Enterprise reform in China : ownership, transition, and performance

Enterprise Heform in China Ownership, Transition, and Performance A World Batik Researpch PTbticlantitor  Enterprise Reform in China  Enterprise Reform in China Ownership, Transition, and Perfrmance Edited by Gary H. Jefferson and Inderjit Singh Published for The World Bank Oxford University Press  Oxford University Press OXFORD NEW YORK ATHENS AUCKLAND BANGKOK BOGOTA BUENOS AIRES CALCUTTA CAPE TOWN CHENNAI DAR ES SALAAM DELHI FLORENCE HONG KONG ISTANBUL KARACHI KUALA LUMPUR MADRID MELBOURNE MEXICO CITY MUMBAI NAIROBI PARIS SAO PAULO SINGAPORE TAIPEI TOKYO TORONTO WARSAW and associated companies in BERLIN IBADAN C 1999 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Published by Oxford University Press, Inc. 198 Madison Avenue, New York, N.Y. 10016 Oxford is a registered trademark of Oxford University Press. 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, photocopy- ing, recording, or otherwise, without the prior permission of Oxford University Press. Manufactured in the United States of America First Printing December 1998 The findings, interpretations, and conclusions expressed in this study are entirely those of the author and should not be attributed in any manner to tlhe World Bank, to its affili- ated organizations, or to members of its Board of Executive Directors or the countries they represent. Cover: The Chinese characters that appear on the cover and that open each chapter (pro- nounced chigai) mean enterprise reform. Library of Congress Cataloging-in-Puiblication Data Enterprise reform in China : ownership, transition, and performance / edited by Gary H. Jefferson, Inderjit Singh. p. cm. Includes bibliographical references and index. ISBN 0-19-521120-0 1. Industries-China-1976- . 2. Industrial productivity-China. 3. Government ownership-China. 4. Government business enterprises- Government policy-China. I. Jeffereson, Gary H. II. Singh, Inderjit, 1941- HC427.92.E58 1998 338.0951-ddc2l 97-36569 CIP  Contents Acknowledgments vii Contributors xi Overview 1 Gary H.Jefferson and Inderjit Singh 1. Ownership Change in Chinese Industry 23 Gary H.Jefferson and Thomas G. Rawski 2. Structure, Authority, and Incentives in Chinese Industry 43 Gary H.Jefferson, Zhang Ping, andJohn Z. Q. Zhao 3. China's Industrial Innovation Ladder: A Model of Endogenous Reform 65 Gary H. Jefferson and Thomas G. Rawski 4. Innovation and Reform in China's Industrial Enterprises 89 Gary H.Jefferson, Thomas G. Rawski, and Zheng Yuxin 5. Reforming Property Rights in China's Industry 107 Gary H.Jefferson, Lu Mai, and John Z. Q. Zhao 6. China's Industrial Performance: A Review of Recent Findings 127 Gary H.Jefferson, Inderjit Singh, XingJunling, and Zhang Shouqing 7. Are China's Rural Enterprises Outperforming State Enterprises? Estimating the Pure Ownership Effect 153 Gary H.Jefferson v  8. Wage and-Employment Behavior in Chinese Industry 171 Gary H.Jefferson, Inderjit Singh, Albert G. Z. Hu, and Wang Benzhou 9. Employment and Wages in Township,Village, and Other Rural Enterprises 197 Mark M. Pitt and Louis Putterman 10. Industrial Investment, Finance, and Enterprise Performance in Chinese Industry 217 Gary H.Jefferson, Albert G. Z. Hu, and Indedit Singh 11. Export Performance and Enterprise Reform in China's Coastal Provinces 241 Frances C. Perkins Appendixes A. A Model of Economic Reform 265 Gary H.Jefferson and Thomas G. Rawski B. Development of the Hengdian Township Enterprise Group:A Case Study 279 Chen Jianbo and Gary H.Jefferson Index 289 VI CONTENTS  Acknowldgments We can only attempt to thank the many people who contributed to the preparation of this volume. The list is long, in part because this pro- ject, which began nearly a decade ago, entailed the collection and analy- sis of multiple sets of survey data and withstood several administrative reorganizations within both China and the World Bank. Chief among the contributors to this volume are those who prepared papers for the book and who exhibited the enormous patience along the way required to see their work through an extended process of several reviews and revisions.Their names and institutional affiliations appear on the contributors list. Many other individuals and their organizations have substantially helped in the research that went into this volume. This project original- ly began as a joint venture among the World Bank, the Institute of Economics in the Chinese Academy of Social Sciences (CASS), the State Economic System Reform Commission, and the Rural Development Research Center. Along the way the project expanded to include numerous other collaborating organizations. Specifically, we wish to thank: * The Institute of Economics (CASS) and its Director Lin Qingsong and staff Zhang Ping and Zhang Xuejun * The Institute for Technical and Quantitative Economic Research (CASS), including its director Li Jingwen and our colleague, Zheng Yuxin, and his staff * The State Economic System Reform Commission including Peng Zhaoping * The Rural Development Research Center including Du Ying, Qiu Jicheng, and Luo Xiaopeng; * The Development Research Center of the State Council, including Chen Jianbo and Li Guodu Vil  * The State Statistical Bureau including Lu Chunheng, Li Qiming, Xing Junling, Zhang Shouqing, and former staff member, Liu Li * The Ministry of Metallurgy, including Zhang Xinchuan and Xue Chuanzhao. In the World Bank we are grateful to Mr. Javid Burki, former director of the China Department, for his early and continuous support for the project, which was initiated in his department; ShahidYusuf, former chief economist of the department; and Richard Stern, former director of the department's Industry and Energy Division. Chen Kang, Xiao Geng, Zheng Decheng, and Dilip Ratha provided enthusiastic and capable research support. We also appreciate the generous and able administrative support of Cecilia Guido-Spano andVesna Petrovic. The efforts of those at the publishing stage who helped to bring this volume to publication are deeply appreciated.These include Nancy Levine of the World Bank and Meta deCoquereaumont, Barbara Karni, Daphne Levitas, Glenn McGrath, and Donna McGreevy of Communications Development Inc. Several anonymous inside and outside reviewers con- tributed invaluable criticisms and suggestions that made this volume far better than it appeared in its initial incarnations. The overall quality of the book also benefited gready from persons who prepared working papers or offered review and comment on earlier man- uscripts.These include VK. Chetty, Richard H. Day, Zvi Griliches,Thomas Rawski, Wang Zhigang, Calla Wiemer, and Zou Gang. We also appreciate the extensive and conscientious research assistance of Xu Wenyi, Wang Benzhou, John Zhao, Albert Hu, and Bai Huamao, all currently or formerly at Brandeis University. Also, Claire Cincotta and Mary Smith in the Department of Economics at Brandeis helped in innumerable ways to support the logistics of this project; they specifical- ly helped to arrange the visits of numerous Chinese collaborators to the States and to Brandeis. We would like to thank Alan Gelb, former chief of the Transition Economics Division of the World Bank, who not only provided valuable guidance, but also participated in the research and was a constant source of patient support and encouragement for research in a comparatively new area of transition economics. The division allowed the project to benefit from the wealth of intellectual interaction with other colleagues and scholars