Download pdf - Chinaâ€™s Ownership Transformation: Process, Outcomes, Prospects

China’s OwnershipTransformation

China’s OwnershipTransformation

Process, Outcomes, Prospects

Ross Garnaut, Ligang Song, Stoyan Tenev, and Yang Yao

International Finance CorporationAustralian National University

China Center for Economic ResearchPeking University

2005

© 2005 The International Finance Corporation and The International Bank forReconstruction and Development/The World Bank2121 Pennsylvania Ave., N.W.Washington, D.C. 20433USATelephone: 202-473-3800Internet: www.ifc.org; www.worldbank.org

All rights reserved

1 2 3 4 5 07 06 05

The findings, interpretations, and conclusions expressed herein are those of theauthors and do not necessarily reflect the views of the Executive Directors ofthe International Finance Corporation or of the Bank for Reconstruction andDevelopment/the World Bank or the governments they represent.

The World Bank does not guarantee the accuracy of the data included in thiswork. The boundaries, colors, denominations, and other information shown onany map in this work do not imply any judgment on the part of the InternationalFinance Corporation or the World Bank concerning the legal status of any ter-ritory or the endorsement or acceptance of such boundaries.

Rights and PermissionsFor permission to photocopy or reprint any part of this work, please send a re-

quest with complete information to the Copyright Clearance Center Inc., 222Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com.

ISBNs: 0-8213-6237-2 978-0-0-8213-6237-2DOI:10.1596/978-0-8213-6237-2

Library of Congress Cataloging-in-Publication DataChina’s ownership transformation : process, outcomes, prospects / Stoyan

Tenev ... [et al.].p. cm.

Includes bibliographical references.ISBN 0-8213-6237-21. China—Economic policy—1976–2000. 2. China—Economic policy—

2000– 3. Industrial policy—China. 4. Privatization—China. 5. Free enterprise—China. 6. Government ownership—China. 7. Government businessenterprises—China. 8. Unemployment—China. 9. Corporate governance—China. 10. China—Economic conditions—1976–2000. 11. China—Economicconditions—2000– I. Garnaut, Ross. II. World Bank.

HC427.92.C42874 2005330.951–dc22

2005047501

Contents

FOREWORD, Javed Hamid vii

PREFACE, Stoyan Tenev xi

ACKNOWLEDGMENTS xv

ABBREVIATIONS AND ACRONYMS xvii

1 INTRODUCTION 1

Overview of SOE Reform in China 2Focus and Empirical Approach of the Study 11The Structure of the Study 24

2 THE MAIN PLAYERS IN GAIZHI 25

How Far Has Gaizhi Progressed? 25The Key Participants 31Theories on Incentives for Gaizhi 38Empirical Tests 42Conclusion 45

3 THE GAIZHI PROCESS 46

Forms of Gaizhi 46Sample Distribution of Forms of Gaizhi 50Trends and Geographic Variations 54What Determines the Form of Gaizhi Chosen? 59The Process and Main Issues Surrounding the Transfer

of State Assets 62Asset Valuation 67Dealing with Enterprise Debts and Other Obligations in

the Gaizhi Process 76Land-Use Rights and Gaizhi 81Conclusion 86

v

vi

CONTENTS

4 IMPACT OF GAIZHI ON LABOR 87

China’s Emerging Social Security System 87Government Policy toward Unemployment

in the Gaizhi Process 89The Impact of Gaizhi on Employment

and Labor Force Structure 95Gaizhi and Obligations of Firms to Workers 103Gaizhi and Changes in Compensation Schemes 108Conclusion 111

5 IMPACT OF GAIZHI ON CORPORATE GOVERNANCE 113

Changes in Ownership Structure 114Ownership and Control 121Gaizhi and Traditional Stakeholders 127Gaizhi, Managerial Autonomy,

and Managerial Incentives 133Gaizhi and Changes in the Relative Influence

of Stakeholders 138The Role of Outside Investors in Corporate Governance 139Conclusion 143

6 IMPACT OF GAIZHI ON FIRM BEHAVIOR AND PERFORMANCE 145

Gaizhi, Internal Restructuring, and Financial Discipline 145Gaizhi and Firm Performance 158Gaizhi and Time Trends in Performance 170The Impact of Other Factors on Performance 172Conclusion 174

7 TOWARD A FAIRER AND MORE EFFICIENT GAIZHI PROCESS 175

The Public Debate about Privatization in China 175Regulating Gaizhi 181Strengthening Enforcement 185Reducing Transaction Costs for Outside Investors 191Enhancing the Role of the de Novo Private Sector

in China’s Transformation 197Conclusion 201

REFERENCES 205

INDEX 217

Foreword

China’s emergence as a global economic player has been accompaniedby a major internal transformation. Over the past decade or so, theeconomy has made the transition from complete reliance on state-owned and collective enterprises to a mixed economy where privateenterprise plays a leading role. This remarkable transformation hasbeen accomplished through the dynamic growth of the de novo privatesector and more recently through privatization.

IFC has been an active participant in this transformation processthrough investments and technical assistance for private companiesand pioneering research on private sector development and enterprisereforms. In 2000, IFC published one of the first studies on the emerg-ing domestic private sector in China. The study analyzed the structureof private enterprise, the enabling environment for its development,and access to financing. It outlined an agenda for entrepreneurs, thegovernment, and the financial sector for addressing constraints to pri-vate sector development. In 2002, IFC jointly with the World Bankpublished a study on the status and evolution of corporate governanceand enterprise reforms in China. The study explored the main corpo-rate governance issues that China has encountered during the course ofcorporatization and ownership transformation of its enterprise sector.

While China has been implementing reforms in its state enterprisesector over the past two decades or so, reforms have accelerated andhave acquired new features since the start of the present century. First,the scale of change has expanded to affect almost every kind of state-owned enterprise—small, medium, large, and very large—under bothcentral and local control. Second, ownership diversification has beenso extensive that the role of the wholly state-owned nonfinancial com-pany has declined substantially in many areas. Third, the range of re-structuring mechanisms being used has expanded dramatically to include bankruptcies, liquidations, listings and de-listings, debt-for-

viivii

equity swaps, sales to private parties (domestic and foreign), and auc-tioning of state firms, their assets, or liabilities. Finally, large layoffs,something unheard of just five or six years ago, have become a regu-lar phenomenon in corporate restructurings and privatizations.

There has been no systematic study of the magnitude, forms, andconsequences of this stage of enterprise restructuring. This book aimsto fill this gap by looking at the process, the main players involved, andthe outcomes. The empirical analysis is based on a survey of close to700 enterprises in 11 Chinese cities. The study is a joint venture amongthe Australian National University, Beijing University, and theInternational Finance Corporation of the World Bank Group. Theformer State Economic and Trade Commission of the Chinese gov-ernment facilitated the study for its successful implementation.Funding was provided by AusAid and IFC.

The study sheds new light on the progress that China has made inenterprise restructuring and privatization and on the challenges that en-terprises, investors, and governments are facing in the process. An im-portant finding of the study is that among the various forms of enter-prise restructuring, privatizations involving outside investors have hadthe strongest positive impact on firm performance. Furthermore, theanalysis finds that outside investors deliver improvements in perfor-mance more quickly than other forms of restructuring. The study showsthat the private sector is emerging as an important player in the re-structuring of SOEs but argues that its role could be enhanced further.

The International Finance Corporation has been playing an activerole in supporting the growth of the private sector in China. Our cu-mulative investments in China since 1985 are approaching $2 billionin over 80 companies. The size and the breadth of IFC’s program inChina are in many ways a function of the level of development of theprivate sector in the economy. When the private sector was mostlysmall and informal, and the industrial and financial sectors were dom-inated by SOEs and joint ventures with foreign private investors, IFC’sChina program consisted largely of industrial projects sponsored byforeign investors. A number of these projects were in effect restruc-turings of state-owned enterprises through the injection of funds andmodern technologies from foreign investors. IFC had an importantrole to play as a provider of long-term project financing that was nototherwise available for private projects.

The emergence of the domestic private sector has given us new op-portunities to broaden our program to include support for local fi-

FOREWORD

viii

nancial institutions, indigenous industrial and infrastructure enter-prises, and small and medium businesses. The dynamic growth of thedomestic private sector is creating the jobs needed to absorb laid-offworkers from restructuring state enterprises. It is therefore a majorpositive force in the restructuring process.

Today we see the biggest investment opportunities in China withdomestic private companies. These businesses are driving the rapidgrowth of the economy as they strive to expand and become more so-phisticated. Increasingly, domestic private companies in China arelooking at acquisitions of SOEs as their main growth strategy. A num-ber of our projects demonstrate how a privately managed companycan transform an ailing state enterprise into a profitable business thatcontributes to the local economy.

The study has benefited from the knowledge accumulated throughour investment and technical assistance experience in China. Its find-ings also provide us with new ideas on how to continue to support theprocess of enterprise restructuring. I hope that investors, policy mak-ers, opinion leaders, journalists, and all those interested in the statusof China’s enterprise reform can also learn from the study. And I hopethat this study contributes to the further progress of enterprise re-structuring in China.

Javed HamidDIRECTOR

EAST ASIA AND PACIFIC DEPARTMENT

INTERNATIONAL FINANCE CORPORATION

FOREWORD

ix

xi

Preface

Gaizhi, a Chinese term meaning “transforming the system,” has be-come a major phenomenon in most parts of China. The restructuringof state enterprises has accelerated in recent years to include bank-ruptcies, liquidations, listings and delistings, debt-for-equity swaps,sales to private parties (domestic and foreign), and auctioning of statefirms and their assets or liabilities. In many cases gaizhi has involvedfull privatization. Gaizhi programs in China have been gradual andlow profile, but in many ways as far reaching as, and generally eco-nomically more productive than, privatization measures in EasternEurope and the former Soviet Union.

Reforms have been most dramatic in the industrial sector wherethe number of state-owned enterprises has declined from 114,000 in1996 to 34,000 in 2003. According to our estimates, about half of thedecline is due to privatizations. Privatization in China has not beenlimited to small enterprises only: the average size of privatized SOEs isabout 600 employees. The process has been socially painful: around30 million SOE workers have been laid off since 1998. A dynamic denovo private sector has been able to absorb most of the laid off work-ers, thus alleviating the social cost of restructuring.

Gaizhi and the growth of the de novo private sector have trans-formed the structure of the Chinese economy. Over the past decade orso, the economy has made the transition from complete reliance onstate-owned and collective enterprises to a mixed economy where pri-vate enterprise plays a leading role. We estimate that the private sec-tor, narrowly defined, has become the largest sector of the Chineseeconomy, accounting for about 37 percent of gross domestic productin 2003. Overall, the nonstate sector accounted for two-thirds ofChina’s GDP in 2003.

Gaizhi is not a one-off event but a continual process of reformsand restructuring. As of October 2004, about 40 percent of SOEs were

making losses, compared with 18 percent for the nonstate sector.Enterprise reforms in China still have a long way to go.

There has been no systematic study of the magnitude, forms, andconsequences of gaizhi. This book aims to fill this gap by looking atthe process, the main players involved, and the outcomes of gaizhi. Theempirical analysis is based on a survey of close to 700 enterprises in11 Chinese cities.

Because gaizhi involves a comprehensive transformation of thestate sector, which had been the foundation of the Chinese economy,a number of players have stakes in the process. Our analysis showsthat local governments and enterprise managers have been the mostactive proponents of reforms with managers assuming a leading rolein later rounds of restructuring. The government considers preservingsocial stability and protecting the welfare of state employees a top pri-ority in SOE restructuring. Therefore, concerns about the social andfiscal implications of redundancies tend to constrain the pace of pri-vatization. Consistent with this finding, firms with greater net assetsare easier to privatize because they can compensate workers. Similarly,cities with stronger financial capacity tend to privatize more aggres-sively because they can absorb a greater portion of the social costs ofrestructuring. As a result, better-performing SOEs are likely to be pri-vatized first, and there are significant regional variations in the paceand scope of enterprise reforms. A dynamic de novo private sectormakes it easier to absorb redundant workers and therefore reduces thecost of restructuring. We find that the level of development of the denovo private sector has been the most important macroeconomic fac-tor leading local governments to release SOEs into private hands.

While gaizhi can be held back by fear of labor redundancies, it isoften the only way to check job losses in the state enterprise sector. Animportant result of our analysis is that gaizhi and privatized firms havemaintained a lower rate of employment reduction and a higher rate ofwage growth than non-gaizhi and fully state-owned firms. Consistentwith the conventional belief, gaizhi firms discharged more workers inthe year when gaizhi was implemented, but in subsequent years andoverall they were able to retain more workers than non-gaizhi firms.

Gaizhi firms were able to limit job losses because restructuring hasbrought efficiency gains. We find that gaizhi has a positive impact onfirm profitability, although a weak or insignificant impact on unit costand labor productivity. Privatizations involving outside investors havethe strongest positive impact on firm performance. Furthermore, out-

xii

PREFACE

side investors deliver improvements in performance more quickly thanother forms of gaizhi: their positive impact tends to appear early in thereform process.

Interestingly, we find that SOEs as outside investors also tend tohave a strong positive impact on firm performance. China’s experienceshows that SOEs with a relatively high degree of autonomy in the mar-ket process may have difficulty in putting reforms into effect in theirown enterprises but can be effective agents of change in other state en-terprises.

Gazhi has brought efficiency gains by aligning incentives and re-allocating decision-making powers within the firm. We find that gaizhifirms are more likely than non-gaizhi firms to provide managers withshares and bonuses. Shareholder representation on the board of direc-tors of gaizhi firms has improved, and power sharing among the share-holders’ conference, the board of directors, and the management hasbegun to occur. The influence of the Communist Party over the firmtends to decline after gaizhi, but the role of the labor union in collectivewage bargaining is more clearly defined and enhanced. We find that thegovernment is retreating from the privatizing firms by reducing its own-ership share, while the share of insiders has been increasing rapidly.

The dominance of managers appears to be the main corporate gov-ernance issue of gaizhi firms. Managers tend to be overrepresented atthe boards of directors and maintain decisive influence on key issues.What players and institutions are emerging to control the agency costsof managerial autonomy? Survey results indicate that outside investorsare more likely to use and rely on the new mechanisms of corporatecontrol, to provide effective checks and balances on managerial discre-tion, and to offer high-powered incentives to senior managers. In gen-eral, the presence of outside investors is associated with a reduced rolefor traditional stakeholders such as the government, the Party, and thelabor union. These traditional stakeholders have less significant roles inoutsider-controlled firms than in insider-controlled firms.

Thus one important result of our analysis is that privatizations in-volving outside investors are generally more productive than otherforms of privatization and gaizhi. Yet, privatization in China does notexhibit a clear trend in the direction of a greater role for outside in-vestors. While on average the ownership share of insiders has grownrapidly in recent years, outsiders’ share has remained largely stagnant.In our sample of firms, insiders held 5 percent of privatizing firms’shares in 1995. In 2002, their share had risen to 32 percent. Over the

xiii

PREFACE

xiv

PREFACE

same period, the combined share held by domestic and foreign privatecompanies has remained at about 20 percent.

Insider privatization could be subject to greater conflicts of inter-ests than other forms of gaizhi, especially given the major role that en-terprise managers play in initiating and implementing restructuringprograms. Media reports on irregularities in insider privatizations andparticularly management buyouts (MBOs) have raised public concernsabout lack of fairness and transparency of the privatization process inChina. In response, the government has promulgated a host of regula-tions aimed at establishing an orderly process of ownership transfor-mation, and at expanding the role of outside investors. A policy pri-ority is to enhance the involvement of the private sector, both domesticand foreign, in the restructuring and privatization of SOEs.

We already observe a change in the role that the domestic privatesector is playing in China’s state enterprise reform. Historically, theprivate sector has been supporting restructuring largely indirectly bycreating the jobs needed to absorb laid-off workers. While this indi-rect role will continue to be important, domestic private enterprises areemerging as significant players in the privatization process. A growingnumber of de novo private firms have begun to look at acquisitions ofSOEs as their main growth strategy. These private companies havebeen injecting capital and dynamism in moribund state enterprises thushelping to preserve jobs. While private enterprises are becoming moreactive in acquiring and restructuring state-owned enterprises, they stillaccount for a small share in all gaizhi cases.

China’s approach to state enterprise reform has been extremelypragmatic. Ownership change is not seen as an end in itself nor is itseen as the automatic solution to inefficiency problems in the state en-terprise sector. Local governments are primarily interested in aspectssuch as tax revenues, growth and employment. Looking for ways toobtain these results, they have been experimenting with institutionalreforms. In the process, local governments have found that the way todeliver tax revenues, growth and employment to their constituenciesis by opening more room for private enterprise. Enhancing the role ofprivate companies in SOE reform will require, however, sustained ef-forts from both the government and the private sector to improve thebusiness environment for entrepreneurship and move private enter-prises toward global best practice.

Stoyan TenevINTERNATIONAL FINANCE CORPORATION

xv

Acknowledgments

The State Economic and Trade Commission (SETC, now part of theNational Development and Reform Commission) facilitated this study.Wang Hailin, director, and other SETC staff provided valuable guid-ance and support throughout. Extensive assistance was provided byKarin Finkelston, IFC’s associate director for China in Beijing. IFC staffJianguo Cui and Wenqin Zhu played a key role in organizing and co-ordinating the contributions of the various parties involved. RanaGanguly managed the project from the Australian National Universityside. The study was funded by AusAid and IFC’s Trust Funds.

Xiaolu Wang (Australian National University and the NationalEconomic Research Institute), Yu Sheng (Australian NationalUniversity), and graduate students from the China Center for EconomicResearch at Beijing University contributed to the technical report on thefield work.

Survey findings were presented and discussed at a workshop orga-nized by the International Finance Corporation in March 2003 inBeijing. Wang Hailin and Tian Chuan (SETC), Chunlin Zhang andWilliam P. Mako (World Bank), Omar Chaudry (IFC), and RobinScott-Charlton and Michael Willcock (Australian Embassy in Beijing)provided valuable comments at the workshop.

A draft of the study was presented and discussed at a workshopin Beijing organized by the International Finance Corporation inMarch 2004. At the workshop, An Chongli (Asian DevelopmentBank), Liu Xiaoxuan (Chinese Academy for Social Sciences), DavinMackenzie (iVentures L.L.C.), Ping Xingqiao (China Center forEconomic Research), Tian Chuan (National Development and ReformCommission), Richard Yu (AusAid), Strahan Spencer (Department forInternational Development), Wang Liming (National Developmentand Reform Commission), Wang Xiaolu (National EconomicResearch Institute), Wang Zhongjing (Ministry of Finance), Yang

Yiyong (Central Party Academy), Zhang Chunlin (World BankGroup), Zhang Shuguang (Chinese Academy of Social Sciences), andZhang Weiying (Guanghua School of Management, Peking University)made useful comments and suggestions. Andy Rothman of CLSA pro-vided valuable comments that enriched the final study.

The study also benefited from comments and insights from IFCand World Bank staff, including Bernard Sheahan, Sanjay Grewal,Sunita Kikeri, and Peter Taylor. Udayan Wagle, Mwaghazi Mwachofi,Mariko Higashi, Maria Cussianovich, Aminata Mbodj, MichaelO’Neill, Bayo Oyewole, Amber Turner, Wai-Keen Wong, andFrederick Wright supported the study through funding from IFC TrustFunds. Lixing Li from the University of Maryland provided valuable re-search assistance. Robyn Flemming edited the text, and Garry Cousinsprepared the index. Special thanks to Dana Lane for her excellent man-agement of the publication process.

xvi

ACKNOWLEDGMENTS

Abbreviations and Acronyms

ABC Agricultural Bank of ChinaAMC assets management companiesBOC Bank of ChinaCCB China Construction BankCCP Chinese Communist PartyCEO chief executive officerCSRC China Securities Regulatory CommissionDCD discounted dividendsDCF discounted cash flowDRC Development Research CenterETC economic and trade commissionsFDI foreign direct investmentFIE foreign-invested enterpriseGDP gross domestic productICBC Industrial and Commercial Bank of ChinaIPO initial public offeringM&A mergers and acquisitionsMBO management buy-outMOF Ministry of FinanceNPC National People’s CongressNPL nonperforming loanPAYE Pay As You EarnPER price-earningsR&D research and developmentROA return on assetsSARS Severe Acute Respiratory SyndromeSASAC State-owned Assets Supervision and Administration CommissionSASMC State Asset Supervision and Management CommissionSETC State Economic and Trade CommissionSME small and medium enterprisesSOE state-owned enterpriseSPC Supreme People’s CourtTVE township and village enterprise

xvii

1Introduction

The recent emergence of China as a global manufacturing powerhouseand the world’s top destination for foreign direct investments (FDI) isattracting considerable attention from policy-makers and the media.In fact, China’s rise to economic prominence is viewed as the economicevent of our age (Wolf 2003). Chinese companies are expanding theirpresence abroad, building internationally known brand names and trans-forming global production networks.

China’s emergence as a global economic player has been accom-panied by a major internal transformation. Over the past decade, theeconomy has made the transition from complete reliance on state-owned and collective enterprise to a mixed economy where private en-terprise plays a leading role. This is an ongoing process, during whichthe economy will remain structurally a very diverse one. While someChinese companies are in the vanguard of globalization, many arestruggling with old legacies of planning. And to use a military metaphor,it is dangerous for the vanguard of the army to get too far ahead of therear-guard. Thus, in these times of optimism and dynamism, increasedattention must be given to solving the problems of the state-ownedsector.

Gaizhi, a Chinese term meaning “transforming the system,” hasbecome a major phenomenon in most parts of the country; in manycases it has involved full privatization. Unlike the mass privatizationprograms that have occurred in Eastern Europe and the former SovietUnion since the late 1980s, gaizhi programs in China have been grad-ual and low profile. The significance of the Chinese reforms shouldnot be underestimated, however. In many ways they have been as far-reaching as, and generally more economically productive than, thosein Eastern Europe and the former Soviet Union.

1

Overview of SOE Reform in China

Reform of China’s state-owned enterprises (SOEs) has been a majoraim since urban reforms began in 1984. Although there were calls toprivatize the SOEs, the government’s initial emphasis was on boostingperformance by changing the internal governance of SOEs and im-proving the market environment in which they operated.

Inspired by the success of the rural household responsibility system,the government introduced a contracting system into the state sectorthat required SOE managers to meet various performance targets—including targets for sales, profitability, and capital accumulation—inreturn for a share of the profits. The success of the enterprise thusdepended on the efforts that managers were prepared to make. Themain problem with this system was that managers were rewarded fortheir successes but not credibly penalized for their failures.

By the late 1980s the government had decided that the best wayto reform small SOEs was to lease them out, with the manager payingthe state a fixed proportion of the firm’s profit. The first significantlease contract was to the Wuhan Motor Engine Factory in 1986, whenthree people put up RMB34,000 as collateral to lease the factory. InMay 1988 the State Council issued a regulation on the leasing of smallSOEs.1 A direct consequence was that managers could be recruitedfrom outside the enterprise. In many cases, leasing led to the de factoprivatization of township and village enterprises (TVEs). After severalyears the accumulated and contributed capital of the manager wouldoutweigh that of the local government, and the firm would be effectivelyowned by the manager.

Incorporation was another significant measure that led to priva-tization. At first the government restricted incorporation to the exchangeof shares among the SOEs; soon, however, private shareholding wasallowed. The first cases of private shareholding were in three GuangzhouSOEs in 1986, when the employees bought 30 percent of the shares oftheir firms. The first large SOE to be incorporated was the ShenyangMotor Corporation, which became Shenyang Jinbei Motors when itissued shares to the public in August 1988.

CHINA’S OWNERSHIP TRANSFORMATION

2

1. The Tentative Regulations on the Lease of Small State-Owned IndustrialEnterprises, State Council, May 20, 1988.

The opening of the Shenzhen Stock Exchange in 1990 and theShanghai Stock Exchange in 1991 enabled SOEs to issue shares to thepublic. The Chinese government ensured that it would not lose con-trol of listed SOEs, however, by requiring that a proportion of thestate’s shares in the firm could not be sold.

Privatization started in earnest after a visit by Deng Xiaoping tosouthern China in 1992. As with many other reform initiatives, priva-tization started at the local level and was later sanctioned by the centralgovernment. The most important impetus for privatization in the lo-calities was the large amount of debt built up by the state sector. Thelevel of debt was a more pressing problem in small cities. For example,in Zhucheng city, Shandong province, 103 of the 150 SOEs were in thered at the end of 1992, with losses amounting to RMB147 million—equivalent to the revenue of the city government over 18 months (Zhao1999). The Shunde government, in Guangdong province, also encoun-tered a debt problem when it first started privatizing its SOEs in 1992.Most local governments decided that it would be possible to privatizeonly small firms, but Shunde and Zhucheng went further by privatiz-ing almost all of their state and collective firms (Huang and Wei 2001;Yao 2003).

In 1995, after extensive discussion, the central government de-cided on the policy of zhuada fangxiao, or “keep the large and let thesmall go.” The state decided to keep between 500 and 1,000 large statefirms and to allow smaller firms to be leased or sold.2 There were goodreasons for this decision. In 1997 the 500 largest state firms, most ofthem controlled by the central government, held 37 percent of thestate’s industrial assets, contributed 46 percent of the taxes collectedfrom state firms, and earned 63 percent of the profits of the state sec-tor. Small firms owned by local governments had been performingpoorly. In 1995, 72.5 percent of local firms, but only 24.3 percent ofcentral government firms, were unprofitable (Zhao 1999). As VicePremier Wu Bangguo said in a speech in December 1997, “Control of

INTRODUCTION

3

2. In 1994 the ministry in charge of government economic affairs, the StateEconomic and Trade Commission (SETC), sent a report entitled “Suggestions onRevitalizing Small State-owned Enterprises” to Vice Premier Wu Bangguo, whowas in charge of enterprise reforms. In September 1995 the policy was formallyannounced by the Central Committee of the Chinese Communist Party (CCP) inone of its plenaries and went forward as a suggestion for the ninth five-year plan.

the [500] largest firms means we have a control of the largest chunk ofthe state economy” (Zhao 1999).

From the “let the small go” part of the policy came the term“gaizhi,” meaning “transforming the system.” An important part ofgaizhi was the order issued in March 1998 for “red hat” firms—thatis, firms registered as collectives but in reality privately run3—to “takeoff their red hats” by the following November.

Privatization first commenced in rural areas. Many localities—including those renowned for the success of their collective enterprises,such as Shunde and southern Jiangsu—implemented privatization ona massive scale. Early in the reform era, the impressive growth of theTVEs had been hailed as proof that, contrary to conventional eco-nomic theory, clearly defined property rights were not essential for de-velopment (for example, Weitzman and Xu 1994). As the growth ofthe TVEs slowed, the disadvantages of their vaguely defined propertyrights became clearer. Like their urban counterparts, the TVEs facedsoft budgets, and the resulting build-up of nonperforming loans (NPLs)placed a considerable burden on local governments (Zhang 1998).

By the end of 1998, more than 80 percent of state and collectivefirms at the level of the county or below had gone through gaizhi,which involved direct privatization in most cases (Zhao 1999). Mostof these firms were TVEs. The pace of reform has accelerated in recentyears, and it is likely that TVEs will soon disappear, proving to havebeen an important but transitional institution in China’s march tomarket.

In the cities, gaizhi has occurred in two waves. Reform started inthe mid-1990s and followed the model of employee shareholdingadopted by Zhucheng. When Zhucheng abandoned this model andmoved toward concentrated ownership through management buy-outs(MBOs), other cities followed suit. The trend reflected the belief that,for an enterprise to be truly transformed, it is necessary for manage-ment to own the majority of shares. Forms of MBO have been the mostcommon model in the second wave of gaizhi and have spread to verylarge firms, such as the SOEs listed on the stock market. Privatizationhas been accepted as the key for urban reform, and the slogan “thestate retreats and the private sector moves forward” has become com-mon in many cities.


4

3. On the “red hat” phenomenon, see Gregory, Tenev, and Wagle (2000).

Since the start of the present century the reform of China’s stateenterprise sector has accelerated and acquired some qualitatively newfeatures. First, the scale of change has expanded to affect almost everykind of SOE—small, medium, large, and very big; under both centraland local control. Second, ownership diversification has been so ex-tensive that the wholly state-owned nonfinancial company has becomean endangered species in China’s business ecology. Third, the range ofrestructuring mechanisms being used has expanded dramatically to in-clude bankruptcies, liquidations, listings and de-listings, debt-for-equityswaps, sales to private parties (domestic and foreign), auctioning of statefirms and their assets or liabilities, standard corporate governancetechniques, and so on. Finally, mass layoffs—unheard off just four or fiveyears ago—have become a widespread phenomenon.

At the grassroots level, local governments have been particularlyactive in implementing corporatization and ownership diversificationin SOEs—the most decisive actions for grappling with the financialburdens and unemployment pressures resulting from loss-makingSOEs. A national survey in 1998 showed that a quarter of China’s87,000 industrial SOEs had been through some phase of gaizhi, whileanother quarter planned to take some measure of gaizhi. Among thegaizhi firms, 60–70 percent had been partially or fully privatized.4 A2001 national survey of industrial SOEs estimated that 86 percenthad been through gaizhi by the end of 2001 and about 70 percenthad been partially or fully privatized.5 The research in this book willshow that half of the six cities chosen for interviews planned to gothrough gaizhi by the end of 2003. In addition, the share of partiallyor fully privatized firms had risen to more than 70 percent. If this per-formance typifies that of the rest of the country, then privatization inChina has already gone further than in many Eastern European andformer Soviet Union countries.

In terms of larger SOEs, around 1,400 companies have been listedover the last decade. Their market capitalization is about 40 percentof gross domestic product (GDP),6 and their 65 million or so individ-ual shareholders have become increasingly assertive over the years.

INTRODUCTION

5

4. Unpublished report of the National Bureau of Statistics.5. Unpublished report of the SETC.6. Standard & Poor’s, “Emerging Stock Markets Review: Performance, Valu-ations, and Constituents,” Emerging Markets Database, December 2004.

The regulator, the China Securities Regulatory Commission (CSRC),has stepped up efforts to strengthen corporate governance practicesamong Chinese listed companies. Of systemic importance for China isthe process of divesting state assets through initial public offerings(IPOs) to fund the national social security system.

Some restructuring of SOEs is occurring through the four assetmanagement companies (AMCs) that have been created to take morethan $170 billion7 in nonperforming loans from the big four state-owned banks.8 As part of their program, 580 SOEs, accounting forabout 40 percent of the state sector’s assets and sales, have been se-lected for debt-equity swaps. The AMCs have emerged as important,and often majority, shareholders in a number of large SOEs. TheAMCs have multiple objectives; thus the desire to maximize immedi-ate financial returns to government may conflict with the need formeaningful restructuring of the firms in their portfolios (Steinfeld2001). Recent regulatory measures, however, that make it easier forforeign investors to buy significant and controlling shares in SOEs arelikely to expand the exit options for these AMCs and to create moreopportunities for diversification of state ownership through them.

The Chinese government has implemented industry rationaliza-tion programs in a number of industries, closing obsolete plants andreducing capacity to combat oversupply and deflationary pressures.The initiative has so far shown mixed results. In many industries, foreach unit of closed capacity several additional units have been created.For example, in the glass industry, in the past few years the govern-ment has closed down 240 small production lines, reducing capacityby 28 million-weight cases. In the meantime, however, 39 bigger andmore modern production lines have been built that will increase ca-pacity by 88 million-weight cases, or 32 percent of total production in2000. Similar stories are reported in textiles and other sectors.

In the strategically important infrastructure and energy sectorswhere the regulatory framework is still evolving, monopolies havebeen broken and competition has been introduced. Many companieshave been corporatized, and some have been listed on local and inter-


6

7. All dollar amounts are U.S. dollars.8. The “big four” are the Agricultural Bank of China (ABC), the Bank of China(BOC), China Construction Bank (CCB), and the Industrial and Commercial Bankof China (ICBC). Together they account for about two-thirds of China’s financialassets.

national exchanges. China has nurtured over 20 giant corporationsand conglomerates that have proven competitive in the inter-national market. Some of these companies are laying off tens—or evenhundreds—of thousands of employees, not because they are in finan-cial distress (some of them are hugely profitable) but because they wishto position themselves as important international players. As of 2002the top 12 Chinese transnational corporations, mainly SOEs, con-trolled over $30 billion in foreign assets and had some 20,000 foreignemployees and $33 billion in foreign sales.

To address the issue of fragmented management of state assets(Tenev et al. 2002, 26), China established the State-owned AssetsSupervision and Administration Commission (SASAC) in March2003. SASAC acts simultaneously as a shareholder, regulator, man-ager, and supervisor of state assets.

Thus both the scale and scope of transformation have been ex-traordinary. Table 1.1 illustrates the changes that have taken placein China’s industrial sector and in the relative position of SOEs in theeconomy. Since 1999 the number of SOEs has been reduced dra-matically and they now account for only 15 percent of all industrialenterprises. The decline in the SOEs’ share in total industrial assetshas been much less dramatic, however. As of October 2004, SOEsstill accounted for more than half of total assets. Although fewer innumber, SOEs have become bigger in China. Already larger than theaverage nonstate enterprise, SOEs have grown much faster over theperiod—2.5 times, versus 1.4 times for nonstate enterprises. Theirshare in total enterprise assets has remained roughly constant overthe period, at about 50 percent. One notable achievement of enter-prise reforms has been the improved profitability of the SOE sector.The return on state-owned assets has improved, and the profitabil-ity gap between state and nonstate enterprises has narrowed some-what. Reforms still have a long way to travel, however: as of October2004, about 40 percent of SOEs were making losses, compared with18 percent for the nonstate sector.

This process of transformation of China’s state sector has led tosocial unrest and will continue to be socially painful. Around 30 millionSOE workers have been laid off since 1998 and, according to the officialgovernment statistics, 8.7 million of these have not found new jobs. Thenumber of labor disputes of all kinds rose by 12.5 percent in 2000, andby another 14.4 percent in 2001 to reach 155,000. In 1999 there were6,767 collective actions (usually strikes or go-slows with a minimum

INTRODUCTION

7

8

TABLE 1.1NUMBER OF STATE-OWNED ENTERPRISES IN CHINA’S INDUSTRIAL SECTOR AND RELATED FINANCIAL DATA, 1999–2004

Loss-making

Average Loss- nonstateSOE asset size Average SOE ROA making enterprises

Total Total SOEs as assets as nonstate asset size profits as nonstate SOEs (% ofNo of assets profits % of all % of total enterprise SOEs % of total enterprise ROA (% of all non-

Year enterprises RMB bn RMB bn enterprises assets (RMB mn) (RMB mn) profits sector SOEs all SOEs) SOEs)

1999 154,882 11,238 220 37 68 36 134 44 3.5 1.3 41 222000 158,749 12,398 426 34 67 39 155 56 4.6 2.9 35 192001 168,799 13,418 466 28 65 39 184 50 5.0 2.7 36 192002 178,876 14,479 562 24 62 40 210 47 5.4 2.9 35 172003 193,483 16,707 815 19 57 46 260 46 6.1 4.0 36 152004 212,648 18,984 913 15 53 50 317 49 5.2 4.5 40 18

NOTES: Data for 2004 are until October 2004; ROA is Return on Assets in percent.SOURCE: CEIC.

INTRODUCTION

9

of three people taking part) involving 251,268 people, an increase of 900 percent relative to 1992. Since 1999 the number of collective disputeshas been increasing by about 20 percent per year (see figure 1.1). In thefirst half of 2004 alone, labor dispute arbitration committees at variouslevels accepted 135,000 labor dispute cases and handled 6,440 collectivelabor disputes, which involved 184,000 persons—considerably morethan in the past.9 Given the magnitude of the layoffs, however, the levelof labor unrest is not extraordinary. In fact, labor seems to be acceptingthese changes. State employees—perhaps the most powerful interestgroup in China—refused for some time to accept layoffs but now acceptthe severance packages. What might explain the change in attitude?

Two economic factors have emerged as being of critical impor-tance in alleviating the social cost of restructuring: the developmentof the national social security system (there has been a significant in-crease in central budget expenditures in social security in recent years:from 1 percent in 1997 to 6.3 percent in 2002), and promotion of the

Collective labor disputes

SOE employment

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0

20

40

60

80

100

120Labor disputes Million workers

1995 1997 1999 2001 2002 2003

FIGURE 1.1Number of People Employed and Number of Collective Labor

Disputes in State-Owned Enterprises in China, 1995–2003

Note: Data for 2003 are extrapolated from half-year data.Sources: CEIC, National Bureau of Statistics of China; People’s Daily.

9. “Labor disputes on the rise,” People’s Daily, November 1, 2004, http://english.people.com.cn/200411/01/eng20041101_162341.html.

10 TABLE 1.2COMPOSITION OF CHINA’S GDP BY OWNERSHIP TYPES, 1998–2003

(percent)

Domestic DomesticState Collectives private Foreign Private private Private

Year controlled (official)a (official)b (official) (official)c (real)d (real)e Agriculture Nonstate

1998 41 22 12 8 20 26 31 18 591999 40 20 13 10 23 26 33 17 602000 39 18 14 12 26 28 36 17 612001 38 17 16 13 29 29 38 16 622002 36 14 19 14 33 31 40 16 642003 34 13 22 15 37 34 44 15 66

a. Official collective firms include “red hat” firms as well as gaizhi firms that are actually private firms.b. Domestic private firms include formally registered private firms and getihu.c. Includes official domestic and official foreign private firms.d. Includes collective firms that are in effect private, assuming they account for half the official collective firms (Gregory, Tenev, and Wagle,2000) and foreign firms that are in reality domestic private firms, assuming they account for a third of the official foreign firms.e. Includes real domestic and real foreign private firms.f. Includes agriculture, which is almost 100 percent private, collective and private firms.SOURCE: Calculated according to a sector-based approach, which derives ownership shares in GDP based on their shares in each of the fol-lowing sectors: farming, forestry, husbandry, and fishery; mining and quarrying; manufacturing; production and supply of electricity, gas, andwater; geological prospecting; transportation and communication services; retail trade and catering services; finance and insurance; real estate;social services; education; scientific research; government services; and others.

Estimates about ownership shares in the GDP of the respective sectors are based on available official data on ownership: shares in sector em-ployment and/or sector output, in effect assuming similar levels of productivity across ownership types. All the original data are from StasticalYearbook of China, 2004.

The methodology and approach are similar to the ones used in Gregory, Tenev, and Wagle, 2000, who provide estimates for ownershipshares in GDP for 1998. Here, calculations for 1998 are redone using latest available data. Differences between the current estimate for 1998and the estimates in Gregory, Tenev, and Wagle are also due to some changes in assumptions; for instance, shareholding firms were includedin the private sector share in the 2000 estimates, while here we count shareholding firms as state controlled.

INTRODUCTION

11

growth of the new private enterprises to absorb workers laid off fromthe state sector.

Gaizhi and the growth of the de novo private sector have trans-formed the structure of the Chinese economy. Table 1.2 presents esti-mates on the composition of China’s GDP by ownership types for theperiod 1998–2003. Over the period the share of the official privatesector in GDP increased from 20 percent in 1998 to 37 percent in 2003.The private sector is now the dominant sector of the Chinese economy.The share of the private sector is probably even larger if we take intoaccount that a significant percentage of the collective firms are in effectprivately controlled and that the private sector is in general more pro-ductive than the other sectors of the economy.

The main question posed by this major, and still unfolding, trans-formation is whether China will succeed in its efforts to reform the statesector and to maintain the delicate balance between tackling its legacyproblems while promoting the new sources of growth.

Focus and Empirical Approach of the Study

To shed some light on this question we need a better understanding ofthe gaizhi process. There has been no systematic study of the magnitude,forms, and consequences of gaizhi. This book aims to fill this gap by look-ing at the types of reforms, their evolution, the key players and the motiv-ations for their decisions, and some outcomes of the gaizhi process.

The analysis is based on a survey of close to 700 enterprises in11 Chinese cities. The survey was conducted from December 2002 toApril 2003. The latest accounting data that most enterprises were ableto provide, however, was for year 2001. The survey adopted a researchstrategy that combines the use of a structured questionnaire given tofirms selected by random sampling, with face-to-face interviews withgovernment officials and enterprise management selected to provideinsights into the privatization process and its outcomes. While the sam-pling provides quantitative data, the interviews reveal rich qualitativeinformation regarding different government policies and various prob-lems that have been encountered in gaizhi.

The Sample Cities. A major feature of the gaizhi process is that it islargely decentralized.10 Therefore, a significant regional variation in theforms and outcomes of gaizhi was to be expected. To account for this

10. This is changing with the establishment of SASAC.


12

variation, cities of different sizes and from different regions were selectedfor the survey and interviews. These 11 cities were, from north to south:Harbin, Fushun, Tangshan, Lanzhou, Weifang, Xining, Zhenjiang,Huangshi, Chengdu, Hengyang, and Guiyang (see figure 1.2).

Harbin and Fushun are located in the northeast, China’s indus-trial powerhouse in the central planning era. Harbin is the capital ofHeilongjiang province, and Fushun is a medium-size city in Liaoningprovince. Tangshan is an old industrial city, the second largest inHebei province, about 120 kilometers north of Beijing. It survived asevere earthquake in 1976. Xining and Lanzhou are two westerncities, the capitals of Qinghai province and Gansu province, respec-tively. Lanzhou is an important industrial base in the northwest,renowned for its heavy chemical industries, while Xining is less in-dustrialized. Chengdu and Guiyang are the capitals of two south-

FIGURE 1.2Geographic Location of the Eleven Sample Cities

INTRODUCTION

13

western provinces, Sichuan and Guizhou, respectively. Weifang is a medium-size city in Shandong province. Its industry is mainly madeup of small and medium enterprises (SMEs). Zhucheng, one of thetwo cities that initiated privatization in China, is located in thisprovince. Zhenjiang, in Jiangsu province, Hengyang, in Hunan province,and Huangshi, in Hubei province, are medium-size cities. Figure 1.3shows the urban population of the 11 cities and their peripheral counties.

Levels of economic development and growth varied among the11 cities. Figures 1.4 and 1.5 compare per capita incomes in 2001and per capita growth rates between 1995 and 2001. With a percapita income of RMB18,900, Zhenjiang was the most affluent city. Harbin was the second wealthiest, with a per capita income ofRMB11,800. Fushun, Lanzhou, Weifang, Guiyang, and Chengduhad incomes in the range of RMB8,000 to RMB11,300. Tangshan,Xining, Hengyang, and Huangshi had the lowest incomes, betweenRMB5,000 and RMB6,000.

Per capita GDP grew at an impressive pace in some of the citiesover the period. Harbin, Hengyang, Guiyang, and Huangshi registered

Total populationUrban population

0 2 4 6 8 10 12

SOURCE: Survey data, provincial bureaus of statistics.

Chengdu

Harbin

Weifang

Hengyang

Tangshan

Guiyang

Lanzhou

Zhenjiang

Huangshi

Fushun

Xining

Millions

FIGURE 1.3Population of the Eleven Sample Cities, 2001


14

double-digit growth rates between 1995 and 2001. Per capita GDPgrowth was also high in Zhenjiang, Fushun, and Chengdu, with growthrates in the range of 9 to 9.7 percent. Tangshan, Lanzhou, and Weifanggrew considerably more slowly, with growth rates less than half thoseof the fastest-growing cities. Xining was the slowest-growing city, withan annual average growth rate of 1.5 percent.

Other economic conditions also varied significantly. Per capitagovernment revenue is an indicator of a city’s fiscal capacity and willaffect how local governments are able to assist workers laid off throughgaizhi. Chengdu had the healthiest per capita government revenue,easily surpassing the second-wealthiest city, Guiyang (see table 1.3). Percapita government revenue was similar across Guiyang, Zhenjiang, andHarbin. Revenue was lower in Fushun, Xining, Tangshan, and Weifang,and lower still in Huangshi and Hengyang. The Hengyang governmentcollected only RMB200 per person in revenue.

The provincial capitals generally had larger services sectors thanthe other cities. The services sector was particularly underdeveloped in

0 5,000 10,000 15,000 20,000


Zhenjiang

Harbin

Fushun

Lanzhou

Weifang

Guiyang

Chengdu

Tangshan

Xining

Hengyang

Huangshi

RMB

FIGURE 1.4Per Capita Income in the

Eleven Sample Cities, 2001

INTRODUCTION

15

Tangshan and Weifang. SOEs still provided most of the employmentin all cities except Xining and Zhenjiang, although the data are notstrictly comparable across localities. The data on industrial output aremore reliable. They show the dominant role of SOEs in the industrialsectors of Fushun, Guiyang, Hengyang, and Huangshi.

The data show that large numbers of workers had been laid offin Fushun and Lanzhou. The category of xiagang represents a typeof interim unemployment whereby an SOE worker has been laid offbut maintains the relationship with the firm. A massive 44 percent ofFushun’s SOE workers and 25 percent of Lanzhou’s SOE workerswere classified as xiagang in 2001. The situation was also of concernin Tangshan, Chengdu, Harbin, and Weifang, where xiagang rates wereall above 7 percent.11

11. Caution needs to be taken when interpreting data on xiagang or unemployedworkers. It is possible that many of these workers have found other jobs but con-tinue to register as xiagang or as unemployed in order to obtain benefits.

0 2 4 6 8 10 12 14


Harbin

Hengyang

Guiyang

Huangshi

Zhenjiang

Fushun

Chengdu

Tangshan

Lanzhou

Weifang

Xining

Percent

FIGURE 1.5Per Capita Growth of Gross Domestic Product in the

Eleven Sample Cities, 1995–2001

16

TABLE 1.3SELECTED ECONOMIC INDICATORS FOR THE ELEVEN SAMPLE CITIES, 2002

Per capita Employment State share in State share in Registered Xiagang Unemployment Realgovernment in services employment industrial output unemployment ratec + xiagang unemployment

revenue (RMB) sector (%) (%)a (%) rate (%)b (%) (1,000) rate (%)d

Chengdu 1,424.8 33.9 63.7 10.8 3.5 7.9 261.3 4.7Fushun 574.4 33.9 55.7 82.0 2.7 43.5 177.1 43.8Guiyang 832.9 35.0 69.1 72.7 4.0 4.0 95.5 4.7Harbin 698.0 33.4 53.0 10.7 3.0 8.8 263.1 5.3Hengyang 199.7 34.3 40.0 53.1 2.6 3.1 111.4 3.0Huangshi 330.7 33.7 60.9 61.2 1.4 n.a. n.a. n.a.Lanzhou n.a. 47.2 80.9 n.a. 2.6 25.0 146.1 22.8Tangshan 511.2 2.3 73.0 28.6 2.8 10.0 81.1 9.1Weifang 482.3 7.3 50.1 13.9 3.0 7.0 144.8 3.4Xining 540.5 58.8 48.0 20.9 3.1 n.a. n.a. n.a.Zhenjiang 813.6 28.8 35.2 18.3 3.6 n.a. n.a. n.a.

a. There are no consistent data for state share of employment for all the cities. The table reports three sets of data for the figure. One is thestate share in total urban employment. Harbin, Hengyang, Lanzhou, and Weifang use this definition. The second is the state share in all wageemployees. Fushun, Huangshi, Tangshan, and Xining use this definition. The third is the state share in the manufacturing sector. Chengdu,Guiyang, and Zhenjiang use this definition.b. The unemployment rate is registered unemployment as a percentage of total city employment.c. The xiagang rate is the number of xiagang workers as a percentage of total SOE employees.d. The real unemployment rate is the sum of registered unemployed and xiagang workers in the city labor force.SOURCES: National Bureau of Statistics of China, China’s Statistical Yearbook: 2002; authors’ estimates.

INTRODUCTION

17

The official unemployment rates (the share of registered un-employment in the total workforce) in these cities were lower than thexiagang rates. Unemployment is undesirable for both workers and thegovernment, because it severs the links workers had to the old danwei(work unit) and often leads to social unrest. Most unemployed workersare therefore classified as xiagang and placed in a reemployment cen-ter (see chapter 4 for details). To the extent that “xiagang” and “un-employment” both refer to the situation of being laid off, it is sensible tocombine those two categories to estimate the real unemployment rate ina city. This is reported in the last two columns of table 1.3.

The large cities of Chengdu and Harbin registered the highestnumbers of total unemployed, with more than 260,000 unemployedworkers. Fushun had the worst unemployment rate, at 43.8 percent,followed by Lanzhou with 22.8 percent. This was consistent with thelarge numbers of xiagang workers and the dominance of the state sec-tor in the two cities.

Sampling Methodology and Interviews. The reform of China’s SOEsbegan in earnest in 1996 after the experiences of Shunde and Zhuchengwere supported in the central government policy of “keep the large andlet the small go.” For this reason, it was decided to survey all the indus-trial firms that were owned by the 11 city governments at the end of 1995before gaizhi began. The survey was administered by the SETC of theState Council and its counterparts in the sample cities. The city economicand trade commissions (ETCs) sent out around 1,100 questionnairesto firms and 683 were returned.

The intention of the survey was to sample current and formerindustrial SOEs owned by municipal governments; however, be-cause district and county governments were involved in the surveyin some cities, some district and county firms, including collectives,were included in the sample. These firms were retained in the sam-ple because many of them employed urban workers and had similarpractices to the SOEs in terms of relationship with the local govern-ment, internal governance structure, employment, and remunera-tion. A small number of questionnaires were also sent to trading andservices firms.

Many of the firms had changed considerably since 1995. All sig-nificant gaizhi events in a firm’s history, such as bankruptcy, mergers,acquisitions, and spin-offs, were recorded. In terms of data collected, ifthe enterprise had merged with another firm, the new firm was surveyedand data were recorded from the year of the merger. If the firm had


18

been split up, the largest of the new active firms was surveyed and datawere recorded from the year of the split. It was often the case that theold firm had moved production to a new spun-off firm, leaving the oldfirm only with the name, debts, and the burden of retirees. As a result,the new firm was typically the only active firm. The survey strategy re-sults in an unbalanced panel of firm data, because some firms had ashorter life span.

