Mongolia

2332

Chongqing

3051

Sichuan

8235

Guizhou

3525

77

Yunnan

4236

75

Guangxi

4385

Tibet

262

88

97

78

58

EPI

126 132 120 144 136 123 -

-

-

-

-

-21

76

63

-67

-60

39

97

-

-

44 123

39 144

19 247

-20

103

78

60

99

51 132

34 133

83

69

77

60

81

66

80

35 144

11

72

121 124

-

16 199 69

69

30 139

-17

1

24

21

-45

29 163

-1

24 -113

22

-50

Shaanxi

3537

115 122

98 113

258 172 135 122 53

85

35

67

-18

-18

Gansu

2512

84 100

92 129

67 119

27

68

-40

-51

76 137 237 236 122 181

Qinghai

482

76

58

-46

-123

Ningxia

549

153 119 207 128 111 125 102

48

-9

-77

Xinjiang

1846

140 150 233 190 396 105 224

25

-172

-80

last two columns show the change in the IP and EPI between the two periods of 1990–95 and 1995–2000. A positive entry in the ΔIPI column implies an increase in the intensity of attraction of that province. As can be seen, the provinces with positive changes in the immigration preference index include Zhejiang, Fujian, Guangdong, and Yunnan, suggesting that they have become increasing attractive

324

The Chinese Economy after WTO Accession

to migrants over the past decade. In contrast, the provinces with positive changes in the emigration preference index (ΔEPI) include Anhui, Jiangxi, Hubei, Guangxi, and Sichuan, suggesting that the pressure for outward migration (i.e. emigration) has increased over time. Combing this analysis of the migration preference index with our statistical analysis, we can identify the prime national centers of attracting migrants. Figure 17.1 shows the percentage of net migration to regional population by provinces with four national centers of attraction for immigration. In the next section, we will use these economic centers as the foundation of one of the main explanatory variables in our empirical model aimed at identifying the major determinants of both immigration and emigration from those provinces in the west region. Determinants of Inter-regional Migration: A Revised Narayana Model On the aggregate level, the size and direction of the migration ﬂow may be affected by many factors such as economic growth, population growth and distribution, education, population policies, as well as many other local attributes and neighborhood effects. To identify the impacts of those factors on inter-regional migration, we use OLS regression on a model derived from Narayana’s Multi-nominal Logistic Regression (see Appendix) on data of the 12 provinces in the Western region. The reason for our focus on the West is due to a multitude of facts. First, the migration ﬂow to the East is well documented, well researched, and well understood. While, in contrast, the migration issue in the West has not drawn much attention. Furthermore, in response to the accumulation of the blind migration to the East, the overburdening of the urban infrastructure, combined with the desire for a more balanced growth, the national strategy since the late 1990s has been “Go West.” The effect of this shift in national strategy has been the net gain of migration in the Northwest and Plateau areas of the West (Table 17.6). Before we proceed with the empirical model, it is worthwhile to once again look at the magnitude of the migration, the decomposition between intra- and inter-regional movements, and the size of the regional populations. This is shown in Table 17.11. The dependent variables are emigration and immigration respectively. The explanatory variables include: 1. Population pressure in the local province (Pop96): measured by the provincial population at the end of 1996 (in 10,000 persons); 2. Job market quality in the local province (NonAgriEmp): measured by the percentage of non-agriculture employment to total population in the province in 2000; 3. Minority culture inﬂuence in the local province (Minority): measured by the percentage of minority to total population in the province in 2000; 4. Education attainment of the population (Illiterate): measured by the percentage

Table 17.11 The 2000 Population and Migration by Region No.

Region

1 Fujian, Guangdong, Hainan 2 Shanghai, Jiangsu, Zhejiang 3 Beijing, Tianjin, Hebei East Coast Region (9)

Population By Region (person) 126881989 135381962 90102344 352366295

Population By Region (%) 10.21 10.89 7.25 28.36

Intra-province Migration (person) 32193689 23083100 11700866 66977655

Intra-province Migration (%) 22.30 15.99 8.10 46.39

Inter-province Migration (person) 17591886 9360662 4128705 31081253

Inter-province Migration (%) 41.47 22.07 9.73 73.27

4

Inner Mongolia, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang 5 Guangxi, Chongqing, Sichuan, Guizhou,Yunnan, Tibet Western Region (12)

112581568

9.06

11774270

8.15

2929124

6.91

236939710

19.07

19026146

13.18

3049183

7.19

349521278

28.13

30800416

21.33

5978307

14.09

6 7

104864179 435860474

8.44 35.08

13199973 33412704

9.14 23.14

1740411 3618591

4.10 8.53

540724653 1242612226

43.52 100.00

46612677 144390748

32.28 100.00

5359002 42418562

12.63 100.00

Liaoning, Jilin, Heilongjiang Shanxi, Anhui, Jiangxi, Shandong, Henai, Hubei, Hunan Central Region (10) National

326

The Chinese Economy after WTO Accession

of illiterate population to total population in the province in 2000; 5. Local economic development (PcGDP00): measured by the average per capita GDP (in current Yuan) in the province in 2000 (=2*GDP/(1999 population+2000 population); 6. Intensity of inter-regional migration (DWPcGDP): this variable attempts to measure the level of attractiveness of the national centers in the east region (refer to the previous section) on the outward migration (emigration) of the local province. This is measured by the distance-weighted per capita GDP of Beijing, Shanghai Guangdong and Xinjiang in 2000. More speciﬁcally: DWPcGDPi = ∑ wijPcGDPj , wij=1/dij, j=1,…,4 j

where wij is the is the “weight” and dij is the distance between province I to the national center j. The modiﬁed Narayana models for immigration (min) and emigration (mout) are deﬁned as: Log(min)= a + a1 Pop96 + a2 NonAgriEmp + a3 Minority + a4 Illiterate + a5PcGDP00 + a6DWPcGDP + μ Log(mout)= b + b1 Pop96 + b2 NonAgriEmp + b3 Minority + b4 Illiterate + b5PcGDP00 + b6DWPcGDP + μ For the emigration (out) model, population is expected to be positively correlated with the scale of the outward migration as the local population size is indicative of the competitive pressure for jobs. The effect on the immigration (in) model is less deterministic. In addition to the competition effect, it may also signify the market size of the economy as many inward migrants could be seeking a place to start a business. The non-agriculture employment variable measures the economic development of the local economy. We expect this to have a negative effect on emigration and a positive effect on immigration. The effect of the minority culture on emigration is hard to gauge, but we postulate a positive effect on immigration. The reason being that government has speciﬁc subsidies for minority groups, and this is furthered by the fact that many of the minority groups lack many basic service industries, which in turn may provide a unique opportunity for immigrants with service business experience. The local per capita GDP is expected to have a negative effect of emigration and a positive effect on immigration. The attraction of the national economic centers in the East/coastal regions will deﬁnitely have a negative effect on immigration. It should have a positive effect on emigration, but we suspect that it may be insigniﬁcant due to the long distance (cost of migration) and the lack of job skills (lower education level). The OLS estimates of these two modiﬁed Narayana MLR models are presented in Table 17.12. Panel A is the OLS results of the log linear regression. As none

Migration and Regional Development in China

327

Table 17.12 OLS Estimates of the Modiﬁed Narayana Model 12a Estimates of the Log Linear Models of Immigration and Emigration Dependent Variable Independent Variables C Pop96 Non-AgriEmp(%) Minority (%) Illiterate (%) PcGDP00 DWPcGDP R-Squared

Log (Emigration %)

Log (Immigration %)

0.6260 (0.2396) 0.0002 (1.0105) -0.0007 (-0.0048) 0.0006 (0.0453) -0.0113 (-0.3193) -0.0002 (-0.3866) 0.0014 (0.5870) 0.6604

-0.1569 (-0.1232) -0.0001 (-0.9041) 0.0971 (1.4268) 0.0119 (1.8213) -0.0015 (-0.0879) 0.0000 (-0.0701) -0.0015 (-1.2595) 0.9115

12b Estimates of the Linear Models of Immigration and Emigration Dependent Variable Independent Variables C Pop96 Non-AgriEmp(%) Minority (%) Illiterate (%) PcGDP00 DWPcGDP R-Squared

Emigration (%)

Immigration (%)

-0.4580 (-0.0671) 0.0010 (2.2109) 0.1055 (0.2892) 0.0166 (0.4746) 0.0057 (0.0613) -0.0005 (-0.4361) 0.0017 (0.2716) 0.7375

1.7893 (0.6369) 0.0000 (-0.2498) 0.2855 (1.9007) 0.0329 (2.2806) -0.0304 (-0.7987) -0.0001 (-0.2012) -0.0073 (-2.7606) 0.9412

of the coefﬁcients are signiﬁcant, we turn our attention to the results of the linear regression model for immigration and emigration, which is presented in panel B. For the emigration model, the only variable that is statistically signiﬁcant is the population variable (Pop96). The sign of the estimated coefﬁcient is positive indicating a positive relationship between the size of local population and the scale of inter-province emigration. The southwest region has been the primary agriculture base for grain production in China. There is a high population density in those provinces, especially Sichuan and Chongqing. Given that none of the other variables are signiﬁcant, we conclude that the primary cause of outward migration or emigration is the over-crowding of the local province. For the immigration model, the ﬁt is much better. The local population variable (Pop96) is not signiﬁcant, which is not unexpected as the local population measures the competitiveness (difﬁculty) in seeking jobs, but also is indicative of opportunities. The non-agriculture employment (Non-AgriEmp) is positive and signiﬁcant, which indicates that a local job market with more opportunities is important in attracting immigration from other provinces. The minority variable (Minority) is positive and signiﬁcant, which suggests that those provinces with higher minority population are more attractive to immigration from other provinces. This is as we expected since