passing through it. Research by others in the division also provided us with a useful and comparative frame of reference from other country experiences and continuously challenged us to show how China's experience was different and why it mattered. In addition to the financial support of the World Bank, we deeply appreciate support from the Henry Luce Foundation, the American Vill ACKNOWLEDGEMENTS  Council of Learned Societies, the Chiang Ching-Guo Foundation, and the Brandeis Mazer Fund. Finally, and most important, we would like to thank our families for their encouragement, patience, and support over the many years when it seemed that our work would never finish. ACKNOWLEDGMENTS IX   Con ributors The affiliations shown are those pertaining at the time of writing. Chen Jianbo Rural Development Research Center, State Council, China Albert G. Z. Hu Department of Economics, Brandeis University Gary H.Jefferson Department of Economics, Brandeis University Lu Mai Harvard University Institute for International Development and Developmenit Research Center, State Council Frances C. Perkins East Asia Analytical Unit, Department of Foreign Affairs and Trade, Canberra Australia Mark M. Pitt Department of Economics, Brown University Louis Putterman Department of Economics, Brown University Thomas G. Rawski Department of Economics, University of Pittsburgh Inderjit Singh Transition Economics Division,World Bank Wang Benzhou China Everbright Financial Holdings, LTD, Hong Kong Xing Junling State Statistics Bureau, China Zhang Ping Economic Research Institute, Chinese Academy of Social Sciences Zhang Shouqing State Statistics Bureau, China John Z. Q. Zhao Department of Economics, Brandeis University ZhengYuxin Institute of Quantitative and Technical Economics, Chinese Academy of Social Sciences xi  Gary H. Jeffer,n ay dnderjit Singh The transformation and growth of China's economy, which raised living standards for one-fifth of the world's population and provided indirect benefits for many more, stands among the most significant economic developments of the late twentieth century. This book rests on two premises. The first is that the key to China's dramatic economic trans- formation is that country's industrialization. The second is that the real story of China's industrialization is unfolding at the level of the individ- ual Chinese enterprise and factory. This volume seeks to document the impact of economic reforms on China's industrial sector and to explain why China's reforms, which appear meager relative to the more ambitious reform programs of Eastern Europe and even Russia, have had such wide-reaching effects. This book differs from much of the research and analysis of China's economy in several ways: * It provides the first in-depth comparative look at different ownership systems, going well beyond the standard state/nonstate contrast to illustrate the real organizational, behavioral, and performance differ- ences among China's systems of ownership. * It analyzes the extent to which specific reforms have altered the behavior and performance of Chinese enterprises across different forms of ownership. * It examines how the millions of enterprises in China interact with their institutional and international environments to create a dynam- ic reform process that has greatly magnified the impact of reform ini- tiatives from the center. * It attempts to distinguish, evaluate, and reconcile various perspectives or schools of thought regarding the nature of China's reform. The story of the industrial enterprise, the most basic unit of organized production, affords the deepest and most comprehensive insight into the 1  progress and problems of China's economic transition. Through its con- nection to other parts of the economy, the enterprise provides a window through which the reform of China's labor markets, banking system, social insurance system, systems of foreign trade and investment, and legal and regulatory system can be viewed and understood. Because the enterprise stands at the nexus of so many other economic institutions, reform of the industrial enterprise is synonymous with China's economic reform. This book is the outgrowth of aWorld Bank project that began in the late 1980s.That project commissioned two large-scale enterprise surveys. One survey covers 1,200 state-owned enterprises and collective-owned enterprises in the "urban system"; the other includes 300 township and village enterprises. Both surveys span the period from the early 1980s through 1991-92 and include detailed information about the manage- ment systems of the participating enterprises. A third data set, produced by China's State Statistical Bureau, includes about 5,000 of China's large and medium-size enterprises for the period 1988-92, nearly one-third of the total during that period. For one year, 1993, the data include the entire population of 18,500 large and medium-size enterprises, spanning 14 different ownership types.This State Statistical Bureau data set enables us to extend our analysis to the important foreign sector, which contains both foreign-funded and overseas Chinese-funded enterprises, some- times collectively referred to as foreign-invested enterprises. Two addi- tional surveys were designed to investigate the dynamics of technical innovation and of foreign trade in China's industrial enterprises. One or more of these five data sets provides the empirical foundation for most chapters in this book. They also provide a rich source of statistical mate- rial that is being made available to a wide range of scholars. These survey data, which capture the experience of more than 7,000 enterprises, create a record from which researchers can document the ways in which China's enterprises, both old and new, have responded to changes in their institutional and policy environment. The statistical records of these enterprises document the experience of each individual enterprise and allow us to construct a broad picture of the program, process, and out- come of China's overall reform experience. While the data upon which most chapters in this book are based will be "antiquated" before long, these data and the analysis that follows provide a critical record of a period that is likely to define the Chinese economy for decades to come. In addition to this overview, the book contains 11 coordinated stud- ies that document and analyze changes in the governance and perfor- mance of China's industrial enterprises. Chapter 1 presents an overview of the structure of China's industrial economy. Chapter 2 extracts key aspects of the survey data to describe, compare, and contrast the systems 2 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  of governance of key ownership types. Chapters 3 through 5 should be read together. Chapter 3 describes the special features of China's eco- nomic and institutional structure that create a kind of dynamic endoge- nous reform process, characterized by the authors as an "industrial innovation ladder."The ladder has two interactive dimensions.The first is technical innovation, which is modeled and empirically tested in chap- ter 4. The second is institutional change, or enterprise reform, which is viewed in chapter 5 as arising from both conditions within the industri- al system and the efforts of central reformers. Chapters 6 and 7 assess the performance of Chinese industry-chapter 6 through a general review of the literature on Chinese enterprise reform and chapter 7 by com- paring the performance of state-owned enterprises and