Sample selection has some special complications in government-assisted surveys. Self-selection by firms, and selection on the part of thecity ETC, can bias the results. For instance, local officials may put pres-sure on firms with close ties with the government, usually larger en-terprises and nonprivatized SOEs, to fill in the questionnaire, causingoversampling of nonprivatized SOEs. Where bias may be a problem,national data will be used to support the discussions on the extent ofgaizhi, or adjustments will be made to the data.

In addition to the survey, research teams interviewed governmentofficials, bank managers, and firm managers in seven cities: Harbin,Tangshan, Chengdu, Guiyang, Chongqing, Zhenjiang, and Hengyang.12

In each city the research team met with officials from governmentagencies such as the ETC and the bureaus of finance, labor and socialsecurity, and land administration. This gave the teams valuable in-sights into government policies and the methods of gaizhi in each city.The information also served as a reference point for the firm inter-views. The teams also met with the local managers of the four majorbanks and the city commercial banks.

A total of 270 firms were interviewed across the seven cities. Thefirm interviews provided important qualitative information regardinggaizhi. They revealed the different attitudes of SOE managers in dif-ferent cities, and provided examples for some of the patterns analyzedin the study.

Sample Distribution. The distribution of the sample firms, shown intable 1.4, varies across the cities because of differing structures of en-terprise ownership and differing response rates. Response rates werepoor in Fushun, which returned only 10 questionnaires, Weifang, andLanzhou. In Xining the total number of firms was small. It is importantto keep in mind the uneven distribution of the sample whenever geo-graphic distribution is important to the interpretation.

12. Chongqing was not on the list of the sample cities, but a few firm interviewswere conducted there.

INTRODUCTION

19

Most of the sample firms are categorized by Chinese statistics asSMEs.13 A comparison with national statistics, however, suggests thatthe sample average was nearly four times the national average over theperiod 1998–2001 (see figure 1.6). This might be a result of the re-structuring brought about by gaizhi but could also reflect a sample biastoward larger firms. This is likely to be because the survey includedmainly industrial firms owned by city governments, and these firms areusually larger than firms owned by district and county governments,which would be included in the national survey. There might also havebeen a higher response rate from the larger firms with connections tocity governments. Within the sample firms, there was a slight declinein the average number of workers employed by firms from 1995 to2001, as shown in figure 1.6.

Most of the sample firms were in manufacturing, especially in petro-chemicals, electronics, machinery, and textiles (see figures 1.7 and 1.8).This distribution agrees with the national data (see figure 1.9).

13. According to a recent SETC document, an industrial firm is an SME if oneof three conditions is met: (1) annual sales revenue is less than RMB100 million;(2) registered capital is less than RMB20 million; or (3) fewer than 2,000 workersare employed

TABLE 1.4DISTRIBUTION OF SOES AND FORMER SOES SURVEYED

IN THE ELEVEN SAMPLE CITIES, DECEMBER 2002–APRIL 2003

Number of Percent of firms sample firms

Harbin 120 17.6Fushun 10 1.5Tangshan 59 8.6Weifang 30 4.4Lanzhou 39 5.7Xining 26 3.8Huangshi 79 11.6Zhenjiang 70 10.2Hengyang 57 8.3Guiyang 149 21.8Chengdu 44 6.4Total 683 100.0

SOURCE: Survey data.


20

A comparison with a national survey by the SETC, released in early2002, suggests that the 11-city survey oversampled non-gaizhi SOEs.In the 11-city survey, 370 firms (54 percent) reported having gonethrough gaizhi by the end of 2002. The data reported to the SETC by54,644 small and medium-size industrial firms at the end of 2001showed that 86.1 percent of firms had finished gaizhi by the end of 2001.These firms made up approximately 62 percent of all registered indus-trial SOEs and can be regarded as representative of the SOE population.

In other ways the 11-city survey is representative of the experi-ences that SOEs have had with gaizhi. The extent of different forms ofgaizhi reported by the survey matches the findings of national surveysin 1998 and 2001 (see figures 1.10, 1.11, and 1.12). The 1998 survey,carried out by the National Bureau of Statistics, sampled 57,881 firms,or approximately 67 percent of registered industrial SOEs.

The 11-city survey showed that 30 percent of gaizhi cases involvedeither a public offering or an internal restructuring. A similar share(27 percent) had become employee shareholding firms, 28 percent offirms had been sold or leased out, 11 percent were bankrupt, and the re-maining 4 percent had become joint ventures. The 1998 survey re-vealed broadly similar patterns. The two categories of “introducingnew investment” and “spinning off” were largely equivalent to the sur-

0

200

400

600

800

1,000

1995 1996 1997 1998 1999 2000 2001

Sample average National average

Number of workers

SOURCES: Survey data and China Industrial Yearbook, various years.

FIGURE 1.6Average Number of Workers in the Sample Firms

and in Firms Nationally, 1995–2001

21


FIGURE 1.7Percentage of Sample Firms in Primary Industries,

Manufacturing, Utilities, and Services, 2001

Primary5%

Manufacturing81%

Utilities1%

Services13%


FIGURE 1.8Distribution of Sample Firms by Industrial Sector, 2001

Food and tobacco10%

Textiles14%

Furniture andprinting

6%

Petrochemicalindustry

25%

Machinery19%

Electronics16%

Steel and metalproducts

6%

Medicines4%

INTRODUCTION

FIGURE 1.9Distribution of Firms by Industrial Sector

in 2001 National Survey.

Food and tobacco12%

Textiles16%

Furniture andprinting

10%

Petrochemical industry25%

Machinery15%

Electronics10%

Steel and metalproducts

10%

Medicines2%

Note: Including all SOEs and any other firm with a sales volume of more than RMB5 million. Source: China Industrial Yearbook: 2002.

FIGURE 1.10Forms of GAIZHI in 375 Sample firms, 2001

Publicoffering

7%

Internalrestructuring

23%

Bankruptcy andreorganization

11%Employeeshareholding

27%

Leasing15%

Jointventure

4%

Open sales13%

Source: Survey data.

22


23

FIGURE 1.11Forms of GAIZHI in 2001 National Survey

Merger14%


18%

Bankruptcy11%

Employeeshareholding

19%

Leasing14%

Joint ventureand other

13%

Open sales11%

Source: SETC.

Newinvestors

23%

Spinoff7%

Merger10%

Bankruptcy5%


21%

Open sales7%

Leasing20%

Jointventure

3%

Other4%

Source: SETC.

FIGURE 1.12Forms of GAIZHI in 1998 National Survey

INTRODUCTION

vey’s categories of “internal restructuring” and “public offering.”Together they made up 30 percent of all gaizhi cases, which is what isfound in this study. The figure for employee shareholding, 21 percent,was also comparable to the 27 percent reported here.

The share of sales and leases (27 percent) was almost identical tothis study. The share of joint ventures was 3 percent, close to the 4 per-cent found in this study. Incidences of bankruptcy differed: only 5 per-cent of the gaizhi cases were bankruptcies in the 1998 survey, comparedwith a share of 11 percent in this study. The difference might be becausethe 1998 survey had a separate category for mergers, and 10 percent ofthe gaizhi cases were reported as mergers, while this study combinesmergers with bankruptcies. The results of the 2001 survey differed again,but the reporting of sales and leases matched both of the other surveysand the incidence of bankruptcy was similar to the 1998 survey.

The Structure of the Study

This study applies descriptive and econometric analysis to survey andofficial statistical data to examine the progress of gaizhi over the yearsand across regions. Chapter 2 discusses the main players in the process,their motivation and incentives. Chapter 3 examines the process: theforms, scope, and timing of gaizhi. It also looks at the incentives for firmsand local governments to undertake a specific form of gaizhi.

Chapters 4, 5, and 6 look at the outcomes of gaizhi. Chapter 4 fo-cuses on employee issues. It outlines China’s social security system anddiscusses its function in the gaizhi process. Chapter 5 examines changesin the corporate governance of gaizhi firms. It studies the dynamics ofthe share structure and the changing roles and influence in the courseof gaizhi of various stakeholders. Detailed statistics are presented onthe distribution of control rights inside the firm, comparing gaizhi andnon-gaizhi firms. Chapter 6 looks at the impact of gaizhi on firm per-formance. Chapter 7 concludes the study by discussing issues relatedto the fairness and efficiency of the gaizhi process, particularly concern-ing the role of management buy-outs and outside investors. These is-sues are at the center of a lively public debate in China, which is likelyto influence future Chinese policies toward gaizhi.


24

25

2The Main Players in Gaizhi

Gaizhi1 is a comprehensive transformation in which many parties havea stake. The process is inevitably complicated and involves compromisesamong different stakeholders. This chapter looks at the main players ingaizhi. It examines their incentives, concerns, and objectives, and con-siders how they determine the timing and scope of the process. Varioushypotheses on the interactions between incentives and the scope of gaizhiare expounded and tested empirically.

How Far Has Gaizhi Progressed?

National data show that there has been a rapid decline in the number ofstate-owned or state-controlled enterprises across all regions in China.Between 1996 and 2001 the number of SOEs declined at an average an-nual rate of 9.1 percent, from 207,166 to 128,445 (see figure 2.1). Thesharpest fall was in 1998. An SOE may cease to exist for three main rea-sons: because it is privatized, merged with another SOE, or liquidated.The data show that around 42 percent of the SOEs that existed in 1996were privatized, merged, or liquidated over the period.

The western provinces were the most active in the restructuringprocess. Approximately half of the region’s SOEs had been reformed bythe end of 2001, 22.4 percent of them in 1997 alone. Chongqing wasthe main engine behind this performance, with 76 percent of the city’sfirms being reformed in that one year. By the end of 2001, only 15 per-cent of Chongqing’s SOEs were still fully state-owned or majority con-trolled. The eastern provinces were slightly below the national averagebefore 1999 and then slightly above after 1999. The central provincestracked the national trend throughout the period.

1. The term “gaizhi” has a broader meaning than simply “privatization,” in thatgaizhi includes privatization but also many other methods of restructuring.


26

The average rate of reform of industrial firms was faster than theaverage for all firms (see figure 2.2). A total of 59 percent of industrialfirms were privatized, merged, or liquidated over the period 1996–2001.By 2003, the percentage had increased to 70.

The two national surveys of industrial firms undergoing gaizhi,mentioned in chapter 1, used the end of 1996 as their reference point,providing data that can be compared with the national statistics onSOE reform. Among the 57,881 firms sampled in the 1998 survey,24 percent were reported to have finished gaizhi by the first quarter of1998, 27 percent were in the process of gaizhi, and 26 percent wereplanning to start gaizhi in the coming year. As shown in figure 2.2, thenumber of industrial SOEs in 1997 was 87 percent of the number in1996. Therefore, 13 percent of industrial SOEs had been privatized,merged, or liquidated by the end of 1997. From the results of the 1998national survey reported in chapter 1 we know that 15 percent of allgaizhi were mergers and bankruptcies. This implies that of the 13 per-cent reduction in the number of SOEs, 3.6 percent was due to mergersand bankruptcies and 9.4 percent to privatizations. Therefore, close

40

60

80

100

Percent

1996 1997 1998 1999 2000 2001

East

CentralWest

FIGURE 2.1State-Owned Enterprises in China as a Percentage

of 1996 Number, by Region, 1996–2001

Note: Number of firms fully owned or controlled with dominant shares by the government. East: Liaoning, Beijing, Tianjin, Shandong, Shanghai, Jiangsu, Zhejiang, Fujian, and Guangdong. Central: Heilongjiang, Jilin, Inner Mongolia, Shanxi, Hebei, Henan, Hubei, Anhui, Hunan, Jiangxi, Hainan, and Guangxi. West: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, Tibet, Yunnan, Guizhou, Sichuan, and Chongqing.Source: China Industrial Statistical Yearbook: 2002.

National

THE MAIN PLAYERS IN GAIZHI

27

to 40 percent of all the gaizhi cases had been privatizations by the timeof the 1998 survey.

The share of gaizhi firms would have reached 80 percent by 1999if the gaizhi plans revealed in the 1998 survey had all been carriedout. This prediction seemed to have been fulfilled, as the 2001 surveyshowed that 86 percent of the 54,644 sample firms had completedgaizhi by that year. Of all the completed cases of gaizhi, 25 percentwere through mergers and bankruptcies, as reported in chapter 1.This estimate implies that 21.5 percent of all industrial SOEs had gonethrough mergers or bankruptcies by 2001. The national data presentedin figure 2.2 show that 59 percent of all industrial SOEs had been pri-vatized, merged, or liquidated by 2001. Therefore, about 38 percentof all SOEs had been privatized by that year. This, in turn, implies that44 percent of all gaizhi cases were privatizations. The estimates about

0

20

40

60

80

100

120

1996 1997 1998 1999 2000 2001 2002 2003

National

EastCentral

West

Percent

FIGURE 2.2Industrial State-Owned Enterprises in China as a

Percentage of 1996 Number, by Region, 1996–2003

Note: Number of firms fully owned or controlled with dominant shares by the government. East: Liaoning, Beijing, Tianjin, Shandong, Shanghai, Jiangsu, Zhejiang, Fujian, and Guangdong. Central: Heilongjiang, Jilin, Inner Mongolia, Shanxi, Hebei, Henan, Hubei, Anhui, Hunan, Jiangxi, Hainan, and Guangxi. West: Xinjiang, Qinghai, Ningxia, Gansu, Shaanxi, Tibet, Yunnan, Guizhou, Sichuan, and Chongqing. Source: China Statistical Yearbook: 2004.


28

the extent of privatization are lower-bound estimates, as often bank-ruptcies do not involve liquidations. In any event, the estimates sug-gest that privatization has been significant in China and that gaizhi hasbecome more radical in recent years.

The 11-city survey reveals a similar pattern. Figure 2.3 shows thepace of reform in four cities and the sample average over the period1995–2001. Just as the national data illustrated, restructuring and pri-vatization had barely begun in 1996 and 1997, but they acceleratedconsiderably in 1998 and 1999. The most dramatic case was Huangshi,where nearly 70 percent of SOEs disappeared in 1999 alone. The mostsluggish city was Lanzhou, where reforms and privatization startedto take off in 1997 but then stalled, leaving about 80 percent of SOEsstill in state hands.

Generally, the southern cities privatized at a faster rate than thenorthern cities. With the exceptions of Hengyang in the south, all thesouthern cities had a lower percentage of nonprivatized SOEs thanthe average and all the northern cities had a higher percentage thanthe average (see figure 2.4). On average, the number of SOEs in thesample cities decreased by 50 percent in the period 1995–2001.

The firm survey shows that the timing of gaizhi broadly followeda similar pattern to that described in the national statistics and surveys.There were 370 firms that had undertaken gaizhi by the end of 2002

Huangshi

Chengdu

HarbinLanzhou

Average

0

20

40

60

80

100

120Percent

1995 1996 1997 1998 1999 2000 2001

FIGURE 2.3Privatized State-Owned Enterprises as a Percentage of

1996 Number in Four Chinese Cities, 1995–2001



29

(see figure 2.5). Less than 10 percent of them had completed gaizhi by1994, but by the end of the period about 80 percent of firms had doneso. Although the pace slowed in 2002, more than 30 firms completedgaizhi that year.

The 11-city survey shows geographic differences in the pace ofgaizhi. Figure 2.6 groups the 11 cities into three regions: northern(Harbin, Fushun, and Tangshan), western (Xining, Lanzhou, Chengdu,and Guiyang), and southern (Weifang, Zhenjiang, Huangshi, andHengyang) and shows the number of gaizhi cases in each region eachyear. Although the southern region did not start gaizhi as early as theother two regions, its cities moved much faster in terms of numbers offirms and the scope of gaizhi. In the southern cities, gaizhi grew steadilyto reach a peak in 2000. The pace in the northern and western cities hasbeen slower and more variable.

0 4020 60 80 100


Hengyang

Lanzhou

Guiyang

Fushun

Average

Harbin

Zhenjiang

Tangshan

Weifang

Xining

Chengdu

Huangshi

Percent

FIGURE 2.4Number of State-Owned Enterprises as a Percentage of

1996 Number in Sample Chinese Cities, 2001


30

0

20

40

60

0

20

40

60

80

100Number of firms Cumulative percentage

FIGURE 2.5Number of Firms Undertaking GAIZHI in

Sample Cities, Before 1990–2002


Before1990

1992 1994 1996 1998 2000 2002

Cumulative percentage

Number of firms

NorthWestSouth

0

10

20

30

40Number

FIGURE 2.6Number of GAIZHI Cases by Region, Before 1990–2002


Before1990

1992 1994 1996 1998 2000 2002


31

The Key Participants

Because gaizhi involves a comprehensive transformation of the statesector, which had been the foundation of the Chinese economy, a numberof players have stakes in the process. The seven key participants in gaizhihave been the central government, the local government, the manage-ment, employees, creditors, outside investors, and the public. The inter-actions among these players shape the form of gaizhi undertaken.

The Central Government. In the early stages of urban reform, gaizhiwas a bottom-up process. During his tour of southern China in the springof 1992, Deng Xiaoping encouraged several localities to experimentwith SOE reform. Gaizhi was given greater prominence at the end of1992 by the 14th Party Congress, when the decision was made to builda socialist market economy with Chinese characteristics. Although thecentral government had no grand plan to reform the whole SOE sector,more localities were encouraged to undergo gaizhi.

For several years the central government did not make officialpronouncements on the reform initiatives, but simply observed theconsequences. The zhuada fangxiao policy in 1995, which sanctionedlocal initiatives, was a result of those observations and marked the be-ginning of active support by the central government for gaizhi. The15th Party Congress in 1997 formally acknowledged the role of theprivate sector by stating that private enterprises were an indispens-able part of the national economy. In 1999 the National People’sCongress (NPC) changed the Constitution to endorse the decisionmade at the 15th Party Congress. Further amendments to China’sConstitution, including explicit clauses on protecting human rightsand private property, were passed at the full session of the NPC inMarch 2004.

The central government’s support for gaizhi came from its desireto transform the planned economy into a market economy. The Chinesegovernment realized in the early 1980s that price liberalization andSOE reform were the two fundamental tasks of that transformation. Priceliberalization was completed in the early 1990s (the most significantevent was perhaps the abolition of food quotas and price controls in1993), after which SOE reform became the government’s foremostconcern. Although efforts to improve SOE efficiency had continuedsince the mid-1980s, the government realized that gaizhi had the mostchance of solving the SOE problem. The central government also real-


32

ized at the beginning of gaizhi that “letting the small go” would improveits fiscal situation. Under the decentralized fiscal system, the centralgovernment has a strong incentive to transfer the management of smallerfirms to local government.

The use of the term “gaizhi,” instead of “privatization,” illustratesthe cautious approach taken by the central government. Even that termdoes not appear frequently in the official media, however, because of adesire to avoid an ideological backlash. The central government’s mainconcerns are that gaizhi may lead to the loss of state assets (bank loans,in particular), result in massive unemployment, and cause social unrest.

In terms of state assets, the real problem is the conflict of interestbetween the central government and local governments. While mostSOEs belong to local governments, the four major commercial banksbelong to the central government. These four state banks are by far thelargest creditors to the SOE sector. Since the Chinese fiscal system ishighly decentralized, and local and central budgets are largely separate,the central government is able to guard its own interests if it protectsthe banks.

The central government is very keen to maintain social stabilityand keep unemployment to a minimum. It therefore pressures localgovernments to adopt every possible means to maintain employment,and thus social stability, during gaizhi. This creates a conflict of inter-est between the central and local governments over responsibility forthe costs of gaizhi.

The interests of the various government agencies involved in gaizhimay differ and in some cases be contradictory. Prior to the last gov-ernment reorganization, the SETC was in charge of SOE operations. Itfavored a faster pace of privatization and more generous deals for themanagement. The Ministry of Finance (MOF) was more concernedabout whether the value of the assets could be fully recovered. The mostsignificant conflict is between the Ministry of Labor and other gov-ernment agencies. The Ministry of Labor is responsible for unemployedand xiagang (temporarily laid-off) workers, so it resists rapid privati-zation. The trade unions are allied with the Ministry of Labor and alsoemphasize the rights of workers. These conflicts translate at the locallevel as well. The slow progress of gaizhi in some cities can be attributedto the conflicts of interest among the different government agencies,especially in cases when the government needs to provide compen-sation to redeploy workers. The establishment of the State-OwnedAssets Supervision and Administration Commission (SASAC) had as


33

one of its objectives the creation of a government agency that wouldassume ownership over SOE reform. While the new agency addressessome of the issues related to fragmented authority over the manage-ment of state assets, it cannot eliminate all the potential conflicts be-tween various government agencies regarding SOE reform.

In summary, the central government sees gaizhi as a necessary steptoward a market economy, but its interests are more complicated thanthey appear. The conflict of fiscal interest between the central govern-ment and local governments plays an important role in shaping themethods and speed of gaizhi in the country.

The Local Government. The local government plays three roles in re-gard to local SOEs: as the original owner of the SOEs,2 as an em-ployer and therefore the provider of social stability, and as a collectorof tax revenues used to provide public services. These three roles arenot always complementary, but the top priority of local governmentis stability.

As owners of state firms, local governments have a big stake inSOE assets. The wage arrears, bank liabilities, and tax arrears of poorlyperforming SOEs place a burden on local governments, however. Inaddition, with progress in market liberalization local governmentshave less control over firms as managers have become more powerful.Managers are more prone to steal from firms if they have significantcontrol rights but no ownership rights.

Local governments must weigh the costs and benefits of re-taining ownership against those of letting the firm go into privatehands. If the firm is retained, there is a risk that its assets will dissipatethrough poor management and theft, but if the firm is privatized thegovernment may obtain some value out of the assets. More importantly,the government can shake off the firm’s liabilities after privatizationand may gain from increased tax revenues if the firm’s performance im-proves. Although these incentives may vary among the cities, many localgovernments have made the pragmatic choice to privatize the SOEs.

In many cases of gaizhi, the local government is unable to cashin on the sale of state assets. Many SOEs already have negative net

2. In theory the central government owns every SOE, but in reality SOEs are ef-fectively owned and controlled by local governments. The official document of the16th Communist Party Congress signaled a formal transfer of ownership to lo-calgovernments.

assets at the time of gaizhi. Many owe wages, social security pay-ments, taxes, or health insurance premiums. These debts have to besettled before the firm is privatized. Gaizhi often involves layoffs,and layoffs create obligations on local governments to support work-ers who lose their jobs. With a shortage of cash, the local governmenthas to use the firm’s assets to meet these obligations. A constraint ongaizhi practices is the government’s wish to maintain employmentand reduce the risk of social instability. This is the overriding con-cern of local governments.

The Management. The firm’s managers usually have a strong incen-tive to support privatization if they believe that they are likely to bebeneficiaries in terms of increased autonomy and investment gains.Prior to gaizhi, the authority of managers rests on their ability to pleaseboth the government and the firm’s employees. Although managers’control rights may be strong, they are not backed by any ownershiprights and this creates a sense of insecurity. After gaizhi, managers mayhave, in addition to their decision-making power, ownership stakesthat could bring rewards if the firm is run successfully.

Managers may face conflicts of interest when they perceive them-selves as prospective buyers of the enterprise. Some managers may de-liberately run the firm down before privatization in order to lower theprivatization price, or they may strip the assets out of the firm. Questionshave been raised about whether managers should be allowed to buy thefirm at a discount to firm valuation, as those same managers may havebeen responsible for the failure of the SOE.

Managers bear a significant risk when they buy a firm, especiallywhen they pay cash. Increasingly they are competing against outsidebuyers, who have the financial resources to redeploy redundant work-ers and inject new investment into the firm but may not want to retainthe old management. If an outside buyer is competing to purchase thefirm, the managers may oppose gaizhi.

Employees. Many SOEs have tended to overhire, despite the fact thatthey operate in traditional industries where markets have been shrink-ing. Privatization is an opportunity to downsize the workforce, but thenlaid-off workers become a burden on the government. Thus employeesoften have mixed feelings toward gaizhi. It may be seen as the only wayto revive the firm, but there is also the risk of redundancy. If gaizhi must


34


35

proceed, redeployment after gaizhi becomes the employees’ centralconcern.

Workers have considerable bargaining power because they areable to resort to collective, sometimes violent, action to force thegovernment to satisfy their demands. Their position has also beenstrengthened by the government’s aversion to social unrest. The cen-tral government requires that every gaizhi plan must be approved bya conference of the employees. The Supreme Court requires that thesettlement of labor claims is the first priority when a firm undergoesgaizhi or bankruptcy. Employees therefore are in the position of secured creditors vis à vis their firms. The top priority given to work-ers’ claims tends to dampen employees’ incentives to monitor man-agement for asset stripping during and before gaizhi.

In most localities, employees who will be continuing with thefirm receive compensation, usually a number of free shares, for los-ing their status as state employees. As such, they receive a share ofthe profits on top of their regular wages and thus can be a positive forcepushing for gaizhi.

Creditors. Banks have a huge stake in gaizhi, as most gaizhi firms arein serious debt. Because many firms attempt to evade the repayment ofdebt—for instance, by falsely declaring bankruptcy and registering anew firm—banks often see gaizhi as being synonymous with attemptsat debt evasion. Banks—and the national banks, in particular—havean intrinsic interest in monitoring gaizhi to ensure that the value oftheir collateral is not eroded in the process. In this sense, their intereststend to coincide with the central government’s concerns about the lossof state assets during gaizhi.

Although the four major state commercial banks are owned bythe central government, they are vulnerable because the firm and thelocal government often collude to squeeze their interests. Furthermore,the central bank leaves little room for commercial banks to write offloans or renegotiate payments. Banks’ quotas for write-off are thereforea constraint on the pace and scope of gaizhi.

The debt problem does not lend itself to a simple solution, suchas strengthening the role of the banks during or after gaizhi. The ac-cumulation of nonperforming loans is a consequence not only of thesoft budget constraint faced by SOEs and their poor performancebut also of failed government policies. The central government hasdeliberately channeled bank loans to priority sectors, and many of


36

the selected projects ended in failure. Local governments also per-suaded banks to issue loans to favored firms. When state firms aredeclared bankrupt, social obligations usually have priority over securedcreditors, and commercial debts are seen as a problem of secondaryimportance.

Outside Investors. Private business owners, other SOEs, and foreigncompanies did not play a large role in the early stages of gaizhi, butthe participation of outside investors—particularly private businessowners—has grown in recent years. A large number of private entre-preneurs have acquired SOEs in the belief that they can make moneyif the firms are properly managed.

Local governments used to play an active role as far as the in-volvement of outside investors in the gaizhi process is concerned. Theycan engage in matchmaking or instruct successful SOEs to take overloss-making SOEs. According to unofficial estimates, about half of allmergers involving SOEs as outside investors are administered mergers.This practice has diminished in recent years, partly as a result of re-forms and a more liberal mergers and acquisitions (M&A) regime, andpartly due to the fact that local governments have to compete moreand more with each other to attract outside investors of all stripes andcolors.

Outside investors are more likely to care about efficiency, and towant to lay off workers and strengthen internal management, thanare existing managers who take over firms. Because they bring newinvestment into the firm, outside investors are in a strong position tonegotiate with the government on redeploying the employees. The oldmanagement typically has a more benign attitude toward employeesbecause of their prior relationship with the firm.

The Public. The central government has maintained a low profile overprivatization, and the general public is not well informed about thegaizhi process. Most people, however, associate gaizhi with unemploy-ment problems and with the need for structural adjustments in thestate sector. The public—and intellectuals, in particular—worry aboutequity issues and the loss of state assets. Heated debates have brokenout in several cases and have influenced the attitude of the central gov-ernment toward gaizhi. The most notable occasion was the 1997–98debate following the 15th Party Congress. Following the rapid privati-zation of TVEs throughout the country, many of which were sold to their

managers, the intelligentsia expressed concerns as to whether such mas-sive sales to the elite were fair to ordinary workers (Qin 1997). Theirdoubts were quickly echoed by the central government, which cooledthe discussion of gaizhi and began to talk about preventing the loss ofstate assets.

All seven players discussed above exert an influence on gaizhi. Thecentral government has been a strong force in pushing for gaizhi sincethe mid-1990s, because of its desire to establish a market economy toease the burden on the state budget. Aware of the potential loss of stateassets, especially bank loans, it has also been keen to regulate the gaizhiprocess through laws, directives, and other regulations.

At the local level, gaizhi promises to exempt the local governmentfrom responsibility for loss-making SOEs and gives managers con-trolling ownership, which explains why both parties support gaizhi.Employees are often divided. Those who expect to stay in the firm usu-ally favor gaizhi because they gain an ownership stake, while thosewho expect to be laid off often oppose it. Outside investors welcomegaizhi because it provides them with new investment opportunities.The intelligentsia is actively involved in debates over gaizhi and can beinfluential with the public and the central government. Creditors aremostly passive in the gaizhi process, despite their large financial stakein gaizhi firms. This is mainly because of China’s banking regula-tions, bankruptcy law, and current government policies. A salient fea-ture of China’s reform policies has been the focus on avoiding losses. Inthe case of gaizhi, this has meant socialization of losses through thebanking system and the budget.

It is important to note, however, that gaizhi is a dynamic processand that the incentives of the parties involved may shift as the processevolves. The identity of the party initiating restructuring reveals a lot about who is supporting gaizhi at various stages of the process.According to survey results, local governments initiated around two-thirds of the first round of gaizhi, while enterprise managers initiatedaround one-fourth (see figure 2.7). In the most recent round, however,enterprise managers were the leading party, initiating about half of allgaizhi cases (see figure 2.8). Interviews revealed that local governmentssupported strengthening the position of managers in order to addefficiency and to raise tax revenues from the firm.

The fact that the employees seldom initiated gaizhi is not sur-prising, given that it often results in massive layoffs. Despite the gov-ernment regulation that any gaizhi plan must be approved by the


37


38

employees or their representatives, around one-quarter of gaizhi firmsreported that employees did not play a significant role in the gaizhiplan. In more recent rounds of gaizhi the incidence of employee ini-tiatives was higher (see figure 2.8), perhaps because prior experiencewith gaizhi has reduced uncertainties for workers who have pre-served their employment status.

Theories on Incentives for Gaizhi

In contrast to Eastern Europe and the former Soviet Union, where pri-vatization was initiated in a top-down fashion to speed the politicaltransformation (Boycko, Shleifer, and Vishny 1996), privatization inChina emerged spontaneously. Gaizhi was initiated by several locali-ties and later sanctioned by the central government. Different regionshad different experiences with gaizhi. An interesting question is, whichpolitical, social, and economic conditions are most conducive to gaizhi?And how did the interactions among the main players influence thecourse of gaizhi?

Firmmanagement

24%

Employees2%

Localgovernment

68%

Centralgovernment

1%

Other5%


FIGURE 2.7Party Initiating the First Round of GAIZHI


39

Four theories have been advanced to explain the incentives forgaizhi: the efficiency hypothesis, the financial crisis hypothesis, the fiscaldecentralization hypothesis, and the market liberalization hypothesis.

Efficiency Hypothesis. The efficiency hypothesis asserts that privati-zation is initiated by the government’s desire to improve firm effi-ciency. Under this theory, gaizhi is promoted by the desire of thecentral government to encourage a market economy and improve firmefficiency, thereby increasing tax revenue. Glaeser, Johnson, and Shleifer(2001) argue that rational political systems lead to the selection ofefficient economic institutions. This can explain the central government’ssupport for gaizhi since the mid-1990s, but it cannot explain the timingof gaizhi: if efficiency was the government’s main concern, why did itnot start gaizhi in the 1980s when the inefficiencies in the SOE sectorbecame widely known?

One way of measuring the efficiency of a state firm is by compar-ing the average value-added per unit of assets of an SOE against those

Firmmanagement

49%

Employees7%

Localgovernment

40%

Centralgovernment

0%

Other4%


FIGURE 2.8Party Initiating the Last Round of GAIZHI


40

of private firms in the sector (Li and Liu 2004). The larger the effi-ciency gap between the two, the more likely the SOE is to gain fromprivatization and so the more likely it is to undertake gaizhi.

Financial Crisis Hypothesis. In examining privatization programs inindustrialized countries, some researchers have found that privatiza-tion was often induced by a financial crisis (see, for example, Northand Weingast 1989). China’s experience suggests a similar rationale.Financial crisis was cited by local officials in Shunde as one of the mostimportant reasons why SOEs in that city were privatized (Yao 2004).Gaizhi took off in the mid-1990s when profitability in the state sectorwas declining rapidly. The profit rate of SOEs declined from 3.8 per-cent in 1994 to 1.5 percent in 1996 and 1997, and the return to capi-tal fell to around 0.7 percent.3 This theory does not explain whyZhucheng and Shunde were such early reformers (Huang and Wei2001; Yao 2004), or why progress was so slow in the northeast andnorthwest where there were a large number of failed SOEs.

The financial situation both of the local government and of itsfirms is likely to be important to this hypothesis. The per capita gov-ernment revenue of a city is adopted to indicate a city’s financial ca-pacity, and two lagged firm performance indicators—the debt-equityratio and value-added output per worker—control for the financialperformance of firms. If the hypothesis holds, firms with poor finan-cial performances and those in cities with a weak financial situationwill be privatized first.

Fiscal Decentralization Hypothesis. Other theories emphasize the im-portance of fiscal decentralization in inducing privatization (Li, Li,and Zhang 2000; Dougherty and McGuckin 2002). The decentral-ization of fiscal power creates strong incentives for local governmentsto compete for tax bases and leads to the privatization of inefficientSOEs. Dougherty and McGuckin (2002) find evidence of greater pri-vatization at the level of the county and below. This does not explain,however, why privatization did not occur in the 1980s when the in-troduction of the fiscal responsibility system allowed local governmentsto retain more of their tax revenues.

3. Although SOE profitability has increased since 1998, the petrochemical sectorcontributed more than half of the increase.


41

Fiscal decentralization does not necessarily lead local govern-ments to be more responsible and may create adverse effects such asregional protection. When comparing fiscal decentralization in Chinaand Russia, Blanchard and Shleifer (2000) find that China performsbetter because central government control mitigates the adverse effectsof fiscal decentralization. Blanchard and Shleifer do not, however, con-sider the problem of the central government’s incentives to pass downdifficult tasks to local governments.

The same top-down relationship exists between provincial gov-ernments and county governments. Yao and Yang (2003) use the ex-amples of rural tax reform and privatization to show how higher-levelgovernments pass down responsibilities, inducing lower-level govern-ments to behave more predatorily toward the private sector.4 Thisresults in the underprovisioning of public goods on the part of higher-level governments (Wang and Wang 2001). Chen, Hillman, and Gu(2002) found that local governments became more predatory after the1993 fiscal reform that established an American-style central–localsystem of tax sharing.

Market Liberalization Hypothesis. A fourth set of theories links pri-vatization to market liberalization. Tian (2001) shows that marketliberalization changes the cost-benefit balance of various forms ofownership relative to central planning. Central planning creates a sit-uation where local government ownership is optimal because the localgovernment has the advantage over entrepreneurs in accessing re-sources. As the planned economy gives way to the market economy,entrepreneurs and entrepreneurship become more critical and priva-tization more necessary. Li, Li, and Zhang (2000) find that marketliberalization results in more intense product market competition,squeezing profit margins and improving incentives inside the firm.Their analysis suggests that a fiscally independent local governmentwill be more willing to privatize its SOEs.

Guo (2003) points out that when the market plays a greater rolein resource allocation, governments have less information with whichto assess SOE performance and the manager acquires more control

4. Rural tax reforms aim to unify the taxes and fees into a uniform 5 percent taxon households’ average net income. The tax rate is currently higher, making it dif-ficult for local governments, especially township governments, to implement thenew tax.


42

over the firm. The government transfers ownership to the manager tobetter align control and ownership. Sonobe and Otsuka (2004) findthat firms that are more involved in market transactions are more suc-cessful after privatization, providing indirect evidence for this theory.

Complex measures have been developed to measure the degree ofmarket liberalization.5 This study uses the lagged employment share ofthe private sector in a province as the indicator, because a larger pri-vate sector in surrounding areas implies a more liberalized market(Gregory, Tenev, and Wagle 2000; Li, Li, and Zhang 2000).

Empirical Tests

In this section we test for the efficiency, financial crisis, and market lib-eralization hypotheses. We do not test for the fiscal decentralization hy-pothesis, as this test would require a different sample and a differentapproach.6 We are looking for empirical evidence regarding the impactof factors such as the efficiency gap between state-owned and privatefirms in the respective sectors, the fiscal capacity of local governmentsand the financial strength of firms, and the effect of the degree of private-sector development on the extent of gaizhi. We measure gaizhi by theshare of government ownership at the firm level. We assume that thelower the level of government ownership, the further the firm is alongthe path of restructuring.

In addition to analyzing the impact of the above-mentioned fac-tors on the extent of gaizhi, we consider other variables that may playa crucial role, such as the number of retired and redundant workers, as-sets per employee, firm size, market competition, and time and regionalvariables.

A firm is more difficult to privatize if it has a large number of re-tirees and redundant workers who need to be redeployed. AlthoughChina is moving from a firm-based pension system to a new unified

5. For instance, the National Economic Research Institute publishes a market lib-eralization index for Chinese provinces. This index includes both policy indicatorsand performance indicators and thus suffers from possible endogeneity problemswhen used in regressions.6. Ideally one would like to compare situations before and after the reform of fis-cal relationships between the center and the localities, or compare localities withdifferent degrees of fiscal independence. Since 1995, however, the relationship ofthe center with the provinces and the other administrative units is more or lessuniform.


43

system, the system that is evolving is in a significant deficit. A priva-tizing SOE is often given the option of paying all retirement premiumsin advance or of continuing with the old firm-based system. In eithercase the retirees are a social burden on a privatizing firm. Redundantworkers include workers dismissed from the firm, xiagang workers,and those who have been “internally retired” (see chapter 4).

It is a common practice for local governments to discount theprice of state assets to compensate either the privatized firm for takingcare of redundant workers or the workers for giving up their state em-ployee status. The net worth of a firm can be a determining factor ingaizhi, so, to account for this possibility, the variable of net assets perworker is added to the regression.

We also control for firm size and industrial competition in theanalysis. Casual observation suggests that larger firms are less likelyto be privatized. The industrial competition variable controls for sec-toral specificities and is represented by the average profit rate in a spe-cific industry.

Finally, year dummies and a regional dummy distinguishing thesouth and the north are added into the regression. All the explanatoryvariables (except the regional and year dummies) used in the analysisare three-year lagged averages.7

The findings are presented in table 2.1. Contrary to the predic-tion of the efficiency hypothesis, the government share does not de-cline when there is a larger potential gain to a firm from privatiza-tion. The result shows exactly the opposite with a high statisticalsignificance: better-performing firms are more likely to be privatized.The financial crisis hypothesis is likewise rejected. A firm with ahigher net return to its assets tends to have a lower government share.In addition, in cities with strong public finances, local governmentstend to hold a lower share in their firms. A firm’s debt-equity ratiodoes not play a significant role in the regression. This result mightbe related to the weak position of the creditors in the privatizationprocess.

There may be several reasons behind the rejection of the two hy-potheses. First, governments tend to privatize the better-performingSOEs first. Second, it is easier for a city with strong public finances tosupport privatization, because it can provide financial concessions and

7. The use of three-year lagged averages partially addressed any endogeneityproblems.


44

TABLE 2.1FACTORS INFLUENCING GAIZHI AS DETERMINED BY

REGRESSION ANALYSIS

Expected Actual sign andVariable sign statistical significance

Efficiency hypothesis variableAverage private-sector value added

in the sector as a percentageof sample firm’s value addedper unit of assets

Financial crisis hypothesis variablesPer capita local government

revenues (RMB)Debt-equity ratio (%)

Net return on assets (%)

Market liberalization hypothesis variableShare of private-sector

employment in provincialemployment

Control variablesWorker redundancy (%)

Net assets per worker(RMB10,000 )

Number of workers (1,000)

Industrial profit rate (%)

Northern city

Year dummies

Negative

Positive

Negative

Positive

Negative

Positive

Negative

Positive

Negative

Positive, statistically significant

Negative, statistically significant

Positive, not statistically significant








Positive, statistically significant. 1995 isthe reference year

NOTE: Sample size is 1,100 cases. Method of estimation is weighted least squaresto correct for the oversampling of unreformed SOEs relative to gaizhi firms.Dependent variable is share of governement ownership.SOURCE: Authors’ estimates.


45

subsidies to redeploy redundant workers and to compensate the buyerfor purchasing state firms.

Of the three hypotheses, the market liberalization hypothesis isthe only one that is not rejected. This result is consistent with the ob-servation that provinces that have a larger private sector have beenmore active with gaizhi. The level of private-sector development ina city may influence gaizhi in several ways. It is presumably easierfor redundant workers to find other jobs in a city with a larger pri-vate sector, so the problem of worker redeployment is less severe.Furthermore, the public’s acceptance of privatization is likely to behigher in a city with a large private sector, and this can reduce theobstacles to privatization.

As for the other variables, all but the variable for industrial profitare significant and hold the expected signs. A higher level of redun-dancy does slow privatization, but a greater net assets position accel-erates it. In addition, larger firms, and firms in northern China, areharder to privatize. A firm in the north of China has on average abouta 7 percent higher share of government ownership.

Conclusion

Gaizhi has changed China’s economic landscape. The number of SOEshas been reduced dramatically. Nationwide, about 80 percent of SOEshad gone through some type of gaizhi by the end of 2001, and 70 per-cent reported that partial or full transfer of ownership had taken placefrom the government to private hands. Gaizhi has been assuming moreradical forms, with privatizations accounting for an increasing sharein recent years.

A number of players have a stake in the gaizhi process. Local gov-ernments and enterprise managers have been the most active propo-nents of reforms, with managers assuming a leading role in later roundsof restructuring. The empirical analysis shows that market liberaliza-tion is an important macroeconomic factor leading local governmentsto release SOE shares into private hands. The analysis also shows thatthe fear of redundancies hinders the pace of privatization. Consistentwith this finding, a firm with greater net assets is easier to privatizebecause it can compensate workers, and a city with a better financialrecord is more likely to privatize because it can afford to pay concessionsand subsidies. Cities tend to privatize the better-performing SOEs first.

46

3The Gaizhi Process

The central issue of gaizhi is the reallocation of property rights overenterprise assets and liabilities. A variety of forms and mechanismshave emerged to accomplish this task. In this chapter we look at thedifferent forms of gaizhi and at their structure, evolution over time,and geographical distribution. The chapter also discusses the factorsinfluencing the choice of different forms of gaizhi in China. It focuseson the common issues that most types of gaizhi have to address, suchas valuation of assets and liabilities, particularly of enterprise debt andland-use rights.

Forms of Gaizhi

The term “gaizhi” is used to mean any structural change to a firm,including public offering, internal restructuring (incorporation,spinning off), bankruptcy and reorganization including throughdebt-equity swaps, ownership diversification by introducing new in-vestors, employee shareholding (limited liability companies or co-operatives), open sale (to management, employees, outside privatefirms, or another SOE), leasing (to management, employees, outsideprivate firms, or another SOE), joint ventures, or a combination ofthe above.

Internal Restructuring. Internal restructuring does not change theidentity of the owner. A form of internal restructuring is incorpora-tion, which started en masse soon after the Company Law came intoforce in 1994. According to the Company Law, if an SOE is to be re-organized into a company, it must change its operating mechanism,identify and verify its assets and determine their respective owners,settle its creditors’ rights and liabilities, conduct an assets appraisal,

and set up standard internal management organs. The law requires acompany to have at least two shareholders, but it allows SOEs to reg-ister as limited liability companies with the state as the sole owner.Therefore, while incorporation may not involve any change in own-ership, it does change the formal relationship between the state andthe enterprise through the concept of limited liability. Incorporationdraws a line, typically not very thick, between the state and the com-pany. It also prepares the ground for ownership diversification, be-cause the firm can now avail itself of a legal framework for bringingin new investors.

Another common way to revitalize an SOE has been to split thefirm into several smaller firms that begin to manufacture new prod-ucts. The old firm becomes a holding company that owns the newspun-off firms and maintains a contractual relationship with themby charging a fee for the use of buildings and equipment. Anotherpractice of spinning off has been to set up a new company that takesthe good assets of the old firm, including buildings, equipment, andcapable personnel, and leaves it with the nonperforming assets, bankand commercial debts, retirees, and redundant workers. It is this formof gaizhi that creditors, particularly banks, fear most. Since spinningoff does not change the ownership, it can be seen as a form of inter-nal restructuring, although reform measures within the new firms canbe quite radical. Initial public offerings by state-owned companies inChina have been typically preceded by the types of internal restructuringdescribed above.

Bankruptcy and Reorganization. Although China’s Bankruptcy Lawcame into force in 1988, it applied only to SOEs and was not widelyapplied until the mid-1990s, when the central government began toadopt bankruptcy as a means to restructure SOEs (Gao and Yao 1999).Since that time the program of “policy-oriented bankruptcy” has re-sulted in 3,377 bankruptcy cases, RMB223.8 billion in write-offs, and6.2 million layoffs. The program is expected to be phased out by theend of the decade, when the last batch of 2,000 cases of policy bank-ruptcy is closed. A new Bankruptcy Law is reportedly nearing comple-tion after 10 years of drafting.

In many instances, the framework of policy bankruptcy has beenabused by SOEs, which, often in conjunction with the local govern-ments, have viewed the procedure as a mechanism for evading debtpayments. Cases of abuse involving collusion between local govern-

THE GAIZHI PROCESS

47

ments and local banks have also been reported.1 Bankruptcy often oc-curs with gaizhi that diversifies the ownership of the firm. Therefore,bankruptcy and subsequent reorganization often imply partial or fullprivatization.

Debt-equity swaps were introduced by the central governmentin 1999 to alleviate the huge nonperforming loan problem. NPLs wereofficially estimated at 25 percent of outstanding loans, but the trueshare is likely to be higher. In 1999, four asset management companieswere established to tackle the problem. One task of these companieshas been to implement the debt-equity swaps. While the merits of theplan are debatable, it has alleviated the burden on participating SOEsand, in some cases, has improved their corporate governance. By theend of March 2004 the four AMCs had disposed of RMB528.7 billionof NPLs (excluding policy-oriented debt-equity swaps) in total, with acash recovery rate of 20 percent. The government plans to allow theAMCs to exist beyond the initial 10-year period. After the disposal ofNPLs is completed, the AMCs will undergo a transformation in thedirection of “commercialization.”

Ownership Diversification. Ownership diversification involves bring-ing in outside investors while maintaining majority state ownership.Diversification could occur by means of an initial public offering, a pri-vate placement, or a private offering. IPOs in China have been used asa mechanism to bring new owners in, rather than enabling existingowners (in this case, the state) to exit. Following the IPO the firm’sownership becomes more diversified, but the public offering typicallydoes not change the dominant position of the state as an owner. Thestate, directly or indirectly, still holds about two-thirds of the sharesof Chinese listed companies. State-owned enterprises have been themain participants in Chinese stock markets as, until recently, sales of


48

1. A report issued by the National Auditing Office disclosed a fraud case involvingcollusion between court officials and a bank in Jinzhou, Liaoning province. Localjudges and managers of the local branch of the Bank of Communication fabricated346 false cases with the names of 285 debtor firms in order to obtain approval forwrite-offs of RMB532 million in loans. RMB221 million had been approved. Whenthe debtor firms paid interest and principal amounts to the bank, the money wentto a private “treasury.”

shares were a low-cost way of financing SOEs. Tenev, Zhang, and Brefort(2002) provide an overview and analysis of the corporate governanceaspects of the IPO process in China.

Employee Shareholding. Employee shareholding has been by far the mostpopular form of gaizhi throughout the country. Although proven to bea suboptimal arrangement in other transitional countries, this form ofgaizhi entailed the least political risk in the early stages of urban reforms.Indeed, many employee-shareholding firms were still registered as col-lectives in the early 1990s. Perhaps it is also the most politically feasibleform of privatization in terms of the power structure within an SOE.The central government requires that each gaizhi plan be approved bya conference of employee representatives. Because it allows ordinaryemployees to have more voice within the firm and to share in its futureprofits, employee shareholding is the form of privatization most likelyto be approved by employees.

To be registered as a limited liability company, the maximum num-ber of shareholders must be below 50, as stipulated by the CompanyLaw. Firms with a larger number of shareholders, but which cannotmeet the requirements for a joint stock company, can assume the sta-tus of employee shareholding cooperatives. Shareholding coopera-tives have been an innovative mechanism for the transformation oftownship and village enterprises in China. Some employee-ownedfirms with more than 50 shareholders have been able to register ascompanies by forming block shares; that is, groups of employees electa representative or trustee and register all the group’s shares underthat person’s name.

Many employees gained shares through compensation given bythe local government in exchange for the removal of their state em-ployee status (see chapter 4). As a result, in the early stages of gaizhi,shares were widely dispersed throughout the firm. In recent years,managers have been able to buy a larger number of shares in newlyprivatized firms. Some gaizhi firms have gone through second andthird rounds of gaizhi, further increasing the number of shares ownedby managers. Management buy-outs have become quite controver-sial in China, given the growing number of cases of reported abuses.In response to public opinion, the SASAC has issued regulations dis-couraging MBOs of large companies.

THE GAIZHI PROCESS

49

Open Sale. This form of gaizhi has become more popular in recent years.The firm is openly sold to insiders or outsiders, perhaps through auction.This is the most radical form of privatization because it can involve thetransfer of the firm to a single private owner or a management group.

Leases. The lease contract now commonly used in gaizhi is quite dif-ferent from that adopted in the early years of SOE reform. The earlyleases acted as incentives within the SOE, but leases are now used tobreak up the SOE. Under current leases the lessee is a legal entity inde-pendent of the government. Some lessees are outsiders and own theirown firm, while others are former employees who have set up new com-panies and lease the buildings, land, and equipment from the govern-ment. Leasing is often adopted in cases where the lessee does not haveenough money to buy the firm. It is another radical form of gaizhi.