328

The Chinese Economy after WTO Accession

minority communities are usually highly agriculturally oriented and are lacking in many commercial and service economies. Most immigrants to the West regions are attracted by the opportunities in the service sector or agriculture sector, and these employment opportunities are exactly what the Minority variable captures. In addition, many of these minority cultures are either tourist attractions themselves due to their cultural heritage, or they are located in remote areas where the natural scenery is also a great attraction (Jones, 2004). The illiterate variable (Illiterate) and the local per capita GDP (PcGDP00) are both negative and insigniﬁcant. The attraction imposed by the coastal economic centers (DWPcGDP) is negative and signiﬁcant, which indicates that they pose a strong alternative to the Westward migration. Concluding Remarks: Policy Implications for Regional Development The primary goal for regional development of China is to improve the standard of living for the people, and boost the local economy while sustaining the environment. Given the many differences between the West, Central and East, a policy that encourages bi-directional migration may be needed. On the one hand, the policy should promote emigration (labor force with low skills) from the traditional agriculture regions (such as the Southwest) to reduce the population pressure; on the other hand, it should induce immigration (especially those with higher education and better skills) from other regions, especially the Northwest area. Both the central and local governments need to play an active role. The emigration to other regions, especially to the East region, can help relieve the population pressure in the west and help sustain the unique environment. People in the West and some remotely located areas still face many more migration barriers than the rest of the country. Cultural differences make it harder for migrants from those regions to adapt to new areas and to assimilate into the community. The lower education level makes it harder for migrants from the less developed areas to obtain jobs in the competitive urban cities of the East. The long distance between the remote regions increases the cost of migration and makes it harder to maintain family connections for those who would like to ﬁnd job opportunities in the coastal economic centers. To ease these difﬁculties and to encourage rational migration, the government needs to design policies that would create a friendlier environment for migration from the West and some other remote areas. These policies should be aimed at building migrants’ conﬁdence and assisting them to adapt/assimilate into the new environment. Active steps also need to be taken to improve the education level in those regions, which has been a hindrance to the mobility of the population in the regions. These barriers towards emigration have prevented China from developing a new spatial equilibrium in terms of population distribution. Our study also indicates that new job opportunity in the coast regions has not been able to induce emigration from the West. There are many reasons for that, and most of them have been addressed above. As China enters WTO, those traditionally agriculture-oriented areas will face

Migration and Regional Development in China

329

much more pressure on their product and labor market. Proactive policies should be introduced in the East regions to help relieve the population pressure from those regions. For example, preferential treatment may be given to investments that give priority towards employment of migrants from the West. The government should continue to herald its “Go West” strategy. More speciﬁcally, it needs to help the West attract skilled workers, especially in professional ﬁelds, service sectors and specialized agriculture. The actions could include direct subsidization of the provincial/local governments of the West, or grants and loans to targeted migrants from the Center or East to encourage their westward move. These skilled migrants will be vital in improving the software side of the social structure, and can only help the overall economic development of the West. It is not our claim that our policies are the solution, nor that the complex issue can be described by our simpliﬁed discussion. Our intention is to draw attention to the issue, as we believe it is important to the regional development of China, but is neglected by most scholars. Much like the old Chinese adage, we are throwing out the brick in the hope of drawing out the jade. If successful, we hope to generate active and open discussion on government policies on migration and investment, especially towards the West China region. References Bao, Shuming and Wing Thye Woo (2004), ‘Migration Scenarios and Western China Development: the Evidence from 2000 Population Census Data,’ in Ding Lu and William A.W. Neilson (eds), China’s West Region Development: Domestic Strategies and Global Implications (World Scientiﬁc), pp. 345–372. Bao, Shuming, Gene Chang, Jeffrey D. Sachs, and Wing Thye Woo (2002), ‘Geographic Factors and China’s Regional Development Under Market Reforms, 1978–98,’ China Economic Review, Vol. 13, pp. 89–111. Central Committee (1979), ‘Regulations on the Work in Rural People’s Communes,’ reprinted (in English) in two parts in Issues and Studies, August (pp. 100–12) and September (pp. 104–15). Central Committee (1984), ‘Decisions of the Central Committee of the CCP on Reform of the Economic structure,’ reprinted (in English) in the Beijing Review, October 29. Chan, Kam Wing (1994), Cities within Invisible Wall: Reinterpreting Urbanization in Post-1949 China, Oxford University Press, Hong Kong. Chan, Kam Wing (1999), ‘The Hukou System and Rural-urban Migration in China: Processes and Changes,’ China Quarterly, pp. 819–55. Chan, Kam Wing and Li Zhang (2003), ‘The Hukou System and Rural-urban Migration in China: Processes and Changes,’ Mimeo, Department of Geography, University of Washington, Seattle. Cheng, Tiejun and Mark Selden (1994), ‘The Origins and Social Consequences of China’s Hukou System,’ China Quarterly, No. 139, pp. 644–68.

330

The Chinese Economy after WTO Accession

Christiansen, F. (1990), ‘Social Division and Peasant Mobility in Mainland China: The Implications of the Hu-K’ou System,’ Issues and Studies, Vol. 26, No. 4. Fei, John C.H. and Jack W. Hou (1994), ‘The Comprehensive Economic Reform of The PRC (1978-),’ in Shao-chuan Leng (ed.), Reform and Development in Deng’s China, Volume IV, The Miller Center Series on Asian Political Leadership, University Press of America, New York, pp. 19–64. Fuess, Scott M., Jr. and Jack W. Hou (2004), ‘Rapid Economic Development and Job Segregation: The Case of Taiwan,’ working paper. Goldstein, Sidney and Alice Goldstein (1990), ‘China,’ in Charles B. Nam, William Serow, and David F. Sly (eds.), International Handbook on Internal Migration, Greenwood Press, New York, pp. 63–84. Hou, Jack W. (2004a), ‘Rural Reform and the Welfare Impact on Urban Workers: An Analytical Approach,’ in Aimin Chen, Gordon G. Liu, and Kevin H. Zhang (eds), Urbanization and Social Welfare in China, Ashgate, Aldershot, pp. 107–37. Hou, Jack W. (2004b), ‘Taiwan Direct Investment in China: Economic Development and Policy Effects,’ working paper. Hou, Jack W. (2004c), ‘Taiwan Direct Investment in China: Economic Development and Policy Effects,’ working paper. Hou, Jack W. and Kevin H. Zhang (2000), ‘Taiwan’s Outward Investment in Mainland China,’ in Hung-Gay Fung and Kevin H. Zhang (eds.), Financial markets and Foreign Investment in Greater China, M.E. Sharpe, New York, pp. 182–203. Hou, Jack W. and Kevin H. Zhang (2001), ‘A Location Analysis of Taiwanese Manufacturing Branch-Plants in mainland China,’ International Journal of Business, Vol. 6, No. 2, pp. 53–66. Hou, Jack W., Robert W. Mead and Hiro Nagahashi (2004), ‘Evolution of China’s US Policy (1965–72: Prelude to the Economic Reform?’ American Journal of Chinese Studies, Vol. 12, No. 1, pp. 1–24. Hu, Angang. 㚵䵡䩶 (2001),᧪ഄऎϢথሩ˖㽓䚼ᓔথᮄ៬⬹ˊЁ೑䅵ߦߎ⠜ ⼒. Hu, Zhaoling (1994), ‘The Eight Laws of Migration and the Migration in China,’ Yunnan Geography and Environment, Vol. 1, pp. 45–53. Jones, Marion (2004) ‘Tourism for Sustainable Development in Southwest China: A Double Edged Sword,’ working paper. Liang, Zai and Michael J. White (1996), ‘Internal migration in China, 1950–88,’ Demography, Vol. 33, pp. 375–84. Liang, Zai (2001), ‘The age of migration in China,’ Population and Development Review, Vol. 27, No. 3, pp. 499–524. Lovely, William (2001), ‘First impression from the 2000 census of China,’ Population and Development Review, Vol. 27, No. 4, pp. 755–70. Mallee, Hein (1995), ‘China’s household registration system under reform,’ Development and Change,Vol. 26, No. 1.

Migration and Regional Development in China

331

Mallee, Hein. (1996), ‘Reform of the hukou system,’ Chinese Sociology and Anthropology, Vol. 29, No. 1. Narayana, M. R. (1990), ‘Policy and Non-policy Economic Determinants of Interregional Migration of Workers in a Developing Country: Some New Evidence Based on a Polytomous Logit Model for India,’ Population Research and Policy Review, Vol. 9, pp. 285–302. Ping, Huang and Frank N. Pieke (2003), ‘China Migration: Country Study,’ presented at the Regional Conference on Migration, Development and Pro-Poor Policy Choices in Asia. Website: www.livelihoods.org. Population Census Ofﬁce of the State Council and the National Bureau of Statistics (2002), Tabulation on the 2000 Population Census of the People’s Republic of China, Vols 1–3. China Statistical Press. Shryock, Henry S., Jacob S. Siegel and Associates (1976), The Methods and Materials of Demography, Academic Press, New York, pp. 394–95. Yan, Shanping. ₴⠓㄂ (1998), ₼⦌⃬◐㄃ⅲ⦿◉梃ⅉ♲扐䲊䤓⸭㊐♙␅㧉 Ⓟᇭ䯍↩ⷵ䪣䴅, Vol. 2, pp. 67–74. Zhen, Peiyen.᳒෍♢ (2003). ೑ᆊ㽓䚼ᓔথ᡹ਞˈЁ೑∈߽∈⬉ߎ⠜⼒.

The Chinese Economy after WTO Accession

332

Appendix Roberto Bachi’s Migration Preference Index Iij =

Mij K ( Pi / Pt )[ Pi /( Pt − Pi )]∑ Mij i, j

where Iij is the migration preference index, Mij is the total migrations from place i to j, Pi is the population at place i, Pj is the population at place j, Pt is the total population of all places, K is a constant, usually assigned to 100. Respectively, the regional immigration preference index is deﬁned as I•j=

M•j ( Pj / Pt )∑ M

K • j

j

and the emigration preference index is deﬁned as Ii • =

Mi • K ( Pi / Pt )∑ Mi • i

If a region’s immigration preference index is greater than 100, it implies that the region has an above (national) average attraction for immigrants. Naturally, the higher the immigration preference index (IPI), the stronger the attraction of the region. In contrast, the higher the emigration preference index (EPI), the greater the pressure on the emigration in the region or the higher the population mobility in the region. Modiﬁed Narayana Logistic Linear Model Let pij denote the probability of a potential migrant’s move from place i to j, and Zij is a vector derived from a function of attributes of places i and j. We can then deﬁne pij =

e Zij ∑ eZil

i, j = 1, …n

l

where Zij = a + b ∑ Xki + c ∑ Xkj + dDij , k is the number of attributes at places i k

k

and j. This is an outline of the original Narayana model. To serve our objective, we revise the Narayana model by assuming a migration model between places a and b, with a as a particular province and b representing all other provinces.