township and village enterprises. By controlling for policy and firm-specific differ- ences, chapter 7 seeks to identify the "pure ownership" effect on pro- ductivity. Like chapter 7, chapters 8 through 11 present formal models and tech- nical econometric work. Each of these chapters seeks to examine the behavior of one or more samples of enterprises with respect to a particu- lar enterprise function. Chapter 8 focuses on the comparative wage-setting and employment behavior of state-owned enterprises, urban cooperatives, township and village enterprises, and foreign-invested enterprises. Chapter 9 targets township and village enterprises and some private rural enter- prises to investigate whether enterprises in this important and poorly understood sector overemploy, underemploy, or choose more-or-less prof- it-maximizing levels of employment. Examining the investment process, chapter 10 asks a key question, perhaps the most important for any transi- tion economy: are the most successful enterprises capturing scarce invest- ment resources in order to expand their share of production? Chapter 11 examines the impact of various open door policies on export performance among state-owned enterprises and foreign-invested enterprises. For the reader focused on the "idea" of China's economic reform, three "high-concept" essays in this volume may be of particular interest. Appendix A develops the analogy between the optimal reform strategy and the investment decision of the firm, looking at reform from the per- spective of the social welfare-maximizing leader or reformer. Chapter 3 looks at China's industrial economy as a system. It extends the product cycle and quality ladders of Vernon (1966) and Grossman and Helpman (1991) to China's domestic industrial economy to portray a dynamic process of institutional change as well as technical innovation. The sec- tion in chapter 5 entided "Implications: the commons, coase, and a prop- erty rights market" extends the notion of a public good and social externalities to state enterprises and argues for the application of the OVERVIEW 3  logic of the Coase theorem to the reform of Chinese industry. This pre- scriptive material identifies the establishment of a property rights market as the central policy priority implied by the two other essays. Highlights of China's reforms Volumes have been written describing and characterizing China's eco- nomic reform program.' This account cannot substitute for these excel- lent descriptions of China's reform program. The brief discussion below is intended simply to orient the less familiar reader to key elements of China's reform program and to alert the more knowledgeable reader to those initiatives that constitute the essential ingredients of China's dynamic reform process. Much of the impetus for China's industrial transformation originated with two highly successfiil reform initiatives-the Open Door policy and the rural reform program-the outline of which became clear in the late 1970s and early 1980s.The massive impact of these complementary ini- tiatives was not anticipated by China's early reformers. The Open Door policy opened China's coastal provinces to foreign trade, investment, and an influx of new technologies, managerial initiative, and foreign exchange. The rural reform program led to immediate productivity gains in rural agriculture that, in turn, vastly expanded the supply of domestic savings and surplus labor while unleashing individual and local govern- ment entrepreneurial initiative. Together the Open Door policy and the rural reform program created a surge of new industrial enterprise fornia- tion, intense competition within many product groups, and pressures for both technological and institutional change.These conditions have forced China's centrally controlled trading system, its institutions of factor allo- cation, and its industrial enterprise system to adapt continuously to the requirements of a modern market-oriented economy. The Open Door policy and the rural reform program created a fertile environment that allowed for the formation and entry of new capital, labor, managerial initiative, and firms, thereby creating effective competi- tion throughout most of Chinese industry.Two other reforms-the devo- lution of managerial autonomy to the firm, and the two-track price system-created the necessary incentive structure. Allowing firms to retain profits and provide bonuses established a powerfuil, if uneven, set of incen- tives for workers and managers to expand and accumulate financial sur- pluses. By exposing decisions at the margin to market forces, the dual-track price system served two functions. First, it enabled smaller new firms to secure the necessary market access and producer goods from established firms to enter the market and grow. In addition, it afforded firms already 4 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  operating in the state sector the product and market space into which to expand and restructure their operations in search of financial gain. A critical prerequisite for the efficacy of these meta- and microinitia- tives has been the creation of a comparatively stable macroeconomic environment, which Chinese authorities have maintained through a mix of direct and indirect instruments. Recollection of the high political and economic costs of rising inflation during 1989 are likely to motivate China's political leaders and policymakers to sustain macroeconomic sta- bility into the foreseeable future. China's industrial performance China's industrial economy created $340 billion of value added in 1995, about a fifth of the U.S. figure for that year (World Bank 1997). By other measures, however, the size of China's industrial economy overwhelms that of every other nation. By the mid-i 990s more than 7 million indus- trial enterprises in China employed more than 140 million workers, up from about 50 million in 1978 and nearly equal to the combined indus- trial work forces of the 28 OECD economies. China's economy is adding more industrial production each year than any other economy in the world, including the United States. Most of the average annual industrial growth of 14 percent during the 15-year period 1980-95 was powered by three factors: the growth in the indus- trial work force by 90 million workers, two-thirds of whom have moved from China's agricultural sector; a doubling of the industrial sector's fixed assets every four to five years, fueled in large part by the nation's domes- tic savings rate, which rose from 25 percent in 1965 to 35 percent by 1980 and now stands at more than 40 percent; and overall industrial pro- ductivity that has grown by 3-4 percent a year (chapter 6, this volume; World Bank 1997). While growth in China's state-owned industry has been rapid by world standards, averaging nearly 8 percent between 1980 and 1995, two emerging sectors have served as the principal engines of China's indus- trial growth and transformation. The first is the rural township and vil- lage sector, which reported annual real rates of growth of nearly 22 percent over this 15-year period, increasing its share in industrial output to more than one-quarter by 1995 (see table 1.3).While township and village enterprises served as the engine of industrial growth during the 1980s, the so-called "other" sector, comprising individual enterprises, joint stock companies, domestic joint ventures, and foreign-invested enterprises (both joint and wholly owned ventures), emerged as the principal source of growth in the 