Joint Ventures. Forming a joint venture with a domestic or foreign firmis another approach to gaizhi. It may involve separating assets fromthe existing firm and forming a new entity; that is, a spin-off. Mergerscan also be included in this category.

Firms undergoing gaizhi may adopt a number of the above mea-sures. The various forms of gaizhi form a spectrum incorporating dif-ferent degrees of change of ownership. At one end of the spectrum isinternal restructuring, where no change of ownership takes place. Atthe other end is open sale. In between are: ownership diversification,where there is a change in the ownership structure but assets are nottransferred from the state to the new owners; bankruptcy, reorgani-zation, and leases, which can lead to significant ownership changesinvolving the transfer of assets from the state to the new owners; andemployee shareholding and joint ventures, where new owners are in-troduced and this is accompanied by a transfer of assets from thestate to the new owners.

Sample Distribution of Forms of Gaizhi

Of the 375 cases of gaizhi in the sample, 20 percent (74 cases) wentthrough internal restructuring (see figure 3.1). About 8 percent of thegaizhi firms (30 cases) went through ownership diversification, in-cluding 25 cases of public offering and 5 cases of private placement tooutside investors. A total of 27 percent (103 cases) introduced em-


50

ployee shareholding, 28 percent of firms (105 cases) were sold orleased out, 13 percent (49 cases) went through bankruptcy and debt-equity swaps, and the remaining 4 percent (14 cases) became joint ven-tures. Therefore, more than 70 percent of gaizhi cases involved thetransfer of ownership from the state to private hands.

Of the 74 cases of internal restructuring, 53 were simply incor-porated, 16 involved spin-offs, and 5 were unclassified. Of the 103 casesof employee shareholding, 53 percent of the firms became limited liabil-ity companies, 34 percent became shareholding cooperatives,2 and theremaining 13 percent were unclassified.

THE GAIZHI PROCESS

51

2. Chinese law does not grant legal status to cooperatives. Interviews with thefirms found that most cooperatives wanted to register as limited liability compa-nies, but the limit on the number of shareholders (under 50) prevented them fromdoing so. As indicated above, employees in some firms have solved this problemby pooling their shares under the name of an entrusted member.

FIGURE 3.1Distribution of Forms of GAIZHI in Sample Firms

Publicoffering

8%


20%


13%Employee

shareholding27%

Leasing15%

Jointventure

4%

Open sales13%


The significant presence of employee shareholding firms in the11-city survey indicates the prevalence of insider control in China’sprivatization. Insiders also took over a large proportion of those firmsthat were sold or leased out: 16 of the 49 sales and 15 of the 56 leasesinvolved insiders, mostly managers (see figure 3.2).

Only 16 sales and 20 leases were exclusively directed to a privatefirm. Overall, sales or leases to an outside private firm comprised only10 percent of the 375 cases of gaizhi. There were 10 sale and 19 leasecases that were classified as “other types” of sales and leases. Thesecases are probably a blend of insider control and outsider participa-tion. If these cases are included, the share of gaizhi cases that involvedat least partial outsider participation rises to 15 percent. This showsthat the private sector has become an active player in gaizhi. Box 3.1provides evidence of this trend from Tangshan and Hengyang.

Although the classifications of the forms of gaizhi used by the twonational surveys are different from those adopted by this study, thereare some comparable results. The 1998 survey revealed broadly simi-


52

Sold exclusively to theold management

26%

Sold to anotherinsider(s)

9%

Sold to aprivate firm

35%

Sold toanother state-controlled firm

9%

Other21%


FIGURE 3.2Structure of SOE Sales by Type of Purchaser

(percent)

THE GAIZHI PROCESS

53

BOX 3.1PARTICIPATION OF THE PRIVATE SECTOR IN GAIZHI

Among the six cities where interviews were conducted, the privatesector was very active in participating in gaizhi in Tangshan,Chengdu, Guiyang, and Hengyang. Open sales and leasing made up33 percent, 43 percent, 37 percent, and 55 percent of gaizhi cases inthose cities, respectively. Many of the leases and sales were to a pri-vate company.

In Tangshan, many private entrepreneurs accumulated theirwealth in the coal mining industry. As the central government beganto limit the development of small coal mines, private owners lookedfor investment opportunities in SOEs. Many SOEs in Tangshanwere poorly performing or had ceased production several years ear-lier. One factory producing mining machinery had been out of pro-duction for three years and had accumulated large bank and wagedebts. The local government found a private businessperson whowas willing to purchase the factory if half the 1,000 workers couldbe laid off. The government agreed and sold the factory for a sub-stantial discount on book value. In 2002, most of Tangshan’s 15gaizhi SOEs were sold to private businesspeople.

In Hengyang, leases were more common than sales. An exampleis a factory that had produced industrial alcohol for 45 years. Thefactory had been renting out part of its buildings and equipment toemployees to produce by-products such as carbon dioxide, but in2001 it still carried a large debt, with a debt-equity ratio of 200 per-cent. In June 2002 a private businessperson took over the rental ofmost of the factory facilities and started new production. Sales dou-bled in the two months after the lease began.

Selling or leasing to an outside private firm has increasingly be-come the main choice of local governments, for three reasons. First,private firms offer cash payments to buy an SOE, so the governmentis able to redeploy workers relatively easily. Second, private buyersexempt the government from having future obligations to the work-ers. And third, private owners are typically able to improve perfor-mance quickly.

lar patterns to the 11-city survey (see figures 1.10 and 1.12). Its twocategories of “introducing new investment” and “spinning off” werelargely equivalent to “internal restructuring” plus “public offering” inthis study. Together they made up 30 percent of all the gaizhi cases,which agrees with the finding in this study. The figures for employeeshareholding were also comparable: 21 percent in the 1998 survey

compared with 27 percent in this study. The share of sales and leaseswas almost the same in the two surveys: 27 percent in the 1998 surveyand 28 percent in this study. The share of joint ventures was also sim-ilar: 3 percent in the 1998 survey and 4 percent in our sample. Theincidences of bankruptcy differ. Only 5 percent of gaizhi cases weredescribed as bankruptcies in the 1998 survey, but in our sample theshare was 11 percent. This may be because the 1998 survey had a sep-arate category for mergers, and 10 percent of gaizhi cases were reportedas mergers, while this study combined mergers and bankruptcies.

The 2001 national survey had some significant differences from the1998 national survey and this study, although in some categories therewere similar results (see figures 1.11 and 1.12). The incidences of opensales and leases were similar to the findings of the other two surveys,and the incidence of bankruptcy was comparable to that found by the1998 survey. The 2001 national survey, however, reported more merg-ers and bankruptcies than either the 1998 or the 11-city survey.

Trends and Geographic Variations

There has been a clear trend toward selling or leasing out SOEs in re-cent years (see table 3.1). Before 1995 no firms were sold openly. In2001 and 2002, open sales, leasing, and employee shareholding werethe three most common forms of gaizhi, together making up morethan two-thirds of gaizhi cases. More than one-fifth of gaizhi caseswere open sales.

The popularity of employee shareholding has declined some-what. It was the dominant form of gaizhi in the late 1990s, repre-senting nearly one-third of gaizhi cases in 1999. That share declinedto less than one-quarter in 2002. Cases of internal restructuringwere common in the 1990s but have declined rapidly since 2000.Joint ventures were popular before 1996 but have since declined.Public offering was a significant form of gaizhi in the mid-1990s but has declined considerably since then (see figures 3.3 and 3.4).The decline of public offering is perhaps due to the poor performanceof China’s stock market. One measure of government support hasbeen to control the supply of new shares by restricting the numberof new IPOs.

Bankruptcies were quite common in the mid-1990s, coincidingwith the passing of the Bankruptcy Law and the provision of centralgovernment funds to write off the bank debts of bankrupt SOEs.


54

55

TABLE 3.1FORMS OF GAIZHI BY YEAR, BEFORE 1995–2002

(percent)

Number Public Internal Employee Open Jointof gaizhi offering restructuring Bankruptcy shareholding sales Leasing venture

Before 1995 30 6.7 20.0 3.3 30.0 0.0 16.7 23.31995 13 0.0 23.1 23.1 7.7 7.7 7.7 30.81996 40 15.0 22.5 27.5 17.5 7.5 10.0 0.01997 32 18.8 37.5 12.5 25.0 3.1 3.1 0.01998 50 8.0 34.0 2.0 28.0 12.0 16.0 0.01999 53 9.4 34.0 1.9 30.2 5.7 17.0 1.92000 54 5.6 20.4 7.4 31.5 14.8 18.5 1.92001 64 3.1 7.8 18.8 29.7 20.3 20.3 0.02002 34 2.9 11.8 11.8 23.5 23.5 23.5 2.9


Many SOEs abused the bankruptcy system to avoid paying backbank debts. The common practice was to file for bankruptcy and thenestablish a new company under a different name on the same site (Gaoand Yao 1999). The central government tightened up bankruptcy reg-ulations, and incidences decreased in the late 1990s. As the number ofcases of gaizhi rose again in 2000–01, however, so did the number ofbankruptcy cases. The 11-city survey found that 19 percent of gaizhicases in 2001 and 12 percent in 2002 were bankruptcies. It becameclear during the field interviews with government and bank officialsand enterprise managers that evading bank debts was still a significantmotivation for gaizhi. The central government issued several decreesto discourage this trend, and the Supreme Court issued a decree thatdefined the evasion of bank debts through bankruptcy as an offense.

There were significant variations among the cities in the forms ofgaizhi (see table 3.2). Generally, internal restructuring was more popu-lar in the northern and western cities. This was particularly true in Harbin


56

Public offering9%


30%


20%


18%

Open sales5%

Leasing18%


FIGURE 3.3Forms of GAIZHI in the First Round of GAIZHI

and Fushun, where more than 40 percent of firms undertook this formof gaizhi. More than 50 percent of gaizhi cases in Xining and Lanzhouwere bankruptcies.

Employee shareholding was popular in Chengdu, Weifang,Huangshi, and Zhenjiang. Chengdu and Huangshi had high incidencesof open sales. Leasing was heavily used in Hengyang. Most of the leaseswere signed between an SOE and an outside private company. (See box3.1 for an example of leasing.) In general, gaizhi was more likely toresult in real privatization in the fast-growing south and southwest (forexample, Chengdu and Guiyang), while in the lagging north andnorthwest (for example, Xining and Lanzhou) gaizhi was more aboutinternal restructuring.

Restructuring is not a one-off event, but rather a continuousprocess. A total of 64 firms (24 percent) undertook two or three roundsof gaizhi. Among these 64 firms, 39 percent decided on public offer-ing or internal restructuring in the first round of gaizhi (see figure3.3). Another 20 percent went through a bankruptcy. All of the bank-rupt firms were later sold or leased out. The remaining 41 percent

THE GAIZHI PROCESS

57

Public offering11%


14%


4%


19%

Open sales21%

Leasing31%


FIGURE 3.4Forms of GAIZHI in the Most Recent Round of GAIZHI

58

TABLE 3.2GAIZHI BY CITY AND METHOD OF FORMATION

(number and percent)

Public Internal Employee Open Joint Number offering restructuring Bankruptcy shareholding sales Leasing ventureof gaizhi (%) (%) (%) (%) (%) (%) (%)

Harbin 40 7.5 47.5 2.5 22.5 10.0 10.0 0.0Fushun 10 0.0 40.0 20.0 20.0 0.0 20.0 0.0Tangshan 57 8.8 26.3 10.5 15.8 15.8 17.5 7.0Xining 18 11.1 22.2 50.0 16.7 0.0 0.0 0.0Lanzhou 11 0.0 9.1 54.5 27.3 9.1 0.0 0.0Guiyang 41 4.9 26.8 7.3 19.5 19.5 17.1 4.9Chengdu 21 4.8 14.3 0.0 38.1 38.1 4.8 0.0Weifang 34 14.7 14.7 11.8 50.0 2.9 0.0 5.9Huangshi 54 7.4 11.1 5.6 38.9 22.2 13.0 1.9Zhenjiang 45 8.9 26.7 2.2 44.4 2.2 6.7 8.9Hengyang 49 4.1 16.3 16.3 6.1 6.1 49.0 2.0


adopted employee shareholding, or were sold or leased out. In the sec-ond and third rounds of gaizhi the combined share of public offering,internal restructuring, and bankruptcy declined to 29 percent (see fig-ure 3.4). The share of employee shareholding was about the same asin the first round, but sales and leases made up 52 percent of all thecases. This shows that gaizhi has become more radical in recent years.

Table 3.3 presents a matrix that compares the most recent roundof gaizhi with the first round. The decisions taken in the most recentround were skewed toward the northeastern part of the matrix, mean-ing that they were more radical than in the first round. Of the 20 casesof internal restructuring in the first round, 16 firms adopted a moreradical form of gaizhi in the second round, as did all 11 of the bank-ruptcy cases and 5 of the 12 employee shareholding firms.

What Determines the Form of Gaizhi Chosen?

In this section we analyze the factors that may have influenced thechoice of different types of gaizhi. We group the forms of gaizhi intofour categories: internal restructuring; employee shareholding; sales,

THE GAIZHI PROCESS

59

TABLE 3.3TRANSITION MATRIX BETWEEN TYPES OF GAIZHI IN THE

FIRST AND MOST RECENT ROUND OF GAIZHI

Bankruptcyand Employee

Public Internal reorgani- share- OpenFirst round offering restructuring zation holding sales Leasing

Public 4 1 0 0 0 0offering

Internal 0 4 1 4 5 6restructuring

Bankruptcy and 0 0 0 2 2 7reorganization

Employee 1 1 0 5 4 1shareholding

Open sales 0 0 1 0 1 0Leasing 2 0 0 1 2 6


Last round

leases, and bankruptcies; and IPOs and joint ventures. Sales, leases,and bankruptcies are grouped together because they represent themost radical form of gaizhi. One difficulty with a regression analy-sis of firms that have undertaken gaizhi is that multiple rounds ofgaizhi have occurred, with second and third rounds being usuallymore radical than the first. We deal with this difficulty by treatingeach round of gaizhi as a separate case. Internal restructuring is usedas a reference group.

The choice of the form of gaizhi results from the interaction oflocal governments and SOEs under their jurisdiction. Therefore, asexplanatory variables we use sets of government/city and firm charac-teristics. The firm characteristics we use in the analysis are produc-tivity gap,3 social burdens, net assets per worker, firm size, debt-equityratio, and return on assets. The use of these variables in regressionanalysis will allow us to shed some light on questions regarding thefactors influencing the forms of gaizhi. Does higher productivity ofprivate firms in the same industry stimulate more radical forms of re-structuring? How does the extent of redundant labor influence theform of restructuring undertaken? Are bigger firms more or lesslikely to undertake radical restructuring? What are the effects of cap-ital per worker, financial leveraging, and profitability on the form ofgaizhi chosen? Among the local government/city characteristics, weinclude fiscal capacity, level of private-sector development in the lo-cality, and regional dummy. Are richer municipalities more likely toadopt radical forms of gaizhi, presumably because of higher capac-ity to absorb the short-term costs of layoffs and higher unemploy-ment? Does a higher level of private-sector development make deeperrestructuring more or less likely? Are regional factors important indetermining the choice of gaizhi forms? The results are summarizedin table 3.4.

The productivity gap between private and state firms seems to bepositively associated with more radical forms of gaizhi; however, theresults are not statistically significant. More profitable firms are lesslikely to adopt employee shareholding relative to internal restructur-ing. They are more likely to adopt a form of gaizhi that falls in the cat-


60

3. For definitions see the notes to table 3.4. Productivity gap combines firm andsectoral characteristics.

61

TABLE 3.4FACTORS INFLUENCING THE CHOICE OF FORMS OF GAIZHI

Lease, sales,Employee and IPOs and joint

shareholding bankruptcies ventures vs. internal vs. internal vs. internal

Factor restructuring restructuring restructuring

Productivity gapa

Debt-equity ratio

Return on assets

Social burdensb

Net assets per worker

Firm sizec

Private-sector developmentd

Municipalityfiscal strengthe

North dummyf


Negative, notstatistically significant


Positive, notstatistically significant


Negative,statisticallysignificant

Positive,statisticallysignificant

Positive,statisticallysignificantat 10%

Negative,statistically significant









Negative,statisticallysignificantat 10%



Negative,notstatisticallysignificant

Negative, not statistically significant

Negative, not statistically significant





NOTE: All explanatory variables are lagged three-year averages. Method of esti-mation: multinomial logit.a. Productivity gap is equal to the relative difference between an SOE’s gross prof-itability and the average gross profitability of private firms in the same industry.b. Social burdens are worker redundancies, defined as the share of retirees andxiagang workers in the total labor force.c. Firm size is in thousands of employees.d. Private-sector development is measured as a percentage of private-sector em-ployment in total employment in the respective locality.e. Fiscal strength of local government is in terms of government revenues per capita.f. The north dummy assumes a value of one if the firm is in the northeast of China.SOURCE: Authors’ estimates.

egory of sale, leasing, and bankruptcy. Neither result is statistically sig-nificant, however.

Firm size has the opposite type of association with gaizhi. Largerfirms are significantly less likely to choose employee shareholding orother forms of gaizhi over internal restructuring. Net assets per workerhas a similar effect: it makes it less likely that more radical forms ofgaizhi will be adopted.

While the coefficients for both financial leverage and social bur-dens are statistically insignificant, it is interesting to note that theyhave opposing signs. It seems that debt burdens and social burdenshave contrasting effects on the extent of restructuring and the formof gaizhi chosen.

An important result is that the level of private-sector develop-ment makes it more likely that deeper and more meaningful forms ofrestructuring will be adopted. Municipal fiscal capacity has an im-pact on the forms of gaizhi. The stronger the fiscal position of thelocal government, the more likely it is for firms to go through employeeshareholding, sales, leases, and bankruptcy relative to internal restruc-turing. Only the result for employee shareholding, however, is statis-tically significant at the 10 percent level. Finally, northern cities are morelikely to restructure a firm internally than to adopt any other form ofgaizhi, which is consistent with the geographic variations presentedin table 3.2.

In summary, it seems that government/local factors tend tohave a larger impact than firm characteristics on the form thatgaizhi will take. In particular, the regional dummy and the level ofprivate-sector development in the region seem to have an importantinfluence on the choice of gaizhi. Size is the most important factoramong firm characteristics.

The more radical forms of gaizhi involve the transfer of stateassets to private owners. The rest of the chapter focuses on the as-pects of the gaizhi process that deal with valuation and transfer ofstate assets.

The Process and Main Issues Surrounding the Transfer of State Assets

“State assets” refers to all productive and nonproductive assets thatbelong to the Chinese state, including natural resources such as un-


62

explored land, mining, and forestry.4 At the end of 2002, state assetswere estimated to be around RMB12 trillion,5 excluding natural re-sources. Of this amount, about 65 percent (RMB7.7 trillion) wereproductive assets.6 Of the total, the central government’s share was48 percent and that of the local governments 52 percent.7

The government normally requires nonproductive assets such asguest houses, hospitals, schools, and dining halls to be separated fromproductive assets by handing them and their employees over to localgovernments to manage. Local governments can turn nonproductiveunits into business entities through gaizhi after operating them forthree years. In some cases, enterprises separate out nonproductive as-sets by establishing independent services firms with little help fromlocal governments. The gaizhi programs can then focus solely on deal-ing with the firms’ productive assets.

Firms beginning the process of gaizhi are required by local gov-ernments to follow an application and reporting process. Gaizhifirms must have their assets valued by an asset valuation agency andverified by the local asset management agency. A proposal shouldthen be submitted to the local office of state enterprise ownership re-form (see box 3.2).

Local governments have used a number of methods to handle thetransfer of state assets in gaizhi. These variations can be attributedmainly to their discretionary power, rather than to differences in pol-icy frameworks, although there are some policy differences acrossregions. What follows is a stylized description based mainly on pre-vailing practices in the localities visited and existing regulations.

The main approach for better-quality SOEs has been to packagethem for listing on the stock market. Firms must comply with stringentrules to qualify for an IPO, meaning that only the better-performingones can avail themselves of this method. Until recently, a quota sys-

THE GAIZHI PROCESS

63

4. The definitions of central government and local government assets are vague.It is difficult to determine which assets resulted from central government invest-ments and which from local governments’ subsequent investments (China BusinessWeekly, May 27, 2003).5. GDP at market prices was RMB10.48 trillion at the end of 2002.6. People’s Daily (overseas edition), June 5, 2002.7. The real value of state assets could be higher because of the existence of extra-budgetary assets held by various levels of governments, as these are not includedin official statistics.


64

BOX 3.2A TYPICAL GAIZHI PROPOSAL

A gaizhi proposal usually has the following elements:

1. A summary of the enterprise

Name, sector, affiliation, main product, number of employees(including laid-off and retired workers), occupied land area, assetstructures.

2. Performance

Key indicators of performance during the past three years.

3. Enterprise assets: Current status and proposal for disposal

Total assets, debts and net assets (both book value and marketor current valuation), proposal for dealing with the net assetsand the land.

4. Redeployment of workers

Proposal for redeploying workers, contract for redeployingworkers, pension, and medical coverage for retired workers, etc.

5. Form of gaizhi

Form, scope, and content of gaizhi (e.g., governance structure).

6. Post-gaizhi business operation

7. Post-gaizhi development planning and forecast of economic efficiency

Investment plan, governance structure, efficiency and risk analysis.

tem for listing was in place, which implied that only well-connectedcompanies had a chance of accessing the market.8 The government hasalso encouraged strong SOEs to take over poorly performing SOEsthrough mergers and acquisitions, which have been occurring mainlyamong small and medium-size SOEs.

If firms have negative net assets and have not qualified for bank-ruptcy because some of their assets are still valuable, they are allowed

8. The quota system began to be phased out in 1999 with the implementation ofthe Securities Laws. Many elements of the system continue to be in place, however(Cooper 2003).

to separate out their operating assets and set up a new legal entity. Thenew firm takes over part of the debts of the old firm in agreement withthe major creditors.9 In most cases the new firm will raise funds for de-velopment through issuing shares to employees, managers, and outsideinvestors.

The government encourages managers to buy the majority of theirfirm’s shares.10 A government guideline, for example, suggests thatshares sold internally should be distributed according to rank in theorder of senior management, technicians, and ordinary employees. Asa measure to encourage the purchase of shares, employees are able todefer payment of two shares for each share that they buy. If sharehold-ers defer payments, they are able to receive dividends and have votingrights, but they have no rights of heritage or transfer over those shares.The deferred payments have to be settled within five years.

Shares have also been distributed to employees in lieu of unpaidwages or retirement benefits. Government regulations also allow in-ventions and other intellectual property to be converted to shares afterthey have been properly valued and priced, dependent on the agree-ment of the shareholder conference. The principal managers of the newstate-controlled shareholding companies can receive loans to buy upto 50 percent of the shares allocated to them from the state asset exitfunds (see below) and are obliged to pay back the loan and the inter-est within five years.

The new firm can start production by leasing some of the assetsof the old firm. The leasing fees are used to pay allowances to laid-offworkers, pensions, and medical expenses, and to repay debts. The oldfirm is allowed to go bankrupt when certain conditions are met.

When an SOE is sold under the condition that the buyer takesover the debts and the responsibility for redeploying all the employ-ees, there are several ways of deciding the price of the assets.11 Whennet assets are positive, the sale price will be determined according tothe value of the net assets. (See the next section on asset valuation.)

THE GAIZHI PROCESS

65

9. During interviews it became apparent that some firms have tried to rid them-selves of old debts, especially bank loans, either by establishing new firms or bygoing through bankruptcy (discussed above).10. Recently, the central goverment has made pronouncements discouragingmanagement buyouts of large firms.11. Although the government prefers firms to sell their assets through auction,with a view that competitive bidding can determine more appropriate prices, inreality auctions are rare.

When a firm’s assets and debts are more or less equal, the firm could besold based on a zero asset price. When net assets are negative, the debtmay be reduced by the amount it would take to make total assets anddebts more or less equal. The debt that has been deducted could befinanced or compensated from part of the sale of land-use rights (seechapter 4), or by using the returns from state assets managed by thestate asset management agency. Another method is to exempt the gaizhifirm from income tax until the zero net asset value is reached.

The main priority of the local government for the sale of stateassets is to use the proceeds to redeploy workers (see chapter 4), butin many cases the SOE’s net assets are negative. Three government-sponsored funds exist to facilitate and reduce the costs of gaizhi:state asset exit funds, SOE bankruptcy provisional funds, and fundsto assist retailers to prepare for enterprise reform. State asset exitfunds provide a loan to the gaizhi firm to allow it to meet the costsof redeploying its workers. The loan must be repaid within five years.The funds are managed by authorized state asset management agen-cies or by finance departments of local governments. The fundingcomes from income earned from transferring ownership rights andselling or leasing state assets.

The bankruptcy provisional funds are operated by the finance de-partments of local governments. They aim to meet the funding gapcaused by delays in redeploying workers and disposing of assets dur-ing bankruptcy. Funds are normally for one year and are required tobe repaid after the assets are eventually sold.

The funds to reform retailing enterprises can be used to coverpart of a firm’s negative net assets, to pay part of the cost of re-deploying workers that cannot be met through selling the firm’s assets,or to compensate for any shortfall from selling and auctioning thefirm’s assets. The funds come from the sale or auction of the net as-sets of state retailing enterprises, from the transfer of state share-holdings in retailing industries, from lease of the remaining part ofstate assets and dividends from state shareholdings in retailing in-dustries, and from public finances.

Not all regions have the ability to set up these funds. Wealthierprovinces such as Guangdong have been better able to handle the costsof gaizhi, although local officials are still concerned about their abil-ity to bear the costs, especially the costs of redeploying workers. Thoseregions where average incomes are low and the number of SOEs is highlack adequate government funding to support gaizhi programs, many


66

of which have stalled.12 Local governments have come under increasedpressure to reform their SOEs as gaizhi programs have increased incoverage and extent, but some have demanded greater central govern-ment support for poorer regions or those with a higher concentrationof SOEs. Social security contributions by the national governmenttripled between 1999 and 2002, while the overall level of governmentexpenditures increased 1.9 times. Social security, however, still accountsfor 4.1 percent of central government expenditures. At the same time,the level of local government subsidies by the national government hasbeen flat at about 30 percent of total government expenditures (CEIC).

A particular thorny issue is the problem of guarantees (collateral)that SOEs have extended to other SOEs. Gaizhi firms have been re-quired to put aside, in a special account, a share of net assets equiva-lent to the amount required to cover their exposure. The account issupervised by the state asset management agencies. If a trigger eventoccurs, the gaizhi firm-guarantor can use the assets in this fund to per-form under the guarantee arrangement. If the actual losses are largerthan the amount the gaizhi firm has reserved to cover the losses, it isresponsible for paying the difference. If no losses occur, or losses aresmaller than the reserved amount, or the firm is able to recover somelosses from acquiring assets from the bankrupt firm, the surplus willgo to the original owner of the assets. During the interviews, a num-ber of chief executive officers (CEOs) indicated that loan guaranteesmade by pre-gaizhi SOEs are a potential source of dispute between var-ious stakeholders in gaizhi, because of fears that such commitmentscould add to enterprise debts.

In employing these different methods for selling or transferringstate assets, one of the state’s main concerns is to prevent the furtherloss of state assets while addressing issues such as the redeployment ofworkers. It is important, therefore, that the process of dealing with as-sets, particularly asset valuation and the policy of discounting the priceof assets, is transparent and fair.

Asset Valuation

When carrying out gaizhi, firms must go through the procedures ofasset valuation, the demarcation of ownership rights, and the verifica-

THE GAIZHI PROCESS

67

12. See chapter 1 for data on per capita incomes and other statistics across differentsample cities.

tion of assets. Valuation is just a starting point, however. While itforms the basis for the sale or transfer of assets, discounts are oftengiven by local governments to compensate the buyer for assumingmore social obligations or to speed up the gaizhi process.

Valuation is not an academic exercise; it is subject to the same in-terests as the gaizhi process. Valuation has to be perceived as fair if itis to be accepted by all the parties concerned. It is safe to assume thatinsiders (managers and employees) have informational advantageswith respect to other players regarding the “true” value of the assets.Outsiders, including the government, can level out the playing field byusing comparisons with other similar companies, and by relying oncompetition (that is, through auctions) to extract informational rentsfrom insiders. Given the importance of procedural features, indepen-dent valuation professionals have become indispensable in the valua-tion process.

China’s asset valuation industry has developed since the late1980s; now more than 90 percent of its businesses deal with state as-sets. The industry is regulated by the 1991 State Council’s Regulationon the Management of State Asset Valuation. The Ministry of Financeis the principal administrative body responsible for the industry, butthe ministries of Construction and of State Land and Resources exer-cise control over the licenses for conducting asset valuations, especiallyproperty and land valuations.

The valuation industry is relatively young and struggling to over-come a number of problems. Local valuation firms often started life asspecialist accountancy departments of central government ministriesor regional administrations. There are layers of administration over-seeing the industry but no uniform system of qualifications. The in-dustry is fragmented because valuers specialize in particular areas,such as assets, property, land, mining rights, or vehicles. Chinese val-uers are always under threat of having their licenses withdrawn if theyundervalue state assets. In other common situations, their local licensehas been granted by a government authority that is also connected tothe state assets being valued. As a result, a proportion of agencies donot act professionally and are influenced by local governments orgaizhi firms who make requests for a favorable valuation. Rules forprofessional ethics and conduct are evolving.

There are several standard methods for valuing business assets:historical cost and realized incomes; replacement cost and business in-come; current cash equivalents and realizable income; and discounted


68

cash flows and economic income (Sterling 1971). Several structuralfeatures of the Chinese economy make the task of valuation difficultand restrict the applicability of some of the standard methods andtechniques.

For example, a major characteristic of the Chinese market is thehigh volatility and significant regional variation that leaves interpre-tation of older information almost meaningless. From inflation ratesof over 25 percent in 1994, we have seen deflation in the late 1990sand early 2000s, and the recent reemergence of inflation. Low volumeof transactions within certain sectors or regions, rapidly changing tar-iffs, unique local taxes, lack of transparency for sales, and an in-builttendency to control centralized information also make useful inter-pretation of available information difficult at best. As a consequence,using market comparables outside of the main cities and industrialareas is of little value, and the state authorities have been reluctant toaccept the transaction approach as a basis for valuation of state-ownedassets.

The high degree of volatility in prices and markets also affects theaccuracy of the income approach, which is based on a future cash flowmodel. As a result the income approach, where the value of an asset islinked to the benefits of future ownership, has not gained wide accep-tance within the valuation community in China. In many cases it is not“sanctioned” by Chinese valuation firms and authorities, particularlywhen it might mean the recognition of redundancy or other permanentdiminution in value of a state asset. Additional problems include thefact that expected rates of return can vary significantly between dif-ferent analysts, and it is very difficult to get accurate historical accountsfor most Chinese businesses. Consequently, this method is almost alwaysrejected by local valuation firms and the authorities.

The approach that is perhaps best understood and most widelyapplied in China is the cost approach. Information on costs is readilyavailable, there are published standard price books for most types ofmachinery, and Chinese valuation firms apply standard depreciationperiods for different types of machinery on a straight-line basis. Aswith any standard formula there is little application of the effect ofindividual circumstances, and economic obsolescence is rarely consid-ered in the depreciation rates adopted. Imported equipment also presentsa problem to local valuers, since they rarely bear in mind the regionalpricing mechanisms of multinationals, differing exchange rates, and con-trary inflation rates.

THE GAIZHI PROCESS

69

In addition, as in Eastern Europe, China’s underdeveloped anddistorted capital markets provide insufficient and unreliable informa-tion to allow sophisticated methods of asset valuation, such as price-earning methods, to be used (OECD 1993). Furthermore, intangibleassets such as brand names, trademarks, and patents have been in-adequately valued. Box 3.3 provides some suggestions and recom-mendations for improving asset valuations, based on the experience ofEastern European countries.

Thus asset valuation in China takes place in a dynamic and com-plex environment. Some aspects of this environment—such as the fast-growing economy, the underdeveloped capital markets, and the young


70

BOX 3.3PROBLEMS AND MEASURES TO IMPROVE ASSET VALUATIONS

1. There is insufficient information to enable a fair valuation:(a) Accounting standards and the presentation of information

need to improve.(b) Training is needed on the preparation of valuations and

business plans.(c) A memorandum should be prepared showing the strengths

and weaknesses of businesses.(d) Accurate costs and prices should be used.(e) A database should be created of past privatization prices to

assist with future comparisons.2. Valuations are uncertain:

(a) Valuations will be higher if the political and economic envi-ronment improves, as investors will have more confidenceand there will be fewer risk factors.

(b) There is a need for policies that develop the profit orienta-tion of companies, provide technical and business education,improve the tax system, assist with export promotion, anddevelop human resources.

(c) Complementary channels of finance should be set up to fundthe initial purchase and further development of companies.International financing should be enhanced through privateinstitutions and international organizations.

3. There is no set methodology for valuations:(a) There needs to be a system for establishing a price using an

“open book” methodology. Valuation agencies should beindependent and objective.

and immature accounting profession—tend to increase uncertaintyand therefore make it more difficult to agree on a “fair” price. Otherfactors, including some of the administrative rules governing asset val-uation in China, may lead to serious overvaluation of state assets. Forexample, Chinese valuation firms work on a set fee scale that is basedon the values reported; a higher value therefore generates higher fees.Furthermore, a valuer faces the risk of his license being withdrawn ifit is determined that he has undervalued state assets. As a result, for-eign companies often complain about serious overvaluation of assetsin favor of the domestic partner, which is usually an SOE (OxfordAnalytica 2002). Finally, there are powerful groups that benefit from

THE GAIZHI PROCESS

71

(b) Valuations need to take into account other elements thanprice, such as:• The percentage of control sold: If a minority sharehold-

ing is sold, consideration should be given to small in-vestors and PER (price-earnings), DCF (discounted cashflow), or DCD (discounted dividends) methods used, asappropriate. In the case of majority shareholding, the con-ditions of market entry, and the synergy that the potentialbuyer will obtain, should be considered.

• The methods of privatization will affect the valuation, asprices vary in cases of public offerings and auctions.

(c) To avoid over- and undervaluation, in certain cases initialprices can be fixed and adjusted by pre-established formu-lae. Another system is that the purchaser is authorized to buya certain percentage, with an option to acquire the remain-ing or a further percentage after a certain period at a fixedprice or at the valuation at that date.

The process of establishing a price should be objective but also takeinto account the positions of both sellers and buyers.

SOURCE: OECD 1993, pp. 100–1.

undervaluation of state assets. Anecdotal evidence suggests that assetappraisals often undervalued the assets to make them cheaper to buy.Although rules exist for registering and verifying the valuation out-comes, many cities had no effective mechanism of supervision over thevaluation of assets. Media reports of cases of erosion of state assets haveattracted a lot of public attention (see box 3.4). Irregularities have beenassociated with situations involving potential conflicts of interest, suchas SOE managers choosing valuers, the practice of giving discounts toappraised value, and insider privatization, particularly MBOs.

Given the potential conflicts of interest surrounding the valuationprocess, it is important to look at how the decision to select a valuer ismade. The choice of a valuer can be made by either the government orthe firm. The 11-city survey found that in the first round of gaizhi un-


72

BOX 3.4FRAUDULENT CASES OF PRIVATIZATION

State media have disclosed a number of fraudulent cases of privatization:

• People’s Daily: Two SOEs in Shanxi province were taken overby a private firm at a cost of RMB230 million. The sale was later in-validated after the intervention of the provincial government, andthe two SOEs were subsequently sold at auction to another privatefirm at a price of RMB64 million, around one-quarter the originalprice. One official was reportedly arrested for corruption.

• CCTV: An SOE manager in Wuhan was fired by a district gov-ernment when he requested a local court to void a privatization con-tract with a private firm and rejected the government’s instructionto withdraw the request. The district government had “helped” alocal private company to acquire the land of the manager’s SOE ata unit price of RMB690,000, when other pieces of land nearby hadbeen sold at a cost of RMB200,000. The unit price was determinedby dividing the total estimated cost of labor settlement by the areaof the land. The announced intention of the district government wasto hide revenue from creditor banks.

• The chairman of Jiahua Company, a Chongqing-based firmand the first case of a medium-size SOE privatization, was detainedin April 2004 for possible theft of state assets during the privatiza-tion process seven years earlier. The success of this privatization hadbeen widely publicized at the time.

dertaken by the sample firms, 38 percent of valuers were chosen bygaizhi firms, 33 percent by local governments, and the remaining 29 percent jointly by local governments and firms13 (see table 3.5).When the form of gaizhi does not involve formal change in ownership—for instance, in the case of internal restructuring and leases—the decisionwas more likely to be made by firms. As figure 3.5 shows, there is sig-nificant regional variation in the choice of valuation firm.

It is common for buyers to receive state assets at a discountedprice.14 Government regulations in Harbin, for example, allow the gov-ernment to offer a discount when selling state assets. Buyers who pur-chase the whole business and make a one-off payment could receive a30 percent discount. Those making a one-off payment covering 60 per-cent of the total assets could receive a 10 percent discount. After 60 per-cent of the value of assets, a half-percentage point discount can be given

THE GAIZHI PROCESS

73

13. Unreported cases are excluded from the calculations.14. Li and Rozelle (2003), in their study on rural privatization in China, find thatclose to 23 percent of surveyed firms paid a price higher than the appraised value.

TABLE 3.5CHOICE OF VALUER BY FORM OF GAIZHI

(percentage in parentheses)

GovernmentOnly Only and

government firm firm Unreported

Listed on 12 (40) 11 (37) 7 (23) 0 (0)the stock market

Internal 30 (24) 35 (28) 28 (23) 30 (25)restructuring

Employee 28 (27) 31 (30) 21 (21) 23 (22)shareholding

Sale 11 (22) 11 (22) 11 (22) 16 (34)Lease 9 (16) 14 (25) 7 (13) 26 (46)Joint venture with 0 (0) 1 (50) 0 (0) 1 (50)

foreign companyOther 1 (8) 2 (17) 6 (50) 3 (25)Total 91 (24) 105 (28) 80 (21) 99 (26)



74

for every percentage point increase in payment. The first paymentshould not be below 30 percent of the decided price, and paymentsshould be made in installments, with guarantees given before the assetsare transferred.

Discounts can also be provided when the operators responsiblefor the transfer of state assets are authorized to do so by the state assetmanagement agencies or government administrative departments towhich the firm belongs. These agencies or departments are allowed toset the sale price above or below the valuation. Any discount is normallyno more than 10 percent below the valuation. Discounts greater than10 percent need the approval of the state asset management agency.

It appears that government discounts on the net assets of gaizhiSOEs result in the loss of state assets, but local governments have toweigh these losses against other objectives, such as the need to meet theirquotas for gaizhi or to maintain employment in the locality. Figure 3.6presents data on the discounts given to gaizhi firms by local governments.

0 20 40 60 80 100

BothFirmGovernment

Chengdu

Fushun

Guiyang

Harbin

Hengyang

Huangshi

Lanzhou

Tangshan

Weifang

Xining

Zhenjiang

Percent

FIGURE 3.5Choice of Valuation Firm in Eleven Sample Cities


The percentage of gaizhi cases that involved a discount on the firm’sassets ranged from 0 percent in Huangshi to 37 percent in Zhenjiang.Local governments often took into consideration the buyer’s ability toupgrade technology and provide employment. Figure 3.7 suggests thatgovernments also favored SMEs.

Table 3.6 reports the distribution of discounts across gaizhi firms.Most firms surveyed (ranging from 74 percent of those firms that un-derwent restructuring to 48 percent of listed companies) did not wantto report whether they had received a discount in gaizhi. Firms thatwere sold were more likely to receive a discount. Regressing the dis-count rate on forms of gaizhi and firm size shows that open sales arelikely to receive a higher discount rate and larger firms a lower dis-count rate (Li 2004). Smaller firms tend to receive discounts more oftenand in relatively larger amounts than do other firms.

As shown in table 3.6, employee shareholding (MBOs) also showsa high propensity to benefit from discounts. Media reporting of irreg-ularities in the process of MBOs has attracted considerable public

THE GAIZHI PROCESS

75

0 5 10 15 20 25 30 35 40


Huangshi

Harbin

Guiyang

Xining

Tangshan

Chengdu

Lanzhou

Fushun

Hengyang

Weifang

Zhenjiang

Percent

FIGURE 3.6GAIZHI Cases with Price Rebates in Eleven Sample Cities

attention to insider privatization in China (see box 3.5). The governmenthas been discouraging the MBOs of large SOEs.

There is clearly a need in China to regulate the situations involvingconflicts of interest in the gaizhi process and to issue laws and guidelinesfor asset valuation procedures and for the valuing profession. The re-cently established State Asset Supervision and Management Commission(SASMC) has taken the lead in this area,15 with a host of new regulationson these issues (see chapter 7).

Dealing with Enterprise Debts and Other Obligations in the Gaizhi Process

The typical SOE is highly leveraged, and a significant portion of itsloans may be overdue. Bank debts make up the main body of enter-prise arrears, but many enterprises also have unpaid taxes, overdue


76

15. The SASMC was created in 2003 to establish a three-level asset managementsystem in which the state, provincial, and metropolitan governments are state assetproviders and claim ownership rights over state assets.

RMB<500million

52%

RMB 500–2,000million

31%

Above RMB 2,000million

17%


FIGURE 3.7Discounts Provided to GAIZHI by Firm Size

(percent of firms)

wages, pensions, and social insurance payments. In the survey, the 493SOEs that were sampled owed a total of RMB25 billion in 2001 (morethan RMB50 million per enterprise), of which bank loans and interestmade up RMB15 billion, or 59 percent of the total. Most overdueloans were owed to the big four state-owned commercial banks andthe city commercial banks.

The NPL problem became apparent following the financial systemreforms of 1992. As the four state-owned banks became more com-mercial, bank loans—instead of government appropriations—becamethe main source of finance for the SOEs. Without effective banking su-pervision, however, the soft budget constraints continued and a largeamount of loans went into arrears. The banks then made more loans tothe SOEs to sustain their operations, as the government did not permitSOEs to become bankrupt. Table 3.7 shows that more than half thebank loans owed by non-gaizhi SOEs were overdue.

Because banks have tried to recover debts by seizing deposits inthe cash accounts of SOEs, firms have withdrawn most of their moneyfrom the banks and retain cash to maintain their cash flows. Manybusiness transactions are conducted in cash. This practice has restricted

THE GAIZHI PROCESS

77

TABLE 3.6DISTRIBUTION OF ASSET PRICE REBATES FOR

STATE-OWNED ENTERPRISES ACROSS FORMS OF GAIZHI

(percentage in parentheses)

PositiveForm of Total discount Zerogaizhi number firm discount Unreported

Internal 123 13 (11) 26 (21) 84 (68)restructuring

Employee 103 28 (28) 14 (14) 61 (59)shareholding

Sale 49 12 (25) 2 (4) 35 (71)Lease 56 0 (0) 7 (13) 49 (87)IPO 30 5 (17) 6 (20) 19 (63)Joint venture 2 0 (0) 0 (0) 2 (100)

with foreigncompany

Other 12 0 (0) 5 (42) 7 (58)


the development of SOEs. The average cash deposits of a non-gaizhiSOE are substantially lower than those of a gaizhi firm (see table 3.7).

Debt problems have hampered the reform of SOEs, but it is noeasy task to resolve the multiple and sometimes competing concernsof gaizhi firms, banks, and local governments. As one CEO in Harbinpointed out, gaizhi firms are more concerned about the debts owed toemployees than about the overdue bank loans, which they see as beingowed to the government by the former SOE. Even if they had origi-nally intended to pay off the debt, most gaizhi firms find themselvesunable to do so, as they would struggle to stay in operation and findit difficult to attract new capital. The government has not gone as faras canceling the debts of SMEs, as it did for the large SOEs. At the


78

BOX 3.5RECENT CASES OF MBOS AT A “DISCOUNT”

In a recent case of privatization, the managers of a local power com-pany acquired their firm at a price of RMB15 million, while the ap-praised value of the firm was RMB65 million. The main discountsgiven were as follows. An amount of RMB33 million was “de-ducted” as compensation to employees, and RMB9.6 million wasgiven to the managers and some employees as “bonuses.” In addi-tion, the managers were given a further 25 percent discount forpayment in cash. The privatization was carried out in late 2003,when SASAC had already called off any privatization of firms in thepower sector.

The SASAC regulation restricting MBOs was also ignored whenthe management of an Anhui-based listed company sealed a dealwith the provincial bureau of finance to acquire its SOE parent at adeep discount. At the end of 2003 the book value of the 27.1 per-cent ownership held by the SOE parent was RMB448 million, whileit was sold to an MBO vehicle at a price of RMB29.5 million. As inmost MBO cases, there was a large deduction that was supposedlyto compensate employees for “losing SOE employees status.” In thiscase, it amounted to RMB55.6 million.

CCTV: A Shandong SOE was partially privatized when a part ofits assets was sold to a private firm. The private firm operates in thepremises of the SOE, competes with the SOE in the same line ofbusiness, and has taken over the best workers of the SOE. It turnsout that the private firm is actually owned by senior managers ofthe SOE.

time of the fieldwork, some medium-size SOEs in Harbin were de-laying their gaizhi programs in anticipation that the governmentwould extend the debt write-offs and debt-equity swaps to medium-size SOEs.

Many gaizhi firms have not begun real gaizhi because they haverefused to take over the debts of the old firm. Firms sometimes takean indirect route to avoid responsibility for these debts—for in-stance, by setting up a new private firm to lease the assets and em-ploy most of the workers, leaving the old firm to carry the bankdebts. In many cases, the old firm is left with only a few employeesengaged to look after the workshops and equipment, and is likely togo bankrupt eventually.

The largest Chinese bank, the Industrial and Commercial Bank ofChina (ICBC), lost a case recently against a Shandong SOE borrower.The key issue was that the borrower had transferred all of its performingassets to a new company created by a debt-equity swap with two AMCs,Orient and Cinda. The ICBC insisted that the new company shouldshare the liability with the old one, but the local court rejected this re-quest based on what was believed to be an instruction from centralSASAC. The legal question it raised is the protection of other creditors

THE GAIZHI PROCESS

79

TABLE 3.7COMPARISON OF FINANCIAL DATA BETWEEN GAIZHI AND NON-GAIZHI

STATE-OWNED ENTERPRISES, 1995–2001(percent)

Outstanding loans/ Overdue loans/ Cash deposits/total assets outstanding loans total assets

Year Gaizhi Non-gaizhi Gaizhi Non-gaizhi Gaizhi Non-gaizhi

1995 48 58 13 41 3 41996 44 54 35 41 7 41997 47 54 17 46 7 41998 51 54 17 52 11 51999 51 55 20 54 15 52000 53 53 24 62 15 52001 60 52 24 66 18 5Average 50 54 21 52 11 4


when some creditors negotiate with the borrower for a debt-equityswap transaction.

Many gaizhi firms use bankruptcy to evade paying back bankdebts. In the survey, 90 percent of CEOs of gaizhi firms believed thatbankruptcy was a feasible option to resolve the problem of their en-terprise’s debts. With only a limited allocation for debt write-offs,banks have been trying to avoid losses, preferring that firms not be re-structured if that means they will go bankrupt. In 2002 the SupremePeople’s Court ruled that the people’s courts will not process bank-ruptcy cases if the main intention is to escape debts.16

Chinese banks have been asked to act as true commercial entities,making it difficult for them to give up their rights as creditors. Bankscan sue enterprises to recover loans, but lawsuits often end up in astalemate, with the banks not able to recover the loans and the firmsunable to continue with restructuring. Several firms in Harbin said thatgaizhi programs had stalled pending a court decision over the repay-ment of bank debts. As one bank manager stated, “We can sue thedebtors in court, but it would not do much to get our money back be-cause we can never get a single cent from the gaizhi firms, whether wewin the lawsuit or not.” Banks are also reluctant to seize the assets ofgaizhi firms, because often the revenue would not cover the cost of auc-tioning the assets. Their only recourse is to refuse to make new loansto firms that default on loan repayments.

When dealing with the debts of gaizhi firms, local governmentsusually give the greatest priority to employees. They also favor thegaizhi firm over the banks, because firms can provide tax revenues andemployment opportunities, both key indicators of the performance oflocal government. Despite the fact that it is neither a debtor nor a cred-itor, the local government is directly involved in gaizhi as the owner ofstate assets, and both gaizhi firms and banks prefer to resort to thelocal government rather than the court when there is a dispute.

It seems that many local governments have bought back, or areprepared to buy back, the debts of their SOEs from AMCs andbanks. In Chongqing, the municipal SASAC has reached an agree-ment with the ICBC to buy back RMB17.2 billion of debts of itsSOEs at a cash price of RMB4.3 billion. An AMC has been createdunder the municipal SASAC to hold these debts and to organize re-


80

16. Clause 12(1), Regulations on Issues of Processing Applications for EnterpriseBankruptcy, Supreme People’s Court, September 1, 2002.

structuring of the debtor SOEs involved. In Hunan province, the cap-ital city, Changsha, has done the same.

Central government regulations have been put in place to strengthenthe management of debt liabilities and to prevent firms from avoidingtheir debts.17 To prevent firms from using gaizhi to discharge theirbank debts, a register has been set up of firms that have tried to avoiddebts. Firms are given a schedule for measures to be taken to repay thedebt. If they fail to do so, they will be unable to apply for new loansor to open new bank accounts. Financial institutions can reduce thecredit rating for certain regions and suspend the reviewing of new proj-ects and the granting of new loans, which will directly affect local gov-ernment interests.

The survey found that other debts, such as overdue wages andpensions, were sometimes harder to deal with during gaizhi than over-due bank loans. Many SOEs owing pensions and wages have failed toobtain approval from the employee committee for their gaizhi pro-posals. To deal with these issues, it has become common practice forgaizhi firms to compensate employees first by selling assets and, if nec-essary, land.