Migration and Regional Development in China

333

We then assume that probability of a migration from place a to b, pab, has the following function: pab =

e Zb e Za + e Zb

where Zk = a + bWXk , k=a,b. Xk is the vector for attributes at place a and b. W is a weight matrix with wii=1, which is deﬁned by distance between a and b. We can then derive: e Zb pab = Za = a '+ b 'WXk 1− pab e For simplicity, we assume: mab pab ≈ 1− pab ma where mab is the total migration from place a to b and ma is the population at place a. Taking the log transformation, we can then apply an OLS estimate of the following model: Log (

mba e Zb pab ) ≈ Log ( ) = Log ( ) = a '+ b 'WXk 1− pab ma e Za

Similarly, we can deﬁne the probability of a potential migration from place b to a as: pba =

e Za e + e Zb Za

e Za mba mba ma mba pba = Zb ≈ = • =α 1− pba e mb ma mb ma mba e Za ≈ α ' Zb ma e where mba is the total migration from place b to a and mb is the population at place b. Again, after a log transformation, we can do an OLS estimate of the model: Log (

mba e Za pba ) ≈ Log ( ) = Log (α ' ) = a "+ b "WXk 1− pba ma e Zb

334

The Chinese Economy after WTO Accession

In this study, mab represents the emigration from province a to all other provinces, mba represents the immigration from all other provinces to province a, and ma is the total population of province a.

Chapter 18

Taiwan Cross-Strait Economic Relations in the Era of Globalization Yongjun Chen

Introduction The major factor that affects economic relations between the two sides of the Taiwan Strait is the development of economic globalization in the new century. This chapter discusses the prospect of cross-strait relations against the background of economic globalization and analyzes the impacts of various factors in the early twenty-ﬁrst century on the cross-strait relationship, such as the economic recessions in the US and Japan, and China’s accession to the World Trade Organization. The chapter is organized in four sections. The ﬁrst section examines the major factors of the twenty-ﬁrst century that affect cross-strait economic relations. The second section discusses the trend of development of cross-strait economic relations. The third section analyzes the structural changes and policy characteristics of the two sides of the Taiwan Strait. The fourth section gives suggestions on how to enhance economic cooperation and development across the Taiwan Strait. Major Factors Affecting Cross-strait Economic Relations The problems related to the development of the cross-strait economic relationship can be attributed to two major factors. On the one hand, the problems have accumulated from the internal contradictions that failed to be handled efﬁciently over the years. On the other hand, the problems are closely related to the changes in the external environment brought about by economic globalization. The changes in external economic environment as well as the needs of internal economic development give pragmatic importance to the favorable and dynamic interactions between the two sides. The factors of the early twenty-ﬁrst century affecting the economic relationship between the two sides include the following impacts, among others. The Impact of Global Economic Integration Global economic integration, the basic feature of world economic change, is the major factor that affects the cross-strait economic relationship. Economic

336

The Chinese Economy after WTO Accession

globalization is not a sudden incident, but a developing tendency and a process that has been continuing for many years. It is true that the globalization process of its current stage is the continuation of historical development over the past 100 years, but the current stage began in the 1980s. The integration is mainly the combined result of the development of international trade, the rapid increase of foreign direct investment (FDI), the accelerating ﬂow of capital between states, the spread of economic liberalization policies, and the advance of modern science and technology. It pushes every country or region to comply with the trend of economic globalization, participate actively in international division of labor and cooperation, and increase the competitiveness of their products. The outside illustration of economic globalization includes strengthening regional cooperation and enhancing the role of dispute settlement mechanisms concerning international trade. Regional Cooperation is Strengthening Regional economic integration is not only the outcome of uneven development of the world economy, but also the inevitable result of the increasingly ﬁerce international trade competition. The prospect of economic growth of a region or a country, to some extent, is determined by regional economic cooperation. In the Asia-Paciﬁc region, because of its large scope and the complicated background of each country, sub-regional cooperation and integration have been slower. Sub-regional integration puts emphasis on the regional, cultural, historical, and civil connections, as well as the existing economic connections. Sub-regional integration has formed a core and spread beyond it, which is reﬂected in the following three aspects. First, since its establishment in 1989, APEC (Asia-Paciﬁc Economic Cooperation) has promoted ‘opening regionalism’ and the advancement of free trade, and they established the goal of realizing regional trade and investment freedom by 2020. Second, the summit meeting between ASEAN (Association of Southeast Asian Nations) and China, Japan, and Korea (‘10 + 3’) proposed to strengthen cooperation in the economy, society, and human resources among these countries. This shows that the cooperation among East Asian countries is in the process of institutionalization and may eventually lead to the formation of the East Asian Free Trade Area. In the meeting of ‘10 + 3’ in 2001, the Chinese government suggested that China and ASEAN will be the ﬁrst countries to sign a free trade agreement. Third, the discussion of setting up a China Economic Zone between mainland China and Hong Kong, Macao, and Taiwan is well under way. Dispute Settlement Mechanisms Being Enhanced The World Trade Organization (WTO) has been accepted as the main mechanism for settling trade disputes. Mainland China and Taiwan had become members of the WTO by the end of 2001. The purpose of the WTO rules is to reduce tariff barriers and to encourage trade and investment between countries and regions, which is helpful in regulating relationships among WTO members. After the two sides join

Taiwan Cross-Strait Economic Relations in the Era of Globalization

337

the WTO, the trade volume between the two sides will increase dramatically. Taiwan will gradually cancel the restrictions on goods imported from mainland China, which will increase the export opportunities for mainland China. Although this may decrease Taiwan’s favorable balance of trade with mainland China, the bilateral trade development under normal market mechanisms will meet the economic development needs of each side. The increase of trade volume and the structural supplement will promote the economic growth of the two sides. In addition, WTO membership is helpful for the increase of investment between the two sides. Currently, the investment between the two sides is still one-sided. Taiwan authorities have set up various obstacles against Taiwan businessmen investing in mainland China. However, after joining WTO, this kind of restriction, in theory, should soon be removed. The Impact of Developed Countries The US, Japan, and other western developed countries have not fully recovered from their economic recessions, which strongly inﬂuence the economies of both sides of the Taiwan Strait. Confronted with challenges brought by economic globalization, Japan’s economy plunged into recession that has lasted for more than ten years, and there is no sign of recovery so far. The Asian ﬁnancial crisis that broke out in 1997 has dragged most countries in this area into economic recession, and the negative inﬂuences of the crisis have not disappeared completely. The American economy managed a hard landing a few years ago, and its economic growth slowed down signiﬁcantly, especially after the September 11 incident. It is impossible for the American economy to adjust itself in a short period. All of these factors will have strong inﬂuences on the cross-strait economies. Asian countries are gradually changing from the traditionally ‘unitary market reliance’ on the US to ‘double reliance’ on the US. Double reliance on the US includes both market reliance and structural reliance, which further casts shadows on the Asian economies. For a long time, Asian countries had taken advantage of the ﬁrst market of favorable balance offered by the U.S. to realize their strategies of exportled economic growth, which resulted in market reliance on American trade. At the same time, Asian countries undertook structural adjustment after the ﬁnancial crisis, trying to participate in the international division of labor within the structure of the new American economy to optimize their own industrial structures. This adjustment thus resulted in structural reliance on the US. In the case of mainland China, with its high degree of reliance on exports to the US, the decrease in American import demand and the devaluation of the US currency will inevitably have a negative impact on China’s economic growth. As for the major export markets of Asian developing countries and regions, the economic recessions in the US and Japan reduced their exports and thus led to re-emergence of trade protectionism. In the short term, the economic growth in Asia depends on the economic recovery in the US and Japan. In the long term however, Asia must experience a change from an extensive to an intensive economy. Asian countries

338

The Chinese Economy after WTO Accession

and regions face similar questions on how to optimize their industrial structures and realize modernization by introducing capital and technology, strengthening macro-management, enhancing system innovation, developing high technology, and transforming traditional industry. The Impact of Cross-strait Economic Characteristics The cross-strait economic characteristics in the initial stage of the twenty-ﬁrst century determined their interests and possible policy choices. Industry in Taiwan is facing transitional obstacles such as resource outﬂow, high land costs, lack of power supply, difﬁculty combining capital and technique, shortage of opportunities for product upgrades, insufﬁciency in research and development input and so on. The orientation and formation of new industries in Taiwan have not caught up with the pace of outsourcing traditional industries to mainland China. Meanwhile, mainland China has maintained a better economic growth impetus. Although its development level is comparatively lower, China’s economic size is considerably larger. However, although mainland China has experienced a higher growth rate of 7 to 8 percent, that growth rate still cannot satisfy the needs of economic development and employment of its huge labor force. It is also problematic to rely on active ﬁscal expansionism on a long-term basis. It is more imperative for mainland China to participate actively in international markets due to its relatively backward economic structure. The economic relationships between the two sides have special meanings. First, mainland China provides Taiwan not only with cheap labor but also export markets. Second, the capital, technique, business management, and market sale ability of Taiwan are in short supply in mainland China. Mainland China and Taiwan can thus complement each other very well. In general, against the background of economic globalization, the cross-strait economic connections are bound to be strengthened gradually with internal impetus and external pressures due to their special historical background, resource complementarities, and the drive of economic interest on both sides. Global Economic Integration: An Analysis of Cross-strait Economic Relationship Development Because of historical reasons, there are considerable differences between the basic conditions and development processes of Taiwan and mainland China. These are mainly reﬂected in the following four aspects: (1) the gap in the level of economic development; (2) the gap in original institutions; (3) the gap in economic openness; and (4) the gap in ideas and human resources adaptable to a market economy. Regarding original institutions, mainland China had been implementing a planned economy for a long time, but it is now in the process of shifting to a market economy. The Taiwan economy is already based on market mechanisms with protection for