1990s. OVERVIEW S  The variety of ownership forms in this "other" sector confounds the conventional public-private dichotomy.While much of the original cap- ital in this sector was publicly owned (most notably the state and collec- tive enterprises that have been converted to joint ventures orjoint stock companies), most of the conversions themselves were made possible by injections of private capital, both domestic and foreign. More than 90 percent of enterprises in the sector are small, wholly privately owned firms classified as "individual" or "private." By the mid-1990s China's industrial structure could be broadly char- acterized as one-third state-owned, one-third collective- and township and village-owned, and one-third joint public-private or wholly pri- vately owned. This burgeoning growth of township and village enter- prises, foreign-invested enterprises, and public-private enterprises, not the privatization of public enterprises, has accounted for the greatest source of change in China's industrial ownership structure. Chinese industry: what needs to be explained? The extraordinary performance of China's industrial economy has been widely acknowledged. There is not, however, unanimity concerning the causes and sustainability of this growth. Many observers are puzzled by three paradoxes: * Relative to other transition economies and to the general notion shared by most economists of what constitutes an effective set of market insti- tutions, China's industrial reforms have been partial and piecemeal. As a result key institutions, including bankruptcy law, the banking system, and the social insurance system, remain undeveloped. Nevertheless, China's industrial growth has been robust and appears likely to continue. * China's economy has been dominated by publicly owned firms, both state- and collective-owned. Although these factories display many of the negative characteristics that economists ascribe to public owner- ship, they accounted for most of the extraordinary growth of Chinese industry during the critical early years of transition. * Most literature on productivity in Chinese industry, principally in state industry, reports that productivity is growing.Yet profitability in state industry is declining, and the number of loss-making enterpris- es is on the rise. One approach to resolving these paradoxes is simply to dismiss them. Sachs andWoo (1997), for example, contend that rising productivity in state industry is illusory and that China's publicly owned enterprises, including its township and village enterprises, are inherendy flawed and destined to fail- 6 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  sooner rather than later. They believe that China's robust economic growth is unsustainable without rapid, comprehensive, and immediate reform. Another approach to resolving these paradoxes is suggested by the proponents of gradual reform. Nolan (1995), for example, argues that incremental reform creates the conditions for effective transition, where- as a "big bang" destroys the institutions needed for transition with growth. The improving performance measures of China's state and col- lective enterprises should not be surprising, he maintains, since "a wide variety of property rights regimes can now be seen to be compatible with effective economic performance" (1995, p. 317). Others, including McKinnon (1994) and Murphy, Schleifer, andVishny (1992) offer "sec- ond best" reform models in which the relaxation of certain government controls leads to instability in the context of incomplete liberalization. Our view is that these paradoxes do exist: China's economy has per- formed impressively despite inadequately reformed economic institu- tions, the intrinsic weaknesses of public ownership, and extensive loss-making in China's state industry. A central purpose of this volume is to clarify and resolve each of these three paradoxes. "hat are the major findings? The conclusions put forth in this volume are organized around several basic questions: * What are the key features of China's economic reform process? What is the motive for continuing and widening the reform process? Why has such limited reform initiative from the center had so much effect? How do we resolve the paradox of rapid growth with limited reform? * What changes in governance and the institutional environment are most responsible for changes in enterprise conduct and performance? What is the role of ownership in accounting for differences in con- duct and performance among enterprise types? Can we account for the paradox of high-performance publicly owned enterprises? v How well are China's industrial enterprises performing? Has the per- formance of state-owned enterprises improved? How can the paradox of rising productivity and falling profitability throughout Chinese industry, most notably in the state sector, be resolved? * In what ways has the behavior of the Chinese firm been altered with respect to employment and wage-setting behavior, investment activi- ty, innovation, and export orientation? Are enterprises increasing effi- ciency by responding to incentives and relative price changes? The key findings of the volume are summarized below. OVERVIEW 7  Growth and reform Reformti strategy. Uncertainty over its vision of the future and aversion to risk help explain China's initial "groping" reform strategy. Advantageous initial conditions and fortuitous early reform initiatives have facilitated China's transition, and success has sustained the continuity of a gradual or evolutionary approach to reform. Success has many fathers. Not surprisingly, numerous interpretations of China's successful experience have been offered. Emphasizing the impor- tance of initial conditions, Sachs andWoo (1997) argue that the dominance of an agricultural sector comparatively free of subsidies and social services allowed China's underdeveloped economy to respond to the meager, inco- herent reform agenda more as a developing country than as an industrial socialist economy encumbered by the weight of ubiquitous subsidies and useless fixed assets. Jefferson and Rawski (chapter 3 and appendix A) also emphasize the importance of China's prereform endowment (that is, its initial conditions) that has allowed for the emergence of intense competi- tion and powerful incentives for managers and local governments. These conditions, which are built into the system, are, in turn, creating a dynam- ic process of endogenous reform. For Groves, Hong, McMillan, and Naughton (1994), the creation of effective incentives, particularly manage- rial incentives, explains China's ability "to grow out of the plan." Although these interpretations may have different policy implications, as explanations of China's rapid industrial transformation none is incon- sistent with the others.We combine these interpretations to create a syn- thesis of explanations that points to initial conditions, competition, and incentives. All of these conditions have enabled limited reform to have substantial impact; all also create pressures for new rounds of reform. Naughton (1994) observes that there is a "generally consistent logic to the way [planned economies] dissolve" (p. 306).The view that emerges from this volume is that within a competitive economic setting, there is also a consistent logic to the process by which China is replacing its planned economy with a market economy. A more rapid reform scenario may have been possible, but it would have exacted additional costs and risks. Day,Wang, and Zou (1994) devel- op several policy reform scenarios based on a dynamic adaptive model of enterprise behavior. They conclude that under a rapid reform sce- nario, in which prices were freed more rapidly and greater financial resources were redirected to facilitate more rapid restructuring, China's economy would have been more susceptible to inflation and fluctuations and would have experienced more layoffs. In their general theory Jefferson and Rawski (chapter 3 and appendix A) liken the reform process to the firm's investment decision, in which the 8 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  pace of reform cuin investment depends on technical relationships between (reform) inputs and (social welfare) outputs, uncertainty, risk, and costs of adjustment. In their view reform, like investment in fixed assets, cannot be costlessly reversed. Hence, in a world of uncertainty in which there is learning by doing, something less than a "big bang" is optimal. From this perspective the lack of a clear vision of the outcome of the reform process in China demanded a "groping" strategy. (In contrast, in Eastern Europe, where the desired state of the world-that is, Western Europe-was clear, "big bang" reform was adopted in order to reach the desired state more rapidly.) Moreover, in China various initial conditions magnified the marginal returns to these partial reform instruments. Finally, the 9-10 percent real rates of annual growth are widely viewed within China as sufficient. Thus far China has been able to achieve high growth with stability. The expectation is that continued high growth will require ongoing reform but not a radical big bang solution. Technical change. Market-based technical change is widespread in China where competition and market incentives foster widespread prod- uct innovation. Moreover, emerging markets in R&D resources (funds and engineers) are generating high and surprisingly uniform returns across state-owned and township and village enterprise industry. Jefferson and Rawski (chapter 3) portray a dynamic process through which innovation spills over into China's economy from international markets through joint ventures and China's larger publicly owned enter- prises.Through a process of imitation by less technologically sophisticat- ed, lower-cost producers, new products and processes filter down the ladder, creating competitive pressures that move up the ladder, thereby forcing new rounds of innovation by firms at the top of the ladder. A central prediction of this industrial innovation ladder is that competition motivates a rising incidence of technical change. Moreover, R&D resources are expected to move down the ladder, where they can gener- ate competitive, or superior, rates of return. In chapter 4 Jefferson, Rawski, and Zheng demonstrate how, at least in the late 1 980s, China's township and village enterprises were engaged in an intense process of technical catch-up with state enterprises, which were widely viewed as the technological leaders across a range of prod- uct groups. In terms of the sheer volume of innovation, township and village enterprise spending on R&D and technicians generated higher returns than comparable activity by state-owned enterprises. In terms of the quality of innovation, measured by the profitability of new products, the allocation of R&D spending and technicians within China's indus- trial economy appears to have been efficient. OVERVIEW 9  Institutional innovation. De facto property rights have devolved to China's publicly owned enterprises, creating a coherent set of decision- making rights and incentives for some enterprises. While the devolution of rights to enterprises and managers varies substantially, so that many state enterprises enjoy more extensive rights than do some township and village enterprises, property rights are generally more devolved and con- gruent within the township and village enterprise sector. Relative to the prereform period, most of China's publicly owned industrial enterprises have enjoyed a substantial expansion of property rights, although these changes are often specified through implicit rather than explicit contracts. Changes in property rights include expanded deci- sionmaking authority for the enterprise, designated authority for a central contracting agent within the enterprise, and a redistribution of residual claimant authority toward the enterprise and its central agent or manager. If these rights are not congruent (if, for example, there is substantial enter- prise autonomy but fragmented internal authority and weak incentives), reform may be counterproductive.Within the enterprise samples studied, decisionmaking rights and the incentive structure were found to be more devolved and coherent within the township and village enterprise sector, which may help explain its relatively robust performance. Clarifying property rights in both state and collective industry, includ- ing township and village enterprises, is a top reform priority among aca- demics and Chinese policymakers. The case study of the Hengdian Company (appendix B), a township conglomerate, demonstrates how an intricate contracting system that embodies a strong incentive structure can, in some settings, effectively substitute for conventionally defined pri- vate property rights. Friction between the Hengdian town government, which continues to claim ownership of the company, and company offi- cials and workers has been muted, in substantial part by the extraordinary force and success of a single individual and the rapid growth of produc- tion and incomes. The disappearance of one or more of these conditions is likely to reveal the cost of the company's poorly defined property rights. Concern about poorly specified property rights is motivating thousands of enterprises within the township and village enterprise sec- tor to "individualize ownership," that is, to distribute shares to managers, workers, and local residents, sometimes leaving a substantial portion of the shares in the hands of local government, thus reducing but not eliminat- ing the ambiguity of formal ownership. Causality in the reform process. Virtually all of the literature on enter- prise reform examines the impact of reform on performance. Causality also operates strongly in the other direction. Poor or declining perfor- 10 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  mance, particularly among enterprises operated at lower levels of gov- ernment, motivates new rounds of reform. Indeed, the industrial inno- vation ladder predicts that causality should run from enterprise performance to reform. The industrial innovation ladder portrays competition squeezing financial performance and motivating a search for more efficient forms of corporate governance, as well as new technologies and markets. This process is at the heart of the concept of endogenous reform. Using enter- prise data,Jefferson, Lu, and Zhao (chapter 5) find evidence in support of the endogenous reform model, particularly among state-owned enter- prises that display poor productivity and financial performance and face a competitive environment and supervisory body at a lower level of gov- ernment.These circumstances motivate new rounds of enterprise reform. These findings reinforce the view that China's industrial reform should be viewed as a process that is driven by forces within China's industrial system as well as a sequence of reforms inspired by officials at the center. But jurisdictions converting larger firms to joint ventures