Land-Use Rights and Gaizhi

Land in China is owned by the state or, in rural and suburban areas,by peasant collectives. Before 1995 all land-use rights for SOEs wereunder government control. The state being the owner of both enter-prises and land, the land-use rights were not defined and the land wasused free of charge by the SOEs. As a result, much of the land takenby SOEs was underutilized or misallocated. This situation continueduntil the 1998 Land Law, which defined land as a scarce resource andstipulated that it should be used to generate revenue.

At least four methods are used to deal with land-use rights ingaizhi: maintaining the appropriation by the firm of the land-use rights;leasing; buying; and converting land-use rights into equity. Maintainingthe appropriation and buying the land-use rights correspond to the twotypes of land-use rights in China: allocated land-use rights and grantedland-use rights.

THE GAIZHI PROCESS

81

17. See People’s Bank of China, “Notice on Strengthening Management of DebtLiabilities and Building a System of Preventing and Penalizing the Behavior ofAvoiding Financing Debts,” January 7, 1997.

Granted land-use rights are limited in time, cost the holder a sig-nificant amount proportionate to the market value of the land, andmay be held by private individuals and entities; allocated land-userights are usually given without the exchange of consideration andwithout time limitations, but may not be held by private individualsand entities. An allocated land-use right is not transferable, and tech-nically cannot be leased or mortgaged, while the holder of grantedland-use rights has a “user right” similar to that of an owner. Thegranted land is transferable and can be leased or mortgaged. Theholder of a granted land-use right needs to pay a significant land-grantfee to the government for acquiring it. The practices in land-use rightscan vary significantly by localities. In some localities, land users bothof allocated and granted land do not pay annual fees. Where they dopay, the fees paid by the holders of granted land are significantly lessthan those paid by the users of allocated land.

Despite the government’s active encouragement for firms to buyland-use rights, many gaizhi firms prefer to maintain their appropria-tion of the land-use rights. The Regulation on Appropriated Land-UseRights in SOE Reform, and the document entitled “StrengtheningLand Resource Management and Promoting SOE Reform,” issued bythe Bureau of National Land Resource Management, have stipulatedthat some gaizhi firms that currently occupy land can use it free ofcharge on conditions set by the local government.

The usual conditions are: gaizhi firms must be operating in sec-tors stipulated by the government; land in use must be assessed, andany extra land must be returned to the government; the land must beused for production, and the firm’s owners cannot change the landuse or lease the land to others without the permission of the relevantagency; and land-use tenure is usually set for between 5 and 10 years,depending on the land use. Although land appropriation has restric-tions, it can save gaizhi firms a great deal of money.

Leasing is a suitable method for gaizhi firms that cannot afford tobuy granted land-use rights. To reduce the financial pressure on gaizhifirms, firms are allowed to pay around 30 percent of the fee up frontand the remaining amount on an annual basis. Leasing has many ad-vantages, but there are debates over the arrangements. In some cities,gaizhi firms with leased rights were able to mortgage or use the land-use rights as collateral for a loan, while such practice was not allowedin other cities.


82

The method of converting land-use rights into equity is rarelyused but can complement other methods; for instance, when the gov-ernment wants to retain shares in gaizhi firms and to take over themanagement rights, or when joint ventures are formed with foreigncompanies.

The survey of 11 cities found that maintaining appropriation wasthe dominant method of acquiring land-use rights in all the cities ex-cept Xining, Harbin, and Huangshi (see table 3.8). Of the 199 gaizhifirms that replied to this question, 45 percent favored this method. Thesecond most common method was buying the land-use rights (31 per-cent), followed by leasing (13 percent) and converting land-use rightsinto equity (6 percent). Table 3.9 reports different forms of land-userights across different forms of gaizhi. The majority of the samplefirms, particularly listed companies and restructured firms, obtainedland-use rights for free. A high proportion of firms that were sold orleased out purchased the land-use rights. Unreported cases were highacross all forms of gaizhi firms.

THE GAIZHI PROCESS

83

TABLE 3.8METHOD OF OBTAINING LAND-USE RIGHTS

FOR 199 GAZIHI FIRMS, BY CITY

(percent)

Governmentconverted land

Allocated Leased Bought into equity Other

Harbin 37 5 42 5 11Fushun 67 0 0 33 0Tangshan 50 17 25 8 0Weifang 39 17 33 0 11Xining 36 18 45 0 0Huangshi 21 9 62 6 3Zhenjiang 52 18 15 9 6Hengyang 71 4 11 7 7Guiyang 47 18 29 0 6Lanzhou 100 0 0 0 0Chengdu 0 13 75 0 12Total 45 13 31 6 5


Although some regions have different procedures, the main stepsfor dealing with land issues are similar. In Hengyang, as an example,five steps must be followed:1. The gaizhi firm must produce documents and certificates to prove

property and land-use rights. It is also necessary to report how theland has been used.

2. The land resource authority should then estimate the value of theland.

3. A final appraisal must be made by an asset appraisal agency, and theresults are given to the gaizhi firm and the land resource authorityfor audit.

4. The gaizhi firm drafts a proposal for dealing with the land-userights, and adds any other gaizhi plans given to the local govern-ment. In other areas, this can be done by the local government, thegaizhi firm, or a third party.

5. The local government will call a meeting of all interested parties—such as the land resource authority, the gaizhi firm, and the banks—to discuss the draft gaizhi proposal. If it is approved, the proposalwill be enacted. If not, it will be returned for revision and furtherdiscussion.


84

TABLE 3.9METHOD OF OBTAINING LAND-USE RIGHTS, BY FORM OF GAIZHI

(number and percent)

Governmentconverted

Form of land intogaizhi Allocated Leased Bought equity Other Unreported

Internal 38 (31) 8 (7) 20 (16) 7 (6) 5 (4) 45 (36)restructuring

Employee 24 (23) 11 (11) 15 (15) 5 (5) 5 (5) 43 (43)shareholding

Sale 11 (22) 5 (10) 10 (20) 0 (0) 0 (0) 23 (48)Lease 6 (11) 3 (5) 13 (23) 0 (0) 1 (2) 33 (59)IPO 17 (57) 1 (3) 6 (20) 0 (0) 0 (0) 6 (20)Joint venture 1 (50) 0 (0) 0 (0) 0 (0) 0 (0) 1 (50)

with foreigncompany

Other 1 (8) 1 (8) 1 (8) 0 (0) 1 (8) 8 (66)


There have been a number of problems with the way land issueshave been dealt with during gaizhi. It has been difficult for the localland resource authority to manage the process of pricing the land, andits decision may not reflect the true value of the land. The decision tooffer a discount on the price to the gaizhi firm has often been made ar-bitrarily. Also, there are no standard regulations for dealing with landissues, leading to inconsistencies and uncertainties over the law. In adispute, the various parties will cite different regulations and it cantake a long time to settle the matter. A number of recent scandals re-lated to shady real estate transactions have increased the public’s demand for transparency and fairness in allocating land-use rights.18

The different regimes of land-use rights may also create a disin-centive for firms to privatize. Private firms will have to pay a signifi-cant fee to convert the allocated land-use rights into grant-use rights.Allocated land-use rights also limit the ability of gaizhi firms to obtainbank financing or to enter into joint venture arrangements with domes-tic or foreign partners. There have been numerous instances where for-eign prospective joint venture partners found late in the process thatthe land the local partner was planning to contribute as in-kind equityto the joint venture was in fact an allocated land and therefore notmarketable.

Over the years, China has evolved a reasonably reliable system ofregistration of land-use rights that permits identification of the own-ers. In an important recent development, it has been established thatthere shall be unrestricted public access to these land records nation-wide. This is an important development that investors, and particu-larly Western investors, have been seeking for some time.

The ability of the local governments to monetize the value of theland is a key factor in implementing gaizhi programs at the local level.Many local governments rely on the proceeds from the sale of land-userights to cover the social obligations related to restructuring. The cur-

THE GAIZHI PROCESS

85

18. The Beijing municipal government issued a decree in February 2004, knownas “Decree No. 4.” The decree caused extensive media attention because it requiresreal estate developers to compete for land in open bidding, rather than through“negotiation” with government officials, as was the case previously. More shock-ing to the public was the media discovery that in the four months prior to DecreeNo. 4, 100 million square meters of land was contracted out. This was equivalentto the total area contracted in the previous 10 years. As such, Decree No. 4 hasbeen described as a measure to “fix the fence after all the sheep have gone.”

rent attempts by the central government to fight overheating by exer-cising a tight control on land sales is having an effect on the pace ofgaizhi programs.19

Conclusion

Gaizhi has taken a variety of forms. The analysis of the factors deter-mining the choice of form of gaizhi finds that firm characteristics seemto be less important. Regional factors such as level of private-sectordevelopment and unobserved city characteristics are more important.The more radical forms of gaizhi involve the transfer of state assets toprivate owners. A properly managed transfer of assets from the stateto private owners is a crucial part of the gaizhi process, with strong im-plications for preventing losses of state assets and maintaining socialstability. There is a need to build an effective and transparent asset val-uation system to achieve a fair and smooth transfer of state assets. Theadministrative capability to handle state assets, both within the govern-ment and within firms, is often inadequate and needs to be strengthened.


86

19. In Qionglai city, Sichuan province, an urban utility privatization programstalled when the central government tightened up its land policy. The program wasdesigned with advisory services from one of the best-known NGO (nongovernmentorganization) economic research institutes, Unirule.

4Impact of Gaizhi on Labor

The restructuring of SOEs in China often involves the need to reduceredundant labor. A large number of SOE employees have already losttheir jobs as a result of restructuring programs. This chapter examinesgovernment policies regarding the redeployment of employees in gaizhiin different cities, as well as the instruments in place to ease the burdenon enterprises and employees. It looks at the issue of compensating SOEworkers, and examines how China’s nascent social security system hassupported those who have lost their jobs.

China’s Emerging Social Security System

The foundations of China’s new social security system were establishedin the 1990s to replace the enterprise-based system of a social safety net.The enterprise-based social security system included only SOEs. The newsystem covers state as well as non-state employees, although a significantnumber of self-employed, private, and joint venture companies remainto be covered. At this stage the system covers mainly urban areas1 andexcludes the rural, as well as most of the migrant, labor force. Financingand management are decentralized, which results in fragmentation andlimitations on labor mobility (World Bank 2003). The new social secu-rity system includes a pension system, a medical insurance system, anunemployment insurance system, and a minimal living safeguard system.

The new pension system was shaped by two State Council regu-lations in 1991 and 1995. The regulations replaced the SOE pension

87

1. The social security system provides only 2 percent coverage in the countryside.Expenditures on social security of farmers, who account for 70 percent of the totalpopulation, only account for 11 percent of total expenditures for the country(Dong and Ye 2003).


88

scheme with a unified urban employee pension plan that stipulatedthat urban employees and employers should both pay into pensionfunds. The 1991 decision of the State Council introduced universal pool-ing of pension insurance. It reduced, to some extent, the differen-tial pension burdens among enterprises; however, it could not cope withthe rapid aging of the employees within the pooling system. To dealwith the emerging cohort imbalance of the workforce, the 1995 regu-lations initiated a transition from Pay As You Earn (PAYE) to partialaccumulation of funds, and introduced the concept of individual con-tributions for pensions. The new system encourages enterprises andemployees to set up complementary pension funds and savings schemes.A 1998 State Council circular accelerated the transition to pooling ofbasic pensions at the provincial level and the transfer to localities ofthe industry-based pooling of basic pension insurance in 11 sectors.Provincial governments are able to decide on the details of the pensionsystem, which allows for different standards in different regions. As aresult of these reforms, China’s pension system is undergoing funda-mental changes: the PAYE system has been transformed into a partialaccumulation system with individual accounts (Zhang 2004); pensionpayments are being transformed from defined benefit to a combinationof defined benefit and defined contribution; and individuals now takepartial responsibility for their pensions through their contributions tothe pension premiums.

Medical insurance is also undergoing radical changes. The freemedical insurance of the past is being replaced by a new medical in-surance system. Similar to the pension system, the new medical insur-ance scheme combines social pooling with individual accounts, andindividuals are required to pay a portion of their medical expenses.

The foundation of the present unemployment insurance schemewas established with the Bylaw on Unemployment Insurance (StateCouncil 1999a). The bylaw states that employees of urban enterpriseswho have paid the unemployment insurance levy for at least one year,including those in SOEs, collectives, foreign enterprises, private enter-prises, and other urban enterprises, are covered by unemployment in-surance. Unemployment insurance is financed by enterprises, employees,interest from the insurance funds, government subsidies, and other funds.Urban enterprises should pay unemployment insurance levies equal to2 percent of their wage bill, and employees should contribute 1 percentof their wages.

IMPACT OF GAIZHI ON LABOR

89

Large cities that administer districts should place unemploymentinsurance funds under a unified management at the municipal level;otherwise the level of management should be regulated by the provin-cial government. The fund pays unemployment benefits and meets themedical costs of unemployed workers for a maximum of 24 months.Migrant rural workers work under contract in urban enterprises and donot pay the levy, receiving only a lump sum if their contracts are termi-nated or broken after one year of work.

The social relief system, which guarantees a minimum standard ofliving for urban residents, was established through the Bylaw on UrbanResidents’ Minimal Living Safeguard (State Council 1999b). The reg-ulation sets a minimal living standard for urban residents. Those whohave a family per capita income lower than the minimal living stan-dard can receive a basic payment from the government. The minimalliving standard varies according to the local cost of living. Applicationsfor payment are reviewed by the community administration and ap-proved by the county government.

The social security system provides a basic safety net for most urbanworkers. The basic pension system and the unemployment insurancesystem each cover around 104 million people. The unemployment ben-efits are quite low, paying only around RMB200–300 ($24–37) a month.Fundamentally, unemployment insurance as a portable right availableto all workers does not exist in China. The current pension system isseverely underfunded, and deficits are increasing in many regions. In2002, according to official statistics, a total of 95 percent of redundantworkers and 98 percent of retired employees received their allowanceson time. The minimal living safeguard system has been established in30 out of China’s 31 provinces. Currently, the share of pensions andsocial welfare relief amounts to 10.8 percent of total governmentexpenditures (IMF 2004). The government intends to further expandthe social security system and to increase the share of the budget spenton social security to 15–20 percent (MLSS 2001).

Government Policy toward Unemployment in the Gaizhi Process

Between 1995 and 2003 the number of state-owned and state-controlledindustrial enterprises fell from 118,000 to around 34,000, and total em-ployment in the SOE sector fell by around 44 million (see table 4.1). The


90

TABLE 4.1TOTAL EMPLOYMENT IN THE STATE-OWNED ENTERPRISE SECTOR,

SELECTED YEARS, 1995–2003

Reduction inNo. of SOEsa Employmentb employmentc

Year (thousand) (million) (million)

1995 118.0 112.6 0.01997 110.0 110.4 2.21999 61.3 85.7 26.92001 46.8 76.4 36.22002 41.1 71.6 41.02003 34.2 68.8 43.8

a. In the industrial sector, including state-controlled enterprises.b. In the whole SOE sector.c. Compared with 1995.SOURCE: National Bureau of Statistics, various years.

2. Loss of state-sector jobs does not imply an automatic increase in temporary orstructural unemployment, as a significant number of state jobs are reclassified asprivate-sector jobs due to privatization (see chapter 2).

number of lost state-sector jobs totaled 17 percent of urban employ-ment in 2003.2 During this period, official urban unemployment in-creased from 2.9 to 4.3 percent. The “true” unemployment rate ismost likely significantly higher. Giles, Park, and Zhang (2004) esti-mate that the unemployment rate among urban permanent residentsincreased from 6.1 percent at the end of 1995 to 11.1 percent at theend of 2002. The unemployment rate of all urban workers, includingtemporary residents, has increased over the same period from 4.0 per-cent to 7.3 percent. From an international perspective, China is closeto average unemployment levels in transition economies and hasreached rates of unemployment similar to high-unemployment ad-vanced countries such as Italy, Germany, and France.

The rise in unemployment has resulted in some labor unrest.The number of labor disputes is increasing rapidly, growing at about20 percent per annum. Information on social instability caused bygaizhi is limited, but conflicts were observed in the sample cities; insome cities, petitions were frequently presented to the municipalgovernment. Workers complained about being made unemployed


91

while receiving little compensation, and about the unfair transfer offirm assets.

The Chinese government considers protecting the welfare of stateemployees a top priority in SOE restructuring. The underlying princi-ple behind numerous regulations dealing with employment issues dur-ing the restructuring of SOEs is that the legitimate rights and lawfulinterests of employees shall not be prejudiced under an SOE restruc-turing or transfer of state-owned assets. For example, the governmenthas stipulated that departments in charge of SOEs, departments forlabor administration, and the trade unions have obligations to helpSOEs redeploy their excess workers (State Council 1993b). Detailedregulations outline the benefits or compensation that should be givento those who take early retirement, resign, or take unpaid holidays,and to those whose jobs are terminated. Service businesses set up bySOEs to reemploy their redundant workers are given tax holidays andother favorable treatment.

In November 2003, SASAC issued “Several Opinions on Ratio-nalization of the Reorganization of State-Owned Enterprises.” TheOpinions requested that a restructuring plan, and in particular anyplan concerning the resettlement of employees, also be subject to theapproval of the congress of the SOE’s employees or of a meeting of em-ployee representatives before the plan can be implemented. The SOEshall use its existing assets to settle in full all unpaid wages, employees’medical bills, inappropriately used employee housing funds, and unpaidsocial security contributions. The regulations also state that the em-ployees’ resettlement program relating to the transfer shall be examinedand agreed by local labor administrative authorities. Failure to properlyresolve employment issues may result in the transfer’s being stoppedby SASAC or held invalid by the courts.

There are three categories of redundant workers. The first is inter-nal retirement. Employees may take an early retirement but continueto be paid by the factory, typically at a lower rate, until they reach thelegal retirement age, after which they are paid their retirement wageout of the social security system. The second category is xiagang.Literally, it means “losing the position.” As explained in chapter 1, anemployee is called a xiagang employee if he stops working in the fac-tory but nevertheless keeps a nominal tie with the factory throughwork registration. Presumably, a xiagang employee can go back to workin the old factory if it needs people, but in reality that does not happenoften. The third category is unemployment. An unemployed worker

loses his job and no longer keeps a tie with his old factory. In effect,xiagang and unemployment are equivalent in terms of the status of theworker, but government policies are different for the two categories ofworkers. The benefit of being described as xiagang, rather than as of-ficially unemployed, is that people are still counted as employees andmay have the chance to resume working for the enterprise. The centralgovernment has announced that the category of xiagang worker willbe phased out by the end of 2005.

A security fund was set up by a 1993 State Council regulation tocover all those who lost their state jobs, including employees from en-terprises that went bankrupt or were restructuring toward bankruptcy,that were disbanded by the government, or that had halted productionwhile being restructured. The fund also covers those whose employ-ment contracts were terminated, those who were dismissed, and anyothers who qualify under state laws and state or provincial govern-ment regulations (State Council 1993a). This fund was later replacedby two allowances: unemployment insurance, which pays out a benefitto unemployed state and non-state employees; and the xiagang livingallowance. Similar policies have been used throughout the country forredeploying SOE employees, although some policies have been im-plemented differently from region to region.

Each SOE had to set up a reemployment center to take unemployedand xiagang workers. The main purpose of the center was twofold: toprovide the laid-off workers with minimum living expenses; and toprovide training and job information for those workers. Between 1998and 2001, 13 million xiagang and unemployed workers found jobsthrough the centers, according to official statistics. The reemploymentcenters were financed by three parties: central and local governments,collections of contributions from the public, and the contributions ofenterprises themselves. Centers established by SOEs under central orlocal government control were to be financed solely by the central orlocal government (State Council Office 1999). When the state contin-ues to have full ownership or the controlling share in the enterpriseafter gaizhi—for instance, in a joint venture or a firm that has been cor-poratized—usually a large number of xiagang workers remain in thereemployment service center.

The central government requires SOEs to consult with the em-ployee representative committee or the union at least 15 days beforeemployees are made redundant. Xiagang workers are then expected tosign a contract with the center to receive a living allowance for the next


92

three years (two years, in some cases) or until they are reemployed.After this period they become officially unemployed and are entitledto receive unemployment insurance for another two years before theygo on to the minimum living allowance, which covers urban residentswhose family per capita incomes are below the poverty line. In 1999the five state ministries and committees decided that these three pay-ments should be increased by 30 percent. Overdue wages or medical ben-efits were not affected by their xiagang status (CCP Central Committeeand the State Council 1998; MLSS and others 1998). Reemploymentservice centers also paid the basic medical insurance of xiagang work-ers (State Council 1998). In line with government policies of gradu-ally phasing out xiagang, recently there have been closures of thereemployment centers and movement of laid-off SOE workers to theregistered unemployment rolls.

During gaizhi some enterprise managers and local governmentspersuade workers to take early retirement in order to reduce the num-ber of redundant workers, although this is against central governmentpolicy (see State Council Office 1999). The usual retirement age for SOEworkers is 60 for male workers and staff, 50 for female workers, and 55for female staff members.3

Firms rarely discharged workers in the early stages of SOE re-form, but this has become more common, especially when SOEs areprivatized. The central government has accepted the practice, stipu-lating that a lump sum severance should be paid, usually at a rate ofone month’s wage for each year of service (State Council 1993b). Theactual payment varies in different locations. In some places the com-pensation rate is two months’ pay for every year of work.

In Harbin, for example, the municipal government set the lumpsum compensation at two to three years of wages at the level of theprevious year’s city average for all those employed by SOEs before1986. Those made redundant do not receive the unemployment ben-efit.4 Workers employed by SOEs from 1986 onward receive onemonth’s average wages at the level of the average wage of the en-terprise for every working year, as well as the unemployment benefit.


93

3. Staff members are distinguished from workers by their involvement in admin-istrative work.4. In 1986 new SOE workers were employed on contracts and no longer had life-time employment status.

Under the new system, both types of workers receive the basic pen-sion if they continue to pay their pension contributions until they retire.

Some local governments have a policy that workers must decidewhether to accept a lump sum payment and then search for a new job,or continue to work in the privatized firm but face the risk of being firedin the future and receiving a lower or no compensation.

A useful summary of China’s labor policies is the employmentindexes developed by the Doing Business project of the World BankGroup. Figures 4.1 and 4.2 present measures of employment rigid-ity in China versus other countries from the World Bank Group’sDoing Business in 2005 project (World Bank Group 2004). Two


94

0 20 40 60 80 100

FIGURE 4.1Index of Rigidity of Employment for

Various Asian Countries, 2004

Hong Kong*

Malaysia

Australia

Papua New Guinea

China

Philippines

Thailand

Indonesia

Korea, Rep.

Index (0 = least rigid)

Note: The employment regulation index is an average of three sub-indexes: difficulty of hiring, rigidity of hours, and difficulty of firing. Each index takes values between 0 and 100, with higher values implying more rigid regulation. Difficulty of hiring covers the regulation of fixed-term contracts (on duration and use) and the minimum wage relative to the average value added per worker. Rigidity of hours covers restrictions on weekend and night work, working time requirements, and mandated days of annual leave with pay.Source: World Bank Group (2004).


95

measures are presented: an employment regulation index and a costof firing measure.

As shown in figures 4.1 and 4.2, China has a relatively low rigid-ity of employment by regional standards. Its costs of firing, however,are higher than the regional average.

The Impact of Gaizhi on Employment and Labor Force Structure

Excess labor has created a massive burden on SOEs; however, thereare structural obstacles to reducing redundant labor. All the samplecities set a limit, usually 10 percent of a firm’s workforce, for lay-off in the privatization/gaizhi package. In practice, a restructured firmwas often required to take all its employees, with concessions from

0 20 40 60 80 160140120100

FIGURE 4.2Cost of Firing Employees in Various Asian Countries, 2004

(in weekly wages)

New Zealand

Australia

Papua New Guinea

Thailand

Malaysia

China

Philippines

Indonesia

Korea, Rep.

Index (0 = least cost, global)

Note: Difficulty of firing covers workers’ legal protections against dismissal, including the grounds for dismissal, and procedures for dismissal (individual and collective). The cost-of-firing indicator measures the cost of advance notice requirements, severance payments, and penalties due when firing a worker, expressed in terms of weekly wages.Source: World Bank Group (2004).

governments in terms of lower privatization prices and/or the retain-ing of land-use rights.

Nevertheless, layoffs have become a common practice. The averagedischarge rate5 in the sample is about 35 percent for the period between1995 and 2001. Both gaizhi and non-gaizhi firms have shed labor. Infact, we estimate that the discharge rate for non-gaizhi firms is about4 percent higher than that for gaizhi firms when controlling for assetsize, firm obligations in arrears, and pre-gaizhi profitability.

There are differences in the discharge rate by types of ownership.Using wholly state-owned enterprises as a control group, and con-trolling for asset size, overdue obligations, and pre-gaizhi profitabil-ity, we find that state-controlled firms6 have a discharge rate that is14.7 points lower—and privately controlled firms have a dischargerate 12.4 points lower—than the discharge rate for wholly state-owned firms (see table 4.2). We also find that firms with higher over-due payments have a higher discharge rate, while larger firms and firmswith a higher level of pre-gaizhi profitability have a lower dischargerate. All coefficients are statistically significant at the conventional5 percent level.


96

5. Defined as the ratio of xiagang and laid-off employees over working, xiagang,and laid-off employees.6. These are firms with diversified ownership, but still with majority (more than50 percent) state control.

TABLE 4.2REGRESSION ANALYSIS OF FACTORS INFLUENCING DISCHARGE

RATE BY TYPE OF OWNERSHIP

Variable Coefficient Standard error P > t

Constant 0.35 0.02 0.000State-controlled −0.147 0.04 0.000Privately controlled −0.124 0.037 0.001Total assets −0.001 0.00001 0.000Overdue payments 0.000 0.000 0.035Pre-gaizhi profitability −0.23 0.04 0.00

NOTE: Variables are averages over the time period. R2 is 0.2.SOURCE: Authors’ calculation.


97

These are interesting and important findings, in our view, as they goagainst the conventional wisdom that restructuring and privatizationare necessarily associated with more job losses. In the case of China, theopposite holds: over the sample period, when controlling for size andprofitability, non-gaizhi firms and purely state-owned firms sufferedlarger job losses. To explore this relationship further, we look at the timepattern of the discharge rate relative to the year of gaizhi. We estimatethe discharge rates for the years before, during, and after gaizhi usingregression analysis and controlling for size, profitability, and type ofgaizhi. The results are presented in figure 4.3. They show a clear stylizedpattern: the discharge rate begins to grow four years before restruc-turing, reaches a climax in the first year after gaizhi, and then graduallybegins to decline.

Firms that went through a more radical form of gaizhi discharged(xiagang and layoffs) more workers in the year when gaizhi happened(see figure 4.4). The leasing firms discharged almost half of the totalnumber of employees, and the firms that were sold out dischargedmore than half of the employees. Publicly listed firms have almost nolayoffs, but we need to keep in mind that these firms typically undergo

Four Three Two One Gaizhiyear

One Two Three Four Five

FIGURE 4.3Rate of Employee Discharge Four Years before and

Five Years after Year of GAIZHI

Sources: Survey data; authors’ estimates.

0.20

0.16

0.12

0.08

0.04

0


98

significant restructuring in what has come to be known in China as“packaging for listing” (Tenev, Zhang, and Brefort 2002).

Similarly, firms controlled by outsiders discharged more workersthan firms controlled by insiders. The discharge rate in outsider-controlled firms was 37 percent, but the figure was 32 percent in insider-controlled firms. Inside an SOE there is often a set of implicit contractsthat provide protection to all the employees in terms of job securityand other tangible benefits. The more radical gaizhi is, the easier it isfor the firm to fire workers, because radical gaizhi measures introduceinto the firm outsiders who are not bound by these implicit contracts(Shleifer and Summers 1988).

Beyond the first year of gaizhi, however, gaizhi firms have donebetter than non-gaizhi firms in the sample. Figures 4.5 and 4.6 showthe time trends of the cumulative percentages of xiagang and laid-offs for the two types of firms. The contrast in terms of xiagang isdramatic. In 1995 the xiagang rate was under 10 percent of the totalnumber of employees in both types of firms, with the non-gaizhifirms being slightly higher; but by 2001 the accumulated xiagang

0 10 20 30 40 50 60

Publicoffering



Sale

Lease

Other


FIGURE 4.4Rate of Employee Discharge by Type of GAIZHI

in the Year of GAIZHI

Percent


99

0

20

40

60

80

1995 1996 1997 1998 1999 2000 2001

Gaizhi firms

Non-gaizhi firms

Note: The number of accumulated xiagang workers was calculated under the assumption that their average stay in the reemployment service centers was three years (reflecting the difficulty of gaining employment and the provision of a subsidy over the period that acts as a disincentive to seek employment).Source: Survey data.

Percent

FIGURE 4.5Percentage of Accumulated XIAGANG inGAIZHI and Non-GAIZHI Firms, 1995–2001

0

3

2

1

4

5

6

1995 1996 1997 1998 1999 2000 2001


Percent

FIGURE 4.6Percentage of Accumulated Layoffs in GAIZHI and

Non-GAIZHI Firms, 1995–2001

Gaizhi firms

Non-gaizhi firms


100

rate had reached 70 percent in non-gaizhi firms, whereas it was 30 percent in gaizhi firms. The contrast in terms of layoffs, however,is less dramatic. In fact, gaizhi firms have maintained a faster pacethan non-gaizhi firms in accumulating laid-off workers. This showsthat gaizhi firms are more inclined than non-gaizhi firms to termi-nate the employment contract if they want to shake off their excessiveemployment.

The finding that privatization had led to significantly fewer lay-offs and xiagang workers after the first year of gaizhi is an encourag-ing one. Statistics for employment growth show similar results. Overthe years, the size of the on-duty workforce has been shrinking in allthe sample firms, and the reduction was accelerated in more recentyears (see figure 4.7). Fully state-controlled SOEs had the highest rateof reduction in on-duty workers: an average annual reduction of 7.3 per-cent. The negative growth of on-duty workers in privately controlledfirms was somewhat lower, at an annual rate of 5.7 percent. Firms withprivate shares but still controlled by the state had the lowest rate of neg-ative employment growth, at an annual average of 3.1 percent. Therewas a convergence in more recent years, however. In 2001 the reduc-tion rate was about 8 percent for all three types of firms. In short, we donot find that privatization is associated with larger job losses relativeto the alternatives.

When compared by the type of gaizhi, the reduction of the work-force has been the most radical in firms sold out and incorporated. On

1996 1997 1998 1999 2000 2001


Percent

FIGURE 4.7Growth of the Workforce in the Sample Firms, 1996–2001

–12

–10

–8

–6

–4

–2

0

2

Fully state-owned

State controlled

Privately controlled


101

average, the number of on-duty workers decreased by an annual rateof 7.7 percent and 7.4 percent in those two types of firms, respectively(see figure 4.8). Again, unlike what happened in the first year of gaizhi,the reduction of the labor force in the subsequent years was not nega-tively related to the extent of privatization and restructuring. An ob-servation to reinforce this conclusion is that outsider-controlled firmsmaintained a lower rate of reduction than insider-controlled firms. Theaverage reduction in outsider-controlled firms was 3.6 percent, but thefigure was 7 percent in insider-controlled firms.

Regression analysis using the fixed-effect panel method, and con-trolling for asset size and overdue payments, confirms the above findings.We find that gaizhi firms had a workforce growth rate 1.78 percentagepoints higher than non-gaizhi firms, and that a state-controlled firm hada growth rate 1.84 percentage points higher than fully state-owned firms.Privately controlled firms also had a higher growth rate than fully state-owned firms, but the gap is statistically insignificant. The relative goodperformance of state-controlled firms may be related to self-selection; thatis, the state may deliberately maintain control in firms with better growthpotential. The fixed-effect estimation places a control for the selection

–10 –8 –6 –4 –2 0

Publicoffering



Sale

Lease

Other


FIGURE 4.8Average Growth of On-Duty Workforce by Type of

GAIZHI in the Sample Firms, 1996–2001

Percent

problem, however. An alternative explanation is that the state-controlledfirms may continue to get help from the government, yet at the same timeshare in the benefits provided by the introduction of private shares.

In addition to changes in the overall size of the labor force, restruc-turing can have an impact on the structure of the workforce. We lookat the ratio of full-time to casual workers relative to the year of gaizhi.The ratios of full-time to casual workers are estimated using regressionanalysis, and the results are shown in figure 4.9. There is a cleartrend toward a higher share of casual, most likely migrant, work-ers. This trend seems to predate the gaizhi process. Gaizhi seems toreverse the trend temporarily, but the process accelerates followingthe second year after restructuring. Firms that are privately controlledtend to have the lowest ratio of full-time to casual workers.

A concern that is frequently raised regarding China’s industrial re-structuring and privatization is that women may be more likely thanmen to lose their jobs in the process. We do not find evidence in our sam-ple of a systematic bias against women. Anecdotal evidence suggests thatfemale migrant workers are often the preferred type of worker by em-ployers in China, as they are seen as a docile and hard-working laborforce (Marquand 2004).


102

Four Three Two One Gaizhiyear

One Two Three Four Five

FIGURE 4.9Ratio of Full-Time Workers to Casual Workers

Four Years before and Five Years after Year of GAIZHI


70

60

50

40

30

20

10

0

Percent

In summary, our analysis suggests that gaizhi and privatized firmsdischarged more workers in the year when restructuring took place, butthat over the sample period they shed less labor than non-gaizhi andwholly state-owned firms. This is an important finding, as gaizhi andownership diversification have often been viewed in China as causingmassive unemployment. The results obtained here show that within arelatively short period of time, restructured firms have a better labor per-formance than unrestructured firms.

Gaizhi and Obligations of Firms to Workers

During the central planning era, workers and staff were permanentlyemployed by SOEs until they retired and received a pension from theenterprise. In the mid-1980s it was decided that new recruits would nolonger have a job for life and would be hired under a term contract. Inpractice the situation changed very little, as most contracts were auto-matically renewed. In the mid-1990s a large number of SOE workerswere laid off when enterprises went through gaizhi. Jobs in the new pri-vatized enterprises were no longer as secure. For this reason, many arguethat workers deserve to be compensated for the loss of job security andlifetime benefits.

Some cities have compensated all SOE workers when their job con-tracts were terminated during gaizhi, regardless of whether they were tobe made redundant or employed by the new owners. The ability of localgovernments to provide compensation under this system depends onwhether the SOEs and the government are in a strong financial position.

In Harbin, Chengdu, and Hengyang, a share of the SOE’s assetsis used to compensate workers before the firm is sold. Only thoseworkers made redundant, however, receive cash. Those workers whocontinue to work in the privatized enterprise are usually required touse their compensation to buy shares in the enterprise. The enterprisebecomes a shareholding cooperative, although the management com-monly holds the largest amount of shares. In other cases, funds are leftin the firms to enable them to pay compensation in the future. In bothcases it is possible that workers who stay with the firm may not re-ceive the compensation directly. It has become more common for thefunds to be looked after by enterprises, rather than to be given directlyto workers.

In some cases of gaizhi in Tangshan, the new owners received adiscount on the price of the SOE assets on the basis that only 10 per-


103

cent of the workers would be made redundant and need compensation.This has reduced the government’s costs in gaizhi, especially in situa-tions where SOEs have zero or negative net assets. It has also, however,made gaizhi more difficult in firms where there is serious overemploy-ment. Whether the level of discount is acceptable to the new owners, whowill have to bear the cost of any additional compensation, will dependon the attractiveness of other elements of the sales package.

The municipal government in Guiyang has adopted a policy offormally discharging all workers during gaizhi, which is followed by a(re)hiring process. This policy has reduced the pressure on buyers, but,given that the private sector is underdeveloped and there are only limitedjob opportunities, there has been greater resistance from workers.

In terms of the settlement of the redundant workers, non-gaizhienterprises in the sample appear to have more generous social securityprovisions than gaizhi firms, but to be slower in paying workers’ wages.Table 4.3 shows overdue wages and other overdue payments, such aspensions or supplemental pension payments, xiagang allowances, andmedical reimbursements, as a percentage of total wages in 2001.Gaizhi enterprises had a significantly lower proportion of overduewages than non-gaizhi firms, possibly because of a combination ofoverdue wages being settled during the gaizhi process and because ofan improvement in the financial situation of firms following gaizhi. Theshare of other overdue payments was higher in gaizhi than in non-gaizhi enterprises, however. In non-gaizhi firms the amount of overduewages amounted to 23.3 percent of the payroll in 2001, and the amountof overdue social security and other payments amounted to 21 percentof the payroll in the same year. Gaizhi firms have done much better in

104

TABLE 4.3OVERDUE WAGES AND OTHER OVERDUE PAYMENTS

AS A PERCENTAGE OF PAYROLL, 2001

Overdue wages Other overdue payments

Non-Gaizhi 23.3 21.0Gaizhi 5.7 26.3Total 12.7 24.2

NOTE: The category “other overdue payments” includes overdue pensions, xiagangallowances, and medical reimbursements.SOURCE: Survey data.


terms of overdue wages, reducing it to 5.7 percent of the payroll, but havebeen equally poor in paying social security, medical insurance, and otherpayments, as the total amount of these arrears was 26.3 percent of thepayroll.

Local governments have emphasized settling the overdue wagesand other payments, and have often made this a condition in the pri-vatization package. There are three major methods of settling over-due wages and other payments (see table 4.4). The first method is forthe government to give the firm cash with which to pay its workers.The percentage of overdue payments settled by this method was lowin the sample, only 1.6 percent. The second method is for the gov-ernment to give the new owners discounts on the privatization price,usually taken on firm assets. This was quite common in the sample.A total of 22.6 percent of the overdue payments were settled in thisway. The third method is for the firm to take on the task and to di-gest the overdue payments itself. A total of 66.6 percent of the over-due payments were settled in this manner.

Table 4.4 breaks up the figures by city. Chengdu relied heavily ongovernment subsidies to repay the overdue payments. Cash from the


105

TABLE 4.4DISTRIBUTION OF METHODS OF SETTLING OVERDUE PAYMENTS

IN THE ELEVEN SAMPLE CITIES, 2001(percent)

Cash paid by Discount Borne bygovernment on assets firm Other

Chengdu 20.0 40.0 30.0 10.0Fushun 0.0 0.0 100.0 0.0Guiyang 0.0 32.9 53.9 13.3Harbin 0.0 18.2 81.8 0.0Hengyang 0.0 14.3 76.2 9.5Huangshi 4.6 5.3 74.3 15.8Lanzhou 0.0 40.0 36.0 24.0Tangshan 0.1 29.5 60.5 10.0Weifang 0.0 38.4 54.6 7.1Xining 0.0 0.0 100.0 2.0Zhenjiang 0.0 23.1 76.9 0.0Total 1.6 22.6 66.6 9.2


government digested 20 percent of all the overdue payments. Othercities seldom provided cash to repay overdue payments. (Huangshihad a figure of 4.6 percent and Tangshan 0.1 percent.) With the ex-ception of Fushun, Huangshi, and Xining, discounts on assets werecommonly used by all other cities, especially in Chengdu and Lanzhou.In contrast, the burden was 100-percent shouldered by the firm inFushun and Xining.

The gaizhi firms used different methods for deploying redundantemployees. Most redundant workers (38 percent) found another jobby themselves (see table 4.5). This is similar to the findings by Giles,Park, and Cai (2003), who report, based on data from the China LaborUrban Survey conducted in five Chinese cities in 2001, that 34.8 per-cent of individuals experiencing job separation were employed againwithin 12 months. A large share (26 percent) were classified as xia-gang, most likely by enterprises that continued to be owned by thestate after gaizhi. Far fewer redundant workers were reemployed in a


106

TABLE 4.5DISTRIBUTION OF METHODS OF REDEPLOYING REDUNDANT WORKERS

IN THE ELEVEN SAMPLE CITIES, 2001(percent)

Hired byEnter Assigned to other firms,

reemploy- a separate as arrangedFind ment entity in the by the

City new jobs center firm government Other

Chengdu 71.6 26.5 1.9 0.0 0.0Fushun 19.0 67.0 0.0 6.3 7.8Guiyang 29.0 14.1 7.5 10.2 39.2Harbin 26.2 50.6 11.8 2.1 9.3Hengyang 36.6 15.7 21.8 1.0 28.7Huangshi 40.5 24.9 21.6 1.0 14.7Lanzhou 40.7 23.3 10.0 0.3 25.7Tangshan 35.3 37.4 1.9 0.0 25.5Weifang 4.4 37.5 12.2 0.0 43.9Xining 65.6 20.4 8.0 0.0 6.0Zhenjiang 45.1 4.5 9.8 9.7 31.1Total 37.5 25.9 11.6 3.0 22.9


separate entity set up by the enterprise or by the government (12 per-cent and 3 percent, respectively).

Methods varied from city to city. In Weifang only 4 percent of re-dundant workers had to find another job on their own, and 37 percentwere classified as xiagang. In Huangshi and Hengyang, both in centralChina, a large proportion of workers were assigned to a separate en-tity set up by the companies. In Zhenjiang and Guiyang, around 10 percent of redundant workers were reemployed in entities set up bythe local government.

Reemployment is always the best solution for redundant workersand the state, because it lowers the cost of gaizhi. This is why govern-ments give tax holidays and deductions to former SOE employees whostart a business, or to new enterprises whose xiagang workers accountfor 60 percent or more of their workforce. Government departmentsand banks also help these firms with registration and access to bankloans. In some cases the government assists redundant workers in find-ing other work, but this practice accounts for only a small proportionof employee redeployment.

When SOEs are sold, local governments can use part of the assetsto compensate xiagang or discharged employees, or they can reducethe price of the assets (including land) to compensate the new ownersfor taking responsibility for employees who will have to be compen-sated if they are fired in the future. Many enterprises are sold at zerovalue, and in some cases the government pays extra subsidies to coverthe redeployment of the employees. (See chapter 3 for details.)

In Tangshan city, buyers of SOEs must agree to take responsibil-ity for all employees and retired workers.7 The government compen-sates buyers for the future cost of redeploying workers by reducing theprice of the assets on the basis that 10 percent of employees will bemade redundant. The new owners frequently reported that the com-pensation was inadequate, as the cost of redeploying workers was muchhigher. There were also complaints from employees that state assetssuch as land were being sold off too cheaply.8


107

7. Workers and staff members retired from SOEs are covered by the new pensionsystem. Pension payments are low, however; perhaps lower than the common pen-sion rate in SOEs. Enterprises still have to pay the difference between the actualand common pension rate.8. During the field study we found only a few cases in different cities of new ownerscomplaining about the overvaluation of state assets.

In Harbin the municipal government allows enterprises goingthrough gaizhi to sell part of their assets to compensate employees whoare being discharged. Land-use rights can be transferred if funds areinadequate.

In Nan An district, Chongqing city, similar to the practice inGuiyang, the employment contracts of all SOE employees are termi-nated. Compensation is given before the enterprises are sold or de-clared bankrupt, regardless of whether workers will be reemployed bythe new owners of the enterprises. This has made gaizhi much easierand eased the dissensions that occur after gaizhi. A large percentage ofthe district’s SOEs have completed gaizhi. The proportion of enter-prises that have done so is much smaller in the other districts, whichadopted a municipal government policy of redeploying employeesafter the enterprise went bankrupt. In these districts, state assets areused to pay off debt, leaving little for employee redeployment. If gov-ernments wish to compensate the employees before gaizhi, they needthe consent of the creditor banks. Usually there also needs to be a largepositive net value of state assets to cover the costs of employee re-deployment. In Nan An district, land values are high and the sale ofland-use rights has financed redeployment.

In what is a typical practice for China, Guiyang has a policy thatgaizhi plans have to pass a vote of the employee conference or em-ployee representative committee. Employees in this city must agreewith the decision to terminate their contract, and they should receivea lump-sum payment of two months’ wages for every year of service.Despite these relatively generous conditions, which were not commonelsewhere, the gaizhi process in Guiyang was not easy. The private sec-tor in the city was underdeveloped, and the prospects of finding newjobs quickly were poor. As shown in chapter 2, the development of theprivate sector in the local economy helps provide employment oppor-tunities for redundant state workers and is a crucial factor for progressin gaizhi.

Gaizhi and Changes in Compensation Schemes

Approximately two-thirds of gaizhi firms have changed their workerremuneration schemes. Figure 4.10 shows the income shares that em-ployees of gaizhi and non-gaizhi firms received in 2002 from fixedwages and salaries, piece-rate wages, bonuses, dividends, and othersources of income. Employees in both gaizhi and non-gaizhi firms


108

earned most of their income from fixed wages, but the share waslower for those in gaizhi firms (at 46 percent of total income) than forthose in non-gaizhi firms (at 68 percent).

In contrast, gaizhi firms relied more on piece-rate wages to moti-vate their employees. They also gave slightly more in bonuses to theiremployees. Although non-gaizhi firms were not able to give shares totheir employees, in gaizhi firms the practice was not that common—dividends made up only 1.7 percent of employee income. Gaizhi firmspreferred performance-based incentives over rewarding employeesthrough dividends. This can be seen as a rational choice on the partof the management, because the shares were dispersed throughout thefirm, and dividends would not have been a particularly effective tool.There are examples, however, where the practice of “incentivizing”employees has been abused, resulting in assets stripping of SOEs.9

This chapter has shown that gaizhi firms have managed a smallerreduction in their workforce than non-gaizhi firms. It is of interest to


109

FIGURE 4.10Employee Remuneration by Type in GAIZHI and

Non-GAIZHI Firms, 2002


70

60

50

40

30

20

10

0

Percent

Fixed wage Wage by piece Bonus Dividends Other

Gaizhi firmsNon-gaizhi firmsAll firms

9. It was recently reported in the Chinese media that the whole management of aBeijing SOE was jailed after they collectively decided to give away RMB24 millionto their employees as bonuses through various forms, including insurance policies.

know whether they have also improved the welfare of the workers leftin the firm. Overall, gaizhi firms have maintained a higher growth ratein wage payments than non-gaizhi firms. Over the period 1995–2001,the average wage in gaizhi firms increased by an annual growth rate of8.1 percent, whereas it was 5.3 percent in non-gaizhi firms. Figure 4.11presents the time trend of the differences between the two groups offirms. With the exception of 2000, gaizhi firms maintained a highergrowth rate every year than non-gaizhi firms. Econometric analysis ofwage growth, controlling for asset size and profitability, confirms themain findings. Gaizhi firms maintained a growth rate that was 3 per-centage points higher than that of non-gaizhi firms. While privatelycontrolled firms were not statistically different from the fully state-owned firms, state-controlled firms achieved a growth rate 2.9 per-centage points higher.

Within the gaizhi firms, those having adopted more radical formsof gaizhi had achieved a higher growth rate for their average wages.Firms that were sold had a wage growth rate of 10.7 percent in the pe-riod 1995–2001, those with internal restructuring had 6.5 percent, andthose going public achieved only 5.4 percent, almost half of that achievedin the firms sold to private owners (see figure 4.12). Consistent with this


110

FIGURE 4.11Rate of Wage Growth in GAIZHI and Non-GAIZHI Firms,

1996–2001


14

12

10

8

6

4

2

0

Percent

Non-gaizhiGaizhi

1996 1997 1998 1999 2000 2001


111

divergence, firms controlled by outside shares registered a growth rateof 8.2 percent, as compared with 5.1 percent achieved by firms con-trolled by inside shares.

Conclusion

This chapter discussed some of the main issues concerning the impactof gaizhi on labor. A somewhat surprising result offered by this chap-ter is that gaizhi and privatized firms have maintained a lower rate ofemployment reduction and a higher rate of wage growth than non-gaizhi and fully state-owned firms, respectively. Consistent with theconventional belief, gaizhi firms discharged more workers in the yearwhen gaizhi was implemented than non-gaizhi firms, but they wereable to afford to retain more workers in the subsequent years. Therefore,gaizhi and privatization in the long run create a win-win result forboth employment and wage growth.

Local governments face a dilemma in deciding who should bearthe responsibility for state employees. Some governments do not have

–4 0 4–2 2 6 8 10 12

Publicoffering



Sale

Lease

Other


FIGURE 4.12Average Wage Growth in Different Forms of GAIZHI,

1995–2001

Percent

the resources to compensate all SOE employees and so offer discountson state assets to potential buyers of state firms. In places where themunicipal governments have taken little responsibility for state em-ployees, it is likely that the process of gaizhi will meet resistance. Astrong policy implication is that governments need to encourage therapid development of the private economy to lessen the overall burdenof gaizhi.


112

113

5Impact of Gaizhi on

Corporate Governance

Market reforms have resulted in a significant degree of insider control,as SOE managers have gradually acquired considerable discretion overthe use of state assets. Expanding managerial autonomy was a neces-sary complement to market reforms, as managers needed the flexibilityto respond to market signals. It has exacerbated agency problems, how-ever, by enlarging the discrepancy between control and the ownershipstructure. The agency costs of this increased autonomy have manifestedthemselves in various incentives for managers to maintain or acquireprivate benefits of control through on-the-job consumption and otherrents related to investment and expansion. Although much of the reforminitiatives with respect to SOEs can be understood as continuous effortsto alleviate these agency costs (see Tenev, Zhang, and Brefort 2002,20–24), the main issue that gaizhi—and privatization in particular—stillhas to solve is how to control the agency costs of managerial autonomy;in other words, how to improve the incentive structure and strengthenthe accountability of managers.

There are two main approaches to resolving agency costs: (1) through monitoring, and (2) by aligning managerial incentives withthose of the owners, including by making managers owners. Gaizhiincludes both approaches. Gaizhi repartitions rights and obligationsregarding production and the distribution of its results among inputowners. As such, it affects incentives and how the firm is governed. Newowners are often introduced with different incentives and capacity tomonitor managers. New institutions are established for the exercise ofcorporate control. Changes occur in the allocation of decision-makingpowers among various stakeholders, which affects the discretion thatmanagers have to make decisions. Managers and employees becomeowners.

This chapter examines the impact of gaizhi on corporate gover-nance. Its focus is on the effectiveness of gaizhi in providing solutions

to the problem of agency costs of managerial autonomy. It asks sev-eral questions. What changes in ownership are occurring as a result ofgaizhi? What is the relationship between ownership and actual controlat the enterprise level? How does gaizhi affect the allocation of decision-making powers within the firm? In particular, has gaizhi had anyimpact on the role of the enterprise party committee? Are the new in-stitutions for corporate control functioning? How does gaizhi modifythe relationship between the firm and the state?

Changes in Ownership Structure

In theory, all Chinese SOEs are owned by the central government, whichmanages the state’s assets on behalf of the citizens. In reality, SOEs areunder the control of different levels of government. In the mid-1990s,in response to falling SOE profits, higher-level governments started totransfer SOEs to lower-level governments in order to avoid the grow-ing fiscal burdens (Yao and Yang 2003).

The 11-city survey covered only SOEs owned by municipal andcounty/district governments.1Therefore, there was little variation inshare distribution among the various levels of government. Gaizhi, inits more radical forms, brings new types of owners.

There has been a steady decline in government shares in privatiz-ing firms (see figure 5.1). Government shares as a percentage of totalshares declined from 20 percent in 1995 to just above 10 percent in2001, and the shares of other SOEs declined from nearly 50 percent toless than 30 percent. Therefore, both direct and indirect (through SOEs)state ownership has declined. What is surprising is the dominant formof indirect state ownership in the mid-1990s. This indicates that by themid-1990s SOEs were playing a key role in spawning and restructur-ing other SOEs. Qian and Stiglitz (1994) and Qian (1995) describe onemechanism for such a role: a process of organizational transformationconsisting of several rounds of joint ventures established by SOEs andtheir subsidiaries with other government and SOE partners. The processof organizational transformation results in several degrees of separa-tion from the initial supervising authority, and therefore in an expandedmanagerial autonomy. In addition to increasing their independencefrom government, SOE managers were also motivated by a desire to


114

1. The analysis in this section focuses only on those firms with some owner-ship changes—thereafter called privatizing firms.

shift bad debts and overemployment burdens on to parent companies,and to undertake new business opportunities without losing existingconnections to and benefits from the state.

Often the state itself was playing the role of matchmaker betweenprofitable and struggling SOEs. According to some estimates,2 about50 percent of the mergers and acquisitions involving SOEs were “mar-riages” arranged by the government. Profitable SOEs have been less keento take over unprofitable state firms in recent years. The process ofM&As has become more market-based in recent years. We are seeingsome strong SOEs participating in industry consolidation across provin-cial boundaries. Many of these strong SOEs are listed companies thatare funding some of their acquisitions through primary and secondarystock offerings. A number of A-share companies have been acquiringnonlisted companies. Tsingtao Brewery acquired 2 factories in 1997, 5 in1998, and 15 in 1999, all near bankruptcy. Shanghai Jinling, a companythat produces and distributes electronic and communications products,has adopted the growth strategy of acquiring troubled SOEs in the sectorand revitalizing them with improved incentive schemes for management.Some listed mining companies—for example, Yanzhou Coal Mining—have also acquired, or are planning to acquire, existing mines.

IMPACT OF GAIZHI ON CORPORATE GOVERNANCE

115

2. Private communications with officials from various government agencies.

0

20

40

60

80

1995 1996 1997 1998 1999 2000 2001

GovernmentOther SOEsState total

Percent

FIGURE 5.1Shares of Ownership in GAIZHI Firms Held by

Government and State-Owned Enterprises, 1995–2001


With the emergence of nonstate players in the economy, the above-described process of transformation has become a process of hybridiza-tion involving the combination of state and nonstate investments in newventures. This hybridization can lead to very rapid changes in ownershippatterns. Suppose that a parent SOE has a majority control in two SOEswith mixed ownership. Privatization of the parent automatically trans-forms the two SOEs into private companies. Thus, as outside SOEinvestors privatize themselves over time, the process of privatizationaccelerates.

The increase in private shares has been strong in recent years.Among the various types of private owners, insiders have been the mostactive. Figure 5.2 shows the trends in private shares owned by variousinsiders within privatizing firms. In 1995 insiders held 5 percent of thetotal shares in privatizing firms, and most were owned by employees.Since then, the shares owned by middle and top management have in-creased rapidly. Of the 32 percent of total shares held by individuals in2001, top management held 13 percent, just under the total employeeshares of 18 percent. In a more limited study on gaizhi, Tenev, Zhang,and Brefort (2002, 34) find a similar percentage of senior managementownership. They find that in the transformed enterprises in Jianhua,Zhejiang province, natural persons held 76 percent of the shares, of whichsenior managers accounted for 20–30 percent.


116

0

5

10

15

20

25

30

35

1995 1996 1997 1998 1999 2000 2001

Middle managementTop managementEmployeesInsider total

Percent


FIGURE 5.2Percentage of Shares Held by Managers and Employees

in GAIZHI Firms, 1995–2001

Sixty percent of the managers in gaizhi firms held shares in thefirm. This percentage is higher than was found by Tenev and Zhang(2002). Privately controlled firms were more likely than government-controlled firms to give shares to the manager (82 percent, comparedwith 59 percent). On average, the manager owned 20 percent of theshares in 169 firms with valid answers. There were wide geographicdisparities, however (see figure 5.3). Generally, managers in the southerncities owned more shares than those in the northern and western cities.Share ownership was particularly low in Xining, Lanzhou, Harbin,Fushun, and Guiyang. Chengdu granted managers the highest percent-age of shares, on average 31.7 percent of the firm’s total.

Domestic private firms have increased their ownership of shares inprivatizing firms from 11 percent in 1995 to around 16 percent in 2001(see figure 5.4), a much smaller rise than in the percentage of sharesowned by insiders. Foreign investment as a percentage of total equity hasdeclined. Foreign companies held 10 percent of total shares in privatizedcompanies in 1995, but only 3.2 percent in 2001. The decline in foreign


117

Fushun

Guiyang

Lanzhou

Harbin

Xining

Weifang

Zhenjiang

Tangshan

Hengyang

Huangshi

Chengdu

0 5 10 15 20 25 30 35


Percent

FIGURE 5.3Percentage of Shares Owned by Managers in

GAIZHI Firms, by City, 2002

ownership amid the rapid growth of FDI to China can be explained inpart by the growing preference of foreign investors for wholly foreign-owned enterprises as opposed to equity joint ventures. As figure 5.5 in-dicates, there is a very pronounced trend toward 100 percent foreignownership at the expense of joint ventures. For instance, equity jointventures accounted for 53 percent of all cumulative FDI by year 2002.In that year, however, equity joint ventures accounted for 30 percent ofall projects involving foreign investment. The percentage shares forwholly foreign-owned enterprises were 34 and 65 percent, respectively.

Another reason for the low and declining trend of foreign own-ership is the negligent share of M&As in Chinese FDI. According tostatistics of the United Nations Conference on Trade and Development,cross-nation M&A in China is less than $2 billion a year, accountingfor only 5 percent of the country’s total FDI. Globally, M&As have beenthe key driver of global FDI flows since the late 1980s, accounting forabout 80 percent of FDI.3

It is the government’s policy to encourage FDI in the restructuringof SOEs. By selling state assets to foreign companies or inviting for-eign equity participation, SOEs can establish a modern enterprise sys-


118

3. See “Ten Major Trends in China’s Utilization of Foreign Investment,” BusinessAlert China, 6, June 1, 2003, http://www.tdctrade.com/alert/cba-e0306h.htm.

0

5

10

15

20

1995 1996 1997 1998 1999 2000 2001

Foreign companiesDomestic private companies

Percent


FIGURE 5.4Percentage of Shares of GAIZHI Firms Owned by ForeignCompanies and Domestic Private Companies, 1995–2001

tem, improve corporate governance, achieve sustainable development,and increase their competitiveness. In November 2002 the governmentissued the Provisional Regulations on the Use of Foreign Capital forRestructuring State-Owned Enterprises. Along with the Circular onIssues Concerning the Transfer of State-Owned Shares and Legal PersonShares of Listed Companies to Foreign Investors, promulgated onNovember 1 by the SETC, the MOF, and the CSRC, it forms part ofChina’s policy on the use of foreign capital for restructuring SOEs. Theenactment of this new legislation will further improve the legal environ-ment for M&As of mainland enterprises by foreign businesses. As aresult of China’s efforts to improve laws and regulations related to cross-nation M&As, such approaches as equity injection, buy-out, shareswapping, and cross-shareholding will become important modes of in-vesting in China by multinational companies. The trend is expected tobe more prevalent, as more SOEs are undergoing asset reorganizationand being put up for sale. Foreign investors will be able to select theirtarget enterprises and to acquire controlling stakes through these M&Aapproaches.


119

20

40

60

0

Percent

FIGURE 5.5Types of Foreign Direct Investment in

2002 and Cumulative to 2002(percent)

Note: There is also a very small number of FDI projects that are classified as joint exploitation. They were omitted from the calculations. Source: Ministry of Commerce, PRC.

Equityjoint venture

Contractualjoint venture

Wholly ownedforeign enterprise

Year 2002Cumulative to 2002

Overall, the percentage of shares held by domestic and foreign pri-vate companies has declined, while the percentage of shares held byindividual insiders has risen dramatically. The decline in corporate share-holding has caused a decline in share concentration in the privatizedfirms. Figure 5.6 shows the shares of the three largest shareholders inthe period 1995–2001. The share of the largest shareholder declined byabout 20 percentage points from a high of 80 percent, whereas the shareof the second- and third-largest shareholders remained in the range of15–20 percent.

The largest shareholder maintained a controlling position, withshare ownership close to 60 percent in 2001. The ownership concen-tration of privatizing firms is quite high and seems to be higher thanthe ownership concentration of listed companies. In the case of listedcompanies, Tenev and Zhang (2002, 78–79) find, for example, thatthe average shareholding of the largest shareholder is 47 percent.

Thus we observe a tendency toward ownership deconcentrationamong privatizing firms, as a result of reduced state shares and in-creasing individual shares, held mainly by insiders. At the same time,observers (Tenev and Zhang 2002) have noticed the tendency ofcompany managers and outside investors to concentrate ownershipat the expense of employees’ shares. Box 5.1 presents a case in which


120

0

30

60

90

1995 1996 1997 1998 1999 2000 2001

Largest shareholder Second and third largest shareholders

Percent


FIGURE 5.6Percentage of Shares Owned in GAIZHI Firms by

Three Largest Shareholders, 1995–2001

the attempt by an outside investor to buy the employees’ shares ledto conflict that has paralyzed the firm. There is no clear trend withrespect to outside shares.

Ownership and Control

Gaizhi typically involves the establishment of new governance struc-tures. The shareholder conference, the board of directors, and theboard of supervisors are introduced as new institutions for the exer-cise of ownership rights. Some wholly state-owned firms may also


121

BOX 5.1A STALEMATE IN CHENGDU

The first privatized SOE in Chengdu was a company that producedconstruction materials. In 1996 the company was sold to the em-ployees and one outside private investor. The private investor bought51 percent of the shares. There was no clear rule regarding the dis-tribution of shares among employees, so the remaining 49 percentwas grouped into a block share owned by all the employees.

Responding to a call from the municipal government to relocateindustry away from the downtown and surrounding areas, the com-pany moved to a suburban industrial park and converted some oldfactory buildings, which were let out to tenants. The location wassuitable and the company began to make money.

The presence of the block of employee shares was a potentialthreat to the private owner, who tried to further privatize the blockshares on the basis of the original values approved by the municipalgovernment in 1996. Because the company’s value had increased,the employees wanted more for their shares and the plan failed. Theprivate owner and the management persuaded the employee repre-sentatives to agree that the company could buy back the shares atthe original price.

Many employees felt cheated and drove away the management.They demanded rents from the tenant businesses, but many busi-nesses refused to pay. By 2002 the occupation had been ongoing fortwo years, and the municipal government had been unable to medi-ate between the management and the employees. In addition, themanagement demanded that the squatting employees pay back morethan RMB10 million in losses caused by their occupation, a demandthat the employees firmly rejected.

establish the new forms of governance if they decide to incorporatethemselves.

One-third of sample firms reported having a shareholder repre-sentative conference. About 60 percent of gaizhi firms had set up aconference. Forty-four percent of the firms reported having a board ofdirectors. This figure was higher than the percentage of firms that hada shareholder conference, because some SOEs—such as incorporatedSOEs—have boards of directors but not shareholder conferences. Onaverage, the board had been set up for four years. Seventy-eight per-cent of gaizhi firms have established boards of directors. Privately con-trolled firms were more likely than government-controlled firms tohave established a board: 92 percent compared with 88 percent.Among the wholly government-controlled firms, only 21 percent hada board of directors.

Most of the firms (75 percent) elected the directors through ashareholder conference. Firms used different methods: 39 percent setthe number of board seats in proportion to the number of sharehold-ers; 36 percent set them in proportion to the number of shares held.A few firms gave the right to choose board members to either theboard of directors or the management (see figure 5.7). There was nosignificant difference between privatizing and fully state-owned firmsin this regard.

Unlike in the shareholder conference, where the systems of oneshare-one vote and one person–one vote were similarly popular, theboard of directors tended to rely on the latter system. Sixty-eight percentof the firms adopted the one person–one vote principle in decision-making by the board. Only 12 percent of the firms adopted the oneshare–one vote system. A total of 21 percent of privately controlled firmshad instituted a one share–one vote system at their boards, while lessthan 9 percent of state-controlled firms had done so.

In the sample, about half of the firms with a board had decidedthat the manager or the Party secretary would be board chairman(see figure 5.8). This is consistent with the study by Tenev and Zhang(2002). In 30 percent of firms the chairman was from the largestshareholder. There was no significant difference between privatizingand fully state-controlled firms. Managers were in strong positions.Not only was the firm’s manager often the board chairman, but, onaverage, the top management took 53 percent of the seats on theboard and the middle management took another 17 percent (seetable 5.1).


122

The following index describes the relationship between the com-position of the board and the share structure of the firm (Xu, Zhu, andLin 2002):

rsi is the ratio of shares held by the ith party, rbi is the ratio of board seatsheld by the ith party, and N is the number of parties. Seven parties areconsidered: the government, top management, middle management,workers, outside SOEs, outside private firms and foreign firms, andothers. There were 207 firms that provided valid data. A low indexnumber indicates a better match between the composition of the boardand the share structure of the firm. The average for privatizing firmswas 0.127, and the average for fully state-owned firms was 0.155. Thedifference was significant at the 5 percent level. The index also differedaccording to the dominance of government shares. It was 0.129 for

IN

r rsi bii= −( )∑1 2


123


FIGURE 5.7Method of Electing Board Members in China, 2002

By shareholderconference inproportion to

number ofshareholders

39%

By shareholderconference inproportion to

shares held36%

By management3%

By board ofdirectors

4%

Other18%

firms with a dominant government share, and only 0.096 for privatelycontrolled firms.4 In addition, parties in outsider-controlled firms wereslightly better represented than insider-controlled firms, with indices of0.121 and 0.138, respectively. There were considerable differencesamong the cities. Harbin and Lanzhou had the highest degree of dis-crepancy between ownership and control structures, with indices of0.172 and 0.167, respectively. Guiyang and Fushun were the next high-est, with indices of 0.140 and 0.132, respectively. Huangshi, Zhenjiang,Tangshan, and Hengyang were in the middle, with indices in the rangeof 0.128 to 0.121. Weifang, Chengdu, and Xining had the best matchbetween ownership and control structures, with indices of 0.119,0.113, and 0.109, respectively.


124

4. In terms of benchmark, an index of 0 indicates a perfect match between owner-ship and control structures; a firm with a representative ownership structure, butwhere the largest shareholder controls all the board seats, will have an index of0.133.


FIGURE 5.8Method of Choosing a Board Chairman in China, 2002

By manageror partysecretary

49%

By thegovernment

6%

By the largestshareholder

30%

Other15%

Table 5.1 shows that gaizhi firms and firms with greater privatecontrol had a better match between shareholding power and repre-sentation on the board of directors. Boards of directors are becomingmore representative in Chinese firms. Gaizhi has been a catalyst in thisprocess. The variations among the cities largely matched their progressin gaizhi.

Have the new organs of corporate governance—the shareholderconference and the board of directors—begun to play a meaningfulrole in gaizhi firms? Figures 5.9 and 5.10 report data on the percent-age of gaizhi firms that find the role of shareholder conferences andboards of directors, respectively, to be important or very important. Asubstantial number of gaizhi firms regarded the shareholder conferenceas important, especially in decisions on investment and profit distri-bution (see figure 5.9). Even then, though, the percentage of gaizhifirms that regarded the shareholder conference as important was onlyaround 30 percent.

The decision to call a shareholder conference was mostly madeby the board of directors. Fifty-nine percent of firms said that it was acollective decision of a number of directors, while only 14 percent offirms said that a certain minimum number of shareholders could call aconference. A total of 13 percent of firms allowed the managers to call ashareholder conference, and 2 percent allowed the labor union to do so.These practices are not fully consistent with the principles of corporategovernance found in developed market economies. The right of the man-agers to call a conference might be because, in many firms, management


125

TABLE 5.1ALLOCATION OF OWNERSHIP AND CONTROL OF GAIZHI FIRMS, 2002

(percent)

Shareholding Board Seats

Government 12 3SOE 28 9Private firms 17 6Workers 18 7Top management 14 53Middle management 4 17Others 7 5

SOURCES: Survey data; authors’ calculations.

comprises the board of directors. Unions are likely to have greater powerin employee shareholding firms.

The role of the shareholder conference was outweighed by that ofthe board of directors in every aspect of decision-making. Boards ofdirectors were particularly important in decisions on investment andprofit distribution, the two areas in which the shareholder conference hadbeen the most involved (see figure 5.10). Most importantly, the board ofdirectors was the most influential body for managerial appointmentin the gaizhi firms. It did not have a heavy presence in production andmarketing decisions, and only a medium input in employment and wagedetermination decisions. Managers were more visible in these decisions.The board of directors was the most influential governance body relativeto all, including old and new structures, in the gaizhi firms.

Currently in China, companies have two boards: a board of super-visors and a board of directors. There is confusion as to the respectiveroles of the two boards. For example, according to CSRC regulationsand in practice, the audit function is run by the board of directors. Atthe same time, the Company Law gives the board of supervisors the


126

0 5 10 15 20 25 30 35

Employment

Wages

Investment

Managerialappointment

Disposal ofprofits

Production andmarketing


Percent

FIGURE 5.9Percentage of GAIZHI Firms That Find the Role

of Shareholder Conferences Important orVery Important in Six Management Areas, 2002

right to inspect the company’s financial situation. Furthermore, it is notclear who the board of supervisors is representing and to whom it is ac-countable. The Company Law specifies that the board of supervisors iscomposed of representatives of the company’s staff and workers, andof the shareholders, but it does not specify to whom the board of super-visors is accountable. The CSRC’s Code of Corporate Governance makesthe board of supervisors accountable to all the shareholders. Therefore,the existing structure creates confusion as to the exact responsibilitiesand accountability of the two boards. This tends to diminish the role ofthe board of directors.

Gaizhi and Traditional Stakeholders

In China, new market institutions have been introduced without firstdismantling the old practices. The coexistence of new and old institu-tions has been a distinctive feature of China’s process of institutionaltransformation. This is particularly pronounced in the area of corpo-rate governance, where old traditional structures such as labor unions


127

0 10 20 30 40 50 60 8070

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.10Percentage of GAIZHI Firms That Find the Role

of Boards of Directors Important orVery Important in Six Management Areas, 2002

and Party committees exist side by side with new institutions such asboards of directors and shareholder meetings.

Gaizhi and the Influence of Stakeholders. Although gaizhi may not af-fect the absolute power of managers, it may have an impact on their rel-ative influence vis-à-vis other stakeholders such as the government, thelabor unions, and the Party committee. How does gaizhi affect the al-location of decision-making powers among these governance bodies?This section reviews the importance of the government, the Party, andthe labor union in key decisions about employment, wages, investment,managerial appointments, the distribution of profits, and productionand marketing.

Privatization transfers ownership from the government to privatehands, but it does not necessarily prevent the government from inter-fering in firm decisions. Figure 5.11 shows that the government wasregarded as more influential in non-gaizhi firms than in gaizhi firms.The government is perceived to be most influential with respect tomanagerial appointment by both gaizhi and non-gaizhi firms. The gov-


128

0 10 20 30 40 50 60 70 80

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.11Percentage of Firms That Find the Role ofGovernment Very Important or Decisive in

Six Management Areas, 2002


ernment also had an important role in the firm’s investment decisions,and played a significant part in decisions about employment and wagerates. In both gaizhi and non-gaizhi firms the government did not in-terfere unduly with the firm’s production and marketing decisions.This shows that the government has generally retreated from the dailyoperations of the firm. The biggest difference between gaizhi and non-gaizhi firms is in the direct influence of the government over manage-rial appointment. We see a significant decline in this influence in gaizhifirms. This is perhaps the most significant impact of gaizhi on the powerstructure within the firm.

The labor union was quite active in decision-making in non-gaizhifirms, especially on employment and wage decisions (see figure 5.12).In gaizhi firms the role of the union had declined, but it maintainedconsiderable influence over employment and wage determination. Thisreflects the high degree of employee ownership in many gaizhi firms.Nonetheless, the diminished role of the union in gaizhi firms indicatesthat there has been a move toward the dominant international prac-tice of involvement only in wage and employment issues.


129

0 5 10 15 20 25 30 35

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.12Percentage of Firms That Find the Role ofLabor Unions Very Important or Decisive in

Six Management Areas, 2002


Gaizhi and the Role of the Party. Of particular interest is the questionof the role of the Communist Party, particularly in managerial appoint-ments. Influential observers (Qian 2000; Qian and Wu 2003) haveargued that reform of large SOEs has been a failure in China mainlybecause of the continued role of the Communist Party in the appoint-ment of managers.

Observers (Qian 2000; Chan and Wong 2003) have noted thedual impact of the Party on firm performance and incentives. On onehand, the Communist Party’s firm control over executive appoint-ments is believed to have several negative consequences. First, the in-fluence of Party committees on personnel appointments leads toskewed management incentives. Managers tend to spend considerableeffort presenting the right political image and nurturing good rela-tions with superiors in the Party hierarchy. Second, the system of in-ternal checks and balances envisaged by the Company Law cannotdevelop. In particular, the board of directors cannot perform one of itskey functions: to freely select, monitor, and fire the executive managersof the company. Third, the monopolization of executive appointmentsby the Party apparatus sets back the development of the managerialmarket. Managers can be transferred (promoted or demoted) to otherSOEs suddenly and based on political criteria, without taking intoaccount the interests of the company and its shareholders. The end re-sult is that Party management curtails the effective monitoring ofmanagerial behavior, distorts management incentives, and creates a ten-dency toward insider control.

On the other hand, observers have argued that the CommunistParty can help alleviate agency costs. Chan and Wong (2003), for example, find that the decision-making power of the local Party com-mittee is positive for controlling the agency costs of the largest share-holder, and negative in terms of its impact on firm performance, forcontrolling the managers’ agency problems. On net, they find that theexisting level of Party control is excessive and that reducing the decision-making power of local Party committees tends to improve the perfor-mance of China’s listed firms.

The general perception is that Party committees are very influen-tial. For instance, Freedom House, a nonprofit organization founded inthe 1940s by Eleanor Roosevelt, states in its annual report on politicalfreedoms, “All government offices, public schools, and state firms stillhave party committees that handle budgets, political education, and


130

personnel decisions.”5 All Chinese SOEs contain branches of theChinese Communist Party. Gaizhi firms have maintained this arrange-ment, and many private firms have also established Party branches.6

The key indicator of the relative importance of the Communist Partyin making decisions is the percentage of firms that regarded the Partyas “very important” or “decisive.” In the sample the Party branch wasfound to be more active in the non-gaizhi firms relative to the gaizhifirms in terms of all firm decisions, except in appointing the manager (seefigure 5.13). The Party, however, is not perceived to have an important


131

5. Freedom House, 2003, http://freedomhouse.org/research/freeworld/2003/countryratings/china.htm.6. According to Chinese media, only 14 percent of the private firms have establishedParty branches (People’s Daily 2001). With the implementation of the “new theoryof the three represents” the percentage has probably increased. According to TheEconomist (2002), only 17 percent of private firms employed Party members in 1999and just 3 percent had any kind of Party organization. Only 35 percent of the foreign-funded enterprises employed Party members that year, and a mere 17 percent hadParty cells.

0 5 10 15 20 25 30 35

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.13Percentage of GAIZHI and Non-GAIZHI Firms That Findthe Role of the Communist Party Very Important or

Decisive in Six Management Areas, 2002


or decisive direct influence on managerial appointments in either gaizhior non-gaizhi firms. Only in 18 percent of gaizhi and 13 percent of non-gaizhi firms does the branch of the Communist Party play an importantrole in managerial appointment. In non-gaizhi firms, managerial ap-pointment is the type of decision where the role of the Communist Partyis the least important. Therefore, while the gaizhi process does notdiminish Party control over managerial appointment, Party control isalready low prior to the transformation process. In fact, gaizhi firms re-port a greater role of the Party in managerial appointment than do non-gaizhi firms. The situation is different, however, in outsider-controlledgaizhi firms. In those firms the role of the Party in managerial appoint-ment is lower than in gaizhi firms (see Figure 5.14). This can be inter-preted as evidence that the Party acts as a countervailing force tomanagerial power in the case of insider-controlled firms.

The Overlap between New and Traditional Institutions. An importantissue in gaizhi firms is how to divide the firm’s functions between thenew institutions representing the shareholders and the traditional or-


132

0 5 10 15 20 25 30

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.14Percentage of GAIZHI and Non-GAIZHI Firms That Find

the Role of the Communist Party in Insider- andOutsider-Controlled Firms Very Important or Decisive

in Six Management Areas, 2002

Insider-controlledOutsider-controlled

ganizations of social control, such as the workers’ congress, the Partycommittee, and the trade unions. The standard approach has been tocombine the leadership or functions of various institutions. The usualpractice is to combine shareholders’ and workers’ congress meetings,and to have the same person serve as chair of the board of directors andsecretary of the Party committee. Many enterprises hold joint meetingsof various representative bodies.

This overlap provides a sense of continuity, a sort of translationbetween the old institutional language and conceptual framework andthe new ones. It therefore facilitates adaptation to the new rules. Atthe same time, the coexistence of the old and new institutions is creat-ing confusion about the division of labor and the procedures to be fol-lowed in performing various functions.

For example, the combination of shareholders’ conference andworkers’ congress is creating confusion about the voting process, espe-cially in employee-owned firms. Both the one person–one vote and theone share–one vote systems of majority decision-making are commonin shareholders’ conferences. More than 80 percent of the firms thathave established shareholders’ conferences have adopted one of thesesystems. Forty-one percent of gaizhi firms with shareholders’ confer-ences voted on the basis of one share-one vote, and 42 percent followeda system of one person-one vote. As mentioned above, some firms wereallowing labor unions to call a shareholders’ conference.

The overlap between the old and the new corporate bodies is pro-viding a channel for political influence on corporate governance.Under the so-called Concurrent System, top enterprise leaders havetended to act both as Party secretaries and as chairs of the board of di-rectors. In our sample, almost half of the chairmen of the board arealso Party secretaries. As a result, the role of Party secretary is oftenfused with that of the chairman of the board of directors. To achievethe objectives of the Corporate Law and improve corporate gover-nance in China, the practices of the Concurrent System should be dis-couraged and the new bodies of corporate governance permitted tofunction independently.

Gaizhi, Managerial Autonomy, and Managerial Incentives

How is gaizhi affecting managerial autonomy with respect to key deci-sions about employment, wages, investment, managerial appointments,the distribution of profits, and production and marketing? Figure 5.15


133


134

shows the percentage of firms that regarded the manager as importantor decisive in the issues under consideration. The manager played themost important role in decisions related to employment and the dailyoperation of the firm. The manager also had a significant role in deci-sions on investment, but a lesser role in decisions on profit distribution.Only a small proportion of the firms believed that the manager was veryimportant in determining his or her own appointment.

There was no significant difference between gaizhi and non-gaizhifirms with respect to managerial influence. Management has acquiredsignificant discretion and power in non-gaizhi firms. Gaizhi does notseem to affect the influence of management, though it seems to havesomewhat limited the manager’s power over financial decisions—inparticular, with respect to the allocation of profits.

How is managerial influence and control exercised? One channelis through control over the collective bodies of corporate governancesuch as the board of directors. In our sample, in many firms, the powerof managers who were also board members is considerable. Sixty-onepercent of managers were also the chairman of the board. The figure

0 20 40 60 80 100

Employment

Wages

Investment


Disposal ofprofits



Percent

FIGURE 5.15Percentage of GAIZHI and Non-GAIZHI Firms That Findthe Role of the Manager Very Important or Decisive

in Six Management Areas, 2002


was high in gaizhi firms and privately controlled firms (69 percent), butlow in firms that were controlled by outside investors (32.3 percent).

China has a relatively high degree of separation between chiefexecutive officer and chairman of the board in its listed companies.According to the Tenev and Zhang (2002) study, in about three-quartersof the listed companies the CEO and chairman are separate people.Debates continue in China, however, as to whether the positions ofCEO and chairman should be combined or separated.

It is important to note that, despite the relatively high degree ofseparation between CEO and chairman, power is frequently concen-trated in Chinese companies in the legal representative, the so-called keyperson, who used to represent the state and is the chairman of theboard. As a practice of company regulations in China, every enterprisemust appoint a legal representative. By definition, the legal representa-tive represents the enterprise in question in every matter, large andsmall. He or she can then delegate such powers to another person un-less such powers may not be delegated under the law or the articles ofassociation. This can lead to situations where the chairman takes onroles and functions that should belong to the CEO. For improved cor-porate governance, this practice will have to change in favor of an al-location of powers that is in line with the respective roles of the CEOand the chairman of the board.

In the aftermath of recent corporate governance scandals, thereis an emerging consensus globally about the need to split the chair-manship of the board from the chief executive position in a company.It has become clear that if a CEO of a company is also the chairman,the board agenda is set by the management and all the informationavailable to the board is filtered by the management. It is thereforemore difficult for the board to obtain the information that allowsfor independent judgment of the company. Also, the two positions arefull-time jobs that require different skills and experience. In theUnited Kingdom, in more than 90 percent of companies the chairmanof the board and the CEO of the company are separate people. In theUnited States the debate is ongoing, but an influential commissionchaired by Treasury Secretary John Snow has recommended splittingthe CEO–chairman job.

Given the concentration of power in management, what mecha-nisms exist to align the incentives of managers with those of owners andthe interests of the companies? As discussed above, managers are be-coming significant owners themselves. Ownership is just one compo-


135

nent, however, of the overall compensation and incentive structure ofChinese managers.

The shape and form of incentives are important for elicitinggood performance from managers. Overall, 24 percent of sample firmslinked the manager’s performance to the amount of shares held, and25.2 percent linked it to the level of wages. Gaizhi firms were morelikely to provide bonuses to their managers than non-gaizhi firms:39 percent of gaizhi firms, as compared with 14 percent of non-gaizhifirms.

Accordingly, 43 percent of privately controlled firms gave bonusesto their managers, whereas the percentage was 39 percent amonggovernment-controlled firms and only 17 percent among fully government-owned firms. In addition, outsider-controlled firms weremore likely than insider-controlled firms to offer such incentives (41 per-cent, compared with 21 percent). Together with Guiyang, Lanzhou paidthe lowest share of bonuses at only 23 percent of the manager’s totalincome (see figure 5.16). Managers in the other cities relied heavily on


136

Guiyang

Lanzhou

Zhenjiang

Harbin

Huangshi

Hengyang

Xining

Tangshan

Chengdu

Weifang

0 10 20 30 40 50 60 70 80


Percent

FIGURE 5.16Percentage of Bonuses in Managers’ Total Income

in Ten Sample Cities, 2002

such incentives. In Weifang, bonuses made up more than 70 percent oftotal wages on average, with only a small portion of the manager’s in-come coming from a fixed salary.

The firms in the survey were much more willing to provide bonusesto their managers than were the publicly listed firms covered in Tenevand Zhang’s study, which found that bonuses made up only 24 per-cent of managers’ incomes on average (Tenev and Zhang 2002, 92).Whether this is a contrast between publicly listed firms and unlistedfirms is unknown and remains a topic for further research.

There is lively debate in China about the reform of the compen-sation system of company managers. The growing disparities be-tween the level of remuneration of top managers in some largeSOEs and ordinary workers’ wages are also attracting considerableattention.7 The current compensation systems of company managersin China suffer from serious weaknesses. First, significant discrep-ancies exist between the salary levels in different types of companies,indicating an underdeveloped managerial market. Second, the im-plicit benchmarking of management salaries to the compensationlevels in the civil service tends to persist. Finally, there are some legaland regulatory barriers to the introduction of modern compensationmechanisms. For example, the Company Law does not provide anybasis for issuing share options; companies cannot reserve unissuedshares and grant vested rights to acquire such shares in the future.Some companies, however, are able to structure innovative contractsthat link management’s compensation with the firm’s performance(see box 5.2). In this area, China, as a latecomer, can benefit from theextensive debates and analysis done post-Enron regarding manage-ment compensation and stock options and create a framework forcompensation policies that allows for better links of compensationto performance. It is essential, however, that actual compensationpolicies are set by the board of directors and not by management orgovernment agencies.


137


7. According to media reports, the average annual salary of senior managers ofSOEs under central SASAC was RMB350,000 in 2003, or 13.5 times that of allemployees, which stood at an average of RMB24,000. Central SASAC PartySecretary Li Yizhong was reported as saying that any gap exceeding 12 times isnot acceptable.

Gaizhi and Changes in the Relative Influence of Stakeholders

Gaizhi thus leads to significant changes in the relative influence of tra-ditional stakeholders. Table 5.2 summarizes these changes by compar-ing gaizhi and non-gaizhi firms. It shows the rank of influence of eachstakeholder in each area of decision-making in the gaizhi firms, togetherwith the change in rank relative to the non-gaizhi firms. The roles of thegovernment and the Party in all aspects of firm decisions were consid-erably weakened after gaizhi. It is important to note, however, that thegovernment retains significant influence in managerial appointmentswithin gaizhi firms. While the decision-making power of the laborunion diminished in most firm decisions, the union actually increasedits role in employment and wage decisions. Overall, gaizhi seems tobe neutral with respect to the decision-making powers of management.We do see, however, changes in some decisions. There is a somewhatreduced role for management in investment and profit distributiondecisions.

The results on the relative influence of the board of directors andthe shareholders’ conference are encouraging. It seems that these newbodies are becoming functional. The board of directors was the mostinfluential body in exactly the areas where its power is expected to


138

BOX 5.2THE MANAGEMENT INCENTIVE SCHEME OF TCL GROUP

The successful IPO of TCL Group, a Guangdong-based SOE groupspecializing in the manufacture of TV sets and electronics, attractedmedia attention when its senior managers turned themselves into mil-lionaires overnight. Each of the top four managers owns shares in thenewly listed group that were worth over RMB100 million at the timeof the IPO. The group’s chairman, Li Dongsheng, has a stake of9.1 percent in the company, which is worth RMB616 million. TCLmanagement’s innovation was to acquire ownership stakes in theircompany not by MBOs or other forms of privatization, but by meansof a contract with the municipal government of Huizhou, which rep-resents the state in the company. The contract enabled managementto claim bonuses for themselves when the value of the ownershipstake of the state appreciated, and to convert those bonuses intoshares in the company. Before the IPO, the management held 25 per-cent of the company, with another 17 percent held by employees.


139


be the highest: managerial appointment, large investments, and dis-tribution of profits. There is a significant increase in the relevance ofthe shareholders’ meetings in gaizhi relative to non-gaizhi firms (seefigure 5.17).

It is important to note, however, that the power of the board ofdirectors is a derived power. The board of directors is an arena where var-ious stakeholders are represented and are struggling for control. As indi-cated in table 5.1, managers have the strongest type of control over theboard of directors, where they are significantly over-represented andcontrol more than half of all seats. Thus we have a situation in thegaizhi firms whereby the board of directors is influential in managerialappointment, but managers control the board. The government is alsoinfluential in managerial appointment but is not represented on theboard. Thus when we take into account the indirect influence throughthe board of directors, it seems that gaizhi has further strengthened thepower of enterprise managers.

In the next section we consider which players and institutions areemerging to control the agency costs of managerial autonomy.

The Role of Outside Investors in Corporate Governance

Survey results suggest that the presence of strong outside investors isassociated with a reduced role for traditional stakeholders. For in-stance, the government, the Party, and the labor union had less signif-icant roles in outsider-controlled firms than in insider-controlled firms.The most significant difference in influence was in the role of the Party(see figure 5.14). Many outsider-controlled firms were owned by pri-vate firms and businesspeople who were not attached to the old struc-tures. The labor union still had an important role, however, in wagedetermination in these firms: 29.4 percent of insider-controlled firmsand 28.8 percent of outsider-controlled firms regarded the union as im-portant in wage determination.

At the same time, firms controlled by outside investors were morelikely to rely on the new institutions of corporate governance. The share-holder conference was more important in outsider-controlled firms thanin insider-controlled firms. In decisions related to investment, 22 percentof outsider-controlled firms believed that the shareholder conferencewas important; in decisions related to profit distribution, 28 percent ofoutsider-controlled firms believed it to be important. The percentages for


140

TABLE 5.2INFLUENCE ON DECISION-MAKING OF VARIOUS STAKEHOLDERS IN

GAIZHI VERSUS NON-GAIZHI FIRMS, 2002

Government Management Labor

Non- Non- Non-Decision Gaizhi gaizhi Changea Gaizhi gaizhi Changea Gaizhi gaizhi Changea

Employment 5 3 −2 1 1 0 3 4 1Wages 4 2 −2 1 1 0 3 4 1Investment 4 2 −2 2 1 −1 5 3 −2Managerial 2 1 −1 3 5 2 6 3 −3

appoint-ments

Distribution 5 4 −1 2 1 −1 4 3 −1of profits

Production 6 4 −2 1 1 0 4 2 −2andmarketing

Average (net 4.3 (−10) 1.7 (0) 4.2 (−6)change)

Note: The highest rank is 1, and the lowest rank is 6.a. Gaizhi relative to non-gaizhi firms.

–20 –15 –10 –5 0 5 10 15 20 25

Government

Management

Labor

Party

BOD

Sh. Conf.

Cumulative changes in influence ranking

Sources: Survey data; authors’ calculations.

FIGURE 5.17Changes in the Relative Influence of

Various Stakeholders in GAIZHI versus Non-GAIZHI Firms, 2002

insider-controlled firms were 14 percent and 13 percent, respectively.Firms controlled by outside investors were also more likely to adopta share-based voting system than were firms controlled by insiders(50 percent versus 42 percent).

Controlling outside investors are also more likely to restrict man-agerial powers, particularly in areas such as financial management, andto provide high-powered incentives to managers. See box 5.3 for anexample of the role that some outside private investors have playedin the corporate governance of acquired state firms. While outsider- andinsider-controlled firms showed no significant differences in the otheraspects of decision-making, outsider-controlled firms were more cautiousthan insider-controlled firms in granting the manager decision-makingpower over investment and profit distribution. Of the insider-controlledfirms, 38 percent believed that the manager was decisive in investmentdecisions and 37 percent said the same for profit distribution decisions.The figures for outsider-controlled firms were 21 percent and 16 percent,respectively.

Firms that were controlled by outside investors had relatively newmanagers. In those firms, 53 percent of managers had previously beenemployees, but the share was 70 percent for other firms. The managershad been in their posts for an average of 4.3 years, which was a shorter


141

Party Board of directors Shareholders’ conference

Non- Non- Non-Gaizhi Gaizhi Changea Gaizhi Gaizhi Changea Gaizhi Gaizhi Changea

4 2 −2 2 5 3 6 6 06 3 −3 2 5 3 5 6 16 4 −2 1 5 4 3 0 33 2 −1 1 4 3 3 0 3

6 2 −4 1 5 4 3 0 3

5 2 −3 2 5 3 3 0 3

5.0 (−15) 1.5 (20) 3.8 (13)

142

BOX 5.3THE SUCCESSFUL EXPERIENCE OF NEW HOPE GROUP

IN REFORMING SOES

The New Hope Group was founded in the early 1980s by fourbrothers of the Liu family. The group is mainly engaged in animalfeed, food processing, banking, real estate, chemicals, and dairybusinesses. Mr. Liu Yonghao, one of the brothers, is a high-profileentrepreneur whose leadership has propelled this family-ownedbusiness to one of the most prominent success stories in Chinese private enterprise.

While initially New Hope Group grew organically by reinvestingaccumulated earnings, recently acquisitions of SOEs have becomethe main pillar of its growth strategy. In 2000 New Hope, jointlywith the International Finance Corporation, the private-sector armof the World Bank Group, invested in Chengdu Huarong Chem-icals, a bankrupt SOE in the chemical sector. Although the newlyacquired state-owned company was bankrupt, it had advantagesover developing a plant from scratch, due to existing facilities andexperienced employees. After three years the company increasedsales from less than $10 million to more than $30 million, and hasbeen successful in maintaining employment of about 500 regularemployees.

The New Hope Group is also rapidly becoming one of the mostprominent players in China’s emerging dairy industry. Encouragedby the success in acquiring Chengdu Huarong, and enticed by theprospects of the dairy industry in China, the group moved into thesector with the acquisition of a state-owned dairy plant in Sichuanprovince in October 2001. Subsequently, New Hope acquired 12state-owned dairy plants in southwestern, eastern, northern, andnortheast provinces. Most of the acquired companies are leaderswith solid brands and loyal followings in their local markets. NewHope plans to double its production capacity by 2008.

The New Hope Group brings capital, committed management,and strict financial control to its investee companies. Investmentsin target companies are made through a holding company that islisted on the domestic stock exchange. Since the holding companyneeds to present consolidated financial statements, the acquiredcompanies are subject to the discipline of the stock market. Theyneed to provide frequent, detailed financial results to the parentcompany and receive frequent inspection visits from headquarters.Strict control is also evidenced by the fact that it is the policy ofthe group for the financial officers in acquired companies to be

(continues)



143

period than the 5.8-year average for other firms. Since the entry of out-side investors is a relatively new phenomenon, the shorter tenure oftheir managers suggests that these investors changed the managementwhen they became involved with the firm.

Furthermore, outsider-controlled firms were more likely to sepa-rate the positions of CEO and chairman of the board. While managerswere also the chairman of the board in 61 percent of the gaizhi firms,the figure was 32 percent in firms that were controlled by outside in-vestors. At the same time, outsider-controlled firms were more likelythan insider-controlled firms to offer incentives to managers: 41 percentcompared with 21 percent.

Conclusion

The above analysis shows that the government has retreated from theprivatizing firms by reducing its shares in those firms to less than 10 per-cent. The shares of employees and outside investors have increasedrapidly. The influence of the Communist Party over the firms has alsodeclined, but the role of the labor union in collective wage bargaininghas been more clearly defined and enhanced. Gaizhi has improved in-centives for managers, because firms are more likely to provide them withshares and bonuses. Shareholder representation on the board of directorshas improved, and power sharing among the shareholders’ conference,the board of directors, and the management has begun to occur.

The dominance of the managers appeared to be the main weaknessin the corporate governance of gaizhi firms. There was a high degree ofoverlap between the managers and the board of directors. The dominant

BOX 5.3(CONTINUED)

appointed by the board of directors, and not by the general man-ager as is typically the practice in China. The general manager andthe chairman of the board are different persons in the acquired companies.


role of managers, who are over-represented on the board of directors,opens up the risk that the management will manipulate the board ofdirectors in a way that infringes the interests of owners and ordinary em-ployees. This danger is greater in those employee shareholding compa-nies where the manager enjoys considerable discretionary power. Inmost of these firms the management initiated and dominated gaizhi.After gaizhi the previous management and the old power structure weremaintained. In many cases, it was only the manager’s conscience thatlimited his or her private gains (box 5.4).

Survey results indicate that outside investors in controlling positionsare more likely to use and rely on the new mechanisms of corporate con-trol, to provide effective checks and balances on managerial discretion,and to offer high-powered incentives to senior managers.

144


BOX 5.4THE TEMPTATIONS FACING MANAGERS

Mr. K is the young, energetic manager of a company in one of the sur-vey cities. When the company was privatized, K was a deputy man-ager. The then manager made a promise to the government that thecompany’s performance would improve within three years. After twoyears the target did not look achievable and Mr. K proposed thatthe company go through gaizhi. The municipal government approvedthe plan, and the company’s net assets were sold to the employees ata discount. The shares were sold according to tenure in the companyregardless of seniority.

The company quickly took off after gaizhi. The previous managerwas promoted to a post in the municipal government, and K becamethe manager. Under K’s leadership the company expanded and received new investment from a Shenzhen investor. At the time of interview the company had RMB200 million in cash in the bank.K admitted that he was too cautious to expand the company furtherand that a new manager was needed.

Some financial intermediaries approached K and offered to financea management buy-out. It would have been easy for K to profit froman MBO. He declined, however, because this “would have been un-fair to the ordinary employees.”

6Impact of Gaizhi on

Firm Behavior and Performance

The merits of gaizhi should ultimately be judged on how well the re-structured firms perform. While improving the performance of enter-prises was not the only goal of gaizhi, it would have been a failedexperiment if efficiency had not improved. This chapter compares theperformance of gaizhi and non-gaizhi firms. Within gaizhi firms, spe-cial focus is given to assessing the impact of different forms of gaizhiand, particularly, of different forms of ownership change. We are in-terested in the impacts of three aspects of ownership change: privateshares versus state shares, outsider shares versus insider shares, anddifferent forms of gaizhi. The chapter looks at both changes in behav-ior and changes in outcomes. In terms of behavioral changes, the chap-ter focuses on the internal restructuring of firms; for instance, whetherdepartments were reorganized, managers replaced, changes made toemployee remuneration, or new products introduced, and whether gaizhihas hardened the budget constraints of firms. In terms of changes in per-formance, we use indicators that have been frequently studied in theliterature (Megginson and Netter 2001; Djankov and Murrell 2002),such as pre-tax profit rate over assets, cost per unit revenue, and laborproductivity.

Gaizhi, Internal Restructuring, and Financial Discipline

Has gaizhi led to changes in the internal structures of firms? Have firmschanged their management and organizational structures, bought newtechnologies and equipment, or introduced new products? These aresome of the questions that we asked enterprise managers.

The most common changes were departmental restructuring, areduction in the number of managers, and the establishment or re-organization of the board of directors (see figure 6.1). At the govern-ment’s request, many firms had set up offices with roles corresponding

145

to those of government agencies. For example, some firms had set upa statistical office to work with the bureau of statistics, an environ-mental office to work with the bureau of environmental protection, ora family planning office to work with the family planning commission.These organizational changes were not in general associated with anincrease in management personnel. In fact, more than 80 percent ofgaizhi firms have reported a reduction of their management personnel.On average, 36 percent of managers were made redundant. This num-ber is very similar to the overall labor redundancies (see chapter 4), sug-gesting that managers and ordinary workers have equally shared thepain of layoffs. More than two-thirds of the gaizhi firms had recruitednew general managers.

Gaizhi firms have been investing more than non-gaizhi firms.More than 70 percent of the gaizhi firms bought new equipment, intro-


146

0 20 40 60 80 100

Other

Establish new factoriesor branches

Recruit newmanager

Change wageremuneration

Develop newproducts

Purchase new equipment orintroduce new technology

Establish or reorganizethe board of directors

Reduce managementpersonnel

Restructuredepartments


FIGURE 6.1Methods of Restructuring in GAIZHI Firms, 2002

Percent

duced new technology, or developed new products. Figures 6.2 and 6.3show average absolute amounts of Renminbi spent on fixed investmentand research and development (R&D) by gaizhi and non-gaizhi firms.The gap increased over the late 1990s. In 1995 non-gaizhi firms spenton average 80 percent of what gaizhi firms spent on fixed investment,but by 2001 the fixed investment of non-gaizhi firms was only 29 per-cent that of gaizhi firms. In 1995 non-gaizhi firms spent 17 percent ofwhat gaizhi firms spent on R&D. The gap narrowed in the mid-1990s,but had returned to the 1995 level by 2001.

Gaizhi changes the relationship between firms and the government.Has this affected the hardness of firms’ budget constraints? The fourmajor areas where the soft budget constraint problem1 is most likely toarise are bank loans, interest payments, taxes, and social security.

Overdue Loans. Both gaizhi and non-gaizhi firms carry a large stockof overdue loans, although gaizhi firms had less than half the stock ofnon-gaizhi firms (see figure 6.4). The average stock of overdue loans in

IMPACT OF GAIZHI ON FIRM BEHAVIOR AND PERFORMANCE

147

1. In a soft budget constraint (what Kornai [1980] calls the “pure” case) the fol-lowing apply (1) firms make their own prices; (2) firms influence tax rules andobtain exemptions and postponements, and taxes are not strictly collected; and(3) subsidies and soft credits are given to firms.

0

5

10

15

20

1995 1996 1997 1998 1999 2000 2001

Gaizhi firmNon-gaizhi firm

RMB mn


FIGURE 6.2Spending on Fixed Investment by

GAIZHI and Non-GAIZHI Firms, 1995–2001


148

0

2

4

1

3

5

6

1995 1996 1997 1998 1999 2000 2001


RMB mn


FIGURE 6.3Spending on R&D by GAIZHI and Non-GAIZHI Firms, 1995–2001

0

5

10

15

20

1995 1996 1997 1998 1999 2000 2001


RMB mn

FIGURE 6.4Average Value of Overdue Loans Held by

GAIZHI and Non-GAIZHI Firms, 1995–2001


non-gaizhi firms rose steadily to reach RMB18 million by 2001. Gaizhiwill become more difficult for these firms if their financial burdens con-tinue to grow. An example is the Hengdong Ceramic Factory, whichborrowed RMB14.4 million from the Industrial and Commercial Bankfor investment in 1995. The factory performed poorly and the loan be-came overdue in 1997. In 2001, four years and several managers later,the overdue amount, including interest, had risen to RMB15.9 million.

New overdue loans are a better indicator of financial disciplinethan the stock of past overdue loans. Figure 6.5 shows that in mostyears a lower share of gaizhi firms than non-gaizhi firms had problemswith new overdue loans. The difference between the two types of firmswas 10 percentage points in 1996 and 17 percentage points in 1997.In the late 1990s the gap narrowed considerably, and by 2001 gaizhifirms had slightly greater problems with overdue loans.

The convergence between gaizhi and non-gaizhi firms, however,hides significant differences within gaizhi firms. Figure 6.6 shows dataon new overdue loans by different forms of gaizhi. Firms that under-went public offering and employee shareholding had a lower level ofoverdue loans than other gaizhi firms. Less than 10 percent of thesefirms had newly created overdue loans, and most of the time these loanswere negligible. In firms that were sold or leased out, the amount of newoverdue loans fluctuated over the period but converged to the level ofthe public offering and employee shareholding firms in 2000 and 2001.


149

0

10

20

30

1996 1997 1998 1999 2000 2001

Gaizhi firms

Non-gaizhi firms


Percent

FIGURE 6.5Percentage of GAIZHI and Non-GAIZHI Firms

with New Overdue Loans, 1996–2001


150

In contrast, one-quarter of the firms that opted for internal restructur-ing had new overdue loans. The performance of these firms is largelyresponsible for the observed convergence in the percentage of firms withnew overdue loans between gaizhi and non-gaizhi firms. Internal re-structuring is the least drastic form of gaizhi, and the enterprises thathave gone through internal restructuring differ least from the traditionalSOEs. While they had been reorganized to comply with the CompanyLaw and may have restructured their internal management, they werestill fully government-owned.

In addition to overdue loans, it may be instructive to look moreclosely at overdue interest rates. Interviews with banks and enterprises re-vealed that the banks were often prepared to extend the term of the loanbut insisted that the interest be paid. They were especially keen to enforcethis rule on gaizhi firms. Maintaining a steady stream of interest paymentsis an indicator to higher-level authorities that the loan is still active, re-gardless of whether it was overdue. Figure 6.7 shows the new overdue in-terest payments in gaizhi and non-gaizhi firms from 1996 to 2001. Gaizhifirms have performed consistently better than non-gaizhi firms, althoughthe gap has narrowed, as in the case of new overdue loans.

Across most types of gaizhi firms the share of firms with overdue in-terest payments increased over the 1990s to converge on an average ofaround 25 percent in 2001 (see figure 6.8). No firm listed on the stockmarket had any overdue interest payments. Such firms are generally of

Percent

0

10

20

30

40

FIGURE 6.6Percentage of Firms with New Overdue Loans by

Different Types of GAIZHI, 1996–2001


Publicoffering

InternalrestructuringEmployee

shareholding

Open sales

Leasing

1997 1998 1999 2000 20011996


151

a better quality, as they typically undergo a significant amount of re-structuring prior to listing. They are, in addition, subject to monitoringby the stock market and need to retain a sound financial reputation inorder to maintain their share price and continue to raise funds. The abil-ity of these firms to obtain finance from the stock market means they

0

10

20

40

30

1996 1997 1998 1999 2000 2001

Gaizhi firms

Non-gaizhi firms


Percent

FIGURE 6.7Percentage of GAIZHI and Non-GAIZHI Firms with New

Overdue Interest Payments, 1996–2001

Percent

0

10

20

30

40

1996 1997 1998 1999 2000 2001

FIGURE 6.8Percentage of Firms with New Overdue Interest

Payments by Type of GAIZHI, 1996–2001

Note: The public offering firms did not have any new overdue interest.Source: Survey data.


Employee shareholding

Open sales

Leasing

are less dependent on bank loans. Firms that had undergone internalrestructuring were the most likely to have overdue interest payments.In general, the increasing trend of gaizhi firms’ defaulting on interestpayments is of concern.

Overdue Taxes. The trends in overdue taxes matched those of overdueinterest payments (see figure 6.9). The percentage of non-gaizhi firmswith new overdue taxes was very high at above 40 percent in mostyears. The percentage of gaizhi firms with new overdue taxes was lessthan half that of non-gaizhi firms in most years, but the gap narrowedquickly. A total of 24 percent of gaizhi firms had new overdue taxes in1996, compared with 40 percent of non-gaizhi firms. However, theshare of gaizhi firms with overdue taxes increased considerably between1996 and 2001. The average size of new overdue taxes of gaizhi firmssteadily increased over the period, but it was still much smaller thanthat of non-gaizhi firms.

The percentage of new overdue taxes held by all types of gaizhifirms had converged to between 20 percent and 40 percent by 2001 (seefigure 6.10). It is surprising to find that a high percentage of publiclylisted firms were in arrears on their tax payments. Given the nature ofthe listing process in the past (Tenev and Zhang 2002), most publiclylisted firms have a particularly close relationship with the government,and their overall financial situation is relatively strong. It is interesting,therefore, to find that their performance in terms of overdue tax pay-


152

0

10

20

50

40

30

1996 1997 1998 1999 2000 2001

Gaizhi firms

Non-gaizhi firms


Percent

FIGURE 6.9Percentage of GAIZHI and Non-GAIZHI Firms

with New Overdue Taxes, 1996–2001

ments, although improving in the last two years of observations, isamong the worst.

Overdue Social Security Payments. The percentage of firms with newoverdue social security payments was high throughout the period,regardless of whether or not the firm had undertaken gaizhi (see fig-ure 6.11). There was no clear indication as to which type of firmperformed better in each year. Within gaizhi firms, there was someconvergence in the share of firms with overdue payments across all typesof gaizhi, but publicly listed firms became the major contributor to thearrears problem (see figure 6.12). Overdue social security payments inpublicly listed firms were nil until 1998, but from that year this groupof firms had a relatively high arrears rate and the highest average sizeof overdue social security payments. The rate decreased to about one-third in 2000 and 2001, but the average size of social security pay-ments owed by publicly listed firms continued to climb and reachedRMB900,000 in 2001.

Internally restructured and employee-owned enterprises show adifferent pattern of behavior regarding overdue social security pay-ments. These enterprises started with very high arrears rates in 1996,but rates declined over the next five years. While the average size of theamount owed remained considerable in restructured firms, employee


153

Percent

0

20

40

60

1996 1997 1998 1999 2000 2001

FIGURE 6.10Percentage of Firms with New Overdue

Tax Payments by Type of GAIZHI, 1996–2001


Public offering


Employee shareholding

Open sales

Leasing


154

shareholding firms fared somewhat better. They had a lower rate ofarrears than other firms, and the average size of the overdue debt wasthe smallest of all types of firms in most years.

The sudden increase in arrears in publicly listed firms in 1998 maybe related to the introduction of China’s social security system in thatyear. The new system requires firms to pay 20 percent of their payroll

0

20

60

40

1996 1997 1998 1999 2000 2001

Gaizhi firms Non-gaizhi firms


Percent

FIGURE 6.11Percentage of GAIZHI and Non-GAIZHI Firms with

New Overdue Social Security Payments, 1996–2001

Percent

FIGURE 6.12Percentage of Various Types of GAIZHI Firms withNew Overdue Social Security Payments, 1995–2001


Public offeringEmployeeshareholding

0

20

80

60

40

100

1995 1996 1997 1998 1999 2000 2001

Internal restructuringOpen sales

Leasing

into funds that are pooled together at the city level, and then at theprovincial level, and ultimately at the national level. The pooling offunds may have created a moral hazard problem for larger and better-performing firms, inducing them to default on their payments. Theirstrong bargaining power, partly due to the fact that they make large taxcontributions to local governments, may have permitted them to behavein such a fashion. The better performance of the employee-shareholdingfirms as far as social security payments are concerned might have some-thing to do with their ownership structure. Since these firms are ownedby their employees, workers may have a greater incentive to ensure thatsocial security payments are made on time, because their retirementdepends on it.

The above analysis suggests that the financial discipline of gaizhifirms, while somewhat better relative to non-gaizhi firms, has been de-teriorating in recent years. Gaizhi firms have a better record than non-gaizhi firms in paying bank loans and interest—and, to some extent,taxes—but they show equally poor discipline as far as social securitypayments are concerned. The aggregate statistics hide significant dif-ferences within gaizhi firms. In terms of access to bank loans, fully pri-vatized firms appear to face a harder budget constraint than firms thathave undergone only internal restructuring. And as for overdue tax andsocial security payments, publicly listed firms seemed to have a surpris-ingly bad record.

We use multivariate analysis to gain a better understanding of thefactors influencing firms’ soft budget constraints. Three indicatorsof soft budget constraints are examined in the regression analysis:overdue bank payments combining principal and interest, overduetaxes,2 and overdue social security payments.

We use two sets of indicators of gaizhi. In the first set, all firmsare divided into three groups according to their controlling shares:fully state-owned, state-controlled, and privately controlled. Fullystate-owned enterprises are the control group. In the second set, firmsare again divided into three groups, this time by forms of gaizhi: non-gaizhi firms, internally restructured firms, and firms that have intro-duced some private ownership. The fully state-owned enterprises areagain the control group. We control for asset size, firms’ social burdens,per capita municipal government revenues, firms’ performance, andtime trends.


155

2. Not reported because there are no statistically significant coefficients.

The results are summarized in table 6.1. Privately controlled firmsshow a better financial discipline than state-owned firms with respectto bank payments. Their overdue loans as a percentage of assets areon average 5.2 points lower than the corresponding number for non-gaizhi firms. With respect to social security payments, however, thevarious types of firms do not differ in a statistically significant wayfrom each other. In terms of the form of privatization, privatizing firmsshow better results than non-gaizhi and internally restructured firmswith respect to bank loans, but a worse performance with respect tosocial security payments.

The results also show that in cities with higher per capita gov-ernment revenues, firms tended to have higher overdue loans. This re-sult suggests that firms in a city with a stronger financial position aremore likely to default on their bank payments. At the same time, astronger fiscal position of the local government is associated with bet-ter firm discipline with respect to social security payments. Varioushypotheses can be advanced to “explain” these results. Perhaps firmsare more likely to default on bank payments if they believe that a fis-cally strong local government is in a better position to offer a bail-out.In terms of social security payments, firms may be reluctant to entrusttheir social security to municipalities with weak finances. Recent re-ports by the central government have warned against the accumula-tion of risks in the fiscal position of local governments, including as aresult of appropriation of social security funds (see box 6.1). Or thesame latent variable may be behind stronger municipal finances andmore disciplined social security payments—that is, better enforcementof tax and social security payments by the local government. Furtherresearch is needed to shed light on these interesting results.

The level of worker redundancy is positively associated with over-due bank loans, but the results are not significant. Worker redundancyis, however, positively and significantly associated with overdue socialsecurity payments. This suggests that keeping redundant workers onthe payroll and making social security contributions are viewed as sub-stitutes by firms and governments. A 1 percent increase in redundancyleads to a 1.1 percent increase in social security arrears as a percent-age of total assets. As expected, better-performing firms are less likelyto default on their tax and social security payments.

These results are largely consistent with the descriptive statistics.We find, therefore, evidence suggesting that privatization has a positiveand significant effect on bank discipline. The soft budget constraint


156

157

TABLE 6.1FACTORS INFLUENCING FINANCIAL DISCIPLINE OF FIRMS

Dependent variable

Overdue bank Overdue socialpayments as security paymentspercentage of as percentage

Independent variables firm assets of firm assets

State-controlled vs. wholly state-owned

Privately controlled vs. wholly state-owned

Internal restructuring vs. non-gaizhi

Some private ownership vs. non-gaizhi

Firm’s social burdensa

Firm’s asset size

Firm’s performanceb

Local government’s fiscal strengthc

Time trendd

NOTE: Method of estimation: weighted least squares.a. “Social burdens” is the proportion of official retirees plus internal retirees plusxiagang workers in the total number of employees; official retirees are counted as asocial burden because, in many cases, firms are still responsible for directly payingtheir retirement wages despite the fact that China has established a new pension sys-tem that in most cases maintains a centralized system up to the provincial level.b. Firm’s performance is measured in terms of return on capital averaged over theprevious three years.c. Local government’s fiscal strength is measured by per capita fiscal revenues atthe municipal level.d. Time trend is captured by year dummies.SOURCE: Author’s estimates.

Not statistically significant

Negative, statisti-cally significant


Negative, statisti-cally significant at 10% level



Negative, statisti-cally significant

Positive, statisticallysignificant

Declining since 1997




Positive, statisticallysignificant at 10% level



Negative, statisticallysignificant


No clear pattern

problem is severe, however, in cases of tax and social security payments.The descriptive analysis shows that roughly one-third of firms were inarrears on those two payments, regardless of the controlling shares orthe form of gaizhi. Therefore, we find that firms are more likely to evadepayments to local governments than to commercial banks.


Gaizhi and Firm Performance

We look at three indicators of firm performance: profitability, unit cost,and labor productivity. Profitability is defined as the return on assets—that is, the percentage of the pretax profit over the total book value ofassets. Unit cost is the percentage of the material and operational costover the revenue. It does not include wage payroll. Labor costs arepartly reflected by labor productivity. In the same way as the definitionby Frydman and others (1999) does, our definition of the unit cost alsoserves the purpose of accounting for a firm’s passive restructuring mea-sures centered on cost reduction. The third indicator, labor productiv-ity, is defined as the real revenue contributed by an on-duty worker. Asalient feature of the Chinese SOEs is their worker redundancy. A con-siderable portion of the workforce is not active, although it is attachedto a particular firm. Hence, we use the number of on-duty workers asan indicator of firm performance. Labor productivity captures the fea-


158

BOX 6.1FISCAL POSITIONS OF LOCAL GOVERNMENTS IN CHINA

The Development Research Center at the State Council (DRC) re-leased two reports in 2004 warning of the potential risks associatedwith local government debts. The reports found that the forms ofdebts incurred by local governments are “so many” and their debtburdens are “so heavy” that it “goes beyond normal people’s imag-ination.” The study team failed to obtain precise data from local gov-ernments, but it estimated that the total amount of local governmentdebts could be on the scale of RMB1 trillion or more. In their esti-mation, primary items include:

• bank loans borrowed by various “development companies”created by local governments to bypass the budget law

• appropriation of social security funds• wage arrears• SOE losses to be written off (primarily grain trading companies)• payables to construction firms for infrastructure projects • debts incurred by township governments

The DRC team recommends the legalization of local governmentbonds to cope with the challenge.

tures of both passive adjustments and positive expansion by the firm. Itmay reflect more of the expansionary side in the Chinese context, how-ever, because of the rigidities in the labor market.

Figures 6.13 to 6.15 show the time trend of the three performanceindicators by various dimensions of the gaizhi process. Figure 6.13 com-pares gaizhi and non-gaizhi firms. Gaizhi firms outperform non-gaizhifirms on all three indicators. We observe, however, the declining trendin the profitability of gaizhi firms over the sample period. Both gaizhiand non-gaizhi firms show a trend of improving labor productivity. Fig-ure 6.14 compares the performance of sample firms by ownership char-acteristics. As before, we look at wholly state-owned, state-controlled,and privately controlled firms. State-controlled firms tend to be the bestperformers in the sample, followed by privately controlled and whollystate-owned firms. Finally, figure 6.15 compares the performance offirms based on whether they are controlled by insiders or outsiders. Wedistinguish between state-owned firms that are controlled by insiders(managers and employees), privately owned firms that are controlled byinsiders (employee shareholdings), and outsider-controlled firms, whereoutsiders could be either private firms or other SOEs. Here the pictureis more mixed. Outsider-controlled firms dominate most of the time.Insider state-controlled firms tend to be the worst performers.

A notable feature in figures 6.13 to 6.15 is the higher volatility inthe performance of private firms compared with wholly state-ownedand state-controlled firms. The higher volatility of the private sectorrelative to other segments of the economy has been a notable feature ofthe transition process in China. For example, the volatility of the UBSWarburg China Private Enterprise Index is on average 1.5 times higherthan the volatility of the Hang Seng Index3 (see table 6.2). Several fac-tors may account for the higher volatility in the performance of pri-vate enterprises. Particularly for the de novo private firms, the highervolatility may be simply a restatement of the fact that private firms aresubject to exit discipline and are younger and smaller than otherfirms. But it also reflects the fact that other sectors of the economy, andparticularly the state-owned sector, react more slowly, if at all, to busi-ness opportunities and exit pressure. The performance volatility of theprivate sector is in this sense higher than it would have been were marketprinciples to apply equally to all sectors of the economy. The tendencyfor private-sector performance to be more volatile is more pronounced


159

3. The volatility of non-SOE A-shares has also been very high (see Chan 2002).


160

–4

–2

0

2

4

8

6

1995 1996 1997 1998 1999 2000 2001

Percent

54

58

56

62

60

64

66

1995 1996 1997 1998 1999 2000 2001

Percent

a. Profitability

b. Unit cost

Source: Calculated and plotted using the survey data.

FIGURE 6.13Comparison of GAIZHI and Non-GAIZHI Firms on

Three Performance Dimensions, 1995–2001

RMB 1,000c. Labor productivity

20

60

40

80

1995 1996 1997 1998 1999 2000 2001

Non-gaizhi firms

Non-gaizhi firms

Non-gaizhi firms

Gaizhi firms

Gaizhi firms

Gaizhi firms


161

–3

–1

1

3

5

7

1995 1996 1997 1998 1999 2000 2001

Fully state owned

State controlled


Percent

50

55

60

65

70

1995 1996 1997 1998 1999 2000 2001

Fully state owned

State controlledPrivately controlled

Percent

a. Profitability

b. Unit cost


FIGURE 6.14Comparison of Private and State-Controlled Firms

on Three Performance Dimensions, 1995–2001

0

20

80

60

40

100

120

1995 1996 1997 1998 1999 2000 2001

Fully state owned

State controlled




162

–3

–1

1

3

5

7

1995 1996 1997 1998 1999 2000 2001

Percent

30

40

50

60

70

1995 1996 1997 1998 1999 2000 2001

Percent

a. Profitability

b. Unit cost


FIGURE 6.15Comparison of Outsider- and Insider-Controlled Firms

on Three Performance Dimensions, 1995–2001

0

20

80

60

40

100

1995 1996 1997 1998 1999 2000 2001


Insider state controlled

Insider privately controlledOutsider controlled


Insider privately controlled

Outsider controlled


Insider privately controlled

Outsider controlled

in sectors where the degree of competition between private and statecompanies is relatively high (Chan 2002).

The higher volatility of performance of private firms implies thatlending to these companies is viewed by commercial banks as being morerisky. On the other hand, because private businesses are less politicallyprotected, the banks have a freer hand to act if problems arise. Chinesebanks tend to be risk averse, however; as a result, private enterprises stillaccount for a disproportionately small share of credit flows in China. Inthe current regime of monetary tightening, private and smaller compa-nies seem to be disproportionately affected (see box 6.2). Privatized firmsin the sample raised concerns that gaizhi may have a negative impact ontheir access to financing.

Figures 6.13 to 6.15 present indicative results only. To isolate theeffect of gaizhi, we need to control for other factors and address theendogeneity issues—that is, that the timing and method of gaizhi maybe influenced by firm performance. In particular, there may be a selec-tion bias in privatization and gaizhi, in the sense that better-performingfirms are selected to conduct gaizhi or to be privatized first. In the fol-lowing sections we present and discuss some of the results of the regres-sion analysis of the factors affecting firm performance where we controlfor other variables and for the selection bias. For a detailed expositionof the methodology, see Song and Yao (2004).

The Effect of Private Ownership. The results in table 6.3 focus on theeffects of private ownership introduced during gaizhi on firm perfor-


163

TABLE 6.2PRICE VOLATILITY OF VARIOUS CATEGORIES OF COMPANIES AFFILIATED

WITH CHINA AND LISTED ON THE HONG KONG EXCHANGE, MAY 2002

30-day 50-day 100-day Index volatility volatility volatility

Hang Seng Index 16.7 18.1 20.4China-affiliated companies 22.5 24.6 29.5China Enterprise Index (SOEs) 14.3 19.5 19.2China Mainland Index 19.0 21.2 25.8

(Red Chips)China Private Enterprise 22.8 28.4 38.3

Index

SOURCE: Bloomberg.

mance. Wholly state-owned enterprises are the control group. We findthat state-controlled and privately controlled firms have a clear advan-tage over wholly state-owned firms with respect to profitability. Theresults concerning the impact of private ownership on unit costs andlabor productivity are mixed and inconclusive.

Ownership diversification has a significant economic impact onfirm profitability. We estimate that a privately controlled firm will havea return on assets that is from 1.2 to 1.7 percent higher than a fullystate-owned firm. In the case of state-controlled firms, their advantageover fully state-owned firms is estimated at between 1 and 2.7 percent.The finding that privatization only has a significant positive impact onprofitability, but has a weak or insignificant impact on unit cost andlabor productivity, is consistent with the finding of Frydman and others(1999) on three Eastern European countries. It may reflect the interplayof limitations on cost cutting in terms of technological and businessenvironment constraints,4 and a focus on expansion rather than on de-fensive cost-cutting restructuring. It also suggests that firms with di-versified ownership tend to make more economic use of their capitalstock than does an old-style SOE. Chinese SOEs are characterized by a


164

BOX 6.2EFFECTS ON THE PRIVATE SECTOR OF

MONETARY TIGHTENING MEASURES

A private entrepreneur, a chicken farmer in Shanxi province, was im-prisoned for “illegal fundraising” after supplying cheap eggs to hiscustomers in return for extra money. The case turned out to be con-troversial, as local public opinion tended to believe that the govern-ment had punished an honest and creative entrepreneur. Proposalswere made at sessions of the National People’s Congress to amendthe law.

With tightened macroeconomic policies, some private enterpriseshave turned to sources of funds with high interest rates. The realfinancing cost can be as high as 45 percent, according to some mediareports. Some credit guarantee companies have set up direct lendingbusinesses.

4. See chapter 4 on labor policy.


165

TABLE 6.3ANALYSIS OF THE EFFECT OF PRIVATE OWNERSHIP ON

FIRM PERFORMANCE

Dependent variable

Independent Labor variables Profitability Unit cost productivity

State-controlled vs. wholly state-owneda

Privately controlled vs. wholly state-owneda

Firm’s social burdensb

Firm’s size

Debt-equity ratio

Arrearsc

NOTE: Dummy variables are used to indicate privately controlled and state-controlled firms. All control variables are three-year time averages. Sectoral andregional dummies were also included in various model specifications. We controlfor selection bias through lagged performance variables and the fixed-effect panelmethods. Only coefficient signs that are consistent in both estimations are reported.Only results that are statistically significant in both methods are reported as statis-tically significant. For details, see Song and Yao (2004).a. Wholly state-owned firms are the control group.b. “Social burdens” is the proportion of official retirees plus internal retirees plusxiagang workers in the total number of employees; official retirees are counted as asocial burden because, in many cases, firms are still responsible for directly payingtheir retirement wages, despite the fact that China has established a new pension sys-tem that in most cases maintains a centralized system up to the provincial level.c. “Arrears” includes overdue payments related to bank loans, taxes, and socialsecurity.SOURCE: Song and Yao (2004).

Negative, notstatisticallysignificant

Not statisticallysignificant


Positive, notstatisticallysignificant



Positive, not statisticallysignificant




Positive, not statisticallysignificant








much higher capital intensity than private firms, indicating certain levelsof inefficiency (Lin and Tan 1999).

The above results concerning the impact of private and stateownership on firm performance have been corroborated by researchon the performance of Chinese listed companies. Sun, Tong, andTong (2002), for example, find that neither too much nor too littlegovernment ownership is beneficial to a firm’s efficiency, and showthat the relationship between government ownership and firm per-formance follows an inverted U-shaped pattern. They show that gov-ernment ownership (whether in the form of state shares or legal personshares) has a positive and significant impact on firm performance.Among the benefits of government ownership they include the sig-naling effect when a government shows its commitment to the firm byretaining a relatively high portion of the firm’s equity; the monitoringrole that the government plays for the benefit of all shareholders; andthe supporting policies that the government formulates to favor thefirms it owns or partially owns. According to Sun, Tong, and Tong(2002), the findings suggest that when a state-owned enterprise goesprivate, it may be beneficial to build in a stage when a certain degreeof government ownership is maintained, rather than going tooquickly from a state-owned status to complete privatization. They donot differentiate, however, between state shares and legal personshares, and combine both in the category of government ownershipfor the purpose of their analysis.

The Effect of Outsiders. Table 6.4 shows the results of the impact ofinsider and outsider shareholdings on firm performance. The mostimportant result is that outside shareholding has a stronger positive im-pact on firm performance than insider shareholding. The effect of out-side shares is particularly pronounced with respect to profitability andlabor productivity. According to our estimates, the return on assets ofa fully outsider-owned firm would be about 2 percentage points higherthan that of a fully government-owned firm. Respectively, a worker ina fully outsider-owned firm would on average contribute RMB6,200per year more than their counterpart in a fully state-owned firm.5 Botheffects are economically and statistically significant.


166

5. The average labor productivity in the sample was in the range of RMB46,000to RMB62,000 for different years.

TABLE 6.4ANALYSIS OF THE EFFECT OF OUTSIDE OWNERSHIP ON

FIRM PERFORMANCE

Dependent variable

Independent Laborvariables Profitability Unit cost productivity

Insider private shares, including:

• Management shares

• Employee shares

Outsider shares, including:

• Outside state shares

• Outside private shares

Firm’s social burdensa

Firm’s size

Debt-equity ratio

Arrearsb

NOTE: All control variables are three-year time averages. Sectoral and regional dummies werealso included in various model specifications. We control for selection bias through laggedperformance variables and the fixed-effect panel methods. Only coefficient signs that are con-sistent in both estimations are reported. Only results that are statistically significant in bothmethods are reported as statistically significant. For details, see Song and Yao (2004).a. “Social burdens” is the proportion of official retirees plus internal retirees plus xiagangworkers in the total number of employees; official retirees are counted as a social burden be-cause, in many cases, firms are still responsible for directly paying their retirement wages,despite the fact that China has established a new pension system that in most cases main-tains a centralized system up to the provincial level.b. Arrears includes overdue payments related to bank loans, taxes, and social security.SOURCE: Authors’ estimates.































167


When insider shares are broken up into management and em-ployee shares, we find that a higher share of employee ownershiptends to have a negative effect on unit cost and a positive effect onproductivity. The results are not statistically significant, however.Management shares have a negative and significant impact on laborproductivity.

Both outside private and outside state shares contribute to thepositive effect of outsider shares on profitability, but the latter makea more significant contribution than the former. Outside state share-holding shows a positive and statistically significant association withprofitability and labor productivity. Larger outside private share-holdings tend to reduce unit costs, but the effect is not statistically sig-nificant under all model specifications. Thus, while more outside stateshares help a firm improve its profitability and labor productivity,more outside private shares help it to reduce costs. This finding seemsto suggest that a firm is more likely to adopt expansionary measuresto improve its efficiency if it has more outside state shares, and ismore likely to adopt defensive measures to reduce costs if it has moreoutside private shares. In the Chinese setting, SOEs may have a biasin favor of an expansionary strategy because they have easier accessto bank loans and other sources of finance, and face a weaker financialdiscipline than private firms.

The finding that outside state ownership has an important positiveimpact on firm performance has several important implications. First,it suggests that outside ownership matters, irrespective of whether it isa private or state outside ownership. Second, it implies that not all typesof state ownership are created equal. Outside state ownership is dif-ferent from direct state ownership in which the control is typicallyin the hands of insiders. Third, it implies that in the Chinese context,outside state ownership may bring some advantages to firms relative tooutside private ownership.

One such advantage is the superior ability of the state sector to ob-tain external financing and government support. For example, SOEsaccount for 37.5 percent of industrial production in 2003 (CEIC), butthey have absorbed 58.6 percent of new lending in that year (GoldmanSachs 2004). In addition, the Chinese stock market is overly dominatedby state-controlled firms. Often, the private sector relies on the fundsleaked from the state sector to get finance (Lu and Yao 2003). Con-sequently, an outside SOE is in a better position to bring more financeto the recipient SOE than is an outside private firm.


168

Outside state ownership shows a much more pronounced domi-nance over inside state ownership, however, than over private outsideownership. A more pertinent question, then, is, What are the advan-tages of outside versus inside state ownership? Most likely the answerhas to deal with the power of outsiders to shake the implicit contractswithin the firm that entrench managers and employees. As a general rule,outsiders find it easier to introduce changes that affect labor and man-agerial practices. An outside SOE may not be able to establish work dis-cipline with respect to its own employees, but its status as an outsidermay give it more leverage to do so with the employees in the new firm.6

Research on other transition economies has confirmed the powerof outside shareholdings to improve efficiency, although only in thecase of outside private ownership. For example, Djankov and Murrell(2002) find that outside block shares are more helpful than scatteredprivate shares, because block shares make it easier to establish bettercorporate governance.

Our finding that employee shares do not have a tangible positiveimpact on efficiency is largely consistent with the literature on transi-tion economies. Our similar finding about the impact of managementshares, however, differs from some early findings by Groves and others(1994, 1995), Gordon and Li (1995), and Li (1997). These studieshave shown that contractual incentives (such as performance-basedbonus schemes) in the early reform period did help the Chinese SOEsto improve efficiency. It seems reasonable to believe that ownership in-centives should play at least the same role as contractual incentives,because the manager gets more autonomy when he owns part of thefirm. A possible explanation for the observed inconsistency is that wedo not control for managerial autonomy. In the literature (for exam-ple, the above-quoted studies on China, while more comprehensivediscussion can be found in Djankov and Murrell 2002), managementautonomy is found to be as important as managerial incentives inachieving firm efficiency. Privatization may not automatically give themanager autonomy, so control for management autonomy is needed


169

6. One such example is the giant Luoyang Tractor Factory. It has a serious prob-lem in its headquarters in Luoyang, Henan province, but it runs a successfuljoint venture (through a subsidiary registered in Great Britain) with another SOEin Zhenjiang, Jiangsu province. Indeed, one of the purposes of many failing SOEsseeking a joint venture relationship with another SOE is to find an opportunity toget rid of the excessive workforce.

in our regressions to obtain an accurate assessment of managementincentives. We do not have a complete series of data on managementautonomy, however, so we cannot perform this test in the panelframework.7

We also look at the impact of various forms of gaizhi on firmperformance. The only result that is statistically significant under allspecifications is the negative and relatively strong association of in-ternal restructuring and firm profitability.

Gaizhi and Time Trends in Performance

In this section, we examine some of the time dimensions of the rela-tionship between various forms of gaizhi and firm performance. Arethe effects of gaizhi concentrated in the first years after restructuring,or are they spread pretty much evenly post-restructuring? Do we see adiminishing impact of gaizhi as new and different cohorts of SOEsenter the process?

Time patterns in the relationship between gaizhi and firm perfor-mance may arise for a variety of reasons. For example, Frydman andothers (1999) point out two kinds of selection bias that may distort theestimation of the effects of privatization. Insiders may underreport thefirm’s performance (or deliberately perform poorly) before privatiza-tion in order to get a better deal in privatization (or because of a short-ened time horizon), so the better postprivatization performance maybe artificial. In this case, one would expect to see stronger effects earlyon postprivatization.

The positive effects of privatization/gaizhi may diminish or becomestronger over the years, depending also on the forms of privatization/gaizhi and other firm-specific characteristics. For example, privatizationin the form of employee shareholdings may have a significant effecton firm performance in the earlier years of privatization because of en-hanced incentives, but this effect may quickly diminish in the later yearsbecause the shortcomings of employee shareholding overtake the pos-itive effects of enhanced incentives. On the other hand, the effects ofprivatization/gaizhi may also be enhanced over the years because ittakes time for restructuring to take effect.


170

7. We have data on the distribution of decision rights in the firm for 2002, theyear when the survey was conducted, that allow us to perform some form of thetest based on cross-sectional comparison in future research.

In addition, later privatized/gaizhi-ed firms may not perform aswell as early privatized/gaizhi-ed firms. This may be because it is simplyeasier to restructure firms that are in relatively good shape. Difficultcases naturally take longer to go through the process. Furthermore, localgovernments typically want to start with success stories in order to buildpolitical capital. Self-selection also plays a role; insiders with superiorknowledge about the “true” characteristics of firms that are to be pri-vatized may push selectively for the better firms to be privatized soonerrather than later. In all these situations, better-quality firms will be pri-vatized first—or, as the saying goes in China, “The prettier daughtersare married off first.” It is also possible to rationalize why poorly per-forming firms may be privatized firms. For instance, local governmentsin a “market for lemons” type of situation may try to divest themselvesof the poor-quality firms first.

If this “prettier daughters” effect were strong, the privatization/gazhi effect would be overestimated for the early privatized/gaizhi-edfirms. Conversely, the effect of the later privatized /gaizhi-ed firmswould be overestimated if “the prettier daughters” were married off late(Guo and Yao 2003). It is therefore more or less an empirical questionas to what is the time profile of the relationship between gaizhi and firmperformance.

In this section we test for the time distribution of the impact ofgaizhi on performance and for the “prettier daughters” effect. The testis performed for only three aspects of the gaizhi process: gaizhi versusnon-gaizhi firms, private versus state shares, and outsider versus in-sider shares. To analyze the time patterns, it is important to determinethe time of privatization/gaizhi in order to see whether the interactionsof the time elapsed before and after privatization/gaizhi with the threeaspects of the gaizhi process have a significant impact on performance.If the positive impact of gaizhi were concentrated in early years, wewould expect to see a weaker impact as the number of years aftergaizhi increases. We would therefore expect the coefficient of the in-teraction term between the gaizhi variable and the number of yearsafter gaizhi to have the opposite sign to the coefficient of the gaizhivariable. If it were the case that the “prettier daughters were marriedoff first,” this would imply that the coefficient of the interaction termbetween years before gaizhi8 and the gaizhi variable would have to be


171


8. The variable assumes a value of f-1995 for t ≥ f, and a value of zero for t < f,where f is the year of gaizhi.


172

negative, as the smaller the number of years before gaizhi, the strongerthe positive impact on performance.

The results are summarized in table 6.5. In regard to the early im-pact of gaizhi, only outsider shares show a consistent effect across allthree performance indicators. Only the impact on unit costs is statisti-cally significant, however. Private shareholding shows an early impacton unit costs and labor productivity, but only the impact on unit costsis statistically significant at the 10 percent level. The results seem tosuggest, therefore, that outside ownership has somewhat quicker de-livery than other forms of gaizhi as far as positive impact on perfor-mance is concerned.

The “prettier daughters” effect turns out to have mixed results. Wefind some evidence for such an effect in terms of profitability, but notin terms of unit costs and labor productivity. More profitable firmstended to go through the gaizhi process first. On the other hand, late-comers tended to have lower unit costs and higher productivity. Thesecontrasting trends could be explained by different adjustment strategiesemployed by early, as opposed to late, gaizhi firms. It is possible thatearly gaizhi firms had relied more on an expansionary strategy toincrease their profitability, but enjoyed less flexibility than the late-comers in terms of adjusting their employment and cost structures.Comparisons between firms that were privatized /gaizhi-ed firms in1996–98 and those that were privatized/gaizhi-ed in 1999–2001 showthat, in their post-privatization/gaizhi years, the first group of firmsreached an average investment rate of 13.3 percent but reduced their on-duty workforce by an average rate of only 2.7 percent per annum; in con-trast, the latter group obtained an investment rate of only 8.6 percent butreduced their on-duty workforce by an annual rate of 8.0 percent.

The Impact of Other Factors on Performance

The results presented in tables 6.3 and 6.4 show that size and socialburdens tend to be negatively associated with performance. The neg-ative impacts of worker redundancy on profitability are to be expected,as more redundant workers increase the firm’s wage, social security,and healthcare burdens. It is not immediately clear what the reasonsbehind the positive effects of worker redundancy on unit cost and thenegative effects on labor productivity might be. One explanation is thatfirms with excess labor substitute labor for other inputs. This will ac-count for the negative effect of redundancy on labor productivity as

173

TABLE 6.5ROLE OF TIMING ON GAIZHI PERFORMANCE

Early gaizhi effect “Prettier daughters” effect

Labor LaborProfitability Unit costs productivity Profitability Unit costs productivity

Gaizhi vs. non-gaizhi

Private vs. state

Outsider vs. insider

NOTE: “Prettier daughters” effect implies that better-quality firms are the first to undergo gaizhi.SOURCE: Authors’ estimates.

No, not statisticallysignificant

No, not statisticallysignificant

Yes, not statisticallysignificant

Yes, not statistically insignificant



Yes, statisticallysignificant at 10% level

Yes, statisticallysignificant


Yes, not statistically insignificant

No, statisticallysignificant


No, statisticallysignificant at10% level




Yes, significantat 10% level

Yes, statistically significant

well. That is, firms with redundant labor try to minimize the numberof xiagang workers. Any worker whose marginal product will behigher than the difference between the minimum wage and the xiagangallowance will be on the factory floor rather than at home. Firms withmore xiagang workers are more likely to have reached this limit. In theprocess, they will also have exhausted opportunities to substitute laborfor other inputs. The same process may account for the negative im-pact of size on performance. As firms are forced to keep workers onthe books, firm size may be another indicator for worker redundancy.

Conclusion

Privatization has brought considerable efficiency gains to the Chineseenterprise sector. It has a significantly positive impact on firm prof-itability, although a weak or no significant impact on unit cost and laborproductivity. We find that outside ownership, irrespective of whetherprivate or state, has a strong and significant impact on performance. Thisimpact tends to be concentrated in the early years of reforms.

The finding that outside state ownership has a strong positive im-pact on firm performance has important implications for our under-standing of state ownership. The conventional belief is that state-ownedenterprises fail because public ownership provides an inadequate in-centive structure within the firm. China’s experience shows that SOEswith a relatively high degree of autonomy to participate in the marketprocess may have difficulty in putting reforms into effect in their ownenterprises but can be effective agents of change in other state enter-prises. From the perspective of the ultimate owner—the state—this canbe viewed as an illustration of the efficacy of using agents to monitorother agents (Varian 1989).


174

7Toward a Fairer and

More Efficient Gaizhi Process

Gaizhi has brought about efficiency-enhancing changes in China’sstate enterprise sector. While the process is far from over, it is alreadydelivering results in terms of preserving employment and improvingfinancial performance. The potentially harmful social impact associ-ated with a restructuring on such a massive scale has so far been muted,due to the combined effect of a rapidly evolving social safety net, thedynamic growth of the private sector, and a robust macroeconomicperformance.

Although very important, the efficiency aspects of the gaizhi pro-cess most likely will not determine the future of enterprise reform inChina. Fairness and distribution issues1 related to gaizhi have been at-tracting most public attention recently. The concern of the governmentand the general public has been that gaizhi has been accompanied byerosion of state assets, corruption, and “inglorious” enrichment by pri-vate individuals. An ongoing public debate in China puts the main issuessurrounding the future of privatization in the country into sharp relief.

The Public Debate about Privatization in China

An unprecedented public debate on the ongoing privatization policy istaking place in the Chinese media.2 Two views on the future of privati-zation policy have chrystalized in the course of the discussion. One view,which is close to the official stance on privatization, calls for “stickingto the current direction of SOE reform and pushing forward ownership

175

1. We do not imply that efficiency and equity are necessarily contradictory to eachother. On the contrary, more often than not they tend to coincide. For instance,an open and competitive process of sale of state assets is more likely than a non-transparent one to maximize privatization proceeds and identify the best owner.

transformation in a regulated way.”2 The other view expressed duringthe debate can be summarized as follows: (1) the ongoing privatization,especially MBOs, has resulted in the loss of a large number of state as-sets to the benefit of a small number of private enterpreneurs and SOEmanagers; (2) SOEs can be run as efficiently as private firms so long asthe government hires professional managers and imposes fiduciary du-ties on them; and (3) the current direction of SOE reform must bechanged and privatization (“property rights reform”) stopped.

Both views reflect concerns about corruption, nontransparency,unemployment, and inequality associated with the gaizhi process, butthey differ in their prescriptions. What are some of the implications ofthe analysis and the findings presented in this study for the concernsexpressed in the debate?

One concern is that the ongoing privatization, especially MBOs,has resulted in the loss of a large amount of state assets to the benefitof a small number of private individuals and SOE managers. It is clearthat gaizhi has been a messy process often associated with asset strip-ping. It is difficult—perhaps impossible—to quantify the extent of thistype of abuse. The best this study can do is to offer some circumstan-tial and anecdotal evidence, which of course needs to be interpretedvery carefully. The analysis helps, however, to put the debate in a cer-tain perspective.

The main source of loss of state assets has been loss-making in in-efficient state firms. In fact, one motivation for gaizhi, as discussed inchapter 2, has been to stop the bleeding of state firms. Has the curebeen worse than the disease? Based on the empirical analysis presentedin this study, the answer should be “no.”

Economic crimes take place on a significant scale in state firmsthat are not going through gaizhi or before they embarked on a processof restructuring. All the major state banks in China, for instance, havebeen plagued by major embezzlement scandals. In fact, some of thesescandals have been unearthed as part of the appraisal process in prepa-ration for restructuring. The rise of economic crime in China is, tosome extent, an unwelcome byproduct of market liberalization. As dis-


176

2. This view is perhaps best reflected in an article published in the People’s Dailyin September 2004 in the name of SASAC’s research department and in a recentspeech delivered at Tsinghua University by Shao Ning, deputy director of SASAC(The Economic Observer 2004).

cussed in chapter 2, market liberalization tends to exacerbate theagency costs of managerial autonomy under traditional state owner-ship. It therefore changes the cost-benefit balance of various forms ofownership relative to central planning and provides incentives for pri-vatization. The government transfers ownership to the manager oroutside owners to better align control and ownership and, in this way,reduce the incentives to commit economic crimes.

Asset stripping during gaizhi can take place in a number of ways,few of which are directly linked with privatization. One commonly usedchannel in China is through the choice of the valuation firm. As re-ported in chapter 3, we don’t find systematic differences between MBOsand sales and the other forms of gaizhi, although one can argue that theconflict of interest in the case of MBOs, in particular, is more severe.Another channel is the practice of giving discounts to appraised valueto buyers of state assets. Since MBOs and sales are the main forms ofgaizhi involving transfer of ownership, it is no surprise that these formsshow a bias in favor of discounts (see chapter 3, table 3.6). It is impor-tant to note that sales to private firms and individuals account for asmall percentage of gaizhi: 10 percent of the gaizhi cases in our sample.Both these channels are relatively easy to control, so that conflicts of in-terest and, consequently, abuses are minimized. But perhaps the mainchannel for asset stripping has been debt evasion through policy bank-ruptcy or various forms of internal restructuring. This channel is per-haps more difficult to control, as it often involves tensions betweenpowerful local incentives and national interests. This last point is alsoa reminder that every irregularity associated with gaizhi has at least twosides, one of which is typically a government agency. In summary, onecan argue that the irregularities often associated with privatization arenot intrinsic attributes of the MBOs or privatization process, but in-volve other factors as well.

The second point is that SOEs can be run as efficiently as privatefirms, so long as the government hires professional managers and im-poses fiduciary duties on them. We will briefly discuss two issues thathave a direct bearing on this argument: (1) the attributes of heteroge-neous ownership structures; and (2) the relationship between owner-ship structure and regulation.

Debates on ownership change—like the one unfolding currently inChina—typically treat private and state ownership as homogeneous andmonolithic concepts. In fact, the so-called private and state ownership can

TOWARD A FAIRER AND MORE EFFICIENT GAIZHI PROCESS

177

take diverse forms, to the point where it is difficult to tell the differ-ence. For example, shareholders in a corporation can include diversi-fied minority individual shareholders, conglomerates with significantstakes but with a dominant related party interest, institutional investorsof various stripes, employees of the firm, private commercial banks,and so on. Each one of these shareholders faces different incentives, in-cluding incentives to exercise ownership rights, and their interests maybe in conflict. The same applies to state ownership. An SOE can be“owned” by a government ministry, by a local government at the levelof the province, municipality, or county, by another SOE, by govern-ment-owned banks, pension funds, or asset management companies,or by combinations thereof. Again, these stakeholders may face dif-ferent incentives, including incentives to exercise their ownershiprights, and their interests may be in conflict with each other (see box7.1). Heterogeneous and indirect3 ownership structures, includinghybrid structures combining various types of private and state owners,are rather typical in modern economies. In China the hybrid firm, andnot the purely private or the purely state-owned firm, has become therepresentative business entity.

This type of heterogeneity is not trivial; it matters. Diverse owner-ship structures exhibit common features that tend to transcend differ-ences based on whether the state or individuals as ultimate owners havethe majority stake.4 First, multiple owners with conflicting incentivesand interests may provide mutual checks and balances, allowing a com-mon principal (the state, for example) to use agents to monitor eachother. Second, diversity, with the implied conflicts of interest betweenvarious types of owners, invites regulation. In the case of conflicts be-tween various owners, whose rights should prevail? In heterogeneousownership structures, one type of ownership interest is often particu-larly weak in the sense of being more vulnerable to abuse by managersand controlling shareholders. Typically, most vulnerable are diversifiedindividual shareholders who abdicate their ownership functions and be-


178

3. Indirect in the sense that the direct owners are themselves agents and the ultimateowners can be removed several degrees of separation from the managers.4. An often-heard criticism of state ownership is based on the assertion that thestate as owner introduces noneconomic objectives in the firm. This assertion looksincreasingly unconvincing in the light of the global corporate social responsibilitymovement. Societal demands find their way into firms irrespective of the type ofownership.


179

BOX 7.1THE HETEROGENEITY OF STATE OWNERSHIP IN CHINA

China is not unique in having companies with mixed ownership.Countries as diverse as Brazil and Vietnam have also pursued suchdiversification of ownership. But China is probably unique in termsof the diversity of state ownership forms. It is often not clear howclosely associated with the state various shareholders are. In manycompanies the provincial government may be the controlling share-holder; in addition, there could be a large number of SOEs with smallstakes of between 1 and 5 percent of the company. A nontrivial ques-tion is how to treat such SOEs: as minority shareholders, or as partof the controlling shareholder? This question arises, for example, inthe context of the December 2004 rules introduced by the CSRC giv-ing minority shareholders the ability to influence the decisions oflisted companies on (1) the issuance of new shares, (2) the issuanceof convertible corporate bonds, (3) asset restructuring, and (4) over-seas listing of a subsidiary firm of the listed company. Anecdotal ev-idence suggests that SOEs as minority shareholders can, and typicallydo, have different interests from the state and SOEs in controlling po-sitions. Their interests can be more aligned with the interests of otherminority investors, including other private minority investors, thanwith the government as a controlling shareholder.

In a recent case from the Supreme People’s Court (SPC), the high-level administrative body that acted as the domestic party to a jointventure transferred the shares belonging to the Chinese party to oneof its subsidiaries by means of its administrative power. Its argumentwas that since the transfer involved state assets and both partieswere state-owned, the ownership of the shares did not change. Thecourt invalidated the argument and ruled that the joint venture part-ner should have the priority to purchase the shares of the domesticpartner (Xu 2004).

come absentee owners.5 It is absentee ownership that drives regulatoryactivism, which steps in to fill the ownership vacuum. It is the presenceof absentee ownership in heterogeneous and otherwise very diverseownership structures internationally that acts as the common denomi-nator that induces the more-or-less homogeneous regulatory response,which we observe in ongoing global corporate governance reforms.

5. Employees, some types of institutional investors, and the state can all be ab-sentee and passive owners.

Thus highly diverse ownership structures tend to invite a homogeneousregulatory response in terms of corporate governance regimes.

The presence of homogeneous regulations does not guaranteehomogeneous enforcement, however. A regulator can be captured bynarrow private, as well as by narrow state, interests.6 Nevertheless, onecan argue that heterogeneous ownership structures in which the pow-ers of various interest groups are more or less balanced would enhancethe independence and enforcement capacity of the regulator.7

One can argue that in heterogeneous and diverse ownership struc-tures the differences between dominant state and dominant private own-ership tend to be less relevant, and that in such ownership structuresregulation tends to assume greater importance relative to ownership.Our analysis in the preceding chapters lends empirical support to thisthesis. We found that heterogeneous ownership structures show simi-lar properties in terms both of their use of the new corporate gover-nance mechanisms and of economic performance. We found that outsideinvestors, whether state or private, tend to energize enterprises. We alsofound, however, that these new heterogeneous ownership structuresput new demands on the regulatory capacity of the state.

What are the implications of this discussion for the future courseof China’s ownership transformation? If we follow the logic of ourarguments, they do not imply a halt of China’s ownership reforms.What they do imply is the need for regulations and stronger regula-tory capacity. They also imply the need for a competitive environ-ment in which ownership structures evolve so that different interestsprovide checks and balances on each other and prevent too muchconcentration of economic and political power. A critical aspect ofsuch a competitive environment is the ability of outside investors toinfuse capital and energy into enterprises. Given the limited role thatprivate outside investors have played in the ownership reformprocess so far, the implication is that China needs more—not less—outside private ownership at this stage. In the remainder of the chap-


180

6. See Hu Angang on the subject of administrative monopoly and the issue of rentseeking by governments and SOEs in China (Hu and Yong 2003).7. An example is the Chinese companies listed on the domestic stock exchanges.Although in general listed companies in China are dominated by state shareholders,some of these companies have hundreds of thousands of small shareholders. Theexpectations and the concerns of these investors influence the regulator and inmany ways contribute to its resolve to improve the corporate governance of listedcompanies.

ter we will focus on these two aspects: (1) building a regulatoryframework and capacity; and (2) reducing the transaction costs foroutside investors.

Regulating Gaizhi

Gaizhi has so far been largely a bottom-up, spontaneous development.The decentralized, largely unregulated approach has infused tremen-dous innovation and dynamics into the process. It has also been asso-ciated with practices that have raised public concern about the fairnessand transparency of the process. This phase of the ownership transfor-mation process is coming to an end. SASAC has recently put out a flurryof regulations dealing with the transformation of SOEs.8 These regu-lations were partly in response to “observed irregularities” resulting inthe loss of state assets. Their goal is to achieve “standard and orderly”transformation.

The regulations clarify the methods that can be used to reformSOEs, such as reorganizations, mergers, leasing contract operations,joint ventures, transfer of state assets, and the joint-stock and joint-stock cooperative systems. They specify the process to be followed,including the preparation of restructuring plans, the appointment ofauditors and appraisers, the determining of the price of state assets, themethods of payment, and the need to protect the interests of creditorsand employees. Special attention is given to the aspects of the gaizhiprocess that are critical with respect to the protection of state assets,such as asset valuation and discounts on the transfer prices.

The regulations specify that the state property rights of nonlistedenterprises must be transferred on the property rights exchange with-out regard to the region, industry, investment, or affiliation of theenterprise. In addition, the transfer must be made publicly and com-petitively by such means as auctioning, public bidding, agreed trans-fer, or other means that are stipulated in state laws and regulations.


181

8. For example, Notice of the General Office of the State Council ForwardingOpinions of the State-Owned Assets Supervision and Administration Commissionof the State Council on Standardizing the Transformation of State-Owned Enter-prises, General Office of the State Council of the People’s Republic of China,November 30, 2003; and Measures for the Appraisal and Administration of State-Owned Assets (issued under the State Council Order No. 91), Decree of the State-Owned Assets Supervision and Administration Commission of the State Counciland the Ministry of Finance No. 3.

Importantly, where the state-owned assets and equity transaction priceis lower than 90 percent of the results of an asset valuation, approvalmust be obtained before the transaction can proceed. The proceduresalso detail the information required to be disclosed by the assignor, thequalifications required of the assignee, the main contents of the assign-ment contract, and the approval procedure.

A notable feature of these regulations is that they establish stricterstandards and a higher level of scrutiny when state assets are trans-ferred to nonstate entities. For example, according to the regulations,an enterprise should not be entrusted to develop a plan for restructur-ing when assets are being transferred to managers and employees of thesame enterprise. If the enterprise is to be transformed into a nonstateone, a terminating audit of its legal representative must be carried out.While the appointment of a qualified asset-appraising firm can nor-mally be done by the SOE, it is the organization that directly holds thestate property rights in the enterprise that should decide which assetappraiser to appoint if the enterprise transfers its state property rightsto nonstate investors. The regulations explicitly instruct that the baseprice of state property rights to be transferred to nonstate investors,and the price of the shares in existing state-owned assets offered tononstate investors, should be set by the organization that approved theSOE’s transformation and transfer of its state property rights.

The new regulations stress the need to protect the rights of bothcreditors and workers. The protection of workers’ rights is given higherspecificity, however, than the protection of creditor rights. Workers aregiven de facto veto power by specifying that the transformation maynot take place until after the plan for resettling workers is adopted atthe workers’ conference. The regulations require that an enterprise goingthrough gaizhi must obtain the consent of its creditors, and that SOEsmust be strictly prevented from making use of transformation to evademonetary liabilities.

The new regulations devote special attention to MBOs. They ad-dress conflict of interest issues by prohibiting the managers who arebuying the state-owned assets of their own firms from participating inkey decisions of the property transfer, such as those concerning finan-cial auditing, the terminating audit, asset and capital verification, assetsevaluation, and the setting of the base prices of property. The opinionsprohibit practices such as managers purchasing state assets with en-terprise funds, borrowing money from their own company or otherstate-controlled companies, or using their enterprise’s assets to provide


182

a guarantee, mortgage, hypothecation, or discount for their financing.Importantly, according to the opinions, managers responsible for thepoor performance of their company’s operation cannot buy their state-owned assets.

An important recent policy development is that the governmenthas ruled out the MBO of large state enterprises (see box 7.2). This re-action seems to be influenced by the concerns about erosion of state as-sets in MBOs. It probably also reflects concerns about concentration ofeconomic power in private hands that could subvert regulatory capacity.

These measures are sensible responses to the most common con-flict of interest situations that have been arising during gaizhi. They


183

BOX 7.2HAIER’S MBO PLANS PROMPT A BACKLASH

FROM THE GOVERNMENT

A plan by Haier, China’s leading whitegoods maker, to give its man-agers greater control of the company has sparked a backlash fromthe government. The government has announced as a general policythat MBOs of large enterprises will not be permitted. The reactiontriggered an intense debate among policy-makers about the course ofownership reform in China.

Haier is a more complicated case, as the company used to be acollective enterprise. Furthermore, its chairman, Zhang Ruimin, isone of China’s most well known entrepreneurs. The Qingdao branchof SASAC declared in 2003 that Haier and two other whitegoodscompanies based in the city—Hisense and Shuangxing—were ownedby the state. Haier’s restructuring hinges on a backdoor listing inHong Kong, which has been approved by shareholders, including theinjection into the listed vehicle of some of the Chinese parent com-pany’s most valuable assets. The listing is considered essential forHaier’s ambitious expansion plans, as the company will be able touse its Hong Kong–traded shares to buy overseas assets.

Some observers interpret the government’s reaction as interferencewith corporate autonomy and shareholders’ rights, as well as in con-tradiction with the government’s announcement that the state will re-tain control only over strategically important sectors. Others tend toview the government’s response as a legitimate exercise of ownershipfunctions by the main owner.

SOURCE: Financial Times, December 15, 2004.


184

will certainly contribute toward standardizing the transformationprocess and making it more transparent. At the same time the mea-sures discriminate against nonstate outside investors. The implicit as-sumption behind the new regulations is that the transfer of state assetsto nonstate-controlled entities is prone to more severe conflicts of in-terest and should therefore be subject to more scrutiny and bureau-cratic supervision. One can argue that erosion of state assets can occuralso in state-controlled enterprises with diversified ownership. In ad-dition, it has become increasingly difficult in China to distinguish be-tween various types of ownership. The hybrid enterprise combiningvarious types of ownership has become the representative Chinese en-terprise at this stage of China’s institutional transformation. Undersuch conditions it would be very difficult for the state to enforce theseregulations. A simpler and fairer solution would be to universalize theproposed measures to all types of gaizhi irrespective of the ownershiptype of the investor.

In addition to developing new regulations, the government hasstepped up enforcement. Since August 2004, SASAC has issued threecirculars on the reform of SOEs and transfer of property rights. Amongother things, these documents require local governments to carry out athorough inspection of all cases of reform and transfer since January2004. The inspection should cover whether the property right transfercases have been listed on the property right trade market, whether the op-erations are in accordance with official standards, whether the process isfair and transparent, whether the legal rights of employees have beeneffectively protected, and whether illegal operations have been cor-rected in a timely manner. SASAC also handed out a specific noticerequesting governmental agencies to compile more detailed statisticson the sales of state-owned property rights, so as to effectively monitorthe status of state-owned assets.

The central government’s strategy for SOE reform deals withother areas that are particularly prone to conflicts of interest and,consequently, erosion of state assets. At the national working con-ference of central SASAC in February 2004, the strategy for SOE re-form was communicated to local officials. Several new key directionswere announced. Parent companies of SOE groups are encouraged toprepare for IPOs wherever possible. The idea is to avoid a split of thegroup into a listed subsidiary and a nonlisted “shell,” as has beencommon in the past. In this way the parent company’s ownershipstructure will also be diversified and will subject the parent company

to additional scrutiny. “Policy-oriented” bankruptcy, which has gen-erated a lot of abuse related to evasion of debt payments, will bephased out within four years. In addition, managers of large SOEswill be subject to additional control mechanisms: boards of directorswill be created in all large central SOEs within three years, and, start-ing in 2004, a new regime of performance evaluation for managers ofcentral large SOEs will be implemented. Importantly, the new strate-gic directions specify that private firms are encouraged to participatein SOE restructuring.

Strengthening Enforcement

Regulations are hollow if they go unenforced. The dramatic increase ingaizhi regulations will need to be accompanied by a strengthening ofthe government’s enforcement capacity. In particular, the capacityof the central government needs to be strengthened. Many of the situ-ations that are subject to regulatory control involve conflicts of interestbetween the center and the localities. Yet the national agencies typicallyrely on local agencies for enforcement. As mentioned above, centralSASAC requires local governments to undertake inspections and to col-lect statistics on gaizhi. The large discrepancy between central govern-ment expenditures and central government revenues suggests that thisimplementation of national objectives through the local governmentsextends over most functions of the government (see figure 7.1). At pres-ent, only about 3 percent of China’s expenditures on governmentadministration, and only about 1 percent of the government’s expen-ditures on social relief and welfare, take place at the central level(Ahmad, Singh, and Fortuna 2004).

With the growth of uniform national regulations, China will needto establish a regulatory capacity at the local level that is independentof local governments. According to the vice chairman of SASAC, Li Yizhong, the establishment of local SASAC offices is the most im-portant and efficient way to prevent the loss of state assets, as it would“materialize the responsibility of the maintenance and appreciation ofstate assets” (Caijing 2003). The plan was to finish the establishmentof local SASACs in 2004.

Enforcement, and particularly fair enforcement, is expensive.With the rapid growth of regulations, China cannot avoid the growthof government expenditures for enforcing regulations. But the growthin regulation costs can be reduced by adopting a new approach to


185

regulation that utilizes and channels the energy of self-interest of theaffected parties.9 In the gaizhi process, there are stakeholders whoseinterests in particular situations can be aligned with the interests of thestate as an owner to prevent dissipation of state assets. By adding itsweight to such parties, the state can become more effective in enforce-ment and in protecting its interests.

It is not sufficiently recognized in China that the interests of credi-tors are often aligned with those of the owner (in this case, the state) inthe gaizhi process. For creditors, net worth is a cushion above their debtclaims. Hence, creditors, and particularly unsecured creditors, have aninterest in carefully monitoring and reacting to any erosion of thiscushion. Modern finance theory, as well as empirical evidence, suggestthat financial intermediaries such as commercial banks, insurance com-panies, and finance companies monitor and control their borrowers onbehalf of other investors.10 In this role they perform a public service.


186

0

10

20

30

40

50

60

1980 1985 1990 1995 2000

Share of centralgovernment intotal governmentexpenditures

Share of centralgovernment intotal governmentrevenues

Percent

FIGURE 7.1Central Government’s Share of Total Government

Revenues and Expenditures, 1978–2003

Sources: CEIC; authors’ calculations.

9. See Tenev, Zhang, and Brefort (2002, 132) for a short account of the approachtaken by the U.S. Securities and Exchange Commission in the early days of itsexistence.10. Rajan and Winton (1995).

Creditors are among the least utilized monitoring mechanisms inChina’s gaizhi process. They need to become an integral part of the re-structuring process. Banks, for instance, can assume a more importantrole in several ways. One vehicle is through the use of covenants in loanagreements. The standard credit contracts used in China include fewcovenants that permit real involvement. On such matters as majorfinancial or asset restructuring, ownership changes, or changes in busi-ness lines, at best they usually require “information” rather than con-sultation, let alone approval. The new regulations on gaizhi require“consent” by creditors. Banks should add specificity to this generalprovision through covenants in loan agreements that specify the processof bank approval for various restructuring plans of the borrower.

Another vehicle is the ability of banks to initiate and carry throughrestructuring of their borrowers. It has been difficult for Chinese banksto initiate restructuring or even to write off bad debts. Any write-off hasto follow stringent requirements from the Ministry of Finance. More re-cently, as part of the government effort to improve the asset quality of thebanks, the People’s Bank of China and MOF have issued new decrees toallow banks to increase write-offs with more flexible requirements.

Finally, the bankruptcy framework needs to be improved tostrengthen and clarify creditor rights. Currently there is no compre-hensive bankruptcy regime in China. The bankruptcy of SOEs followsa separate framework. The bankruptcy rules governing SOEs do notcontemplate reorganization by agreement or decree, either inside oroutside the bankruptcy proceedings. Most significantly, these rules aredrafted with a view to protecting workers’ rights11 and providing forthe extensive participation of local courts, and of regional and localpolitical authorities. A Bankruptcy Law is under preparation in China.The draft Bankruptcy Law covers private enterprises, and thereforemakes a step toward a comprehensive bankruptcy framework. Ac-cording to the draft, however, SOE bankruptcies would remain sub-ject to State Council regulations.

Similar to unsecured creditors, employees have a strong interest inmonitoring the net worth of the company. Employees have a compar-


187

11. Policy-makers in China are concerned that strengthening creditor rights maybe at the expense of labor. While there are situations where the interests of work-ers and creditors are in conflict, it should be noted that there is no inherent contra-diction between creditors and labor. In fact, their interests tend to coincide.Workers can be viewed as unsecured creditors who have an interest in moni-toring the financial situation of the company.

ative advantage in terms of access to privileged information and cantherefore play an important role in identifying and reporting irregular-ities in the gaizhi process. The role of employees can be enhanced byappropriate legislation protecting whistleblowers. China is probablythe only country in the world where the Constitution itself contains aprovision guaranteeing protection to whistleblowers. Article 41 of theConstitution confers the right on citizens to make “complaints andcharges against, or exposures of, violation of the law or dereliction ofduty by any state organ or functionary,” while prohibiting “fabricationor distortion of fact for purpose of libel or false incrimination.” Besidesempowering every citizen to blow the whistle on authorities, the Con-stitution says: “No one may suppress such complaints, charges andexposures, or retaliate against the citizens making them.”

Despite the constitutional provision, the rights of whistleblow-ers are far from being adequately protected in China. Their consti-tutional rights are not further developed and elaborated upon in aspecial legislative Act. Furthermore, article 41 talks about the rela-tionship between whistleblowers and the state but does not cover thelabor relationship in state and nonstate companies. The developmentof such legislation will be an important part of strengthening workers’rights.

The ability of the government to mobilize the energy of indepen-dent public monitoring is proportionate to the amount of informationavailable to the public. Disclosure of information by corporates and thegovernment is of key importance. While China has made significantprogress in the area of corporate disclosure (Tenev, Zhang, and Brefort2002), the movement toward an open government has been slower. TheChinese State Council is working on a draft of China’s first freedomof information legislation. However, some localities are coming outwith local legislation. In January 2004, Shanghai adopted China’s firstprovincial-level open information legislation (Horsley 2004). Earlier, in2003, Guangzhou, spurred by the SARS episode, issued the first Mu-nicipal Open Government Information Provisions.

Rigorous enforcement of rules and regulations begins with self-discipline by the regulator. One cannot expect a strong regulatorycapacity to develop in China if the government does not play by the rules.A particularly sensitive issue is that of managerial appointment in re-formed enterprises with significant state ownership, where practicesexist that are incompatible with the new corporate governance regimefor gaizhi enterprises.


188

It is still a widespread practice in China for the government andthe Party directly to appoint managers in gaizhi enterprises after by-passing the institutions of corporate governance, particularly estab-lished procedures by boards of directors. Arbitrary and unpredictable(from the point of view of shareholders) transfers of enterprise managersfrom and to the government are a common occurrence (see box 7.3).The implicit, or even explicit, benchmarking of enterprise managers togrades and levels in the civil service persists. If such practices continue,the new institutions of corporate governance will not be able to developas functioning organs of corporate control.

Some influential observers see the control by the Communist Partyover the appointment of SOE managers as one of the main outstandingproblems of SOE reforms in China (Qian and Wu 2003). According tothem, no meaningful corporate governance can evolve while the Partycontinues to have a role in the managerial appointments of large SOEs.They expect corporate governance reform to be the most difficult partof SOE reforms because of the political position of the CCP, and rec-ommend that the issue of the Party’s role be addressed before the goalof “separation of government and enterprise” can be achieved.

A more pragmatic approach, in our view, would be to address therole of the Party in managerial appointments through the corporategovernance framework. An owner in an enterprise has the right to so-licit recommendations for candidates for the board of directors. Thegovernment, as an owner, can solicit and get recommendations fromthe Party as to who can be nominated as a board member or chairman.Private investors with the right to nominate board members can do thesame—they can solicit third parties, including the Communist Party,for candidates. This is not critical. What is critically important is to gothrough the proper governance procedures for nominating and select-ing candidates for senior managerial positions. The focus on processwill provide a safeguard so that the relationship between the CommunistParty and the Chinese government does not interfere with the gover-nance practices set by companies’ shareholders.

To date, the government, as the largest shareholder in many enter-prises, has generally not adhered to good governance standards. It hasallowed the role of the Party to interfere with sound corporate gover-nance standards, since the government has allowed human resourcesdecisions to be made and sometimes implemented prior to the boardmeetings where a government-nominated candidate should be vettedand confirmed.


189

190

BOX 7.3LACK OF DEMARCATION BETWEEN GOVERNMENT OFFICIALS

AND BUSINESSMEN IN CHINA

In March 2004 the Chinese Communist Party’s Central Commissionfor Discipline Inspection and the Organization Department of theCCP Central Committee jointly issued instructions requiring all localgovernments and ministries to make thorough examinations on theissue of Party or government leaders concurrently taking jobs in com-panies. The instructions were triggered by a report of Xinhua NewsAgency, which disclosed that many officials in Wuhu city, Anhuiprovince were concurrently taking jobs in SOEs. The Wuhu city localgovernment reacted without delay, with government officials imme-diately resigning from their positions in the SOEs.

The Provisional Rules of State Public Servants, promulgated in1993, expressly prohibit public servants from engaging in businessactivities. However, the rules prohibit them from engaging in busi-ness activities only as individuals. They still leave open the possibil-ity of “official” employment or concurrent engagement in SOEs.

The practice of concurrent positions is due partly to the historicalpractice of assigning administrative rank to enterprises and linkingmanagerial positions with civil service grades. Some large SOEs es-tablished in the early years of New China held a higher status thanlocal governments. For example, Daqing city came into being afterthe establishment of Daqing Oil Administration. This has resultedin the general managers of some state-owned companies being con-currently the most senior leaders of local Party committees and localgovernments, or standing members of local Party committees anddecision-makers in local government. This is not uncommon. Infact, many state-owned companies set up by governments are led byone or more government-appointed officials who may work full-time or part-time for the companies.

That the practice of assigning administrative rank to enterprisesand linking managerial positions with civil service grades persists isapparent by the transfers between large SOEs and government agen-cies. Li Yizhong, former general manager of China PetrochemicalCorporation and president of Sinopec, has been appointed Partysecretary and vice chairman of the newly established State AssetSupervisory Commission. Wei Liucheng, former general managerof China Off-shore Petroleum is governor of Hainan province. Thusthere is still no clear demarcation between senior managers ofstate-owned companies and government officials.

SOURCE: Li Nan (2004).


The reality on the ground is that, given majority government con-trol, it is likely that government/Party–supported candidates would beelected in any case. Nevertheless, it is important for the government totake candidates to the board per the articles of association and not justaccept the recommendation from the Party as a fait accompli. As out-side shareholders’ shares in enterprises increase, it will be even moreimperative to ensure that the government/Party–recommended candi-dates come to the boards for appointments, since there will be morelikelihood that a block of nongovernment shareholders could object tosuch a candidate.

With the establishment of SASAC, the policy for reforming the per-sonnel system in state-owned companies and for establishing a moderncorporate management system has gained momentum. The central gov-ernment has announced its strategy to withdraw from direct economicactivities and to function mainly as a regulator of the macroeconomyand provider of public goods. According to the modern corporate sys-tem in state-owned companies, the state should act as a stockholder whosupervises the companies, instead of taking part in their day-to-day man-agement. The transfer between senior managers and government offi-cials tends to result in managers who give priority to political objectives.This practice inhibits the development of the managerial profession anda managerial market in China as well as undermines minority share-holders’ rights as laid out in good corporate governance practices.

Reducing Transaction Costs for Outside Investors

The Chinese government has recognized the importance of outside in-vestors in the restructuring of SOEs. A number of policy initiatives andregulatory measures reflect the drive to create a regime that facilitatesthe participation of outside investors in the ownership transformationprocess. For example, the Shanghai government launched its new roundof reforms of state assets management in March 2004 with the aim ofreorganizing the sectoral state holding companies set up some yearsago. The central objective is to create multiple shareholders in virtuallyevery SOE, even with state-owned shareholders. The intention reflectsthe accumulated wisdom that heterogeneous ownership provides notonly checks and balances, and hence facilitates monitoring, but alsoenergizes SOEs.

A host of new regulations aimed at developing a market for cor-porate control have come into force in China. In September 2002 the


191

CSRC promulgated the Measures to Administer the Takeover of ListedCompanies (the “Takeover Rules”) and the Rules for Management ofDisclosure Requirements for Changes for Controlling Shareholders ofListed Companies (the “Disclosure Requirements”). The TakeoverRules and Disclosure Requirements are modeled on similar takeoverregimes applicable in other jurisdictions. They are comprehensive andapply to a broad range of possible transactions and situations, manyof which are currently only theoretical in China—for example, hostiletakeovers.

The Takeover Rules and Disclosure Requirements are likely tohave an impact on the reform of SOEs, as more than 90 percent of alllisted companies are state-owned. Chinese companies listed on the do-mestic stock exchanges have two main classes of shares: nontradableor legal person shares, and tradable shares.12 The new regulations willapply to both, as shares of SOEs traded on the market typically makeup no more than 25–40 percent of a company’s outstanding shares andthe takeover threshold in the Takeover Rules is 30 percent. Accordingto the rules, if the acquisition agreement will result in the investor’s hold-ing 30 percent or more of the target company’s outstanding shares, anoffer must be made to acquire all the outstanding shares of the com-pany. The Takeover Rules set different price floors for the two classesof shares. For nontraded shares, the price should be no less than thehigher of (1) the highest price paid by the purchaser for other non-traded shares of the target company during the last six months, and(2) the most recently audited net asset value per share of the targetcompany. The price offered for traded shares must be no less than thehigher of (1) the highest price paid by the purchaser for such shares inthe previous six months, and (2) 90 percent of the arithmetic mean ofpublicly quoted sales of the stock over the previous 30 days.

A series of regulations were promulgated in China in 2002 and2003 to enhance the role of foreign investors in the reform of SOEs.In November 2002 the China Securities Regulatory Commission andthe People’s Bank of China jointly issued the Provisional Rules on theManagement of Investment in Domestic Securities by QualifiedForeign Institutional Investors (the “QFII Rules”). The QFII Rulesmake it possible for the first time for foreign institutional investors that


192

12. See Tenev, Zhang, and Brefort (2002) for a description of the Chinesestock market.

satisfy certain criteria to acquire A-shares as well as B-shares. Thereare significant restrictions on the size and liquidity of these invest-ments, however. Individual institutional investors may now acquire inthe market up to 10 percent of total outstanding shares (traded andnontraded), and qualified foreign institutional investors may acquire,in the aggregate, up to 20 percent. Foreign institutional investors ac-quiring such shares have restrictions on their access to foreign currencyfrom the proceeds of any sale of those shares for a period of one tothree years (depending on the type of investor). Because of these re-strictions, the QFII Rules alone offer only limited possibilities forM&A activity in China.

Until recently, nontraded legal person shares, which are typicallycontrolling shares, could be transferred between the state and Chinese“legal persons” but were off-limits to foreign buyers. Those ruleschanged in 2003 when the CSRC, the MOF, and the SETC jointly is-sued a Notice on Relevant Issues Concerning the Transfer to ForeignInvestors of Listed Company State-Owned Shares and Legal PersonShares (the “State-Owned Shares Notice”), effective January 1, 2003.The notice contemplates the direct sale to foreign investors of bothstate-owned and legal person shares, and offers the potential for foreigninvestors to acquire, for the first time, sole or shared control of China’spublicly listed SOEs. At about the same time, regulations for the ac-quisition of, and other investments in, nonlisted SOEs were announced:the Tentative Provisions on the Use of Foreign Investment to RestructureState-Owned Enterprises (the “SOE Restructuring Provisions”), whichthe former SETC, the MOF, the State Administration for Industry andCommerce (SAIC), and the State Administration of Foreign Exchange(SAFE) jointly promulgated in November 2002. The tentative provisionsappear to give the workers an effective veto over such restructuring trans-actions and, at a minimum, ensure that employee costs and terms will bea central feature of any privatization investment. The Interim Provisionson the Acquisition of Domestic Enterprises by Foreign Investors cameabout in April 2003. Many pre-existing rules and procedures for foreigninvestments in nonlisted Chinese domestic enterprises were consol-idated and clarified through the M&A Rules that were promulgated inApril 2004. Finally, the options and vehicles for restructuring of foreign-invested enterprises (FIEs) themselves were significantly expanded inChina by allowing two structures in which one FIE may hold an own-ership interest in another: the holding company and the limited liabil-ity company.


193

Partly as a result of these regulations, foreign acquisitions of localChinese companies surged in the first half of 2004. Following threeyears during which foreign acquisitions ran at about $5 billion peryear (10 percent of FDI), acquisitions nearly tripled in the first half of2004 to $7.3 billion and have likely reached 15–20 percent of FDI for2004. Acquisitions of local companies are fast becoming a third waveof FDI in China, following the prevalence of joint ventures in the1980s and early 1990s and the surge of wholly foreign-owned enter-prises during the past 10 years (Woodard and Wang 2004). Even withthis significant increase, however, M&As still account for a relativelysmall proportion of Chinese FDI. Globally, M&As account for morethan 50 percent of FDI.

Despite the issuance of regulations, M&As involving foreign in-vestors are proceeding more slowly than might have been expected. Thereason: an increasingly transparent legal system is needed for M&As toreduce risks and transaction costs. The regulations are a step forward,but not enough to tip the balance.

Impediments exist on both the demand and supply side of the pro-cess. Investors have difficulty finding independent and reliable financialinformation about the companies. Publicly available records on manyaspects of a Chinese company’s business, such as legal title to land-userights, the existence of pending litigation, and priority security interestsover assets, are often either unavailable or unreliable. Corporate ac-counting is also frequently lax, especially by foreign standards. AndChinese companies, particularly SOEs, are accustomed to a cultureof secrecy and may not be forthcoming in disclosing their records.13 As


194

13. Sometimes, even majority control does not guarantee access by the “control-ling” foreign party to a company’s records. In a recent case that went to theSupreme Court, the foreign party took a 51 percent stake with the right to appointthe president of the board of directors of the joint venture and five members of theboard. The domestic party held 49 percent of the company, with the right to rec-ommend the general manager of the joint venture and four members of the board.During operations, however, the foreign party complained that even though it wasthe controlling party it could not check various original certificates and financialaccounts, and it was deprived of management rights. Further, it stated that thegeneral manager of the joint venture also acted as the general manager of the do-mestic party, which is not in compliance with the law. This case illustrates the im-portance of drafting corporate charters and joint venture contracts in which thedivision of powers between the board of directors and the general manager isclearly specified.

a result, it can be difficult to conduct satisfactory due diligence. Underthese circumstances, most foreign investors will want comprehensiverepresentations and warranties, indemnities for breach, and security forthose indemnities. These arrangements are unfamiliar to many Chinesecompanies, and obtaining conditions that incorporate them is oftena challenge.

When investors spend a significant amount in due diligence, theyoften are scared of what they may find: companies with overdue tax andsocial security payments, with allocated instead of granted land-userights, or with no clear boundaries with their parent company or re-lated companies (Meyer and Lu 2005). Practices such as the arbitraryshifting of assets from one company to another by the supervisionauthority or parent company are widespread.14

In addition, numerous obstacles complicate debt financing forM&A transactions in China. Because the procedures for pledging equityinterests or registering security interests in assets are not fully developed,and because enforcement of such interests is difficult at best, banks areoften unwilling to loan funds for acquisitions. Further complicationsexist in the case of acquisitions by foreign investors. FIEs are most oftenthe acquisition vehicle but are subject to maximum leverage ratios thatcap their borrowing capacity. Acquisitions of Chinese companies byforeign investors are further restricted by prohibitions and limitationson foreign investments in specified economic sectors, and by uncertain


195

14. In a recent joint venture case from the Supreme People’s Court (SPC), thehigh-level administrative body that acted as the domestic party to a joint venturetransferred the shares belonging to the Chinese party to one of its subsidiaries bymeans of its administrative power. The court stated that such an administrativemethod of changing ownership is not in line with the law, and that the joint ven-ture partner has a priority right to buy the shares. Without a resolution from theother party of the joint venture and the board of directors, and without the req-uisite approval and registration procedure from the government authority, sucha decision did not have a legally binding effect. Furthermore, the SPC invalidatedan argument from the Chinese party that the applicability of a regulation(Determination of the Title to and the Handling of a Dispute of Ownership ofState-Owned Assets Tentative Provisions, issued in 1993) issued by the StateAssets Administration Bureau held in the case. The Chinese party argued that thetransfer involved state assets, and hence the ownership of the shares did notchange after being transferred under the joint venture laws. The SPC held thatthe Equity Joint Venture Law and its Im-plementing Regulations are the applic-able laws (Xu 2004).

prospects as to when and under what conditions legal person sharesacquired by foreign investors would become tradable.15

Disincentives exist on the supply side as well. Often potential take-over targets enjoy special tax and other advantages and subsidies that ajoint venture or a merger may jeopardize. As a rule, domestic partners/targets and local governments make adjustments so that governmentsupport flows to the domestic or local party only. This involves addi-tional transaction costs that may discourage M&As. Under the recentregulations, SOEs that transfer their shares—even controlling shares—to a foreign investor will not qualify as FIEs. This will make them theonly companies in China with more than a 25 percent foreign invest-ment that do not so qualify. As a practical matter, this will deprive suchinvestments of potentially preferential tax treatment and may serve as adisincentive to some transactions.

An important consideration for outside investors, particularly for-eign ones, is antidilution rights. Outside investors often seek the pro-tection of antidilution rights in order to maintain their proportionateinterest in target companies. The segmentation of China’s capital mar-kets with the presence of different classes of shares makes this difficult.In the case of foreign investors, they cannot subscribe to new issuancesof A-shares, for example, unless they are qualified foreign institutionalinvestors. Although the new regulations have given foreign investorsthe right to buy existing legal person shares, it is not clear whether theyhave the right to subscribe to legal person shares newly issued by pri-vate placement.

In general, the existence of different classes of shares presents amajor obstacle to the development of a market for corporate controlin China. It acts as an antitakeover device. Different classes of sharescreate conflicts of interest between different shareholders, particularlybetween the holders of legal person shares and the holders of tradableshares.16 This complicates the corporate governance of Chinese listedcompanies. The lack of tradability of legal person shares also inhibitsthe development of a private equity and venture capital market, as it


196

15. While there have been some recent examples of governmental approval forconversion of such shares to traded shares, there is no clear policy on this issue,and domestic market pressures will probably dissuade authorities from approvingmany conversions in the near future. This lack of a clear public exit may inhibitmany foreign investors.16. In general, holders of tradable shares value liquidity and capital gains. Holdersof legal person shares prefer dividend payments and value control.

limits exits for outside investors. After several attempts to find a solu-tion to the problem of the legal person shares overhang, the regulatorseems to be moving toward a decentralized approach, basically leav-ing it to the shareholders to decide. The CSRC regulations of December2004, which give minority shareholders the right to influence majorcorporate decisions, seem to have been designed partly with the objectiveof facilitating the process of case-by-case decisions on the tradability oflegal person shares. It remains to be seen how these regulations will workin practice, however.

Enhancing the Role of the de Novo Private Sector in China’s Transformation

As discussed in previous chapters, the domestic private sector is emerg-ing as a significant player in the gaizhi process.While most privateenterprises are small and will continue to rely on organic growth, agrowing number of de novo private firms are looking at acquisitionsof SOEs to speed up expansion. Large private conglomerates such asFosun High-Technology Group, New Hope Group, Wanxiang Group,and Orient Group have been aggressively buying stakes in state-ownedenterprises. Increasingly, such large private groups have been able toimplement their acquisition strategies across provincial boundaries,and some have moved overseas. Smaller private companies all overChina are also becoming more active. The Sichuan Sihai Group, forexample, with operations in the meat-processing business, has beenexpanding by acquiring bankrupt state-owned slaughterhouses andother businesses in a number of locations around Sichuan. These pri-vate companies have been injecting capital and dynamism in mori-bund state enterprises, thus helping to preserve jobs.17

While private enterprises are becoming more active in acquiringand restructuring state-owned enterprises, they still account for asmall share in all gaizhi cases. The policy of the government is to ex-pand the role of the private sector, both domestic and foreign, in the


197

17. In box 5.3 in chapter 5 we described the experience of Changdu Huarong,which was formed in 2000 when New Hope Group, one of the largest private groupsin China, acquired a former state company. The Sichuan Sihai Group, by purchas-ing bankrupt state-owned slaughterhouses, has not only preserved jobs in theseenterprises but has also provided income opportunities for approximately 300 small-scale pig breeders. (See IFC, 2004, “Scaling Up Private Sector Models for PovertyReduction: A Report on the Field Visit to Sichuan and Zhejiang Provinces, China.”)

restructuring and privatization of SOEs. This will require sustainedefforts from both the government and the private sector to improvethe business environment for private economic activities and moveChina’s private enterprise toward global best practice. China is mak-ing substantial progress in improving official attitudes toward theprivate sector.

The fourth revision of the 1982 Chinese Constitution from March2004 includes for the first time protection of private property and humanrights. Article 13 of the Constitution was revised to state that the law-ful private property of citizens should not be violated, and that the stateshould protect the private property and inheritance rights of citizens inaccordance with the law. The articles explicitly recognize the need forcompensation in case the state exercises its rights of eminent domain.The provision is a clear signal that the private sector is to gain moreofficial support—for example, in easier access to credit, which has longbeen a problem for private businesses. The amendment builds on formerPresident Jiang’s “theory of the three represents,” with its emphasison the representation of the interests of capital, and not only of work-ers and peasants. It reflects the new reality of China: a rapidly rising mid-dle class that wants protection of its property rights. Importantly, theamendments emphasize that “lawful” (meaning legally obtained) pri-vate property will be protected. It therefore can serve as a weapon againstcorruption and the use of state assets for personal enrichment.

The amendments might serve as a tool for private businesses topush for further reforms and protection of their position in society, asDonald Clarke points out (in Subrahmanyan 2004). They are expectedto trigger legislative activity in the area of private property rights to,for instance, define the government rights of eminent domain, therights of people to sue the government for taking, or inadequatelycompensating private entities for taking private property, and so on.“Lawful” does not mean much in the absence of a well-developed andconsistent legal framework for private property.

Even with a complete legal framework for private property, un-certainty may continue for private entrepreneurs who have accumu-lated their wealth in a fuzzy or nonexistent framework for privatebusiness. The fact of the matter is that it is perhaps difficult to find a pri-vate entrepreneur in China who can be certain that he or she has neverviolated the law in the past. The Chinese private sector has grown formost of its short history in an inhospitable environment. Private entre-preneurs did not enjoy a clear identity, their property rights lacked pro-


198

tection, and they had to function in an environment of significant legaland political uncertainties. Faced with a plethora of restrictions andbiases, they had to establish close links with the local bureaucracy andoperate under a high degree of informality (Gregory, Tenev, and Wagle2000). It is widely believed in China—as in all transition economies, forthat matter—that many successful private entrepreneur can be foundguilty of something, ranging from violating rules set up in the centralplanning context to tax evasion and bribery. Many private entrepre-neurs in China cannot explain clearly the origins of their accumulatedwealth. The situation creates uncertainty that, if not addressed, may in-hibit the healthy growth of the private sector.

In China the term “original sin” is used to refer to the lack of clar-ity in the ownership of property rights on assets acquired in the earlydays of capital accumulation. The issue of “original sin” has generateda lot of debate in China. The spectrum of opinions on how to deal withthe issue is broad, ranging from confiscation of “illegal” property toamnesty. Many societies have faced similar problems. The Americanexperience with the disposal of public lands during its westward expan-sion may be instructive in this regard (see box 7.4).

Some localities in China are beginning to develop their own solu-tions to this problem. For example, the issue of the “original sin” of pri-vate entrepreneurs became a subject of public attention in January 2004when the Political and Legal Committee of Hebei Party Committeeissued a document known as the “30-point document.” There are twokey articles of the “30-point document,” according to media reports:

• For criminal activities committed by private entrepreneurs inthe early years of their business creation, criminal litigation procedureshould not be launched if the legal period for prosecution has beenpassed.

• For those criminal activities that are still in the legal period forprosecution, the court should consider in a comprehensive mannerfactors such as the nature of the crime, its particular actions and con-sequences, the repentance shown by the entrepreneur, and the currentfinancial condition and future trend of his business, in decidingwhether to give him either no punishment, a reduced term of sentence,or probation.

This document has been widely taken as a signal that the Communistauthorities are offering “pardons” to private entrepreneurs. The inten-


199


200

BOX 7.4DEALING WITH “ORIGINAL SIN” IN AMERICA’S

GREAT PRIVATIZATION

America’s westward expansion followed a framework developed byThomas Jefferson in the great land ordinances of 1785 and 1787. Itwas a process of privatization of communal property. But the actualprocesses of disposal of land had very little to do with the originalplan. The government apparatus was too slow in setting up the sys-tem. The young American nation was impatient, and events movedtoo quickly. Thus, from the survey ordinance of 1785 on, squatterssettled large areas of public land in defiance of the law, ahead of theofficial survey, and without a title. But while they waited for thepublic sale day, these settlers all over the central and Midwesternstates set up local governments in the form of “claims associations,”elected officers with whom to record their land claims and fromwhom to obtain decisions on conflicts, and then generally abidedamong themselves by these records and decisions. Often unlawful inorigin, settlements nevertheless quickly produced an effective demandfor law (Hurst 1956). Except for Massachusetts, state governmentstended to deal generously with the squatters, valuing their pioneeringlabors. The federal government was far less generous, often sendingtroops to burn villages (Hughes and Cain 1998). The American legalinnovation for dealing with the issue was preemption—the right tobuy the land a settler had improved upon before it was offered forpublic sale. In 1830, Congress passed a broad preemption act to berenewed biennially. In 1841 a general preemption act was passed,limiting preemption rights to 160 acres. The principle of preemptionwould become the key to integration of extralegal property arrange-ments in American law (De Soto 2000). In the end the land wasoccupied in form as specified by the original plan. The process wasmessy and inefficient: buyers had to wait as long as five years afterpurchase to get a title, and land officers were commonly corrupt.But farms, cities, railroads, and a great civilization were built in theprocess.

tion of the “30-point document” was to set rules for the courts, theprosecution offices, and the policy force, which are all under the lead-ership of the Political and Legal Committee, to encourage them to helpcreate a favorable “investment climate” for the private sector. Legalprofessionals have criticized this document on the basis that thePolitical and Legal Committee has no power to make law.


201


The official attitude toward the private sector remains positiveand improving, but public criticism of the behavior of private busi-nessmen remains. This criticism is based on the perception that therules of the game do not apply for the rich and successful in China.With the growth of the private sector in China, more and more privateenterprises are adopting socially responsible practices. They put a pri-ority on workers’ welfare, engage in charitable activities, and buildstrong ties with the local communities (see box 7.5). However, occa-sional scandals involving private entrepreneurs tend to have a dispro-portionately strong impact on public perceptions about the privatesector in China. The collapse in 2004 of D’Long, allegedly China’slargest private company, revealed a tangled structure and a shadowymode of operations (Murphy 2004b). The episode, together with rev-elations of a number of scandals involving Chinese private companies,has damaged the reputation of China’s fledgling private sector. Thecombination of growing disparity of wealth and rising unemploymentis potentially an explosive one. Reflecting such social concerns, thenew Chinese government has put a renewed emphasis on social justice,and not only on runaway economic progress.

The Chinese private sector cannot afford to ignore this public criti-cism. The ongoing public debate in China should make private business-men more sensitive to public opinion and more responsive to the needsof the general population. At a poverty reduction conference in Shanghaiin May 2004, World Bank President James Wolfensohn bluntly warnedthat China’s growing income disparity could lead to a political explosion.He urged China’s nouveau riches to build a bridge to “the underclass”for the sake of their own long-term survival (Murphy 2004a).

Conclusion

As a result of gaizhi, a heterogeneous ownership structure involvingmultiple owners, often with diverging interests, has emerged in China.This ownership structure is creating demand for regulations while atthe same time enhancing the options available to the regulators. The re-sults of gaizhi so far indicate that outside investors, both state-ownedand private, have breathed new life into SOEs. While the successes ofgaizhi are significant, it is the “irregularities” that have attracted mostpublic attention. The Chinese public has been concerned that privatecompanies and SOE managers are enriching themselves at the expenseof the state and the society at large. The Chinese private sector, in


202

BOX 7.5DELIVERING MORE THAN PROFITS

Private companies in China are contributing to sustainable devel-opment by building good business practices, a sense of responsi-bility toward workers, and ties with the local communities.

New Hope Dairy, for example, is working in innovative wayswith farmers to improve the quality of its raw materials while in-creasing farmers’ income. It is providing free milk to schools andkindergartens, generating goodwill, increasing awareness aboutthe health and nutritional benefits of milk, and thus enhancing futuremilk consumption.

Shiyin Paper in the Sichuan capital of Chengdu profitably pro-duces environmentally friendly paper cups and bowls. The com-pany is making regular tax payments based on accounts that havebeen audited by an independent CPA firm.

Chint in Wenzhou has donated 60 million Rmb to the public wel-fare and called on 156 enterprises to form the first Aid-the-PoorSociety in China and raised Rmb 300 million for the cause. Ao Kangin Wenzhou is a producer of leather shoes, employing more than5,000 people. The enterprise has contributed about Rmb 23 millionsince 1993 for schooling, developing poor areas, and other causes.

Private enterprises increasingly emphasize workers’ welfare. DaHaoda in Wenzhou has been focusing on improving the workingenvironment. It has also invested 7 million Rmb to build betterfacilities in the headquarters of the company, which can hold 780employees. The company has purchased recreational and sportsequipment to enrich the leisure time of the staff, improving theirliving conditions and increasing cohesion within the enterprise.

Chint in Wenzhou spends Rmb 18 million per year to buy allthe staff insurance including insurance against injury suffered onthe job. The group has set up Chint College in collaboration withShanghai Technological and Polytechnic College to enhance thequality of the staff. It also dedicated a day to listening to concernsand suggestions from staff.

SOURCE: IFC, 2004. “Scaling Up Private Sector Models for Poverty Reduc-tion: A Report on the Field Visit to Sichuan and Zhejiang Provinces, China.”


203

particular, would need to become more sensitive to social expectationsabout how the private sector should conduct its affairs in China.

The policy response has been to introduce regulations on the gaizhiprocess and to strengthen their enforcement. A priority for the govern-ment is to enhance the role of the de novo private sector in the gaizhiprocess. This would require sustained efforts from both the governmentand the private sector to continue to improve the business environmentfor private economic activity and move private enterprise in China to-ward global best practice.

205

References

Ahmad, Entisham, Raju Singh, and Mario Fortuna. 2004. “TowardMore Effective Redistribution: Reform Options for IntergovernmentalTransfers in China.” IMF, Working Paper No. WP/04/98.

Anderson, James, Young Lee, and Peter Murrell. 2000. “Competitionand Privatization amidst Weak Institutions: Evidence from Mongolia.”Economic Inquiry 38(4): 527–46.

Barberis, Nicholas, Maxim Boycko, Andrei Shleifer, and NataliaTsukanova. 1996. “How Does Privatization Work? Evidence fromthe Russian Shops.” Journal of Political Economy 104(4): 764–90.

Blanchard, Oliver, and Andrei Shleifer. 2000. “Federalism with andwithout Political Centralization: China versus Russia.” HarvardInstitute of Economic Research Working Papers 1889. HarvardUniversity.

Boycko, Maxim, Andrei Shleifer, and Robert Vishny. 1996. PrivatizingRussia. Cambridge, Mass.: The MIT Press.

Caijing magazine. 2003. Official on State-Owned Assets, interview withLi Yizhong, Party Secretary and Vice Chairman of the State-OwnedAssets Supervision and Administration Commission of the State Coun-cil (SASAC). December 20. Available on: http://www.caijing.com.cn/emglish/2003/1220/1220official.htm.

CCP (Chinese Communist Party) Central Committee and the StateCouncil. 1998. “CCP Central Committee and the State Council Noticeon Mending Basic Living Security and Reemployment for SOE Laid-Off Workers.” CCP Central Committee Document No. 10.

Chan, Eric, and Sonia Wong. 2003. “Political Control and Performancein China’s Listed Firms.” University of Hong Kong, Working Paper

REFERENCES

206

No. 1067. Available on: http://www.hiebs.hku.hk/working_paper_updates/pdf/wp1067.pdf.

Chan, Vincent. 2002. “Private Sector in China.” UBS Warburg. April 8.

Chen, Kang, Arye Hillman, and Qingyang Gu. 2002. “Fiscal Re-centralization and Behavioral Change of Local Governments: Fromthe Helping Hand to the Grabbing Hand.” China Economic Quarterly2(1): 111–32.

China Law and Practice. 2004. “Breaking New Grounds in State-Owned Assets Transfers.” China Law and Practice March.

Chow, Gregory C. 2002. China’s Economic Transformation. Oxford:Blackwell Publishers Inc.

Churchill, Winston. 2003. “The Cause of the Left-Out Millions.” InNever Give In! The Best of Winston Churchill’s Speeches. New York:Hyperion.

Cooper, Mary. 2003. The Politics of China’s Shareholding System. AsiaPacific Research Center, Institute for International Studies, StanfordUniversity, Stanford, Calif. Available on: http://iis-db.stanford.edu/pubs/20265/Cooper_Shareholding.pdf.

Coricelli, Fabrizio, and Simeon Djankov. 2001. “Hardened Budgets andEnterprise Restructuring: Theory and an Application to Romania.”Journal of Comparative Economics 29(4): 749–63.

De Soto, Hernando. 2000. The Mystery of Capital: Why CapitalismTriumphs in the West and Fails Everywhere Else. New York: BasicBooks.

Demsetz, Harold, and Kenneth Lehn. 1985. “The Structure of CorporateOwnership: Causes and Consequences.” Journal of Political Economy93(6): 1155–77.

Djankov, Simeon, and Peter Murrell. 2002. “Enterprise Restructuring inTransition: A Quantitative Survey.” Journal of Economic Literature40(3): 739–92.

Dong, Keyong, and Xiangfeng Ye. 2003. “Social Security Reform inChina.” China Economic Review 14(4): 417–25.

Dougherty, Sean, and Robert McGuckin. 2004. “The Effects of Fed-eralism and Privatization on Productivity in Chinese Firms.” Eco-

REFERENCES

207

nomics Working Paper Archive at WUSLT, available on http://ideas.repec.org/pdo17.html.

Economic Observer, The. 2004. “Reform of State-Owned Enterprises:Clear Signals from Government.” The Economic Observer Septem-ber 20. Available on: http://www.eobserver.com.cn/english/readnews.asp?ID=246.

Economist, The. 2002. “China: The Withering Away of the Party.”The Economist May 30.

Financial Times, The. 2004. “Haier Sparks Beijing Backlash on StateAssets.” Financial Times December 15.

Fleisher, Belton, and Dennis Yang. 2003. “Labor Laws and Regulationsin China.” China Economic Review 14(4): 426–33.

Freedom House, China. 2003. Available on: http://freedomhouse.org/research/freeworld/2003/countryratings/china.htm.

Frydman, Roman, Cheryl Gray, Marek Hessel, and Andrzej Rapac-zynski. 1999. “When Does Privatization Work? The Impact ofPrivate Ownership on Corporate Performance in the TransitionEconomies.” Quarterly Journal of Economics 114(4): 1153–91.

———. 2000. “The Limits of Discipline: Ownership and Hard BudgetConstraints in the Transition Economies.” Economics of Transition8(3): 577–601.

Gao, Shanwen, and Yang Yao. 1999. “Implementation of SociallyOptimal Outcomes in the Process of Dissolving Public Enterprisesin China.” China Economic Review 10(2): 41–58.

Giles, John, Albert Park, and Fang Cai. 2003. “How Has EconomicRestructuring Affected China’s Urban Workers?” Ann Arbor, Mich.:University of Michigan.

Giles, John, Albert Park, and Juwei Zhang. 2004. “What is China’sTrue Unemployment Rate?” Ann Arbor, Mich.: University ofMichigan.

Glaeser, Edward, Simon Johnson, and Andrei Shleifer. 2001. “Coaseversus the Coasian.” Quarterly Journal of Economics 116(3): 853–99.

Goldman Sachs. 2004. “China Banks.” Goldman Sachs. Global Invest-ment Research, October 5. Thomson One Analytics.

REFERENCES

208

Gordon, Roger, and Wei Li. 1995. “The Change in Productivity ofChinese State Enterprises, 1983–1987.” Journal of ProductivityAnalysis 6(1): 5–26.

Gregory, Neil, Stoyan Tenev, and Dileep Wagle. 2000. China’s Emerg-ing Private Enterprises: Prospects for the New Century. Washington,D.C.: IFC.

Groves, Theodore, Yongmiao Hong, John McMillan, and BarryNaughton. 1994. “Autonomy and Incentives in Chinese State Enter-prises.” Quarterly Journal of Economics 109(1): 183–209.

———. 1995. “China’s Evolving Managerial Labor Market.” Journalof Political Economy 103(4): 873–92.

Guo, Kai. 2003. “Market Liberalization and Spontaneous Privatization.”Masters Thesis. CCER, Beijing University.

Guo, Kai, and Yang Yao. 2003. “Testing among Several Theories ofPrivatization in China.” Center for International Development,Working Paper, Stanford University, Stanford, Calif.

Hansmann, Henry. 1996. The Ownership of Enterprises. Cambridge,Mass: Belknap.

Havrylyshyn, Oleh, and Donal McGettigan. 1999. “Privatization inTransition Countries: A Sampling of the Literature.” IMF, WorkingPaper No. WP/99/6.

Horsley, Jamie. 2004. “Shanghai Advances the Cause of Open Govern-ment Information in China.” Processed. China Law Center, YaleLaw School. Available on: http://www.freedominfo.org/.

Hu, Angang, and Guo Yong. 2003. “Administrative Monopoly, Cor-ruption and China’s Economic Transformation.” In Pamela Mar andFrannk-Jurgen Richter, eds., China: Enabling a New Ear of Changes.Singapore: John Wiley & Sons (Asia) Pte Ltd.

Huang, Shao’an, and Jian Wei. 2001. “A Research Report on State-Owned SMEs in Zhucheng.” Processed.

Hughes, Jonathan, and Louis P. Cain. 1998. American EconomicHistory, Fifth Edition. Reading, Mass.: Addison-Wesley.

Hurst, James W. 1956. Law and the Conditions of Freedom in theNineteenth-Century United States. Madison, Wisc.: University ofWisconsin Press.

REFERENCES

209

IFC (International Finance Corporation). 2004. “Scaling Up PrivateSector Models for Poverty Reduction: A Report on the Field Visit toSichuan and Zhejiang Provinces, China.” Washington, D.C.: IFC.

IMF (International Monetary Fund). 2004. “People’s Republic of China:2004 Article IV Consultation-Staff Report.” IMF, Washington, D.C.

Kornai, Janos. 1980. Economics of Shortage. Amsterdam and NewYork: North-Holland Publishing Co.

———. 1986. “The Soft Budget Constraint.” Kyklos 39(1): 3–30.

Li, David, and Francis Liu. 2004. “Why Do Governments DumpState Enterprises? Evidence from China.” In Takatoshi Ito andAnne Krueger, eds., Governance, Regulation and Privatization.East Asia Seminar on Economics Vol. 12. Chicago: University ofChicago Press.

Li, Hongbin, and Scott Rozelle. 2003. “Privatizing Rural China:Insider Privatization, Innovative Contracts, and the Performance ofTownship Enterprises.” The China Quarterly: 982–1004.

Li, Lixing. 2004. “Firms’ Social Responsibility and PerformanceChange in Ownership Reform: Evidence from Chinese Enterprises.”Processed. University of Maryland, College Park.

Li, Nan. 2004. “How to Distinguish Government Officials from Busi-nessmen?” The Economic Observer, March 29. Available on: http://www.eobserver.com.cn/english/readnews.asp?ID=160.

Li, Shaomin, Shuhe Li, and Weiying Zhang. 2000. “The Road to Capi-talism: Competition and Institutional Change in China.” Journal ofComparative Economics 28(2): 269–92.

Li, Wei. 1997. “The Impact of Economic Reform on the Performance ofChinese State Enterprises, 1980–1989.” Journal of Political Economy105(5): 1080–106.

Lin, Justin-Yifu, Fang Cai, and Zhou Li. 1995. Full Information andSOE Reform. Shanghai: Shanghai Renmin Press.

———. 1996. Sufficient Information and SOE Reform in China,Shanghai: Shanghai Sanlian Press.

———. 1998. “Competition, Policy Burdens, and State-Owned Enter-prise Reform.” American Economic Review 88(2): 422–7.

REFERENCES

210

Lin, Justin, and Guofu Tan. 1999. “Policy Burdens, Accountability andthe Sort Budget Constraint.” American Economic Review 89(2):426–31.

Lu, Susan, and Yang Yao. 2003. “The Effectiveness of the Law, FinancialDevelopment, and Economic Growth in an Economy of FinancialRepression: Evidence from China.” Center for Inter-national Devel-opment, Working Paper, Stanford University, Stanford, Calif.

McFaul, Michael. 1995. “State Power, Institutional Change and thePolitics of Privatization in Russia.” World Politics 47(2): 210–43.

McGregor, Richard. 2004. “Haier Sparks Beijing Backlash on StateAssets.” Financial Times December 14.

McNally, Christopher. 2002. “Strange Bedfellows: Communist PartyInstitutions and New Governance Mechanisms in Chinese StateHolding Corporations.” Business and Politics 4(1): 91–109.

Marquand, Rober. 2004. “Booming China Raises Workers’ Hopes—and Discontent: Low Pay and Poor Conditions Drove Employees atTwo Shoe Plants in the Pearl River Delta to Riot.” The ChristianScience Monitor September 30.

Megginson, William, and Jeffrey Netter. 2001. “From State to Markets:A Survey of Empirical Studies on Privatization.” Journal of EconomicLiterature XXXIX (June): 321–89.

Meyer, Marshall W., and Xiaohui Lu. 2005. “Managing IndefiniteBoundaries: The Strategy and Structure of a Chinese Business Firm.”Management and Organization Review 1(1) (March): 57–86. Avail-able on: http://ssrn.com/abstract=584216.

MLSS (Ministry of Labor and Social Security). 2001. “Essentials of theTenth Five-Year Plan for Development of Labor and Social SecurityProjects.” April 24.

MLSS (Ministry of Labor and Social Security), State Committee forEconomics and Trade, Ministry of Finance, Ministry of Education,National Bureau of Statistics, and All China General Trade Union.1998. “Notice Relating to Enhancing Administration of SOE Lay-Off Employee and Establishing Reemployment Service Centres.”Ministry of Labor and Social Security Document No. 8.

REFERENCES

211

Murphy, David. 2004a. “The Dangers of Too Much Success.” FarEastern Economic Review 167(23) (June 10): 28–32.

———. 2004b. “Entrepreneurship Gets a Bad Name.” Far EasternEconomic Review 167(35) ( June 3): 24–28.

National Bureau of Statistics. Various years. China Statistical Yearbook.Beijing: China Statistics Press.

North, Douglass, and Barry Weingast. 1989. “Constitutions andCommitment: The Evolution of Institutions Governing PublicChoice in Seventeenth-Century England.” Journal of EconomicHistory 49(4).

Norton, Patrick, and Howard Chao. 2003. “Mergers and Acquisitionsin China.” An O’Meleveny & Myers LLP Research Report, April.

OECD (Organisation for Economic Cooperation and Development).1993. Valuation and Privatization. Centre for Cooperation with theEuropean Economies in Transition, OECD.

Oxford Analytica. 2002. “China: Bad Debt Resolution Offers Op-portunities and Risks.” Available on: http://www.oxanstore.com/displayfree.php?NewsItemID=86516.

People’s Daily. 2001. “CPC Branches in Private Enterprises.” People’sDaily July 3 and various issues.

Qian, Yingyi. 1994. “A Theory of Shortage in Socialist EconomiesBased on the Soft Budget Constraint.” American Economic Review84(1): 145–56.

———. 1995. “Reforming Corporate Governance and Finance inChina.” In Masahiko Aoki and Hyung-Ki Kim, eds., CorporateGovernance in Transition Economies: Insider Control and the Roleof Banks. Washington, D.C.: The World Bank.

———. 2000. “The Institutional Foundations of China’s Market Tran-sition.” In Boris Pleskovic and Joseph Stiglitz, eds., Annual WorldBank Conference on Development Economics 1999. Washington,D.C.: The World Bank.

Qian, Yingyi, and Joseph Stiglitz. 1994. “Institutional Innovations andthe Role of Local Government in Transition Economies: The Caseof Guangdong Province of China.” In John McMillan and Barry

REFERENCES

212

Naughton, eds., Reforming Asian Socialism: Evolving MarketInstitutions. Mich.: University of Michigan Press.

Qian, Yingyi, and Jinglian Wu. 2003. “China’s Transition to a MarketEconomy: How Far across the River?” In Nicholas C. Hope, DennisTao Yang, and Mu Yang Li, eds., How Far across the River?:Chinese Policy Reform at the Millennium. Stanford, Calif.: StanfordUniversity Press.

Qin, Hui. 1997. Case Studies of TVE Gaizhi in Jiangsu and Zhejiang(in Chinese). Hong Kong: Hong Kong Chinese University Press.

Rajan, Raghuram, and Andrew Winton. 1995. “Covenants and Collat-eral as Incentives to Monitor.” University of Chicago, Processed.Available on: http://gsbwww.uchicago.edu/fac/raghuram.rajan/research/cov1.pdf.

Roland, Gerard. 2002. “The Political Economy of Transition.” Journalof Economic Perspectives 16(1): 29–50.

Shleifer, A., and L. Summers. 1988. “Breach of Trust in Hostile Take-overs.” In A. J. Auerbach, ed., Corporate Takeovers: Causes andConsequences. Chicago: University of Chicago Press.

Song, Ligang, and Yang Yao. 2004. “Impacts of Privatization onFirms’ Performance in China: China Center for Economic Research,Working Paper No. E2004005, Beijing University, Beijing.

Sonobe, Tetsushi, and Keijiro Otsuka. 2004. “Productivity Effects ofTVE Privatization: The Case Study of Garment and Metal CastingEnterprises in the Greater Yangtze River Region.” In Taketoshi Ito andAnne O. Krueger, eds., Governance, Regulation and Privatization.East Asia Seminar on Economics Vol. 12. Chicago: University ofChicago Press.

State Council. 1991. “State Council Decision on Reforming EnterpriseEmployee Pension System.” No. 33, June 26.

———. 1993a. “Regulations on Job-Waiting Security for Employeesof State-Owned Enterprises.” April 12.

———. 1993b. “Regulations on Redeployment of Redundant Em-ployees in State-Owned Enterprises.” April 20.

———. 1995. “State Council Notice on Deepening Reform of Enter-prise Employee Pension System.” No. 6, March 1.

REFERENCES

213

———. 1998. “State Council Decision on Establishing Urban EmployeeBasic Medical Insurance System.” No. 44, December 14.

———. 1999a. “Bylaw on Unemployment Insurance.” January 22.

———. 1999b. “Bylaw on Urban Residents’ Minimal Living Safe-guard.” September 28.

State Council Office. 1999. “State Council Office Notice Relating toFurther Improving the Distribution of Lay-Off Basic Living Allowanceand Pension.” No. 10, March 3.

Steinfeld, Edward. 2001. “China’s Program of Debt-Equity Swaps:Government Failure or Market Failure?” Paper presented at theconference on Financial Sector Reform in China, John F. KennedySchool of Government, Harvard University, September 11–13.

Sterling, Robert R., ed. 1971. Asset Valuation and Income Determination:A Consideration of the Alternatives. Houston: Scholars Book Co.

Su, Jian, and Garry Jefferson. 2003. “Privatization and Efficiency:Theory and Evidence from China.” Working Paper, Department ofEconomics, Brandeis University.

Subrahmanyan, Arjun. 2004. “Constitutionalism in China: ChangingDynamics in Legal and Political Debates.” China Law and Prac-tice May.

Sun, Qian, Wilson Tong, and Jing Tong. 2002. “How Does GovernmentOwnership Affect Firm Performance? Evidence from China’s Priva-tization Experience.” Journal of Business Finance and Accounting29(1): 1–27.

Svejnar, Jan. 2002. “Transition Economies: Performance and Chal-lenges.” Journal of Economic Perspectives 16(1): 3–28.

Tenev, Stoyan, Chunlin Zhang, and Loup Brefort. 2002. CorporateGovernance and Enterprise Reform in China. Washington, D.C.:World Bank and International Finance Corporation.

Tian, Guoqiang. 2001. “A Theory of Ownership Arrangements andSmooth Transition to a Free Market Economy.” Journal of Insti-tutional and Theoretical Economics 157(3): 380–412.

Varian, H. R. 1989. “Monitoring Agents with Other Agents.” Papers 89-18, Michigan—Center for Research on Economic & Social Theory.

REFERENCES

214

Wang, Shaoguang, and Youqiang Wang. 2001. “Citizenship, In-come Tax, and Fiscal System” (in Chinese), Strategy and Manage-ment 3: 1–14.

Wang, Yueping, and Yang Yao. 2002. “Market Reforms, TechnologicalCapabilities and the Performance of Small Enterprises in China.”Small Business Economics 18(1–3): 197–211.

Weitzman, Martin, and Chenggang Xu. 1994. “Chinese TownshipVillage Enterprises as Vaguely Defined Cooperatives.” Journal ofComparative Economics 18: 121–45.

Wen, Mei. 2002. “Theories of Privatization and Empirical Analysis.”In Tian Guoqiang, ed., Frontiers of Modern Economics and FinancialEconomics. Beijing: Commercial Publisher.

Wolf, Holger C. 1999. “Transition Strategies: Choices and Outcomes.”Princeton Studies in International Finance 85 (June).

Wolf, Martin. 2003. “China’s Awakening.” Financial TimesSeptember 21.

Woodard, Kim, and Anita Qingli Wang. 2004. “Acquisitions in China:A View of the Field.” The China Business Review 31(6) (November/December): 34–39.

World Bank. 2003. “China: Promoting Growth with Equity.” CountryEconomic Memorandum. Washington, D.C.

World Bank Group. 2004. Doing Business in 2005. London andNew York: Oxford University Press.

Xu, Bailey. 2004. “From RMB 42 Million to Nothing: What LessonsCan We Learn from JV Disputes?” China Law and Practice April.

Xu, Lixin, Tian Zhu, and Yimin Lin. 2002. “Political Control, AgencyProblems, and Ownership Reform: Evidence from China.” Processed.

Yao, Yang. 2003. “Small Privatization.” In Ross Garnaut and LigangSong, eds. China’s Third Economic Transformation: The Rise of thePrivate Economy. London: Routledge.

———. 2004. “Government Commitment and the Outcome ofPrivatization in China.” In Taketoshi Ito and Anne Krueger, eds.,Governance, Regulation and Privatization, East Asia Seminar onEconomics Vol. 12. Chicago: University of Chicago Press.

REFERENCES

215

Yao, Yang, and Yang Lei. 2003. “Distorted Institutional Supply andthe Consequences of the Chinese Fiscal Decentralization.” Strategyand Management 3: 27–33.

Zhang, Gang. 1998. “A Study on Township and Village Enterprises.”Ph.D. Thesis. Stockholm: Stockholm School of Economics.

Zhang, Shouqi. 2004. “Injecting Funds into Individual Accounts:Measure to Cope with the Financing Challenge Facing China.” Inter-national Social Security Association. Thirteenth Regional Conferencefor Asia and the Pacific, Kuwait, March 8–10.

Zhao, Xiao. 1999. “Competition, Public Choice and InstitutionalChange.” CCER, Working Paper No. C1999025, Beijing University.

Index

217217

absentee ownership, 179acquisitions. See mergers and

acquisitionsagency costs, 113, 130, 176Agricultural Bank of China, 6nAid-the-Poor Society, 202allocated land-use rights, 81–82,

83, 84, 85AMCs, 6, 48, 79, 80–81antidilution rights, 196Ao Kang, 202appropriation, of land-use rights,

81–82, 83asset management companies, 6,

48, 79, 80–81assets. See state assetsassets stripping, 109, 176–77auctioning, of state firms, 5, 65nauditing, 181, 182Australia, 94, 95

bad debts, writing off, 187Bank of China, 6nbankruptcies

abuse of to avoid debt, 47–48,54, 56, 65, 80, 177, 185

bankruptcy provisional funds, 66in first round of gaizhi, 56as a form of gaizhi, xi, 5, 22–23,

24, 26–28, 46–48, 50, 60lack of comprehensive regime

for, 187

in most recent round of gaizhi,57, 57, 59

priority given to workers’ claimsin, 35

in sample cities, 54, 58in sample distribution, 51, 51

Bankruptcy Law, 47, 54, 187banks

abuse of bankruptcies, 48credit for private enterprise, 163as creditors to the SOE sector,

32, 35–36debt recovery by, 77, 80embezzlement scandals, 176evasion of payment to, 157monitoring of borrowers, 186role in restructuring process, 187socialization of losses through, 37unwilling to lend funds for acqui-

sitions, 195See also loans

benchmarking, of enterprise man-agers, 189

benchmarking, of managerialsalaries, 137

boards of directorschairman of the board, 122, 124,

133, 134–35division of powers between man-

ager and, 194ninfluence of the Party on, 130,

189, 191

Page numbers in italics refer to figures. Page numbers followed by an “n”refer to footnotes.

boards of directors (continued)influence on decision-making,

140, 141management on, xiii, 122, 123,

134–35, 139, 144method of electing, 122, 123in new governance structures,

121–27, 138–39perceived importance of, 127regulation of, 185reorganization of, 146shareholder representation on,

xiiiboards of supervisors, 121, 126–27bonds, 158bonuses, xiii, 108–9, 109, 136,

136–37brand names, 70Brefort, Loup, 116budget constraints, effect of gaizhi

on, 147, 155–57Bureau of National Land Resource

Management, 82Bylaw on Unemployment

Insurance, 88Bylaw on Urban Residents’

Minimal Living Safeguard, 89

Cai, Fang, 106capital intensity, 166casual workers, 102, 102central government

assets of, 63nconcurrent officials and business-

men, 190conflict of interest with local gov-

ernments, 32decreased decision-making after

gaizhi, 138encourages outside investment,

191enforcement of regulations, 185in fiscal decentralization, 41

INDEX

218

funding of redeployment centers,92

influence on decision-making,140, 140

influence on managerial appoint-ments, 128, 128–29, 130,131–32, 139, 188–89

introduces debt-for-equity swaps,48

level of subsidies for local gov-ernments, 67

regulations on debt liabilities, 81role in gaizhi, 31–33, 37, 38, 39share of total government rev-

enues and expenditures, 185,186

unemployment policy, 89–93vetoes MBOs of large enterprises,

183warning on financial position of

local governments, 156central planning, 41CEOs, 135chairman of board, 122, 124, 133,

134–35Chan, Eric, 130Changsha, 81charities, 202Chengdu

bonuses for managers, 136choice of valuation firm, 74compensation of workers, 103economic indicators, 16employee shareholding, 57interviews in, 18location, 12, 12management shareholdings, 117,

117method of obtaining land-use

rights, 83open sales, 57ownership and control struc-

tures, 124

218

per capita GDP growth, 13, 15per capita government revenue,

14per capita income, 14population, 13private sector, 53privatization stalemate, 121privatized SOEs, 28redeployment of redundant

workers, 106as a sample city, 11settling of overdue payments,

105, 106state-owned enterprises, 29, 29types of gaizhi, 58unemployment, 17xiagang workers in, 15

Chengdu Huarong Chemicals, 142China Construction Bank, 6nChina Enterprise Index, 163China Labor Urban Survey, 106China Mainland Index, 163China Private Enterprise Index,

159, 163China Securities Regulatory

Commission, 6, 127, 192, 193,197

Chinese Communist Party14th Party Congress, 3115th Party Congress, 31, 36adopts enterprise reforms, 3ninfluence on boards of directors,

130, 189, 191influence on decision-making,

138, 140, 141, 144influence on managerial appoint-

ments, 130, 131–32, 189orders officials to resign business

positions, 190role in corporate governance,

128, 130–33, 139Chinese State Council, 188Chint, 202

INDEX

219

Chongqing, 18, 25, 80–81, 108Cinda, 79Circular on Issues Concerning the

Transfer of State-Owned Shares,119

Clarke, Donald, 198Code of Corporate Governance,

127collateral, 67, 82collectives, 49, 183Company Law, 46, 49, 126–27,

130, 137, 150compensation, 49, 78, 81, 91,

93–94, 103–8competition, 6Concurrent System, 133Constitution of China, 31, 188,

198cooperatives, 51ncorporate accounting, 194corporate control, 191–92corporate disclosure, 188, 192corporate governance

bypassing of by government andParty, 189

changes in influence of stake-holders, 138–40, 140

Code of Corporate Governance,127

coexistence of old and new prac-tices, 127–33

CSRC as regulator, 6effects of classes of shares on,

196impact of gaizhi on, 113–44new governance structures,

121–27, 123–24, 126–27as a restructuring mechanism, 5role of outside investors, 139,

141–43corporatization, 6–7corruption, 72, 175cost approach, in valuation, 69–70

credit contracts, 187creditors, in the gaizhi process,

35–36, 182, 186–87credit ratings, 81criminal activities, 199cross-shareholding, 119CSRC, 6, 127, 192, 193, 197

Da Haoda, 202danvei, 17Daqing, 190debt avoidance, 47–48, 54, 56, 65,

80, 177, 185debt-equity ratio, 61, 165, 167debt financing, 195debt-for-equity swaps, xi, 5, 46, 48,

51, 79debt problems, 3, 35–36, 76–81,

158decision-making, 122, 133Decree No. 4, 85ndeflation, 69delistings, xi, 5Deng Xiaoping, 3, 31depreciation, 69Development Research Center, 158disclosure, 188, 192, 194Disclosure Requirements, 192dismissal. See firing employees, cost

ofdividends, 108–9, 109Djankov, Simeon, 169D’Long, 201Doing Business in 2005 project, 94Dougherty, Sean, 40downsizing, 34–35

early retirements, 91, 93efficiency, of firms, xii, 39–40, 43,

44, 144, 166, 169, 175efficiency hypothesis, 39–40, 43, 44embezzlement scandals, 176

INDEX

220

employeesmedical bills, 91, 93monitoring of company net

worth by, 186–87role in gaizhi, 34–35, 37–38, 38,

39veto over restructuring, 193See also compensation; employee

shareholdings; redundancyemployee shareholdings

choice of valuer in, 73compared to management share-

holdings, 116, 116–17decline of popularity of, 54effect on efficiency, 169effect on labor productivity, 168effect on unit cost, 168in first round of gaizhi, 56as a form of gaizhi, 22–23, 24,

46, 49, 59, 60, 62geographic variations in survey,

57increase in, 144in lieu of wages or benefits, 65in most recent round of gaizhi,

57, 59obtaining land-use rights, 84overdue interest rates after, 151overdue loans after, 149, 150overdue social security payments

after, 153–54, 154, 155overdue taxes after, 153rate of employee discharge, 98relation to workforce growth,

101in sample cities, 53–54, 58in sample distribution, 50–51,

51, 52shares as compensation for

workers, 103stalemate in Chengdu, 121in TCL Group, 138

wage growth rates under, 111employment

contract status of SOE workers,93n, 103

fall in for SOE sector, 89–90rigidity of, 94–95statistics for sample cities, 16See also labor; redeployment; un-

employmentemployment regulation index, 94enforcement, of regulations, 184,

185–86enterprise-based social security sys-

tem, 87enterprise debts, dealing with, 3,

35–36, 76–81equity injections, 119exit funds, 66

farmers, social security for, 87nFDIs, 1, 118–19, 119, 194female workers, 102FIEs, 193, 195–96finance companies, 186financial crisis hypothesis, 40, 43,

44financial discipline, 147, 155–56,

157, 168financial intermediaries, 186firing employees, cost of, 95firm performance, xii–xiii, 158–59,

160–62, 163–74firm size, 61–62, 75, 76, 157, 165,

167, 172, 174fiscal decentralization hypothesis,

40–41foreign direct investments, 1,

118–19, 119, 194foreign-invested enterprises, 193,

195–96foreign investment, in SOEs, 6,

117–20, 118–19, 193–96

INDEX

221

Fosun High-Technology Group,197

Freedom House, 130freedom of information, 188Frydman, Roman, 164, 170full-time workers, 102, 102Fushun

choice of valuation firm, 74economic indicators, 16internal restructuring, 57interviews in, 18location, 12, 12management shareholdings, 117,


rights, 83ownership and control struc-

tures, 124per capita GDP growth, 13, 15per capita government revenue,

14per capita income, 13, 14population, 13redeployment of redundant

workers, 106response to survey, 18as a sample city, 11settling of overdue payments,

105, 106state-owned enterprises, 29, 29types of gaizhi, 58unemployment, 17xiagang workers, 15

gaizhichanges in influence of stake-

holders, 138–40, 140in the cities, 4–5coexistence with old practices,

127–33debt problems, 35–36, 76–81decentralization in, 11

gaizhi (continued)efficiency of gaizhi firms, 144,

175empirical tests for, 42–45factors affecting choice of form,

59–62forms of, 22–23, 46–50, 55, 56,

57gaizhi firms contrasted with non-

gaizhi firms, 129geographic variations in survey,

56–57governance structures, 121–27,

123–24, 126–27impact on corporate governance,

113–44impact on efficiency, xiiimpact on profitability, xiiimpact on workforce structure,

102, 102–3in the industrial sector, xikey participants in, 31–38land-use rights, 81–86managerial autonomy under,

133–35meaning of, xi, 1measurement of, 42obligations of firms to workers,

103–8outside investment in, 36–37, 48,

50, 53, 139, 141–44, 180progress of process of, 25–31public debate on, 36–37, 175–81rate of employee discharge,

96–98, 97–99role of central government,

31–33, 37, 38–39, 128,128–29

role of employees, 34–35, 37–38,38–39

role of labor unions, 129, 129role of local governments, xii,

xiv, 32, 33–34, 36, 37, 38–39

INDEX

222

role of management, xiii, 34,38–39

role of the Party, 130–33sample distribution of forms of,

50–54, 51–52theories on incentives for, 38–45typical gaizhi proposal, 64undertaking of by sample cities,

30workforce growth rate, 101,

101–2See also bankruptcies; employee

shareholdings; firm perfor-mance; internal restructuring;labor, impact of gaizhi on;leasing; open sales, of firms;ownership; privatization; pub-lic offerings; remuneration ofworkers; restructuring; stateassets; survey, of gaizhi

Giles, John, 90, 106Glaeser, Edward, 39go-slows, 7government. See central govern-

ment; local governmentsgovernment officials, as business-

men, 190granted land-use rights, 81–82, 85Guangdong province, 66Guangzhou, 188guarantees, 67Guiyang

bonuses for managers, 136, 136choice of valuation firm, 74compensation of redundant

workers, 108discharge and rehiring of labor,

104economic indicators, 16interviews in, 18location, 12, 12management shareholdings, 117,

117

method of obtaining land-userights, 83

ownership and control struc-tures, 124


14per capita income, 13, 14population, 13private sector, 53redeployment of redundant

workers, 106, 107as a sample city, 11settling of overdue payments,

105state-owned enterprises, 29, 29types of gaizhi, 58

Guo, Kai, 41

Haier, 183Hang Seng Index, 159, 163Harbin

bonuses for managers, 136choice of valuation firm, 74compensation of workers, 103,

108economic indicators, 16internal restructuring, 56interviews in, 18location, 12, 12management shareholdings, 117method of obtaining land-use



14per capita income, 13, 14population, 13privatized SOEs, 28redeployment of redundant

workers, 106

INDEX

223

sale of state assets, 73as a sample city, 11settling of overdue payments,

105SMEs in, 79state-owned enterprises, 29, 29types of gaizhi, 58unemployment, 17xiagang workers, 15

Hengdong Ceramic Factory, 149Hengyang

bonuses for managers, 136choice of valuation firm, 74compensation of workers, 103economic indicators, 16interviews in, 18land-use rights, 84leasing, 57location, 12, 13management shareholdings, 117method of obtaining land-use



14per capita income, 13, 14population, 13private sector, 53privatized SOEs, 28redeployment of redundant


105state-owned enterprises, 29, 29types of gaizhi, 58

Hisense, 183holding companies, 47, 193Hong Kong, 94Huangshi

bonuses for managers, 136

Huangshi (continued)choice of valuation firm, 74economic indicators, 16employee shareholding, 57location, 12, 13management shareholdings, 117method of obtaining land-use

rights, 83open sales, 57ownership and control struc-


14per capita income, 13, 14population, 13privatized SOEs, 28, 28redeployment of redundant


105, 106state-owned enterprises, 29, 29types of gaizhi, 58

human rights, 31hybrid firms, 178, 179, 184

illegal fundraising, 164incentives, 38–45, 109, 113,

135–38, 169–70, 178income approach, in valuation, 69incorporation, 2, 46–47, 51, 100,

122Indonesia, 94, 95Industrial and Commercial Bank of

China, 6n, 79, 149industrial disputes, 7, 9, 9, 90–91industrial output, in sample cities,

16industrial sector, xiinflation, 69inheritance rights, 198

INDEX

224

initial public offerings, 6, 47–48,54, 63–64, 84, 184

insurance companies, 186intellectual property, 65intelligentsia, 36, 37interest rates, overdue, 150–52, 151Interim Provisions on the

Acquisition of DomesticEnterprises by Foreign Investors,193

internal restructuringbudget constraints, 155–57choice of valuer in, 73debt avoidance through, 177decline of, 54discounted price for assets, 75in first round of gaizhi, 56as a form of gaizhi, 22–23,

46–47, 59–60, 62geographic variations in survey,

56–57hardening of budget constraints,

155in most recent round of gaizhi,

57, 59obtaining land-use rights, 84overdue interest rates after, 151,

151–52overdue loans after, 150, 150overdue social security payments

after, 153, 154overdue taxes after, 153progress of process of, 25–31rate of employee discharge, 98relation to workforce growth,

101rounds of, 57in sample cities, 20, 24, 53, 58in sample distribution, 50–51, 51in sample firms, 22–23wage growth rates under, 110,

111

International Finance Corporation,142

interviews, 17–18investment

comparisons between privatizedfirms, 172

discrimination against outside in-vestors, 184

foreign direct investments, 1,118–19, 119, 194

gaizhi versus non-gaizhi firms,146–47, 147

new investment, 23outside investment in gaizhi

process, xiii, 36–37, 48, 50,53, 139, 141–44, 180

restrictions on size and liquidity,193

SOEs as outside investors, xiiitransaction costs for outside in-

vestors, 191–97IPOs, 6, 47–48, 54, 63–64, 84, 184

Jefferson, Thomas, 200Jianhua, 116jobs-for-life, disappearance of, 103Johnson, Simon, 39joint ventures

choice of valuer in, 73contrasted with wholly foreign-

owned ventures, 118, 119conversion of land-use rights into

equity, 83corporate governance, 194n,

195nas a form of gaizhi, 22–23, 46,

50, 51, 51motivation for, 169nobtaining land-use rights, 84organizational transformation

through, 114in sample cities, 54, 58

INDEX

225

key persons, 135Korea, South, 94, 95

laborimpact of gaizhi on workforce

structure, 102, 102–3industrial disputes, 7, 9, 9, 90–91influence on decision-making,

140, 140productivity of, 158–59, 160–62,

164, 166–68, 172–74workforce growth rate, 101,

101–2See also employment; redun-

dancy; unemploymentlabor unions, xiii, 32, 127–29, 129,

139, 144land, conversion of into equity, 83,

84land, privatization of, 200Land Law (1998), 81land-use rights, 81–86, 108land valuation, 85Lanzhou

bonuses for managers, 136, 136choice of valuation firm, 74economic indicators, 16forms of gaizhi in, 57location, 12, 12management shareholdings, 117,



tures, 124per capita GDP growth, 13, 15per capita income, 13, 14population, 13privatized SOEs, 28, 28redeployment of redundant

workers, 106response to survey, 18

Lanzhou (continued)as a sample city, 11settling of overdue payments,

105, 106state-owned enterprises, 29, 29types of gaizhi in, 58unemployment, 17xiagang workers, 15

layoffs. See redundancyleasing

choice of valuer in, 73in first round of gaizhi, 56as a form of gaizhi, 22–23, 46,

50, 60geographic variations in survey,

57of land-use rights, 81, 82, 83, 84in most recent round of gaizhi,

57, 59obtaining land-use rights, 84overdue interest rates after, 151overdue loans after, 150overdue social security payments

after, 154overdue taxes after, 153rate of employee discharge, 98relation to workforce growth,

101in sample cities, 53, 54, 58in sample distribution, 51, 51, 52of SOEs in 1980s, 2in transfer of state assets, 65, 66trend towards, 54wage growth rates under, 111

legal person shares, 192, 193,196–97

Legal Person Shares of ListedCompanies to Foreign Investors,119

legal representatives, 135‘let the small go’ policy, 3–4, 32Li, Shaomin, 41Li, Shuhe, 41

INDEX

226

Li, Yizhong, 185, 190limited liability companies, 46, 47,

49, 51, 193liquidations, xi, 5, 25, 28listings, xi, 5, 73Liu family, 142living standards, minimum, 89loans

advantages for state sector, 168to buy shares, 65gaizhi versus non-gaizhi firms,

155nonperforming loans, 6, 48, 77overdue bank loans, 76–77, 79,

147, 148–51, 149–52, 155–57restructuring of borrowers, 187state-owned enterprises as guar-

antors, 67local governments

abuse of bankruptcies by, 48assets of, 63ncompensation of redundant

workers, 94, 111–12conflict of interest with central

government, 32dealing with enterprise debt in

gaizhi, 80–81encouragement of outside in-

vestors by, 36enforcement of regulations, 185evasion of payment to, 157in fiscal decentralization, 41fiscal strength of, 61, 62funding of redeployment centers,

92give priority to overdue wages,

105national government subsidies

for, 67obligation to support laid-off

workers, 34, 60regulation of unemployment in-

surance, 89

relationship with local SOEs,33–34

responsibility for SOE losses, 37risks in financial position of, 156,

158role in gaizhi, xii, xiv, 38, 39, 66SOEs owned by, 3, 114use of proceeds from sale of land-

use rights, 85losses, socialization of, 37Luoyong Tractor Factory, 169n

macroeconomic policy, 164Malaysia, 94, 95management

on boards of directors, xiii, 122,123, 134–35, 139, 144

division of powers betweenboard and, xiii, 194n

government officials in positionsof, 190–91

incentives for, 135–38, 136influence of government on ap-

pointments, 128, 128–29, 130,131–32, 188–89

influence of Party on appoint-ments, 189

influence on decision-making,140, 140, 141

influence on use of state assets,113

managerial autonomy, 113,133–35, 169, 176

perceived importance of, 134,134

performance evaluation, 185recruitment of new management,

146reduction of in gaizhi firms, 146,

146redundancies, 146remuneration of, 137, 139, 144,

146

INDEX

227

role in gaizhi, xiii, 34, 37, 38, 39shareholdings by, 49, 65,

116–17, 116–17, 138, 168temptations facing, 143transfer of ownership to, 42

management buy-outs, xiv, 4, 49,72, 75–76, 78, 176–77, 182–83

managerial market, 130market liberalization hypothesis,

41–42, 44, 45market volatility, 69MBOs, xiv, 4, 49, 72, 75–76, 78,

176–77, 182–83McGuckin, Robert, 40medical insurance, 88mergers and acquisitions

administered mergers, 36, 115in Chinese foreign direct invest-

ment, 118, 119debt financing for, 195by de novo private firms, xivdevelopment of New Hope

Group, 142effect of QFII Rules on, 193as a form of gaizhi, 23, 24, 27,

50government encourages

takeovers, 64nonlisted SOEs, 193as a proportion of FDI, 194in sample cities, 54by SOEs, 115statistics, 25, 26Takeover Rules, 192transaction costs, 196

methodology, 11, 17–18migrant workers, 102minimum living safeguard system,

89, 93Ministry of Construction, 68Ministry of Finance, 32, 68, 187,

193Ministry of Labor, 32

Ministry of State Land andResources, 68

money tightening, 164monitoring, for agency costs, 113monopolies, 6moral hazard, 155multivariate analysis, 155–56, 157Murrell, Peter, 169

National Economic ResearchInstitute, 42

National People’s Congress, 31net assets per worker, 61, 62net worth, of companies, 186, 187new enterprises, 107New Hope Dairy, 202New Hope Group, 142, 197New Zealand, 95nonperforming loans, 6, 48, 77nonstate enterprises, 7, 9nontradable shares, 192, 193NPLs, 6, 48, 77

open government, 188open sales, of firms

choice of valuer in, 73discounted price for assets, 75in first round of gaizhi, 56as a form of gaizhi, 22–23, 46,

50, 59–60in most recent round of gaizhi,

57, 59obtaining land-use rights, 84overdue interest rates after, 151overdue loans after, 150overdue social security payments

after, 154overdue taxes after, 153rate of employee discharge, 98relation to job losses, 100relation to workforce growth,

101in sample cities, 53, 54, 58

INDEX

228

in sample distribution, 51, 52, 52trend towards, 54wage growth rates under, 110,

111Orient Group, 197‘original sin,’ 199–200ownership

changes in structure of, 114–21,115–20

composition of GDP by, 10and control, 121–27discharge rates by type of, 96diversification of, 5, 46–50, 164,

178–79effect of private ownership on

firm performance, 163–66under gaizhi, 37heterogeneous ownership struc-

tures, 177–80, 191, 201hybridized ownership structures,

116relation to regulation, 178–80relation to social demands, 178ntransfer of to management, 42

‘packaging for listing,’ 98Papua New Guinea, 94, 95Park, Albert, 90, 106the Party. See Chinese Communist

Partypatents, 70Pay As You Earn (pension system),

88pensions, 77, 81, 87–88, 93–94,

104, 107nPeople’s Bank of China, 187, 192per capita government revenue, 14,

16, 156performance, of firms, 158–59,

160–62, 163–74performance, of management, 185performance-based incentives, 109performance targets, 2

Philippines, 94, 95piece-rate wages, 108–9, 109political systems, 39‘prettier daughters’ effect, 171, 173price liberalization, 31price volatility, 69, 163private ownership, effect on firm

performance, 163–66private property rights, 198private sector

access to finance, 168, 198acquisitions by, xiv, 197budget constraints, 155–57contribution to sustainable devel-

opment, 202dominance of, 11efficiency of, 166firm performance, 159, 160–62,

163growth of, ximanagers as board members, 135official attitudes towards, 198,

201‘original sin’ in, 199–200overdue wages and payments,

104–5as participant in gaizhi, 52, 53,

61, 62Party branches in, 131nscandals in, 176, 201shareholdings in gaizhi firms,

117, 121, 172socially responsible practices,

201, 202, 203uncertainty in, 198–99workforce growth rate, 101See also hybrid firms

private shareholdings, 2privatization

China compared to EasternEurope, 1

effect on bank discipline, 156–57financial crisis hypothesis, 40

INDEX

229

fiscal decentralization hypothesis,40–41

fraudulent cases of, 72government shareholdings, 114hybridized ownership structures,

116impact on labor productivity,

164impact on profitability, 164impact on unit cost, 164by local government, 33–34outside versus inside privatiza-

tion, xiii–xivownership concentration of pri-

vatizing firms, 120‘prettier daughters’ effect, 171,

173public debate on, 36–37, 175–81reclassification of jobs in, 90nrelation to job losses, xii, 100,

100–101, 111in rural areas, 4selection bias in studies of,

170–71of SOEs (1996–2001), 25–28, 28spontaneous emergence of, 38stalemate in Chengdu, 121start of, 3See also gaizhi; investment

product development, 146, 147productivity, of labor, 158–59,

160–62, 164, 166–68, 172–74productivity gap, 60, 61profitability, xii, 158–59, 160–62,

164, 166–67, 172–74property rights, 4, 31, 46, 181,

184, 198Provisional Regulations on the Use

of Foreign Capital forRestructuring State-OwnedEnterprises, 119

Provisional Rules of State PublicServants, 190

Provisional Rules on theManagement of Investment inDomestic Services by QualifiedForeign Institutional Investors,192–93

the public, in the gaizhi process,36–37

public offeringsdecline of, 54in first round of gaizhi, 56as a form of gaizhi, 20, 22, 24,

46, 50initial public offerings, 6, 47–48,

54, 63–64, 84, 184in most recent round of gaizhi,

57, 59overdue interest rates after, 151overdue loans after, 149, 150overdue social security payments

after, 154, 155overdue taxes after, 152–53,

153, 155rate of employee discharge, 98relation to workforce growth,

101in sample cities, 53, 58wage growth rates under, 110,

111public servants, in business, 190

QFII Rules, 192–93Qian, Yingyi, 114Qionglai, 86n

rationalization programs, 6R&D, 147, 148real estate developers, 85nRed Chips, 163redeployment, 17, 35–36, 42–43,

66, 67, 91, 106–8redundancy

compensation for, 93discharge rates, 96–98, 97–99

INDEX

230

gaizhi and non-gaizhi firms com-pared, 96, 99

impact on profitability, 172industrial disputes due to layoffs,

9influence of on form of restruc-

turing, 60mid-1990s, 103relation to overdue social secu-

rity payments, 156relation to privatization, xii, 45,

100, 100, 111resulting from gaizhi, xi, 5, 37,

47setting of limits for, 95settlements for, 104social obligations on local gov-

ernment, 34, 60statistics, 7tax concessions for reemploy-

ment services, 91types of redundant workers, 43,

91–92See also redeployment; unem-

ployment; xiagang workersregression analysis, 60, 96, 101,

102, 102, 155, 163regulation, xiv, 178–85, 191–93Regulation on Appropriated Land-

Use Rights in SOE Reform, 82Regulation on the Management of

State Asset Valuations, 68remuneration of managers, 137,

139, 144, 146remuneration of workers

collective wage bargaining, xiiigaizhi and non-gaizhi firms com-

pared, 108–9, 109, 110,110–11

overdue wages, 76–77, 81, 91,93, 104–5

salary gap with managers, 137ntypes of, 108–9, 109

wage growth rates, 110–11,110–11

See also compensationresearch and development, 147, 148resource allocation, 41restructuring

effect on labor performance, 103employees’ resettlement pro-

grams, 91foreign direct investment encour-

aged in, 118–19range of mechanisms, 5social cost, 9, 60, 61, 62, 85, 175through asset management com-

panies, 6See also internal restructuring

retailing enterprises, funds for re-form of, 66

retirements, 42–43, 91, 93, 157,165

return on assets, 61Roosevelt, Eleanor, 130rural household responsibility

system, 2

safety net, 89, 93, 175sample cities, 11–17, 18–19, 30sample distribution, 18–20, 50–54,

51–52sample selection, 18SARS, 188SASAC, 32–33, 49, 78–79, 91,

181, 184, 185scandals, 176, 201selection bias, in privatization stud-

ies, 170–71service sector, 14SETC, 32, 193Shanghai, 188, 191Shanghai Jinling, 115shareholder conferences, xiii,

121–22, 123, 125–26, 126, 133,138–39, 140, 141

INDEX

231

shareholding cooperatives, 51, 103shares

buy-outs, 119classes of, 196–97deferred payment of, 65holdings by foreign companies,

117–20, 118holdings by government, 114holdings by management, 49, 65,

116–17, 116–17, 138, 168increase in private shares, 116issue of, 137relation of board to share struc-

ture, 123share swaps, 119in SOEs, 191Takeover Rules, 192three largest shareholders, 120,

120See also employee shareholdings;

shareholder conferences‘shell’ companies, 184Shenyang Jinbei Motors, 2Shenyang Motor Corporation, 2Shenzhen Stock Exchange, 3Shiyin Paper, 202Shleifer, Andrei, 39Shuangxing, 183Shunde, 3, 17, 40Sichuan Sihai Group, 197SMEs, 19, 78–79social burdens, of firms, 43, 157,

165, 167, 172social cost, of restructuring, xi, 7,

9, 60, 61, 62, 85, 175social insurance payments, 77socially responsible practices, 201,

202, 203social relief system, 89social security expenditure, 67, 104social security payments, overdue,

91, 104–5, 153–55, 154, 155–57social security system, 9, 87–89,

154–55

SOEs. See state-owned enterprisessoft budget constraints, 147,

155–56, 157spin-offs, 23, 46, 47, 50, 51, 53stakeholders, changes in influence

of, 138–40, 140State Administration for Industry

and Commerce, 193State Administration of Foreign

Exchange, 193state assets

arbitrary shifting of assets, 195concern of central government

over, 37conflict between central and local

government, 32discounts on price of, 43, 74–76,

75–76, 78, 86, 177, 181erosion of, 72, 175, 176, 184exit funds, 66management influence on use of,

113negative net assets, 64–65, 66overvaluation and undervalua-

tion, 71–72productive and nonproductive, 63transfer of in gaizhi, 62–67, 179,

181–82, 184use of to meet obligations to

workers, 33–34valuation of, 63, 65, 67–76,

74–76, 181, 182State-Owned Assets Supervision

and Administration Commission,32–33, 49, 78–79, 91, 181, 184,185, 190

state-owned enterprisesaccess to finance, 168bankruptcy provisional funds, 66budget constraints, 155–57in composition of GDP, 10contract status of workers in,

93n, 103

INDEX

232

debt problems, 3, 35–36, 76–81decline in numbers, xi, 25, 26,

27, 28, 89–90efficiency of, 39–40, 166, 169,

175extension of guarantees by, 67firm performance, 158–59,

160–62foreign investment in, 6, 117–20,

118–19, 193–96fraudulent privatization of, 72government officials working in,

190incorporation of, 2industrial sector, 8jobs-for-life disappear, 103‘let the small go’ policy, 3–4, 32level of government ownership

of, 42as limited liability companies, 47management of reform, 32–33mergers and acquisitions by, 115outside versus inside control,

xiii–xiv, 168–69overdue bank loans by, 76–77overview of reform process, 2–11owned by local government, 3,

114Party branches in, 131privatization of (1996-2001),

25–28, 28profitability of, 7, 8share in total industrial assets, 7,

8SOE Restructuring Provisions,

193trend in ownership structure

(1995–2001), 115workforce growth rate, 101–2See also corporate governance;

gaizhi; hybrid firms; owner-ship; regulation; restructuring;survey, of gaizhi

Stiglitz, Joseph, 114stock exchanges, 180nstrikes, 7Sun, Qian, 166Supreme People’s Court, 80, 179,

195nsurvey, of gaizhi

compared with 1998 survey, 20,24

firms by economic sector, 21firms by industrial sector, 21–22methodology, 11, 17–18number of firms undertaking

gaizhi, 30sample cities, 11–17sample distribution, 18–20, 24structure of study, 24

sustainable development, 202

Takeover Rules, 192Tangshan

bonuses for managers, 136choice of valuation firm, 74compensation of workers, 103economic indicators, 16interviews in, 18location, 12, 12management shareholdings, 117method of obtaining land-use



14per capita income, 13, 14population, 13private sector, 53redeployment of redundant

workers, 106, 107as a sample city, 11

INDEX

233

settling of overdue payments,105, 106

state-owned enterprises in, 29, 29types of gaizhi, 58xiagang workers, 15

taxes, 76, 107, 152–53, 152–53,155–57, 199

TCL Group, 138technology, new, 146, 147Tenev, Stoyan, 116, 117, 120, 122,

135, 137terminations, 100Thailand, 94, 95theory of the three represents, 19830-point document, 199–200Tian, Guoqiang, 41time trends, in firm performance,

170–72, 173Tong, Jing, 166Tong, Wilson, 166total government revenues and ex-

penditures, 185, 186township and village enterprises

(TVEs), 2, 4, 36tradable shares, 192, 196trademarks, 70trade unions. See labor unionstransaction costs, 191–97, 196transnational corporations, 7Tsingtao Brewery, 115TVEs, 2, 4, 36

UBS Warburg China PrivateEnterprise Index, 159, 163

unemploymentcentral government policy, 89–93discharge rates, 96–98, 97–99jobless contrasted with xiagang

workers, 91–92resulting from gaizhi, 34–35,

42–43, 95–103, 97–100

unemployment (continued)in sample cities, 15, 16, 17state security fund for jobless, 92See also redeployment; redun-

dancy; xiagang workersunemployment insurance, 88–89,

92unemployment insurance levy, 88unions, xiii, 32, 127–29, 129, 139,

144Unirule, 86nunit cost, 158, 160–62, 164, 167,

168, 172, 173United Nations Conference on

Trade and Development, 118United States, privatization of land,

200United States Securities and

Exchange Commission, 186nurban utility privatization program,

86n

valuation. See land valuation; stateassets, valuation of

valuation firms, choice of, 177valuation industry, 68valuers, selection of, 72–73, 74

wages. See remuneration of work-ers

Wanxiang Group, 197Weifang

bonuses for managers, 136choice of valuation firm, 74economic indicators, 16employee shareholding, 57location, 12, 12management shareholdings, 117method of obtaining land-use


tures, 124per capita GDP growth, 13, 15

INDEX

234

per capita government revenue,14

per capita income, 13, 14population, 13redeployment of redundant

workers, 106, 107response to survey, 18as a sample city, 11settling of overdue payments, 105state-owned enterprises, 29, 29types of gaizhi, 58xiagang workers, 15

Wei Liucheng, 190whistleblowers, 188Wolfensohn, James, 201women, 102Wong, Sonia, 130workers’ congresses, 133workers’ rights, 182, 187, 188workforce. See laborworking conditions, 202work units, 17write-offs, 35, 47, 187Wu Bangguo, 3–4Wuhan Motor Engine Factory, 2Wuhu, 190

xiagang workersallowances for, 92–93, 104benefits of status of, 92characteristics of, 15, 91impact on firm performance, 174rate of, 98, 99, 100redeployment of, 92, 107in sample cities, 16as a social burden for firms, 43

Xiningbankruptcies in, 57bonuses for managers, 136choice of valuation firm, 74economic indicators, 16forms of gaizhi in, 57location, 12, 12

management shareholdings, 117,117

method of obtaining land-userights, 83

ownership and control struc-tures, 124



workers, 106response to survey, 18as a sample city, 11sample firms, 18settling of overdue payments,

105, 106state-owned enterprises in, 29,

29types of gaizhi in, 58

Yang, Lei, 41Yanzhou Coal Mining, 115Yao, Yang, 41

Zhang, Chunlin, 116, 117, 120,122, 135, 137

Zhang, Juwei, 90

INDEX

235

Zhang, Ruimin, 183Zhang, Weiying, 41Zhenjiang

bonuses for managers, 136choice of valuation firm, 74economic indicators, 16employee shareholding, 57interviews in, 18location, 12, 13management shareholdings, 117method of obtaining land-use





105state-owned enterprises in, 29,

29types of gaizhi in, 58

zhuada fangxiao, 3Zhucheng, 3, 4, 12, 12, 17, 40