Taiwan Cross-Strait Economic Relations in the Era of Globalization

339

private property. As for economic openness, the current degree of dependence by mainland China on imports and exports has reached 40 percent or so, and mainland China is now integrated into the international market. In contrast, Taiwan’s economy is mostly export-oriented, and its external dependence is much higher. Despite these gaps, the two sides not only share a similar historical background in culture, ancestry, and location, but also have strong complementary economic activities. This can be represented as follows: (1) mainland China has advantages with its vast expanse, rich manpower resources, substantial industrial foundation, powerful technical strength, high scientiﬁc level of basic research, huge market capacity, and strong high-tech research and innovation. The major problems mainland China is facing are those of a shortage of construction capital, a lack of belowmiddle scientiﬁc personnel, and a lack of experience in marketing and management. (2) Taiwan has advantages in such aspects as exploiting foreign markets, developing agricultural technology, tapping international markets, promoting industrialization of scientiﬁc accomplishments, and training talents in management. The major problems Taiwan is facing are resource scarcity and a narrow market. (3) Both sides have different advantages in developing new high-tech industries. Taiwan’s advantages lie in industrialization and marketing new technology, especially in the electronic information industry, while mainland China has the advantage in basic science research and market demand for high-tech products. As for mainland China, it may beneﬁt from introducing capital, technology and modern management from Taiwan. In Taiwan, businessmen will promote the growth of the economy and employment through the investment in mainland China. The cross-strait cooperation will help the industrialization process of science and technology. Although mainland China is stronger in basic and applied science and has more abundant research accomplishments, its industrialization of the research accomplishments is less than ideal. Taiwan has more experience and achievement in the aspect of industrialization. As for Taiwan, carrying out resource conﬁguration beyond the island will help to develop its industrialization advantage and characteristic industries and will help Taiwan participate in the international division of labor in a larger scope. The economic connections between the two sides will likely follow the trend of global economic integration and the reinforcement of economic complements. The development is mainly reﬂected in the following aspects. Global Cooperation between Production and Consumption Taiwan is enlarging its trading market towards mainland China. Currently, the total foreign trade of Taiwan with mainland China, including imports and exports, is more than US$32 billion per year. Taiwan’s exports to mainland China are about US$27 billion and its imports from mainland China are about US$5 billion. Altogether, the trade surplus is approximately US$20 billion. Mainland China has become the region where Taiwan has the biggest favorable balance of trade. Taiwan’s total favorable balance of foreign trade is smaller than the volume of that with mainland China. If

340

The Chinese Economy after WTO Accession

it were not for the favorable balance of trade with mainland China, Taiwan’s foreign trade may turn from a surplus into a deﬁcit. Cooperation has promoted the development of Taiwan’s economy. The GDP of Taiwan in 2001 was approximately US$280 billion, and its exports to mainland China were approximately US$27.3 billion or 9.8 percent of its GDP. Therefore, its exports to mainland China have played an important role in Taiwan’s economic development. Strengthening the cooperation with mainland China helps Taiwan realize its goal of becoming a production center focusing on high-tech development. Establishing Taiwan as the high-tech production center is one of the important themes of ‘Operating Center of the Asia Paciﬁc’ that Taiwan has put forward. Mainland China has greater demand for high-tech products, so developing cooperation between mainland China and Taiwan will undoubtedly promote this role. With the promotion of trade and direct investment in Taiwan, Hong Kong, and Macao, the economy in mainland China has achieved maximum development, especially in the manufacturing industry. Further, the dense state of the labor force in China has stronger competitive ability in the international market. Regional Economic Cooperation The cross-strait cooperation has become the most realistic option and has shown strong vitality. ‘The Economic Circle of Chinese,’ with the major actors of mainland China, Taiwan, and the region of Hong Kong and Macao, is accelerating in its formation. The World Bank forecasts that this economic district will become the fourth major international economic region by the ﬁrst half of the twenty-ﬁrst century and will be equally as important as the three economic regions of the US, Europe, and Japan. Professor Robert Mundell, who won the Nobel Prize in Economics in 1999, pointed out in March 2000: ‘Looking forward, the Economic Circle of Chinese will come into being.’ Then on June 12 at the international consulting meeting of Taipei, he said, ‘Cross-strait cooperation may form an economic circle, acting as a secondary economic area.’ Dependence on Each Other’s Economies Ziqing Pan and Zinai Li established a model of ‘the macro-economic relationship model between mainland China and Taiwan,’ and studied Taiwan’s economic reliance on mainland China in the merchandise trading sector. The ﬁndings of their study are as follows. Taiwan has stronger dependence on mainland China for merchandise exports. The favorable balance of trade with mainland China from 1992 to 1998 contributes to Taiwan’s GDP with an average growth rate of 17 percent per year in the same period, exceeding the effect of investment promotion on GDP. If mainland China were to reduce its imports from Taiwan at 100 percent per year, it would not only make the average growth rate of actual and nominal GDP drop by 0.5 and 0.22 percent respectively, but would also cause a decrease in resident income,

Taiwan Cross-Strait Economic Relations in the Era of Globalization

341

nongovernmental consumption, investment, regular capital formation, imports, exports, and employment. The biggest inﬂuence would be on the resident income level and employment conditions. Certainly, if mainland China completely reduces its imports to zero, it would also produce an unfavorable inﬂuence on its economic operation, but this inﬂuence would be relatively much smaller. In the analysis of the same period, the annual average growth rate of actual and nominal GDP in mainland China would drop by 0.01 and 0.02 percent respectively. Stronger Power of Capital over Labor Since Taiwan authorities permitted investment in mainland China in the early 1990s, the investment made by Taiwanese businessmen in mainland China has been developing very quickly. The investment scale has now reached more than 50 thousand from 1,000 in 1990, and the actual investment amount has reached more than US$30 billions. The investments have extended from the industrial production by medium and small businesses at the initial stage to the industrial production by large scale enterprises and enterprise groups. This has led to the equal development of small, medium, and large businesses and the structure of relational enterprise investment among them. The ﬁeld of investment ranges from light-spin manufacturing industry to high-tech estate, and the region of investment has also promptly enlarged. Taiwan’s investment in mainland China has taken the ﬁrst place of its foreign investment. For the part of mainland China, Taiwan’s investment to mainland China has taken the second place of all foreign investment, only next to that of Hong Kong. Weaker Regional Economic Permission Rights After joining the World Trade Organization and because the economic freedom and internationalization of Taiwan is higher than that in mainland China, Taiwan will face huge pressure on its restricted economic policy toward mainland China. According to the non-discrimination principle and national treatment principle of the World Trade Organization, the policy of ‘Jie Ji Yong Ren’ (avoiding hurrying and being patient) by Taiwan authorities will collapse on itself. However, this is a process in development, and it has not fully realized the advantages that it should have. From the reality of the cross-strait economic relationship, some observations are as follows: (1) the disequilibrium of the cross-strait trade; (2) the disequilibrium of the cross-strait investment; (3) the disequilibrium of geography and ﬁeld of investment; and (4) the coexistence of functional integration and institutional non-integration. As for cross-strait investment, China hardly invests in Taiwan. In terms of the geography and ﬁeld of investment, there is more cooperation in traditional industries, and there is more investment in small-scale enterprises but less in large-scale enterprises. Further, there is more investment in the East and less in the West.

342

The Chinese Economy after WTO Accession

Effect of Globalization on Cross-strait Industrial Structures Outﬂow of Industries In the past 10 years, almost all of the traditional industries in Taiwan suffered a slump. Development in the manufacturing industry was chieﬂy dependent on electronic information technology, but under the impact of a global depression in information technology, even the manufacturing industry is suffering from a grim slump. Some claim that the slump is due to the ‘hollowing out’ effect resulting from the outﬂow of industries. Reasons for an economic slump in a region or nation can be broadly divided into two categories: one is a business recession and the other is a decline in longterm competitiveness. According to Longhong Zhang, the manufacturing industry in Taiwan is primarily facing a short-term economic recession. Although there appears to be a downward trend in long-term export competitiveness, the trend is indistinct. According to our analysis, the economic slump occurring in Taiwan’s industry is not because of the ‘hollowing out’ of industries but because Taiwan is in transition from a traditional extensive and export-oriented economy to a modern intensive economy. Outﬂow of Industrial Capital After Economic Development According to the ‘wild goose ﬂight pattern’ theory, Taiwan occupies the front of the line, and when its economic development reaches a certain level, the outﬂow of industries becomes a requirement. The 1990s were a watershed in the acceleration of the outﬂow of Taiwan’s industrial capital. Finn’s Life Cycle Theory of Manufactures explains theoretically the industrial capital transfer among countries of different development levels in different stages, and its features are decided by the pattern of division of labor. In regard to the information technology of Taiwan, there are two characteristics of its development that affect the competition among products: the many products that have entered the maturity stage and the aggregation effect of manufacturers. Based on the Life Cycle Theory of Manufactures in international trade theory, the mature-type manufacturer, with lower technical obstacles, will bring more international comparative beneﬁts to the manufacturer so long as he can produce them with a lower cost. As for the aggregation effect of manufacturers, there is high relevancy among information technological industries. Such relevancy is evident in both the cost of production and technology. Thus comes the aggregation effect of manufacturers, and the outﬂow of industries is a necessity following economic development. Industrial Capital Outﬂow After Economic Restructuring The three industrial structures of Taiwan are in consistent optimization, as is the inner structure of industry—the second industrial structure and the index of productivity

Taiwan Cross-Strait Economic Relations in the Era of Globalization

343

of highly capital-intensive and advanced technology-intensive industries grows the fastest. As for the pillar industries of IT in Taiwan, there are three aims for their investment in mainland China: seeking a manufacturing base, expanding markets, and seeking technological resources. Such an outﬂow is the intrinsic necessity of economic restructuring. Meanwhile, it implies that Taiwan will have a louder voice on the platform. Industrial Capital Outﬂow Will Not ‘Hollow Out’ Industries in Taiwan According to the Life Cycle Theory of Manufactures in international trade theory, countries in the ‘wild goose ﬂight pattern,’ on the one hand, wash out certain industries and transfer them to less developed countries, but on the other hand, they receive industrial capital from the more developed countries. The net imports are advanced technology industries, such as machinery, chemistry, and electrical appliances, while the net exports are low technology industries, such as foodstuffs and textiles. China is the Best Choice for Taiwan’s Transfer of Industrial Capital Difference of economic and industrial structure across the Straits forms the foundation of a mutual beneﬁt relationship and conforms to the transferring requirement of the ‘wild goose lineup’ theory. At the beginning of upgrading its industries, a large number of medium- and small-sized traditional industries ﬂowed out of Taiwan and mostly into mainland China. This contributes to the industrial upgrading development in the 1990s, because without transferring, a great number of industries would have gone bankrupt under the entire economic deterioration. In light of the economic rule, large-sized enterprises followed in the footsteps of medium- and small-sized industries, which led to the phenomenon of large-sized enterprises investing in mainland China in the early 1990s. Comparative Advantages With an open economy, the industrial structure of a country or an area is restricted and inﬂuenced not only by the level of economic development, but also by the industrial structure of the related countries and areas. To analyze the effect of a change in the industrial structure of the two sides, we should take the present industrial structure into consideration and compare the comparative advantages, then analyze the tendency for interaction under economic globalization. Comparing the Stages of Industrial Development The stage of industrial development in Taiwan is close to that of developed countries. Recently, besides the tendency of industrial development, the product mix of production or exports of manufacturing industries in Taiwan have experienced signiﬁcant changes. In the aspect of production, if industries are divided into

344

The Chinese Economy after WTO Accession

conventional, basic, and technology-intensive categories, the contribution of conventional industries to aggregate manufacturing has decreased year by year. It has dropped from 41.95 percent in 1987 to 31.28 percent in 1997, a 10.67 percent decrease over those 10 years. However, the contribution of technology-intensive industries to aggregate manufacturing has increased by 7.86 percent, and the contribution of basic industries has increased by 2.81 percent. Prompted by the changing tendency of the proportion of each industry, the industrial structure of manufacturing in Taiwan has developed toward large scale and concentration, especially the industries of electric power, electronic machinery and apparatus, which constitute more than 20 percent of GDP. With regard to exports, the proportion of exports from labor-intensive industries fell from 43.45 percent in 1989 to 34.55 percent in 1998, while the proportion of exports from capital-intensive industries rose from 26.59 percent in 1989 to 29.30 percent in 1998. The proportion of exports from high-tech labor-intensive industries in 1998 had reached 40.91 percent, much higher than the 24.25 percent level in 1989. Meanwhile, the proportion of exports from chemical products rose from 44.53 percent in 1989 to 63.83 percent in 1998. Similarly, the export proportion of high-tech products rose from 33.92 percent in 1989 to 49.44 percent in 1998. The export structure in Taiwan has been readjusted towards the level of developed countries. Both the manufacturing and export structures in Taiwan have obviously tended toward technology-intensive industries and there has also been an upgrade in manpower structure. All of these data account for the fact that the manufacturing structure in Taiwan has been moving closer towards that of developed countries. The industrial development stage of mainland China is at a lower level than Taiwan but is rising. The economy of China has developed quickly due to many factors such as the reform policy over the past 20 years, cheap labor, rich natural resources, and the speciﬁc background of Hong Kong, Macao and Taiwan. Some competitive advantages have been formed gradually. Seen from the economic and commercial relationship between mainland China and Taiwan, the cooperation is mainly based on trade and commerce brought about by investment. Making use of the resources of mainland China, Taiwan exports many intermediate raw materials and machine replacement parts, which play an important role in the promotion of industrial structure. Therefore, Taiwan considers the economic relationship between the two sides as cooperation. However, new problems are appearing in the process. First, the industrial technology in mainland China is being upgraded. The proportion of technologyintensive products in exports in 2000 rose to 25.86 percent, almost equal to the proportion of Taiwan in 1989. The industrial structures in Guangdong, Shanghai, and Beijing have converged towards that of Taiwan, and industrial groups have been formed. Second, import substitutes in mainland China have developed quickly. The old labor-intensive industries have their development space in the Midwest region, but producing downstream products promotes the development of the production of raw materials. Investment has changed gradually from labor-intensive industries to information industries. In 2000, in the industries that received investment from

Taiwan Cross-Strait Economic Relations in the Era of Globalization

345

Taiwan, manufacturing industries reached a proportion of 91.40 percent, which was the highest proportion. In the thin industry, the industries that have higher proportions include electronic power (28.02 percent), basic metal and metal manufacturing (8.31 percent), plastics (7.83 percent), and food and beverage (7.48 percent). Therefore, in the short term, mainland China is not capable of forming the competitive power which inﬂuences the industries in Taiwan. To sum up, though there is endless space in the vast west of mainland China for the development of labor-intensive industries, the general level of the industrial structure in mainland China is still low. Comparative Advantages of the Industries Looking at static comparative advantages, Mr. Fan Jinming of Research Institute of Chinese Economics compares the labor productivity in the industries of Taiwan and mainland China, and ﬁnds the production advantage of mainland China lies mostly in the following industries: labor-intensive industries, material-intensive industries, and some capital-intensive industries. Mainland China has invested a lot in some heavy industries and has formed economies of scale. Compared with Taiwan, the ﬁrst twelve industries which have the comparative advantages in competition in mainland China are crude oil, natural gas, metal, manufactured goods with grain, feed, coal, household electrical appliances, other chemical products, oil and other industrial chemicals, medicine, other subsidiary crude goods, and petroleum reﬁning products. The comparative advantages in Taiwan come from capital-intensive industries, technology labor-intensive industries and so on. These include information technology, plastic cement and other chemical materials, chemical ﬁbers, soft beverages, chemical fertilizers, cement, ceramic products, electric power and gas, glass and glass products, cement products, saw lumber and plywood, automobiles and locomotives, cast iron, and rough products of iron and steel. In terms of dynamic comparative advantages, there are currently four types of industries in the industrial development and investment process in Taiwan: 1. Both investment and industry grow vigorously. According to their order, there are electronic and electric appliances, oil and coal products, and basic metal industries. 2. Investment is mainly in Taiwan and industries develop steadily. According to their order, there are machinery, chemical products, and metal products manufacturing industries. 3. Investment in mainland China leads industries in Taiwan to grow at a slower rate. These industries include nonmetal and mineral products (such as building materials), food and beverages, and transportation vehicles. 4. Investment outside Taiwan or in mainland China grows vigorously with atrophic investment in Taiwan. According to the order of severity, there are leather and fur products, sundry products, ready-made clothes, bamboo, vine and willow products, precision apparatus, paper and printing, textiles, and plastics products manufacturing industries. The industries in which Taiwan

346

The Chinese Economy after WTO Accession

actively invests in mainland China or are atrophic in Taiwan are exactly the industries developing in mainland China. Other than the development of the respective comparative industrial advantages mentioned above, the industries of the two sides of the strait have great interaction in the following respects. First, the aspect of industries will be examined. Mainland China and Taiwan have their own advantages in areas of knowledge-intensive and new high-tech industries. 1. Taiwan has its advantages in information and communication technology, while mainland China has its advantages in nucleon technology, biological engineering, and aerospace industries, so the two sides can develop their own advantageous industries and complement each other. 2. There are quite a few labor-intensive industries in Taiwan and their developing conditions are becoming worse and worse, so some of them can move to mainland China, especially to western China where development is quickening. 3. There is much space in the transverse and longitudinal coordination of the industrial development between the two sides. 4. In agriculture and service, there is a broad ﬁeld to complement and cooperate. 5. The usage and cooperation of human resources will be more rational. Mainland China needs professional personnel who are qualiﬁed in contemporary management and administration, contemporary ﬁnancial services, market intermediation and so on. Taiwan needs qualiﬁed technical personnel in basic and applied industries. This is exactly where the two sides can complement each other. 6. Mainland China will accept the industrial moves including that of Taiwan and will become the global manufacturing center. The manufacturing industries in Taiwan will develop from the traditional to the exquisite and will produce a small amount of high-quality products. Heavy chemical industries will become the pillar of Taiwan, which will mainly provide supplies for interior markets and the export of downstream products. Technology-intensive industries with high value-added will lead the growth of manufacturing industries. Second, the aspect of agriculture is important. The current agricultural cooperation between the two sides mainly includes indirect trade and investment. On the one hand, there is high homogeneity in the agricultural products of the two sides and there is a short distance between the two sides. On the other hand, the economic development levels of the two sides are different, but the developing conditions of agriculture are complementary to each other. Taiwan can enrich the shortage of capital and technology in mainland China, and mainland China can counteract Taiwan’s insufﬁciency with respect to markets and materials. Thus beneﬁcial agricultural exchanges are propelled.

Taiwan Cross-Strait Economic Relations in the Era of Globalization

347

There are four areas for potential agricultural cooperation between the two sides. (1) There are cooperation opportunities in the aspects of ﬁshery disputes, resource maintenance, products smuggling, and disease infection and so on. (2) Taiwan imports materials which are in limited supply in Taiwan from mainland China, while mainland China mostly imports high-quality aquatic products, local specialties, and manufactured food to meet the high-quality food needs of higher economic development. Agricultural investment can bring capital, technology, variety, and management to the agricultural industries of mainland China. Further, Taiwan can explore the markets in mainland China and use mainland China as a manufacturing and exporting base by taking advantage of its cheap materials and labor. There is much to do in the aspect of technological cooperation. Taiwan has the lead in agricultural and administrative technology of the tropical and subtropical zones, while mainland China has its advantages in the stock of seed. They can be mutually beneﬁcial to increase competitive power. The industrial division between the two sides will exhibit the following tendencies. According to the principle of comparative proﬁts, Taiwan imports the resources of agricultural processing from mainland China and exports agricultural and aquatic products to the high consumption area of mainland China in order to reach ‘vertical division.’ They enlarge the exchange of capital, technology, and natural resources and gradually choose and lift the appropriate items from vertical division to ‘regional division’ and ‘division on the basis of specialization.’ Taiwan mainly produces products of high technology, high competitive power, and high quality, while mainland China produces, transports, and sells products to Taiwan which are not economical to produce in Taiwan. Competitive Advantages To maintain competitive advantages and long-term operation, Taiwan industries need global perspectives and competence. Mainland China provides hope for Taiwan industries. As Taiwan Island is small in area, limited in resources, and restricted in its future development of the domestic market, its industries would have to extend outward to be competitive against other developed countries. Taiwan is built on manufacturing industries, so new industries in the future will also be high value-added manufacturing industries. This will be the pillar for the development of the six supporting systems, including the supporting upperreaches system, i.e., the innovation, research, and development system. This includes manufacture design, function design, and new combinations, which could make Taiwan a platform in the design industry. The lower stream could be built as a management center of supply chains integrating material ﬂow, capital ﬂow, and information ﬂow. An investment system and a management and operating system could make Taiwan a regional transfer station, a transferring center of aviation and navigation, and a highly industrialized ﬁnance center. Taiwan could also develop a ﬁnancial support system.

348

The Chinese Economy after WTO Accession

As for mainland China, its industrial development depends on the continuity of its role in the ‘wild goose ﬂight pattern’ model and continuously taking in the transferred industries from Hong Kong and Taiwan. Due to the pressure of underemployment, developing manufacturing industries in mainland China will be a long-term objective. By utilizing the advantages of the information era, mainland China could develop by leaps and bounds and strengthen its cooperation with Taiwan in advanced technological industries. At the same time, it could use the advantage of labor and resources to improve the competitiveness of manufacturing, as well as participate actively in the international division of labor led by the US. Enhancing Cross-strait Economic Cooperation and Development Within an environment of economic globalization, both sides of the Straits should try to develop their economy and trade across the Straits actively, and at the same time promote development and the division of labor. Both sides should develop their advantageous industries and manufacturing based on their respective social and economic backgrounds. Meanwhile, they should develop by improving the competition of manufacturing and perfecting conditions for trade. Utilizing Advantages The economic development across the Straits shows an apparent ladder-like structure. There exists not only the same structure but also clear differences and complementarities. The practical choice is to allocate resources based on a larger scope to improve their respective competition based on their different advantages under the environment of economic globalization. One approach is to utilize comparative static advantages. According to the above analysis, Taiwan possesses apparent advantages in the areas of information technology, technology-intensive industries, electrical appliances, and services. The market competition of traditional industries has declined recently mainly due to the lagging development of manufacturing and the slowing down of technology improvement. Taiwan should implement technological reconstruction of its industries so as to enjoy the beneﬁt of a mature period and to further advance toward the standard of the developed countries. This would upgrade technological value and actively transfer the economic structure. Another approach is to utilize comparative and dynamic advantages. China and Taiwan should take advantage of this time when America and Japan are under economic restructuring to further transfer the traditional industries to mainland China. They should follow the principle of ‘proceeding in the light of local conditions and in accordance with a rational division of labor, with all the regions exploiting their own particular advantages for mutual beneﬁt and development’ to develop and strengthen the advantageous industries. From now on, two kinds of industrial conductive functions should be exerted: one is to receive the radiant effect of American and Japanese industries

Taiwan Cross-Strait Economic Relations in the Era of Globalization

349

to develop new economic industry and the other is to participate in the international division of labor. As for Taiwan, the developing trends of manufacturing in the twenty-ﬁrst century are: (1) traditional industries are reﬁning and producing products of high quality in low quantity; (2) heavy industries and chemical industries are becoming supporting industries, mainly to supply domestic markets and exporting industries; (3) technology-intensive industries with high added value are becoming the main driving force for manufacturing industries; (4) industries are gradually being liberalized since the labor force is consistently running out while the demand for quality is growing; (5) multinational companies are becoming the backbone for development in response to the changing trend of international trade; (6) attention is being given to economic development together with environmental protection; and (7) innovation and advancement of technology are becoming the driving force of economic development. As for mainland China, it is still in the stage of industrialization, and industries in Western China are especially underdeveloped. Nevertheless, China possesses apparent advantages in labor-intensive and resource-intensive industries such as leather, fur and ware products, mixed manufactures, ready-made clothes, ﬁne appliances, paper-making and printing, textiles, and rubber. It is better to use foreign capital to promote economic development and employment, to accelerate the process of industrialization, and to receive the radiant and transference of Taiwan industries actively. At the same time, mainland China should take advantage of basic science and develop cities in the special economic zone, participate in the international division of labor, and promote advanced technological industries to make certain achievement in new types of industries. Improving the Competitiveness of Products Taiwan can lower costs by transferring to mainland China industries in which it does not have comparative advantages, such as labor-intensive and resource-intensive industries. Meanwhile, Taiwan can enhance the competitiveness of its products by technologically restructuring its traditional industries and developing exquisite manufacturing industries and services. As for mainland China in economic transition, certain approaches are needed to promote the competitiveness of its products. (1) The reform and development of enterprises should be integrated. Emphases should be placed on the potential advantageous enterprises to form a specialized manufacturing system with practical advantages. Enterprises without comparative advantages should be put into the markets, which will determine their survival. (2) Market demand and supply should be emphasized. (3) Competitive strategies should be adopted. Enterprises take part in not only the ﬁeld of price competition but the ﬁelds of non-price competition and amalgamation. From the aspect of market access, preferential treatment can strengthen the cooperation with local sales agencies, clients or dealers. A union relationship could

350

The Chinese Economy after WTO Accession

be made with local industrial and commercial enterprises to purchase local industrial and commercial companies and to set up new sales agencies. This would form the industrial structure with radiant and diffusing effects and competitive advantages and at the same time conform to market requirements. Improving the Environment for Trade Under the principle of ‘one country, two systems’ and the trend of economic globalization, political localization should be broken down and political obstacles should be reduced. The resolution of economic and trade problems should proceed under the WTO system for settling trade disputes. Fujian province should use its regional advantages to promote cooperation across the Straits. Both sides of the Straits form an echo in the economic and geographical distribution pattern, and economic integration is taking shape. Meanwhile, international regional economic cooperation is increasing, which makes cooperation across the Straits necessary in order to conform to the international economic pattern of the twenty-ﬁrst century. Finally, culture, especially the culture of southern Fujian, should be promoted as the link and the driving force to ensure positive economic development across the Strait. References Fang, Xinghai, and Sengfen Song (1998), Raising International Competitiveness – Taiwan’s Experience and Its Implications for Mainland China (Report from the Chinese Economist Society Visiting Taiwan),’ China Economy Press, Beijing. Pan, Wenji and Zinai Li (2001), ‘A Study of the Dependency of Taiwan’s Economy on Mainland China,’ The Economy of Special Economic Zones, Hong Kong, Maco and Taiwan, Vol. 2.

Index

Agricultural Bank of China (ABC) 128n9, 133, 134, 135, 184, 186, 187, 189n6, 190 agriculture sector 14, 162, 212, 214, 259, 274, 277, 280 and the ﬁnancial system 184, 187, 189–94, 196–8, 200 impact of WTO accession on 1–2, 17, 18, 20, 22, 27, 28, 29, 31 and labor 294–301, 303, 305 and migration 309–10, 319, 327–9 and Taiwan 346–7 wheat futures 238–52 Anhui 47, 55, 96, 169, 182 and migration 316, 317, 318, 322, 323, 324, 325 antidumping 22, 25, 209–23 Argentina 94, 210, 211, 213, 221, 240 Asia-Paciﬁc Economic Cooperation (APEC) 336 asset management corporations (AMCs) 128–30, 132–3, 135 Association of Southeast Asian Nations (ASEAN) 336 Australia 94, 98, 210, 211, 213, 240 automobile industry 9, 11, 13, 14–15, 19, 22–3, 27, 28–9, 345 Bank of China (BOC) 128n9, 129, 133–4, 186 banks and banking sector 51, 53, 90, 97, 167, 175, 184–7, 190–91, 194, 197, 199, 202 and ﬁnancial reform 126–50 impact of WTO accession on 2, 18, 21, 23, 27, 30 and state-owned enterprises (SOEs) 97 see also state owned commercial banks Beijing 23, 26, 30, 38, 46, 47, 51, 54, 55, 95, 96, 161–2, 168, 182, 200, 344 and migration 309, 312, 315, 317, 318, 319, 322, 323, 325, 326

beverages industry 10, 11, 86, 214, 265, 266, 267, 345 Brazil 94, 137, 210, 211, 213, 221 building materials industry 10, 11 Canada 94, 98, 210, 211, 213, 225, 238, 240, 258 catering industry 299, 305 Changchun 23, 168 chemicals industry 10, 11, 86, 213, 214, 219, 265, 266, 267 and Taiwan 344, 345, 346, 349 Chengdu 162, 168 Chicago Board of Trade (CBOT) 237–9, 243–52 Chile 94, 135, 137n20, 142, 143, 145 China Banking Regulatory Commission (CBRC) 128–9, 139 China Construction Bank (CCB) 128n9, 129, 133–4, 186 China Dalian Commodity Exchange (CDCE) 239 China Feiyue 33–44 China Zhengzhou Commodity Exchange (CZCE) 237–9, 243–52 Chongqing 23, 47, 55, 93n10, 96, 169 and migration 316, 317, 323, 325, 327 City Bank 23 coal-mining industry 9, 11, 18, 345 collective-owned enterprises 33, 48, 53, 65, 84, 91, 295, 299, 303 commodity futures 237–52 Communist Party of China (CPC) 25, 27, 49, 65 construction industry 12, 29, 162, 294, 299, 304–5 consumer goods manufacturing industry 10, 14, 162 Cuba 81 Cultural Revolution 65, 67, 307, 309

352

The Chinese Economy after WTO Accession

currency 2, 30, 36, 98n13, 128n9, 136, 141, 163, 274, 277, 289, 292 foreign 30, 129–30, 134–5, 145–6, 163, 274 see also foreign exchange deﬂation 114 domestic capital versus foreign capital 89–90, 92, 95, 97–108 Dongguan 15 dual pricing 1, 274 Eastern Europe 136 economic growth and development 9–16, 17, 22, 258, 275, 276, 282, 284 and ﬁnancial development 90, 107, 125, 127, 131–2, 133, 135, 181–2 and ﬁnancial sustainability 111–15, 121–2 and local governments 51–2, 53, 57, 59 and migration 307–29 and Taiwan 335–40, 342, 343, 346–50 Economy and Trade Commission (ETCC) 187n3, 219 electrical appliances industry 10, 11, 12, 86, 213, 214, 266 and Taiwan 343, 345, 348 electronics industry 10, 11, 14, 20, 23, 28, 86, 162 and Taiwan 339, 342, 344, 345 employment, see labor European Union (EU) 37, 112, 163, 240 and anti-dumping 209–10, 211, 213, 215, 216, 220 Feiyue, see China Feiyue ﬁnancial sector 126–50, 162, 175, 183–7 see also banks ﬁscal policy, see government food and food-processing industries 9, 11, 86, 214, 265, 266, 267 and Taiwan 343, 345, 347 see also catering industry foreign capital, see foreign investment foreign exchange 22, 39, 126, 134, 141n26, 149, 273, 274, 276–80, 284, 289 foreign investment 19, 98, 258, 273, 274, 282, 309–10, 319 and ﬁnancial development 125, 137, 148

globalization and 64, 65, 67–8, 70–71, 75–81 impact of WTO accession on 1, 2, 19, 21–2, 23–4, 25–6 and Taiwan 336–7, 339–41, 344, 345, 346 and taxes 90–91, 93–7, 310 see also domestic capital versus foreign capital, venture capital foreign trade 19, 196, 273, 297, 300 and anti-dumping, see anti-dumping balance 89, 91, 100, 101, 103, 104, 106–7, 108, 292 exports 33–6, 37, 39, 42, 125, 192, 277, 280, 282, 287, 295, 297 globalization and 64–9, 71, 75–81, 291–2 impact of WTO accession on 19, 21, 22, 26, 30, 35–6 and taxes 92 and ﬁnancial development 146–8 globalization and 65, 291–2 home market effect on 257–70 impact of WTO accession on 1, 17, 19, 21–2, 24, 26, 30, 35–6, 238–9, 251 imports 19, 21, 22, 30, 36, 257, 262, 291 globalization and 65, 67 impact of WTO accession on 19, 21, 22, 26, 30, 36 and taxes 92 and Taiwan 336–7, 339–40, 344, 346, 348–50 and taxes 91–2, 101, 103, 107, 108 see also commodity futures France 23, 24, 94, 96, 98, 98n13, 225, 228 Fujian 47, 55, 96, 108, 168 and migration 316, 317, 318, 322, 323, 325 and Taiwan 350 Gansu 47, 55, 93n10, 168, 192 and migration 316, 317, 323, 325 gas extraction industry 18, 93, 345 gas supply industry 10, 11, 18, 86 General Agreement on Tariffs and Trade (GATT), 1, 211 Germany 23, 24, 34, 35, 38, 94, 96, 98, 158, 225, 228

Index globalization 63–81, 138, 291–2 and economic relation with Taiwan 335–6, 337, 338–9, 342–8, 350 gold mining industry 27, 86 government, central 49, 155–6, 159–60, 162–70, 171–3, 174–8, 220–22 budgets 100, 111–17, 120–22, 130, 131, 186, 259 debt 65, 111–16, 120–22, 131–2, 148 expenditure 53, 98, 100, 101, 104, 106–7, 112–13, 117–22, 125, 130 intervention, regulation and policy 13, 24–6, 36, 49, 157, 166, 174, 177, 226, 234, 239, 273 and capital tax 95, 108 and ﬁnancial development 139, 181–2, 184, 186–8, 189–91, 193–202 and labor standards 292, 300 and migration 307–11, 328–9 and privatization 66, 74–5 Ministry of Agriculture 39, 187n4 Ministry of Commerce 163 Ministry of Finance 132, 134–6, 138n22, 165–6, 172, 187n3 Ministry of Foreign Trade and Economics Cooperation 39, 219–20 Ministry of Science and Technology 165–6, 167, 172 see also Communist Party of China (CPC), National People’s Congress, State Council government, local 2, 20, 25, 45–59, 155–6, 164, 165, 167–70, 172, 175, 226 budgets 122 debt 122, 131–2 expenditure 53 ﬁnancial intermediation development 182, 187, 189 intervention, regulation and policy 46, 49–51, 74, 187, 200, 262, 328–9 Great Britain 90n3, 94, 96, 98, 228–9, 232 Gross Domestic Product (GDP) 22, 26, 34, 52–4, 65, 67, 112–15, 125, 129–32, 136, 142, 175, 182, 186–7, 191–2, 194, 196, 198, 226, 262, 273, 275– 80, 284, 287–8, 310, 326, 328, 341 Taiwan 340, 344

353

Gross National Product (GNP) 274, 275, 276, 280, 284, 289 Guangdong 15, 46, 47, 55, 95, 96, 108, 169, 344 and migration 311, 316, 317, 318, 322, 323, 325, 326 Guangxi 47, 55, 96, 168 and migration 316, 317, 322, 323, 324, 325 Guangzhou 30, 39, 162, 168 Guizhou 47, 55, 93n10, 96, 168 and migration 316, 317, 322, 323, 325 Hainan 47, 55, 96, 168, 192n12 and migration 316, 317, 318, 322, 323, 325 Hebei 47, 55, 96, 168 and migration 316, 317, 318, 323, 325 Heilongjiang 47, 55, 96, 169 and migration 316, 317, 318, 323, 325 Henan 47, 55, 96, 168 and migration 316, 317, 318, 323, 325 high-tech industry 14, 23, 28, 257 and Taiwan 339–40, 341, 344, 346 and venture capital 155–9, 161, 164–9, 170, 171, 173, 176, 177–8 Hong Kong 15, 39, 84, 90n3, 92, 95, 98, 107, 108, 295, 336, 340, 341, 344, 348 household registration/responsibility system 189, 191–5, 274, 294–5, 300 and migration 307, 308–10, 311–12 housing 10, 13, 14 industry 14 Hubei 47, 55, 96, 168 and migration 316, 317, 318, 322, 323, 324, 325 Huijin Investment Corporation Ltd 134 hukou see household registration system Hunan 47, 55, 96, 169 and migration 316, 317, 318, 322, 323, 325 Hungary 94, 133 IDG 162–3, 169 India 94, 209–10, 211, 213, 221, 273 compared with China 273–84 Indonesia 94, 98, 143, 145, 146n30

354

The Chinese Economy after WTO Accession

Industrial and Commercial Bank of China (ICBC) 128n9, 133, 134, 135, 137, 186 industrial sector 9–16, 17, 18, 28, 29, 198, 200, 202, 277, 280, 287, 294, 304 and migration 309 inﬂation 113–14, 121, 136, 139, 141, 273, 274 Inner Mongolia/Neimonggu 47, 55, 168 and migration 316, 317, 323, 325 Innovation Fund for Small Technologybased Firms (IFSTF) 164–7, 172–3 insurance industry 18, 21, 23, 30 intellectual property 22, 24, 165, 225–6, 229 interest rates 89, 99–105, 107, 112–15, 121, 127, 131, 135, 138, 139, 141, 142–50, 156, 262 International Labor Organization (ILO) 291–2 iron and steel industry 10, 11, 12, 18, 86, 345 Italy 34, 35, 38, 98n13, 225 Japan 23, 34, 35, 37, 38, 90n3, 94, 185, 225, 228, 335, 336, 337, 340, 348 and anti-dumping 213, 219–20 and foreign investment to China 96, 98, 108 and home market effect 258, 263, 267 Jialefu 24 Jiangsu 15, 47, 49, 55, 96, 108, 167, 168, 169, 187n4 and migration 316, 317, 318, 322, 323, 325 Jiangxi 47, 55, 96, 168 and migration 316, 317, 318, 322, 323, 324, 325 Jinlin 47, 55, 96, 168, 182 and migration 316, 317, 318, 323, 325 Korea 36, 94, 98, 137, 142, 143, 145, 146n30, 185, 336 Kunshan 15 labor and employment 35, 71, 74, 79, 176, 215, 222 and capital tax 89, 90–91, 98–101, 104, 107 and home market effect 257, 259–62, 263–5

and migration 307–12, 319, 320, 321, 324, 326–9 and private sector 45, 46–8, 52–3, 54 productivity 2, 69 rural 20, 22, 26, 28, 31, 53, 258, 263, 293–301, 303–5, 307–10, 311–13 standards 291–301 and Taiwan 336, 337, 338, 339, 340, 341, 342, 344–5, 346, 347, 348–9 unemployment 2, 22, 28, 51, 311 urban 18, 26, 31, 53, 293–301, 303–5, 307–13, 328 Latin America 39, 142 Liaoning 47, 55, 96, 168 and migration 316, 317, 318, 323, 325 Macau 84, 90n3, 92, 96, 98, 108, 295, 336, 340, 344 machine-building industry 9, 10, 11, 86, 214, 266 and Taiwan 343, 344, 345 Maidelong 24 Malaysia 94, 137n20, 142, 143, 145, 146n30 manufacturing sector 20, 21, 22–3, 27, 30, 69, 162, 213, 214, 257–70, 294, 299, 305 and migration 319 and Taiwan 340, 341, 342–9 metals and metal products industries 19, 162, 213, 214, 265, 266, 267, 345 see also iron and steel industry, nonferrous metals industry Mexico 39, 94, 215, 221 migration inter-provincial 308, 313–14, 316–17, 318, 321, 323, 325 intra-provincial 313, 317–18, 321, 325 inter-regional 307–8, 309, 310–11, 313–17, 319–20, 322–9, 332–4 rural to urban 14, 200, 293–301, 307–8, 309–10, 311–12 urban to rural 312–13 money market 141, 144, 147, 149 National People’s Congress 24, 49, 93 National Planning Commission of China (NPCC) 187n3 Neimonggu see Inner Mongolia New Zealand 94, 98, 142, 143, 145

Index Ningxia 47, 55, 93n10, 96, 168 and migration 311, 316, 317, 322, 323, 325 non-ferrous metals industry 11, 86 non-metal minerals industry 19, 86, 214, 265, 266, 267, 345 non-performing loans (NPLs) 97, 111, 126–30, 131, 132–6, 138–9, 143, 148, 149 North Korea 81 oil and petrochemicals industry 11, 18, 86, 93, 345 paper and papermaking industries 10, 11, 86, 214, 265, 266 and Taiwan 345, 349 Pearl River Delta 15 People’s Bank of China (PBOC) 134–9, 145–6, 184n1 pharmaceutical industry 10, 11, 86, 225–34 Philippines 94, 98, 137n20, 146n30 plastics and rubber industry 19, 86, 213, 214, 265, 266 and Taiwan 345, 349 Poland 50, 94 power industry 10, 11, 12, 18, 86, 95 and Taiwan 344, 345 precious metals industry 27, 86, 93 private sector 25–7, 31, 33–42, 45–59, 91, 310 privatization 21, 45, 49, 63–81, 127, 132, 142 public expenditure, see government public sector 18, 20, 21, 25–6, 27, 31, 65 see also state owned enterprises (SOE) Qinghai/Qinhai 47, 55, 93n10, 95, 96, 168 and migration 311, 313, 316, 317, 323, 325 Qiu Jibao 33, 36 quotas 20, 28, 31, 35, 218 Rao, Narasimha 273 real estate industry 10, 12, 18, 189–90 retail sector 20, 23, 23, 162, 305 Rural Credit Cooperatives (RRCs) 129, 145, 184, 187, 189n7, 190 rural sector 31 labor and employment, see labor

355

migration, see migration versus urban sector, see urban sector Russia 50, 51–2, 94, 98, 282–3 SARS 9 savings rate 97, 125 service sector 1, 15, 17, 18, 20, 23–4, 27, 29, 30, 69, 162, 198, 200, 202, 284, 294 and migration 312, 319, 322, 326, 328, 329 and Taiwan 346, 348, 349 Shanxi 47, 55, 96, 168 and migration 316, 317, 318, 323, 325 Shaanxi 47, 55, 93n10, 96, and migration 316, 317, 323, 325 Shandong 46, 47, 49, 55, 96, 168 and migration 316, 317, 318, 323, 325 Shanghai 15, 22, 23, 26, 30, 46, 47, 54, 55, 95, 96, 108, 161, 168, 192, 200, 344 and migration 309, 312, 316, 317, 318, 319, 322, 323, 325, 326 Shanghai Pudong Development Bank 23–4 Shantou 310 Shenzhen 15, 161, 163, 310 shipbuilding industry 20 Shunde 46 Sichuan 47, 55, 93n10, 96, 168 and migration 316, 317, 322, 323, 324, 325, 327 Singapore 90n3, 92, 94, 96, 98 Slovakia 50 South Africa 210, 211, 213, 221 South Korea 209, 212, 213, 216, 219, 282–3 Southeast Asia 39, 336 Soviet Union 273 special economic zones (SEZs) 81, 93, 95, 108, 274, 310, 349 State Administration of Foreign Exchange (SAFE) 134 State Council 24, 91, 93, 129, 134, 139, 164, 176 state owned commercial banks (SOCBs) 111, 127–30, 133–5, 138–9, 145, 184, 186, 187, 190, 191 state owned enterprises (SOEs) 2, 45, 84, 91, 264, 274, 295, 299, 303 government ﬁnancial support of 173, 184, 186–7, 191, 194, 198–200

356

The Chinese Economy after WTO Accession

impact of WTO accession on 20, 21, 25–6, 27, 30 and local government 50, 51, 74 and non-performing loans 97, 111, 126, 127, 135 and the private sector 48, 64–6 privatization of 46, 49, 64–6, 70, 71, 74, 76n10, 78, 80, 127, 136, 141–2 reform of 122, 126, 127, 135, 136, 141–2, 274, 311 state sector, see public sector Suzhou 25 Sweden 132 Taiwan 1, 36, 84, 96, 98, 216, 295 compared with China 338–9, 343–6 economic relations with China 335–50 Taizhou 33, 36, 38 tariffs 21, 28, 31, 33, 35–6, 212, 217–18, 336 tax 26, 36, 53, 66, 121, 130, 139, 175 business (BT) 92 capital 89–108 incentives 90–91, 93, 95, 108, 310 revenue from 89–90, 94–6, 100, 104, 107, 112–13, 117–22 value-added (VAT) 92, 95 telecommunications industry 2, 14, 15, 18, 21, 24, 25, 28, 30, 86, 161, 162 textile and apparels industry 10, 11, 19, 20, 22, 28, 33, 35, 42, 86, 162, 213, 214, 215, 266, 292, 310 and Taiwan 343, 345, 349 Thailand 94, 137n20, 219 Tiananmen Square event 48 Tianjin 47, 55, 95, 96, 168, 192, 200 and migration 312, 315, 317, 318, 319, 322, 323, 325 Tianjin Tiangong Sewing Machine Company 35 Tibet 93n10, 95, 96, 192n12 and migration 311, 312, 316, 317, 322, 323, 325 timber and lumber industry 18, 86, 214 tobacco industry 10, 11, 18, 86, 214, 265, 266, 267 township and village enterprises (TVEs) 45, 48, 49, 184, 187, 190, 200–201, 295, 310

trade barriers and restrictions 1, 21, 23, 27, 147, 150, 336 and anti-dumping 211–12, 215, 217, 218 transportation 13, 18, 90, 95 Turkey 135n18 Ukraine 50, 94 unemployment, see labor United Kingdom (UK), see Great Britain United States of America (USA) 23, 24, 28, 30, 36, 37, 39–40, 81, 94, 95, 112, 135, 139, 185, 310, 335, 337, 340, 348 and anti-dumping 209, 211, 212–20 and capital tax 90, 92 and foreign investment to China 96, 98, 108 and home market effect 257–70 pharmaceutical industry 225, 228–33 and venture capital 155, 157, 158, 170–72, 173, 175, 177 wheat futures market in 237–52 Urban Cooperative Bank 27 Urban Credit Union 27 urban sector 31 labor and employment, see labor migration, see migration population 14 versus rural sector 31, 181–202, 309 urbanization 14, 15–16, 31, 192n11, 263 venture capital foreign 162–3, 176 private 155–63, 164, 169, 176–7 public 155–63, 164–70, 175–7 public-private partnerships 160, 170–74, 177–8 Vietnam 50, 51 wage rates 89, 99, 105, 108, 259–60, 270, 274, 291–301, 303–5 Walmart 24, 41 water supply industry 10, 11, 18, 22, 86 Wenzhou 51 World Economic Forum 282, 284 World Health Organization (WHO) 234

Index World Trade Organization (WTO) 1, 17, 292, 297, 350 and anti-dumping 209–16, 220–22 impact of Chinese accession to 1–2, 13, 17–31, 33–4, 35–6, 37–8, 40, 42, 67, 328 and anti-dumping 209, 217–18, 219, 222–3 on economic relations with Taiwan 335, 336–7, 341 on ﬁnancial sector 126, 128, 146–7, 148, 150 on wheat futures market 238, 239, 251 Wuhan 23, 162, 168 Xi’an 162 Xiamen 310

357

Xiaoping, Deng 12, 48, 63, 65, 67, 95 Xinjiang 47, 55, 93n10, 96, 168 and migration 311, 316, 317, 318, 322, 323, 325, 326 Yunan 47, 55, 93n10, 96, 168 and migration 311, 316, 317, 318, 323, 325 Zedong, Mao 64–5, 67 Zhejiang Jili 27 Zhejiang 33, 36, 39, 47, 49, 51, 55, 96, 168, 173, 192 and migration 316, 317, 318, 322, 323, 325 Zhu Rongji 36 Zhucheng City 46 Zhuhai 310

The Chinese Economy After WTO Accession (The Chinese Economy Series)

Chinese Economy in the 1990s (Studies in the Chinese Economy)

China's Rural Economy after WTO: Problems And Strategies (The Chinese Economy Series)

Chinese Economy in the 1990s (Studies in the Chinese Economy)

The Globalization of the Chinese Economy

Changing workplace relations in the Chinese economy (Studies on the Chinese Economy)

The Chinese Economy: Transitions and Growth

The Chinese economy: fighting inflation, deepening reforms

Changing Workplace Relations in the Chinese Economy

The Chinese Worker after Socialism

The rise of the Chinese economy: the middle kingdom emerges

China's Integration With the Global Economy: WTO Accession, Foreign Direct Investment and International Trade (Advances in Chinese Economic Studies)

The Dynamics of China's Rejuvenation (Studies on the Chinese Economy)

The dragon millennium: Chinese business in the coming world economy

The Dragon Millenium: Chinese Business in the Coming World Economy

The Chinese in Britain, 1800-present: Economy, Transnationalism, Identity

GLOBAL BIG BUSINESS AND THE CHINESE BREWING INDUSTRY (Routledge Studies on the Chinese Economy)

The Chinese Banking Industry: Lessons from history for today's challenges (Routledge Studies on the Chinese Economy)

Chinese Citizenship: Views From the Margins (Routledgecurzon Studies on the Chinese Economy)

The Chinese Communist Party in Reform (Routledge Studies in the Chinese Economy.)

China in the World Economy (Routledge Studies on the Chinese Economy)

Restructuring the Chinese City: Changing Society, Economy and Space

International Joint Ventures in China (Studies on the Chinese Economy)

China's Post-Reform Economy - Achieving Harmony, Sustaining Growth (Routledge Studies on the Chinese Economy)

Chinese Coal Industry - An Economic History (Routledgecurzon Studies on the Chinese Economy, 2)

The Indian Economy Sixty Years after Independence

Ethics and Economy: After Levinas

Story of the Chinese Zodiac: English Chinese

The Chinese Century: The Rising Chinese Economy and Its Impact on the Global Economy, the Balance of Power, and Your Job

economy

Economy

The Chinese Economy After WTO Accession (The Chinese Economy Series)

Chinese Economy in the 1990s (Studies in the Chinese Economy)

China's Rural Economy after WTO: Problems And Strategies (The Chinese Economy Series)

Chinese Economy in the 1990s (Studies in the Chinese Economy)

The Globalization of the Chinese Economy

Changing workplace relations in the Chinese economy (Studies on the Chinese Economy)

The Chinese Economy: Transitions and Growth

The Chinese economy: fighting inflation, deepening reforms

Changing Workplace Relations in the Chinese Economy

The Chinese Worker after Socialism

The rise of the Chinese economy: the middle kingdom emerges

China's Integration With the Global Economy: WTO Accession, Foreign Direct Investment and International Trade (Advances in Chinese Economic Studies)

The Dynamics of China's Rejuvenation (Studies on the Chinese Economy)

The dragon millennium: Chinese business in the coming world economy

The Dragon Millenium: Chinese Business in the Coming World Economy

The Chinese in Britain, 1800-present: Economy, Transnationalism, Identity

GLOBAL BIG BUSINESS AND THE CHINESE BREWING INDUSTRY (Routledge Studies on the Chinese Economy)

The Chinese Banking Industry: Lessons from history for today's challenges (Routledge Studies on the Chinese Economy)

Chinese Citizenship: Views From the Margins (Routledgecurzon Studies on the Chinese Economy)

The Chinese Communist Party in Reform (Routledge Studies in the Chinese Economy.)

China in the World Economy (Routledge Studies on the Chinese Economy)

Restructuring the Chinese City: Changing Society, Economy and Space

International Joint Ventures in China (Studies on the Chinese Economy)

China's Post-Reform Economy - Achieving Harmony, Sustaining Growth (Routledge Studies on the Chinese Economy)

Chinese Coal Industry - An Economic History (Routledgecurzon Studies on the Chinese Economy, 2)

The Indian Economy Sixty Years after Independence

Ethics and Economy: After Levinas

Story of the Chinese Zodiac: English Chinese

The Chinese Century: The Rising Chinese Economy and Its Impact on the Global Economy, the Balance of Power, and Your Job

economy

Economy

Recommend Documents