or shareholding firms tend to select the most successful state-owned enterprises.This evi- dence demonstrating that performance drives reform creates vexing simultaneity issues, which make it difficult to discern the direction and magnitude of the relationship between reform and performance. Policy implications. The emergence of a dynamic industrial innovation ladder may be China's most valuable transition resource. To enable indi- vidual entrepreneurs, firms, and local levels of government to respond to pressures and opportunities for reform that are generated within the indus- trial system, the Chinese government should embrace the logic of the Coase theorem. Within a context of competition and incentives for reform, clarifying property rights and lowering transactions costs will enable entrepreneurs, firms, and local governments to more easily negoti- ate efficiency-enhancing exchanges of control rights and industrial assets. China's industrial innovation ladder-an immense population of het- erogeneous firms seeking to survive and grow within a system of sub- stantial and growing competition and incentives-is arguably the most important of China's transitional institutions. It also carries certain poli- cy implications, outlined by Jefferson, Lu, and Zhao (chapter 5) in their discussion of a Coasian approach to reform. From this perspective, by creating weak authority and incentives for monitoring within the firm, dispersed public ownership often displays characteristics similar to those of public goods (nonexcludability, in which managers, workers, and pub- lic officials take more from the enterprise than they give; and nondi- minishability, in which governments employ fiscal and financial subsidies OVERVIEW 1 1  to replenish these exactions). These subsidies, in turn, give rise to social externalities involving publicly financed subsidies, inflation, and ineffi- cient resource allocation. Externalities arising from dispersed public ownership can be linmited by applying the logic of the Coase theorem, which effectively argues for the creation of a property rights market by clearly assigning property rights and reducing transactions costs. Better-specified property rights and lower transactions costs-achieved by increasing competition through tighter budget constraints, improving transparency, and making contracting easier-enable agents to spontaneously negotiate efficiency- enhancing exchanges of control rights and industrial assets. In China where competition and local incentives for financial gain exist, creation of a set of institutions ensuring property rights can facilitate the decen- tralized, spontaneous exchanges between individual entrepreneurs and local governments needed to allocate property rights more efficiently. Already the incidence of mergers, acquisitions, privatization, and man- agerial contracting is growing. The potential for creating productive gains by expanding and reducing the costs of such exchanges is immense. Industrial performance Totalfactor productivity. Across all sectors of Chinese industry total fac- tor productivity has improved. Between 1980 and into the early 1990s, total factor productivity grew at an average annual rate of 2-4 percent in state industry; rates in the collective sector were higher. Some preliminary e.vidence of a slowdown in the early 1990s should be carefully watched to determine whether it reflects cyclical or structural conditions. Jefferson, Singh, Xing, and Zhang (chapter 6) identify 12 studies of productivity growth in China's state sector. Eight of these report total factor productivity growth rates that fall within the range of 2-4 percent; two lie above this range and two lie below it. Another comprehensive survey (Wu 1993) also finds widespread evidence of substantial produc- tivity gain. In this volume Perkins (chapter 11) finds evidence of rising productivity in both state enterprises and foreign-investedjoint ventures. Comparative studies of total factor productivity growth in the state and collective sectors, including those by Jefferson, Rawski, and Zheng (1992, 1996) and Wu and Wu (1994) show higher growth rates of total factor productivity among collective enterprises, particularly in the township and village enterprise sector, where productivity growth appears to have been 1.5-3.0 times that of the state sector. These findings run counter to conventional wisdom, which holds that efficiency at state enterprises cannot increase unless ownership is changed and that it is not possible to improve performance substantially 12 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  at publicly owned enterprises, including China's collectives and township and village enterprises (see, for example, Galal and Shirley 1995). In China the performance of state-owned enterprises and collectives con- tradicts both of these propositions: measured by the growth of real out- put and productivity in physical terms, state enterprises and collectives have performed well. Increasing numbers of studies are linking this improved productivity performance to specific reforms. Hay and others (1994) and Jefferson, Singh, Hu, and Wang (chapter 8) cite the importance of profit retention, bonuses, and various reforms within the wage-employment system. Groves and others (1994) underscore the importance of managerial incentives. Hay and others (1994) and Jefferson, Hu, and Singh (chapter 10) find growing correspondence between performance, investment, and retained earnings. Perkins (chapter 11) demonstrates the impact of auton- omy, incentives, and the trade regime on enterprise export performance. Jefferson, Lu, and Zhao (chapter 5) and Jefferson and Rawski (1997) attribute efficiency gains to ongoing exchanges negotiated within China's fledgling property rights market.The combination of these mutually rein- forcing changes has enabled both state-owned enterprises and urban and rural collective-owned enterprises to substantially improve their perfor- mance. Profitability. In all sectors of the industrial economy profitability has declined. Profitability is lower in the state sector than in the collective, foreign-funded, and shareholding sectors but comparable to the overseas Chinese-funded sector. Assessments of performance must distinguish between productivity, which may rise, and profitability, which may fall, even as productivity rises. While other explanations deserve considera- tion, the central, most fundamental explanation of the industrywide decline in profitability in China is rising competition. Profitability in state industry has fallen more than elsewhere, in part because productiv- ity in that sector is rising relatively more slowly than elsewhere. The simultaneous rise in productivity and fall in profitability is one of the striking paradoxes of China's industrial transition. Measured as the ratio of total profits plus taxes to total capital, overall pretax profitability in Chinese industry fell from about 25 percent in 1980 (Naughton 1994) to 20.5 percent in 1988, 9.5 percent in 1992, and 7.1 percent in 1996 (table 1.9). Much of this decline has resulted from heightened competi- tion. That profit rates are lowest in the small-firm sector, which is most exposed to competition, is evident from table 6.8, which shows 1996 pretax profit-to-revenue rates of 12.8 percent for large firms, 6.4 percent for medium-size firms, and 5.9 percent for small firms.The low rates of OVERVIEW 13  profit reported by the overseas Chinese-funded sector may reflect the concentration of small firms in export-oriented light industry. In state industry the pattern of decline was similar to overall industry until after 1992. By 1995 the ratio of profit plus tax to total assets had declined to 6.6 percent. During the early 1990s about one-third of China's industrial state enterprises were reporting losses. Two types of factors may explain the relative decline in profitability in state industry. The first is the entry of nonstate enterprises into industries that had enjoyed quasi-monopoly profits. Two sets of results, both reported in chapter 6, find that the entry of nonstate producers has driven down profits in the state sector. This phenomenon has been most acute in recent years, as overcapacity has driven down prices, especially in the consumer goods sector. The relative decline in profitability of state-owned enterprises reflects the rise in their productivity, which nevertheless lags behind that of the nonstate sector. Hence, although the long-run average cost curve in state industry is shifting downward, costs in the nonstate sector and prices gen- erally are declining faster. Productivity among state-owned enterprises is not rising fast enough to prevent growing losses that arise from increas- ing domestic and international competition, which is causing demand for China's industrial goods to become increasingly price elastic. Low and falling rates of profit in state industry reflect other factors relating to state industry in itself; Jefferson, Singh, Xing, and Zhang review these in chapter 6. The limited means of exit in state industry restricts the ability of state industry to shed chronic loss-making enter- prises, thereby fattening the loss-making tail of profit distribution in that sector. At the same time conversion of the most successful state enter- prises to forms of nonstate ownership is thinning the tail at the high- profit end of the distribution. Overall, profit rates in state industry are also depressed by the continued concentration of subsidized housing and social services in the sector. Finally, low real rates of interest and a lack of discipline in the financial system appear to be encouraging overinvest- ment in state industry, which is depressing profitability and levels of cap- ital productivity. Enterprise conduct In a variety of ways state-owned, urban collective, and township and vil- lage enterprises demonstrate a surprising degree of consistency with the behavior predicted of the neoclassical firm. During the 1980s and 1 990s wages have become more strongly associated with labor productivity, bonuses have become more strongly associated with profits, investment and innovation activity have become more closely linked with prof- 14 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  itability, and autonomy and export-oriented incentives have been used to spur export activity. Both chapters that examine wage-setting behavior in Chinese indus- try find significant links between reward and effort. In their chapter on wage and employment behavior (chapter 8), Jefferson, Singh, Hu, and Wang concur with a major conclusion of Hay and others (1994) that "the standard neoclassical model seems to fit these firms, even though imperfect adjustment and/or lags are clearly evident" (p. 143). Hay and others limit their analysis to state industry; Jefferson, Singh, Hu, and Wang extend the conclusion to enterprises of all ownership types. They find that state-owned enterprises are more constrained by past values of the wage bill, basic wage, employment, and bonuses, while joint ventures generally demonstrate greater short-run responsiveness to price changes. Yet in the long-run, state-owned enterprises adjust substantially-per- haps more fully than other ownership types-to changes in relative prices.The incentive effect of wages and bonus are most evident in state industry, where the threat of layoff is weak; less so in the collective sec- tor; and hardly evident in the joint venture sector. In chapter 9 Pitt and Putterman examine the widely held view that both state and collective enterprises in China overemploy labor. Comparing actual employment levels to those projected to be profit- maximizing given outside wages and estimated technical parameters, their results suggest that, if anything, township and village enterprises are biased against, rather than toward, employment creation. This striking result, combined with related research on state-owned enterprise (Dong and Putterman 1997), indicates that during the 1980s both state and township and village enterprises experienced monopsony power in labor hiring. The work of Dong and Putterman (1997) also suggests that state industry's monopsony power has ended as the rapid expansion and diver- sification of China's industrial base and relaxation on labor mobility have created more competitive and unified labor markets. Jefferson, Hu, and Singh (chapter 10) find a robust fit between the survey data and various investment functions derived from neoclassical principles. They conclude that their findings "provide encouraging evi- dence that capital's marginal revenue product, enterprise profitability, and relative prices are significantly affecting patterns of capital accumula- tion." Various data sets consistently demonstrate a substantial and grow- ing link between enterprise performance and investment, a pattern that appears to be just as much in evidence in state industry as it is in urban collective enterprises, township and village enterprises, and joint ven- tures. This is one of the most important findings in the hterature on Chinese enterprise reform. Over the long run, robust economic growth OVERVIEW 15  requires a close statistical link between enterprise economic perfor- mance and investment, the principal avenue for expanding production capacity. The authors' conclusion that in general the stronger state- owned enterprises are capturing most of the investment resources in that sector needs to be reconciled with Jefferson., Rawski, and Zheng's find- ing (1996) of an overall decline in capital productivity in state industry beginning in 1988 and continuing into the early 1990s. Perkins (chapter 11) finds that policies intended to encourage export activity are, in fact, raising total factor productivity and expanding exports. The major determinants of export activity appear to be enter- prise decisionmaking autonomy and exposure to freer domestic markets. Provision of duty drawbacks and tax and other privileges enjoyed by firms in the Special Economic Zones expose enterprises to internation- al markets and promote exports.These are a powerful set of conclusions. While Perkins does control, in some measure, for obvious simultaneity problems, the conclusions should nonetheless be viewed somewhat ten- tatively. The role of ownership The collective sector has outstripped the state sector in terms of pro- ductivity growth and profitability, and average levels of total factor pro- ductivity appear to be higher in the collective and joint venture sectors than in state enterprises. Some of this disparity results from policy dif- ferences that are not intrinsic to ownership. These include less marketi- zation, more social overhead, and less reformed management systems among state-owned enterprises. The fact that these factors constrain pro- ductivity and profitability in the sector suggests that further gains are achievable through policy reform that stops short of outright privatiza- tion. Within all five enterprise samples virtually all the enterprises are either wholly publicly owned or include substantial public ownership; the sample of purely privately owned enterprises was not large enough to allow performance by private firms to be compared with performance by the various types of publicly owned enterprises. Such analysis repre- sents the next important frontier of research on China's industrial enter- prises. Jefferson (chapter 7) indicates that in 1989, total factor productivity in township and village enterprises was either greater than or comparable to that of their state-owned enterprise industrial branch counterparts. If, as the evidence suggests, the township and village enterprise sector has continued to experience more rapid productivity growth, the sector now enjoys higher productivity levels than state enterprises. It is difficult to 16 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  appreciate fully the reasons for the superior productivity performance of the township and village enterprise sector. Chapters 6 and 7 attempt to identify sources of differing performance among state industry, township and village enterprises, and other forms of ownership. The statistical work in both of these chapters indicates that part of the disparity in measured productivity can be explained by differences in market exposure and contracting arrangements. Differences in produc- tion scale and industrial branch composition also matter. Disparities in productivity and profitability can also be explained in part by differences in social overhead and pension obligations. Relative to state-owned enterprises, township and village enterprises also generally enjoy proper- ty rights that are more devolved and coherent. While controlling for these conditions tends to narrow the perfor- mance gap between state enterprises and township and village enterpris- es, the gap nevertheless persists. As Perkins notes (chapter 11), "even controlling for policy differences (including decision autonomy), urban collective, township and village, and foreign-funded enterprises are all significantly more successful exporters than state enterprises." It may be that whether the factors that continue to depress efficiency by state enterprises relative to nonstate enterprises are extrinsic to state ownership or not is immaterial. If the state is unable to unshackle state industry from the constraints of social obligations, blocked exit, and impeded competition, the "pure ownership effect" will remain entangled with policy variables. Three views of ownership can be distinguished. The first, held by Nolan (1995) and Rawski (1994), among others, is that the form of ownership is less important than a level playing field, on which state and collective enterprises can hold their own. Others, including Sachs and Woo (1997) and Galal and Shirley (1995), view publicly owned enter- prises as inherently flawed and thus urge their disbandment. A third view is agnostic. From this perspective, held byWeitzman and Xu (1994), who see township and village enterprises as sustained by a "culture of coop- eration," some publicly owned enterprises may be able to compete suc- cessfully in the foreseeable future. Regardless of which of these views one embraces, the appropriate policy response is to focus on construct- ing an efficient property rights market that consistently subjects all firms, regardless of ownership, to a market-based performance test. This view, propounded by Jefferson (1998), argues for leaving the ownership issue to a property rights market that provides clear, unencumbered avenues through which unsuccessful firms and ownership types can exit or be acquired by more successful firms. It suggests that the Chinese govern- ment may wish to retain or assign ownership rights to various units of OVERVIEW 17  government.Whether or not a substantial publicly owned sector survives in China outside of the natural monopoly industries should depend on the ability of these publicly owned enterprises to compete and innovate within China's emerging property rights market and, more generally, its industrial innovation ladder. Sustainability Enterprise performance has improved as a result of a continuous, if grad- ual, sequence of reforms. Expanding competition, encouraging marketi- zation and exposure to international trade and investment, increasing fiscal and financial discipline, and supporting evolving property rights markets while maintaining a stable macroeconomic environment are key to sustaining economic reform and a growing industrial economy. These are also the measures most needed to strengthen and capitalize on China's industrial innovation ladder. Strengthening the industrial inno- vation ladder, in turn, will serve to build in pressures throughout the industrial system for steady, ongoing reform. In the medium term, productivity gains can be achieved through reductions in static inefficiency associated with further marketization and property rights reform under existing ownership arrangements.This potential should not be underestimated Through a continuing process of factor reallocation, bankruptcy, and merger, the resources of the more than 7 million enterprises in China will gradually be consolidated, so that economies of scale and scope comparable to those realized in the industrial economies are achieved. Beyond this vast potential for static efficiency gains, China's high savings rates and increasing openness to international trade and investment are motivating and financing the adoption of improved physical and human capital, which is needed to support ongoing rapid technical progress within China's more advanced industrial sectors. Productivity gains in China's state-owned enterprises have enabled the state enterprises to serve as a useful transitional device by provid- ing a guarantee of social insurance to China's urban population and a source of producer goods to the emerging nonstate sector. The con- version of state-owned enterprises to joint ventures and to large share- holding enterprises has fueled the growth of these new ownership forms, while increasing numbers of smaller state-owned enterprises are being fully privatized. As a declining proportion of state enterprises is able to compete effectively with other ownership forms and as new social insurance arrangements are put in place, the pace of conversion is accelerating, both through the market and at the direct initiative of the Chinese government. 18 ENTERPRISE REFORM IN CHINA: OWNERSHIP, TRANSITION, AND PERFORMANCE  Jefferson and Rawski's model of an innovation ladder (chapter 3) repre- sents one perspective on the sustainability of reform. In this model a com- petitive process of "natural selection" is built into the economy, in which enterprises and local governments must search for new governance struc- tures and new technical innovations in order to remain competitive and sol- vent. The speed and efficiency with which the Chinese economy and people realize the potential gains already embodied in their industrial sys- tem will depend on the capacity of a competitive, well-functioning prop- erty rights market to mediate the millions of exchanges required to match supplies of labor, capital, technology, and entrepreneurship. Clarifying the assignment of property rights within China's industrial system, expanding competition, and reducing transactions costs will increase the efficiency of China's industrial innovation ladder. Even as a property rights market develops, however, various forces will curtail high rates of growth. These include the following: