Rural enterprise development in China, 1986-90

162 * ~~World Bank Discussion Papers China and Mongolia Department Series R-ural Enterprise D evelopmnent ian Ch'ina, 1986-90 Anthony J. Ody L Ct  Recent World Bank Discussion Papers No. 104 Enterprise Reform and Privitization in Socialist Economies. Barbara Lee andJohn NeDlis No. 105 Redefining the Role of Govemment in Agriculturefor the 1990s. Odin Knudsen, John Nash, and others No. 106 Social Spending in Latin America: The Story of the 1980s. Margaret E. Grosh No. 107 Kenya at the Demographic Tuming Point? Hypotheses and a Proposed Research Agenda. Allen C. KeUey and Charles E. Nobbe No. 108 Debt Management Systems. Debt and Intemational Finance Division No. 109 Indian Women: Their Health and Economic Productivity. Meera Chatterjee No. 110 Social Security in Latin America: Issues and Optionsfor the World Bank. William McGreevey No. 111 Household Consequences of High Fertility in Pakistan. Susan Cochrane, Valerie Kozel, and Harold Alderman No. 112 Strengthening Protection of Intellectual Property in Developing Countries: A Survey of the Literature. Wolfgang Siebeck, editor, with Robert E. Evenson, Wiliam Lesser, and Carlos A. Primo Braga No. 113 World Bank Lendingfor Small and Medium Enterprises. Leila Webster No. 114 Using Knowledgefrom Social Science in Development Projects. Michael M. Cemea No. 115 Designing Major Policy Reform: Lessonsfrom the Transport Sector. Ian G. Heggie No. 116 Women's Work, Education, and Family Welfare in Peru. Barbara K. Herz and Shahidur R. Khandker, editors No. 117 Developing Financial Institutionsfor the Poor and Reducing Barriers to Accessfor Women. Sharon L. Holt and Helena Ribe No. 118 Improving the Peformance of Soviet Enterprises. John Nellis No. 119 Public Enterprise Reform: Lessonsfrom the Past and Issuesfor the Future. Ahmed Galal No. 120 The Information Technology Revolution and Economic Development. Nagy K. Hanna No. 121 Promoting Rural Cooperatives in Developing Countries: The Case of Sub-Saharan Africa. Avishay Braverman, J. Luis Guasch, Monika Huppi, and Lorenz Pohlmeier No. 122 Performance Evaluationfor Public Enterprises. Leroy P. Jones No. 123 Urban Housing Reform in China: An Economic Analysis. George S. Tolley No. 124 The New Fiscal Federalism in BraziL Anwar Shah No. 125 Housing Reform in Socialist Economies. Bertrand Renaud No. 126 Agricultural Technology in Sub-Saharan Africa: A Workshop on Research Issues. Suzanne Gnaegy and Jock R. Anderson, editors No. 127 Using Indigenous Knowledge in Agricultural Development. D. Michael Warren No. 128 Research on Irnigation and Drainage Technologies: Fifteen Years of World Bank Experience. Raed Safadi and Herve Plusquellec No. 129 Rent Control in Developing Countries. Stephen Malpezzi and Gwendolyn Ball No. 130 Patterns of Direct Foreign Investment in China. Zafar Shah Khan No. 131 A New View of Economic Growth: Four Lectures. Maurice FG. Scott No. 132 Adjusting Educational Policies: Conserving Resources While Raising School Quality. Bruce Fuler and Aklilu Habte, editors (Continued on the inside back cover.)  Rural Enterprise D evelopment in China, 1986-90  China and Mongolia Department Series Pattuens of Direct Foreign Investment in China, Zafikr Shah Khan, September 1991 Coal Pridig in China: Ises and Reform Stratee, Yves Albouy, October 1991 Refonring Pices: The Fxperkic of China, Hungary, and Poland, Anand Rajaram,January 1992 Deeloping Mongolia, Shahid Yusuf and Shahid javed Burd, Janury 1992 Sino-Japanes Economic Relationships: Trade, Direct Investment, and Fur Stegy, Shuichi Ono, Januay 1992 Th Sec&oral Foudations of China's Dmvlopment. Shahidjaved Budri and Shaid Yusuf editos, Januay 1992 Transport Dmevopment in Southen China. Clell G. Harral, editor, and Peter Cook and Edward Holland, principal contributors, February 1992  162 w X World Bank Discussion Papers China and Mongolia Department Series Rural Enterprise Development in China, 1986-90 AnthonyJ. Ody The World Bank Washington, D.C.  Copyright C 1992 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing May 1992 Discussion Papers present results of country analysis or research that is circulated to encourage discussion and comment within the development community. To present these results with the least possible delay, the typescript of this paper has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, Department F, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France. ISSN: 0259-210X Anthony J. Ody is senior economist in the Country Operations Division of the World Bank's East Asia and Pacific-Country Department II (China and Mongolia). Library of Congress Cataloging-in-Publication Data Ody, AnthonyJ., 1950- Rural enterprise development in China, 1986-90 / Anthony Ody. p. cm. - (World Bank discussion papers ; 162) Includes bibliographical references. ISBN 0-8213-2118-8 1. China-Industries, Rural. 2. China-Economic policy-1976- I. Title. II. Series. HC427.92.034 1992 338.0951'09173'4-dc2O 92-12836 CIP  Abstract This paper updates through 1990 World Bank analysis of China's nonagricultural rural enterprise sector, which has come since the late 1970s to play an increasing role in both industry and the service sectors, and which operates largely outside the framework of central state control. The sector's performance during the 7th Five-Year Plan period (1986-90) is reviewed with respect to output, employment, exports and fiscal contribution, and regional and subsectoral variations are examined. Controversies over the sector's record in areas such as economic efficiency, energy intensity and pollution control are also analyzed. Review of the official policy climate towards the sector covers initiatives intended to improve the legislative and incentive frameworks and attempts to improve public provision of services to rural firms, as well as their treatment during the austerity campaign of 1988-90. The paper concludes that the nonstate sector has a crucial role to play in future development in China, suggests that this may now be receiving increasing recognition by the central authorities and outlines areas where the policy climate still shows room for improvement.  - vi- OFFICIAL EXCHANGE RATE (As of November 1991) Currency unit = Yuan (Y) = 100 Fen Y 1.00 = $0.19 $1.00 = Y 5.39 FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric System ABBREVIATIONS ABC - Agricultural Bank of China BTVEs - Bureau of Township and Village Enterprises CASS - Chinese Academy of Social Sciences EZO - Experimental Zone Office FEAC - Foreign Exchange Adjustment Center FTC - Foreign Trade Corporation GVO - Gross Value of Output ISTIC - Institute of Science and Technology Information of China MOA - Ministry of Agriculture NEPA - National Environmental Protection Agency NGFA - Nongovernmental Financial Association RCC - Rural Credit Cooperative RE - Rural Enterprise RFSU - Rural Financial Service Unit RTOV - Rural Total Output Value SOE - State-Owned Enterprise SPC - State Planning Commission SSTC - State Science and Technology Commission STC - Science and Technology Commission TIC - Township Industrial Corporation TVCE - Township and Village Collective Enterprise TVE - Township and Village Enterprise UCC - Urban Credit Cooperative  - vii - FOREWORD The World Bank's economic and sector work program in China is a very active one ranging over a wide spectrum of topics from macro- economics to health and education. Each year we publish a handful of our formal studies, but thus far most of the background papers and informal reports, many of them containing valuable analysis and information, have remained outside the public domain. Through the China and Mongolia Department Working Paper Series, we hope to make available to a broad readership among the China watchers and development communities a few of the papers which can contribute to a better understanding of China's modernization. The study by Mr. Ody of rural enterprises provides a timely review of China's most dynamic subsector. During the nineteen eighties, rural enterprises have been an important source of growth in employment, manufactured products and, increasingly, of exports. These enterprises now account for about 27% of total industrial output and perhaps 47% of total industrial employment. These shares were 18% and 43% respectively in 1985. Their momentum is unabated in the early nineteen nineties and, if suitably nurtured by credit and industrial policies, rural enterprises can serve as one of the pillars of economic achievement in this decade and beyond. One of the encouraging signs we have seen in recent months is increasing public discussion within China of some of the experiments in rural enterprise reform and development examined in Section II of the present paper. Mr. Ody's paper not only analyzes recent rural enterprise performance but also suggests how their further nurturing might best be conducted. Shahid Javed Burki Director China and Mongolia Department   - ix - Table of Contents Page No. Summary .... . . . . . . . . . . . . . . . . . . . . . . xi I. DEVELOPMENT FROM 1986 ... . . . . . . . . . . . . . . . . 1 Introduction . . . . . . . . . . . . 0. . . . . . . . . 1 Key Characteristics . . . . . . . . . . . . . . . . . . I Scale of Sector.. . . .3 Growth Trends.. . . . . .3 Industrial Subsectors . . . . . . . .5 Regional and Provincial Aspects . . . . . . . . . . . . . . 7 Exports . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Investment, Credit and Factor Supply . . . . . . .. . 10 Labor Costs and Competitiveness . . . . . . . . . . . . . . 13 Profits and Taxes.. . . . .. . . .14 Efficiency and Performance . . .16 Economic Efficiency.. . . . . .16 Relationship with Other Sectors . . . . . . . . . . . . 19 Environmental Aspects ..20 II. REFORMS AND EXPERIMENTS . . . . . . . . . . .. . . 21 Review of Implementation .22 Private Enterprise Regulations . . . . . . . . . . . . . 22 Rural Collective Regulations . . . . . . . . . . . . . . 22 Joint Stock Organizations . . . . . . . . . . . . . . . 23 RE Support Services.. ..24 Export Promotion ..25 Financial Sector Development . . . . . . . . . . . . . . 26 Additional Experiments ..27 III. STABILIZATION, RECTIFICATION AND REVIVAL (1988-90) . . . . . 28 Objectives . . . . . . . . . . . . . . . . . . . . . . . 28 Implementation ..29 Beginnings of Revival . . . . . . . . . . . . . . . . . 31 IV. CONCLUSIONS .... . . . . . . . . . . . . . . . . . . . . 33 BIBLIOGRAPHY . . . . . . . . . . . . . 38 ANNEXES 1. Structure of Rural Enterprise Sector--1988 . . . . . . . . . 41 2. The Urban Nonstate Sector: an Outline . . . . . . . . . . . 42  x Page No. TABLES IN TEXT 1.1 Gross Industrial Output by Type of Ownership . . . . . . . 6 1.2 Regional Differentials ... . . . . . . . . . . . . . . . 8 1.3 Township and Village Enterprise Exports . . . . . . . . . 9 1.4 Composition and Growth of TVE Exports . . . . . . . . . . 10 1.5 Selected TVCE Profitability Indicators . . . . . . . . . . 15 CHARTS IN TEXT 1.1 Outline of Rural Enterprise Sector - 1988 . . . . . . . . 4 3.1 Growth of Gross Industrial Output 1987-90 . . . . . . . . 30  - xi - SUMMARY 1. The present discussion paper updates and extends World Bank analysis of China's rural, nonstate, nonagricultural enterprises (REs), the most eco- nomically dynamic sector of the domestic economy. World Bank work on the sector was initiated in 1986/87, with a major collaborative study of rural industry, whose results were published recently, and this paper brings the account up to 1990. REs comprise both officially sponsored Township and Vil- lage Collective Enterprises (TVCEs) and various "below-village" categories, such as cooperatives, partnerships and private and individual firms. REs are active not only in industry, but also in construction, transport and commerce. ii. In the late 1980s, over 18 million REs provided employment (some- times seasonal or part-time) to over 90 million people, some 17 percent of the national labor force. Their gross value of output was about 21 percent of the national total. TVCEs were fewer than 10 percent of the number of REs, but accounted for half of RE employment and three quarters of the sector's output value. Industry was the largest single RE sector, with 40 percent of the number of firms, 60 percent of the employment and 70 percent of the value of output. Within industry, TVCEs accounted for 20 percent of the enterprises, 60 percent of employment and 80 percent of gross output. Industrial REs cur- rently account for over a quarter of China's total industrial gross output. Though data on real output for REs are not entirely trustworthy, there is no doubt that its growth, as well as that of employment, has been rapid in all sectors following the launching of economic reforms and the relaxation of earlier administrative restrictions since the late 1970s. In the second half of the 1980s, in particular, the private and individual subsector has shown especially fast expansion, though average firm size remains very small. In addition to REs' contribution to output and employment, their trade growth has also been striking and they now provide over one-fifth of China's total exports. iii. TVCEs typically enjoy close ties to specific rural communities, with the supply of both capital and labor overwhelmingly local in origin. Though their managers have considerable day-to-day autonomy, local governments are involved in major strategic decisions. These community links, while in many ways a source of strength and stability, can impose barriers to enterprise size, and have virtually ruled out enterprise relocation or cross-community mergers. They also pose conflicts of interest for lower-level administrations' regulatory efforts, for example in the sometimes troubling area of pollution, indicating the need for careful monitoring by county or higher-level agencies. While industrial REs in less-developed interior regions are heavily concentrated in the production of low-grade building mate- rials for geographically isolated markets, those in more advanced regions produce a very wide range of consumer and capital goods. They are typically highly outward-oriented in their marketing strategies, and have been quick to adapt their product mix to changes in demand. Outputs and physical inputs are largely traded on free and competitive markets, and REs get little help from the formal planning and materials allocation systems. This can lead to diffi- culties in access to energy (especially electric power), long-distance trans- port (especially rail), information and other support services and adequately trained technical and managerial personnel.  - xii - iv. The paper's analysis of the earlier part of the period under study (1986 through late 1988) finds a continuation of the rapid growth of the early 1980s. Subsectoral analysis shows growth to be broad-based, with particularly rapid expansion in relatively high-technology subsectors, such as electronics and chemicals. On a regional basis, growth continued to be faster in the more developed (mostly coastal) regions than the interior. However, all regions displayed wide provincial variations, indicating the crucial difference the local policy environment can make. Within the coastal zone, the original RE pioneers like Jiangsu and Zhejiang provinces, whose lead was based on earlier toleration of REs than in other areas, have recently been falling behind newly dynamic provinces such as Guangdong and Shandong. The latter seem to have been more successful than Jiangsu or Zhejiang in restraining the growth of local taxes and levies on REs, and Guangdong, in particular, has been more tolerant of private REs, whereas Jiangsu continues heavily to favor the TVCE model, while obstructing private enterprise development. Guangdong's private firms seem more likely to hire migrants from poorer regions than the tradi- tional, community-based collective model found in Jiangsu, and this contrast could have significant implications for poverty alleviation policy. v. Despite the continued growth in overall RE output, the paper sees possible cause for concern in the secular decline in (TVCE) profitability, which has so far persisted through all phases of the business cycle, ever since the early 1980s. The paper examines various possible explanations for falling profits. Increases in tax rates have certainly played a significant part, though the trend is also observable in pre-tax results. The TVCEs have, over the years, gradually increased their share of bank credit, and on average have higher ratios of debt to total capital than state-owned enterprises (SOEs), although the more capital-intensive state firms would show higher ratios of debt to sales. In absolute terms, however, the TVCEs' average leverage is still low by international standards, though the extremely short- term nature of bank lending in China can expose firms to financial stress when market conditions and/or credit policies are adverse. The paper does not agree with those who blame rising RE wages for the falling profit trend. In aggregate, RE wages have risen no faster than those in SOEs, and RE wages have declined as a share of total production costs. In absolute terms, RE workers' remuneration generally remains below SOE levels (particularly when account is taken of non-cash SOE benefits such as heavily subsidized housing, basic foods, pensions and medical coverage). In addition, REs' wages are as a rule far more closely tied to their workers' productivity than in the state sector. Fundamentally, much of the explanation for observed profit trends should prob- ably be sought in the increasingly competitive nature of markets over the decade. In retrospect, the conditions of the early 1980s were clearly atypical, with the sudden lifting of earlier administrative restrictions and opening-up of new business opportunities temporarily permitting rather easy attainment of supra-normal profits by the early entrants. Declining profits do, however, suggest that the sector's continued growth may become more dependent than in the past on access to formal credit mechanisms. vi. The period has seen two bursts of policy reform initiatives favoring REs, interrupted by a harsher climate between September 1988 and March 1990. Important legislative steps included mid-1988 regulations establishing a much clearer legal status for private enterprises and a more flawed set of regula- tions on rural collective enterprises dating from mid-1990. Although many  - xiii - aspects of the collective regulations deserve to be welcomed, including clari- fication of arrangements for "contract" management, they leave the ownership rights to such firms and their relationship with the government system still poorly defined. Promised RE bankruptcy rules have yet to be issued. Intro- duction of a joint stock company framework, so far attempted only on an exper- imental basis, could provide a better legal basis for defining clear ownership rights to firms than currently exists. vii. Other supportive government efforts include the "Spark program," launched in 1985 by the State Science and Technology Commission (SSTC), which assists demonstration technology-upgrading subprojects in REs, provides RE staff training courses and also seeks to improve information services for small businesses. One area where the Spark program has apparently failed to achieve meaningful results, however, is the modernization and restructuring of the domestic capital goods sector. Most REs are overwhelmingly dependent upon indigenously manufactured plant and equipment, partly as a result of the ela- borate procedures involved in importing items which might compete with domes- tic suppliers. Many of the criticisms which have at times been made of REs' performance, e.g., as regards efficiency in utilization of energy or raw mate- rials, should probably more appropriately be directed against the lack of competitiveness on quality of protected domestic equipment suppliers. viii. The period between September 1988 and late 1989/early 1990 was in some ways the most difficult one the REs have faced since the current reforms began. The Chinese economy was clearly in need of relatively tight macroeco- nomic policies to contain overheating and squeeze inflation. But these poli- cies, as initially designed, also included discriminatory efforts to "tilt" the supply of credit and other inputs further in favor of the SOEs. Particu- larly in the first three quarters of 1989, the supply of bank credit to REs, formerly rising more rapidly than total credit in the economy in line with REs' faster output growth, was subject to severe quantitative ceilings. REs' investment collapsed and many firms were reported to experience difficulty in financing even routine working capital needs. Problems seem, if anything, to have been most acute in some of the more developed coastal provinces. At the same time, REs were to some degree able to mitigate the effects of the squeeze by tapping nonbank sources of finance, including rural credit cooperatives and their own workers' savings. As it was, growth in output and employment at best ground to a halt, and many contemporary reports spoke of widespread clo- sures and job losses. Published statistics for December 1989, however, sug- gest that net losses in RE establishments and jobs over the previous year may have been considerably less than much of the earlier comment would have sug- gested, and it seems possible that at least some of the contemporary reports may have contained an element of hyperbole. ix. A troubling feature of the austerity period was that the credit squeeze was accompanied by a more critical tone in published policy statements on the sector from Beijing. REs were variously blamed for unfair competition with SOEs, unfair competition with agriculture, poor economic results, poor product quality, disproportionate waste of energy and disproportionate respon- sibility for pollution. The paper attempts to examine the admittedly incom- plete evidence available on the more widely reported criticisms of the sector. It sees the allegations of unequal competition between sectors as largely  - xiv - symptomatic of the partial nature of price reform to date, and the coexistence of low fixed plan prices for many basic raw materials and foods with much higher free market prices for the same goods--the need to complete the process of price reform is emphasized. x. Such studies as exist suggest that industrial REs are comparatively efficient in economic terms. Their capital productivity is far higher than that of the SOEs, which is normal for small, labor-intensive enterprises worldwide. Correspondingly, average labor productivity tends, in aggregate, to be higher in the SOEs. This may reflect industrial SOEs' higher capital intensity (fixed assets per worker about four times those in TVCEs), rather than more effective labor management practices. When TVCEs were compared with smaller SOEs producing a similar product, it was the nonstate firms which showed the higher labor productivity, largely because of much wider adoption of payment by results. There is also some evidence that factor productivity has grown faster in the nonstate enterprises. On the other hand, low RE tech- nology and skill levels frequently translate into unsatisfactory product qual- ity (TVCE exports come from a fairly narrow subset of firms), though here too a comparison with the smaller SOEs favored the nonstate firms. Some aggrega- tive studies suggest the existence of untapped economies of scale, and this may well be true in some cases, at least at the production stage (though the necessary detailed work on well-defined subsectors has yet to be undertaken). Given China's ill-developed transport system, however, comparisons of total delivered cost to consumers may produce different results from those made at factory-gate. xi. No reliable studies yet exist which examine REs' energy consumption against that of comparable SOEs in similar lines of production. China's industrial energy consumption is typically above international comparators, though it has improved significantly in recent years. Most energy is used in SOEs, as the state sector tends to dominate large-scale heavy industry. Exceptions include building materials (especially bricks), where REs have a dominant position, and coke production, where they seem to have grown rapidly in recent years. In both cases, a wide range of technologies is found within the RE sector. Some small, traditional RE plants are undoubtedly wasteful in their use of energy, but survive because transport deficiencies leave them still competitive on a total delivered cost basis against larger, more modern (but more distant) factories. Other REs have proved innovative in their adop- tion of new, energy-saving technologies. REs are more likely than SOEs to face realistic, market-based energy prices, and this fact, together with their profit-oriented nature, will predispose them to adopt efficient techniques if not obstructed by barriers to the supply of finance, information or up-to-date capital goods. Removing such barriers might be a more productive course for public policy than the administrative sanctions apparently now favored. xii. RE pollution is an issue where anecdotal accounts vie for attention with more serious studies. Again, many of the industries with the greatest potential for damage are concentrated in the SOE sector. However, this is clearly one area where the concern over REs' record has some basis in fact. The potential conflict of interest between the regulatory and ownership roles of local governments has already been mentioned. Different regions appear to have enjoyed varying degrees of success in their pollution control efforts, a fact which itself calls for more detailed investigation. In some jurisdic-  - xv - tions, REs' performance is satisfactory in absolute terms, as well as no worse than that of comparable SOEs. In other localities, scattered REs are essen- tially uncovered by any regulatory system, while in southern Jiangsu, in par- ticular, the cumulative effects of concentrated RE development have over- whelmed both regulatory capabilities and the environment's absorptive capacity. Official claims notwithstanding, intermittent central "campaigns" (such as in 1988-90) do not appear to have produced significant lasting results. The paper sees no "quick fix" for these problems, and no alternative to the steady effort of building up properly funded and staffed public regula- tory agencies and backing them with sustained political will at both central and lower levels. Local encouragement of RE clustering, and more effective industrial zoning, could help make satisfactory technical solutions more affordable. xiii. After the criticisms of the RE sector prevalent for much of 1989, the early months of 1990 were to witness a marked change of tone and, with an easing of earlier restrictions, the RE sector resumed over 1990/91 its earlier position as China's leading source of growth. Beijing seems to have arrived at a new and more balanced understanding of the importance of the REs' contribution to the national economy. The paper argues that the change in part reflects an increasing awareness that the cumulated problems of the SOE sector, characterized now not only by sluggishness and inefficiency but also by increasingly open and heavy financial losses, are not to be resolved by a fairly simple "tilting" of credit or supplies, but only by serious and painful restructuring efforts and fundamental policy reforms. Simultaneously the leadership seems to recognize the need for the nonstate sectors to be the leading force in future employment creation--especially in services. xiv. The paper concludes that there still remain many areas of the Chinese economy in which de facto discrimination exists between firms within different ownership categories. It argues for further reforms in the pricing and allocation of basic materials and services, as well as in the functioning of credit markets and the tax system to achieve a more level playing field. Regarding human capital, measures to promote greater mobility of skilled workers--e.g., by allowing portability of pension benefits--need to be accom- panied by efforts to upgrade the effectiveness of secondary education and vocational training within the rural economy. For the inhabitants of particu- larly disadvantaged regions, indigenous development will likely need to be supplemented by a more sympathetic examination of the implications of greater internal migration. Finally, liberalizing markets further and strengthening the definition and protection of property rights could help not only sustain the vigor of the nonstate sector in China's rural areas, but also extend some of its energy to the urban economy.   I. DEVELOPMENT FROM 1986 1.1 Introduction. Over the past 13 years, rural nonagricultural enter- prises (REs), operating outside the framework of state ownership, have emerged as the most dynamic element in the Chinese economy. Such firms are active both in industry and in service sectors such as transport, construction and commerce. They may be owned either privately by one or more individuals, or - as in the case of the Township and Village Collective Enterprises (TVCEs) - collectively by an entire rural community. In 1986/87, the World Bank and the Chinese Academy of Social Sciences (CASS) undertook a major collaborative research project on China's rural industrial sector, which involved intensive fieldwork in four widely varying rural counties, and whose results were pub- lished recently.l/ The present paper seeks to supplement this earlier work by: (i) using currently available material to update the account of RE sector development to cover the period of the 7th Five Year Plan (1986-90); (ii) examining policy directions during the period as revealed by legislation, public programs, official statements, etc.; (iii) synthesizing the findings of recent World Bank economic and sector work and other studies relevant to the RE sector; and (iv) selectively revisiting policy issues.2/ 1.2 Key Characteristics. The community ownership of China's TVCEs has no obvious international parallel. The pooling of capital and skills involved made it possible for enterprises of reasonable size to be organized relatively quickly when earlier restrictions on rural industrialization (and on nonstate enterprise participation in service sectors such as distribution and road and water-borne transportation) were lifted in the late 1970s and early 1980s. Based on international conventions on enterprise scale in developing coun- tries, the average Chinese township collective firm would indeed be considered of medium rather than small scale (i.e., employment of 50-99 persons), and many exceed even this limit. At the opposite extreme, by contrast, the typi- cal individually owned Chinese enterprise, with a workforce counted in single figures, is more immediately recognizable from international comparators in the form of "very small" firms (<10 workers), "cottage shops" (<5 workers) or "microenterprises."3/ 1/ Byrd and Lin Qingsong (eds.), 1990; Byrd and Gelb (1990); and World Bank (1989). 2/ Many sources continue to use the term "township and village enterprises" ("TVEs"), but this creates ambiguity as it is sometimes used to cover all REs (regardless of ownership type) and sometimes only TVCEs. This paper attempts to avoid referring to "TVEs" except where its use in original sources makes it unavoidable. 3/ See e.g., Little, Mazumdar and Page, (1987) and Liedholm and Mead, (1987).  -2- 1.3 TVCEs generally enjoy close links to specific local communities and their governments. Start-up capital is frequently provided from local govern- ment resources, and though subsequent financing may come increasingly from the banking system (as well as retained earnings), the volume of local loans tends to be closely linked to that of local deposits, cross-jurisdictional mobility of capital is low, and bank branches generally expect and receive local gov- ernment guidance on the investment of what are often regarded as "community resources." Once operational, TVCEs usually become, in turn, major sources of revenue for local governments, through both taxation and profit remittances. Communities may use temporary transfers among different community enterprises (often routed through "township industrial corporations" or TICs) to ease the process of restructuring output in the face of market shifts, but overall local budget constraints are hard, and this represents a fundamental differ- ence between TVCEs and China's State-owned enterprises (SOEs). Though labor markets are open to inward migration in a few areas, it is more normally expected that TVCEs will function as a source of employment primarily or solely for the largely immobile population of "local residents." The limited supply of community capital, and the desire to diversify production activities to minimize risk, can constrain the size of individual TVCEs, while employment and tax considerations tend to limit possibilities for cross-community consol- idation or relocation of firms. 1.4 Despite their local ties, however, REs have typically proved out- ward-oriented in their marketing strategies. While "local protectionism" may be an option for some provincially-controlled state firms, lower levels such as counties or townships can seldom control markets of meaningful scale. In Wuxi county, Jiangsu, for example, the World Bank-CASS research found that only 4 percent of TVCE sales were within the home township, and 82 percent of firms sold less than 20 percent of their output within their home county. Large majorities of the firms sampled in all four counties studied sold at least 40 percent of their output outside their home province. Though no more recent data are available, it is considered unlikely that the basically outward-oriented stance of the sector has subsequently been reversed to any very substantial extent. The REs, which sell little, if any, of their output through Plan channels, operate in a largely market-oriented environment, and dedicate substantially more resources to marketing than SOEs, much of whose output is handled through state distribution channels. Similar contrasts apply to input procurement. TVCE managers appear generally to enjoy greater autonomy than their SOE counterparts in day-to-day business matters, though community governments tend to be involved in major strategic decisions. 1.5 Below the level of the village, a wide variety of joint-rural owner- ship forms exists, some the descendants of earlier team enterprises,4/ others more recently formed joint-household firms, cooperatives or partner- ships. The organizational forms of such firms sometimes appear to reflect local improvisation rather than clearly established legal models. Individual firms were afforded new legitimacy by the adoption of the 1982 Constitution, but were then still officially subject to a ceiling of eight workers. 4/ The production team as a distinct below-village economic unit has by now disappeared in most parts of rural China. For accounts of the process whereby many of the team enterprises were contracted out to individuals or groups, see papers by Gold and Oi in Reynolds (ed.) 1988.  Enforcement of this limit clearly varied between different localities--in some areas larger private enterprises became either nominally or de facto collec- tivized (at the national level, the ceiling was eliminated by 1988 regulations on "private" firms). More broadly, local governments' ability to control registration procedures, taxation and the supply of factors and services can make the local policy environment the crucial determinant of private sector development. Commentators point to two polar models. In the Wuxi model (from southern Jiangsu), TVCEs have been heavily promoted and active measures taken to discourage private competition (in the mid-1980s, collective firms accounted for over 95 percent of Wuxi's RE industrial output). The contrast- ing Wenzhou model takes its name from an area in southern Zhejiang, formerly relatively disadvantaged, which experienced particularly rapid development during the 1980s based on a highly favorable environment for individual and private firms.5/ The Wenzhou model has obvious attractions for poorer localities with limited community resources to invest: one such county in Anhui (Jieshou) reported over 50 percent of its RE industrial output in the mid-1980s originating from individual firms or partnerships. 1.6 Scale of Sector. In 1988, an estimated 18.9 million rural enter- prises (Chart 1.1 and Annex 1) accounted for a gross value of output (GVO) equivalent to 21.8 percent of the "total output of society" in the Chinese economy. The bulk of the output (70 percent) came from the industrial sec- tor.6/ In turn, the TVCEs, although less than 10 percent of the number of firms, accounted for 70 percent of total GVO. The share of the RE sector in employment is harder to define, in part because seasonal and part-time work is common in rural areas. Depending on the definitions applied, between 62.0 million and 93.0 million rural inhabitants worked outside agriculture in 1988 (equivalent to 11.4-17.1 percent of the total labor force or 15.5- 23.2 percent of the rural labor force).7/ 1.7 Growth Trends. Earlier World Bank studies have documented the exceptionally rapid growth of rural industrial and other nonagricultural activities between 1978 and the mid-1980s, with real industrial growth esti- mated at 20 percent per annum or above. The development of the RE sector during the 1986-90 Plan period can be divided into several phases. The period from 1986 to the final quarter of 1988 saw continued rapid growth: as shown in Table 1.1, rural firms' share of total industrial output rose from 17.7 percent in 1985 to 25.7 percent in 1988 (while state firms' share fell 5/ See Nolan and Dong Fureng (eds) (1990). 6/ Note, however, that the pervasive use of gross output measures tends to overstate the share of industry, due to double-counting of intermediate goods. 7/ According to officials of the State Statistical Bureau interviewed in February 1991, the major difference between the narrower definition ("rural social labor force") and the broader one ("persons engaged in township and village enterprises") is that the former enumerates individ- uals according to the sector where they work for the longest period of the year, while the latter records nonagricultural employment if it exceeds three months a year. It is possible that there is also some difference in coverage of establishments.  Chart 1 18 OUTLINE OF RURAL ENTERPRISE SECTOR--1988 COMMERCE, FOOD INDUSTRY AGRICULTURE CONSTRUCTION TRANSPORT AND OTHER TOTAL 0.24 wln enter- 0.42 wln prices enterprises TWIISIIIP 14.95 mln 24.90 xln COLLECTIVES workere workers Y 184.67 bln GVO 1.00 mln enterprises 0.23 mln enterprises 0.07 ln enterprises 0.06 enterprises 0.23 *ln enterprises 35.07 mlu workers 2.50 min workers 6.55 sin workers 1.00 mln workers 1.82 mln workers Y 473.48 bln GVO Y 378.21 bln GVO Y 12.38 bin GVO Y 58.21 bln GVO Y 7.65 bln GVO Y 17.03 bln GVO 0.73 mln enter- 1.17 mln prises enterprises VJLLAGE 18.42 mln workers 24.04 min COLLECTIVES workers T 170.36 bln GVO 0 69 wln enter- JOIN? aURL" prises Y 40.01 bln GVO Not iucluded In 21.96 Iin workers enterprise 0.89 mli enterprises 3.67 ml enterprises 6.00 nlu enterprises 17.29 iln enterprises statistics 6.30 wlu workers 5.84 lIn workers 12.41 wln workers 46.51 nln workers 5.70 mla enter- RURAL INDIVIDUAL prises Y 26.67 bin GVO T 40.80 bln GVO Y 39.75 bln GVO Y 228.30 bln GVo V 72.20 bln GVP TOTAL RURAL 7.74 mli enterprises 0.23 mln enterprises 0.96 lIn enterprises 3.73 mln *nterprises 6.23 mln interprles 18.88 mlin enterprise* ENTERPRISES 57.03 mln workers 2.50 mlu workers 14.85 ln i rkers 6.84 xln workers 14.23 nl workers 95.45 wln workers V 452.94 bln GVO V 11.53 bin GVO Y 82.77 bin GVO Y 47.35 bln GVO V 69.50 bli GVO Y 649.57 bin GVO Notes Subcceponents do not alwys add to separately published totals. further details in Annex 1. Sourcet China Statistical Vearbook 1989. China Statistics Abstract 1990. end Vorld Bank staff calculations.  - 5 - from 64.9 percent to 56.8 percent): published real annual growth rates in 1987 and 1988 were over 30 percent. The recorded growth of the rural indi- vidual firm sector in industry was even more dramatic, with annual rates exceeding 50 percent--largely reflecting extensive creation of new firms. Service sector growth was also vigorous during this period. Although there are reasons to believe that the published data overstate rates of real growth, much of the recorded growth was undoubtedly genuine.8/ In the latter part of 1988, in the context of a broader effort both to stabilize the overheated economy and to rectify developments perceived as undesirable, the tone of central policy shifted and REs were singled out for more severe credit con- trols than other sectors and for administrative measures of rectification. The slowdown in this period will be reviewed in more detail in Section III. The first quarter of 1990 saw an apparent retreat from these more restrictive policies, however, and REs have since been leading the recovery of the eco- nomy. 1.8 Rapid output growth during 1986-88 was accompanied by an apparent acceleration of employment creation. On the broader definition (which, as noted, includes part-time and seasonal workers), rural industrial employment grew from 47.6 million to 57.0 million (an average increase of 9.4 percent). Employment growth in industry was significantly slower on the narrower stat- istical definition, for reasons which are not fully apparent. Rural service sectors (construction, commerce, and transport) added 6.5 million new workers (average growth of 10.5 percent). TVCEs exhibited fairly robust employment growth within the industrial sector (on average 7.4 percent), but showed lit- tle or no job creation in the service sectors, whose dynamism almost entirely reflected the impact of individual and other below-village enterprises. Employment in SOEs over the same period grew at an average rate of only 3.4 percent, yielding a total increment of 6.5 million staff and workers. Industrial Subsectors 1.9 Numerous case histories attest to the flexibility REs can show in adapting their output mix to changing market conditions. This represents another contrast to the SOEs, whose subordination to specific sectoral minis- tries or bureaus deprives them of comparable flexibility. At the national level, industrial production of TVCEs is remarkably diversified (no comparable data for below-village firms are available). Out of 40 identified subsectors, TVCEs account for less than 5 percent of output in only three (petroleum pro- cessing and power generation, due to scale economies, and tobacco processing due to regulatory restrictions on entry). The mining and processing of build- ing materials are heavily dominated by rural firms, but TVCEs also account for 8/ Sectoral price deflators for industry in China are computed from firm-level reporting of output in both current and constant price terms, and it has been clear for some time that a significant proportion of rural firms fails accurately to report the constant price data, thus biasing real growth estimates upward, See e.g. Wong, (1988). Note that if the retail price index were substituted for the implied deflators used for rural industrial output in Table 1.1, real growth rates for the sector over 1985-89, though lowered, would still amount to 28.6 percent (85/86), 27.8 percent (86/87), 23.5 percent (87/88), and 4.1 percent (88/89).  Table 1.1: GROSS INDUSTRIAL OUTPUT BY TYPE OF OWNERSHIP (Y MIN.) 86/85 86/85 IMPLIED 87/86 87/86 IWFLIED 1985 f OF TOTAL 1986 S OF TOTAL % GROVTH GROHWH DEFLATOR 1987 % OF TOTAL % GROWTH G ROWTH DEFLATOR (nominal) (nominal) (nominal) (real) (X) (nominal) (nominal) (real) (X) State-Owned Enterprises 630,212 64.9 697,112 62.3 10.6 6.2 4.2 825,009 59.7 18.3 11.3 6.3 Collective-Owned Enterprises 311,719 32.1 375,154 33.5 20.4 14.6 5.0 478,174 34.6 27.5 23.2 3.4 Township Enterprises 76,055 7.8 98,108 8.8 29.0 N.A. N.A. 128,419 9.3 30.9 28.2 2.1 Vil lage Enterprises 66,272 6.8 83,849 7.5 26.5 N.A. N.A. 116,535 8.4 39.0 34.3 3.5 Joint City-Town Enterprises N.A. 0.0 2,056 0.2 N.A. N.A. N.A. 3,031 0.2 47.4 35.7 8.7 Joint Rural Enterprisos 15,175 1.6 22,723 2.0 49.7 N.A. N.A. 28,615 2.1 25.9 21.7 3.5 Other Collectives 1/ 154,217 15.9 168,418 15.0 9.2 N.A. N.A. 201,574 14.6 19.7 14.8 4.3 Individual Enterprises 17,975 1.8 30,854 2.8 71.6 N.A. N.A. 50,239 3.6 62.8 56.6 4.0 Urban 3,339 0.3 2,913 0.3 -12.8 -12.8 0.0 5,027 0.4 72.6 58.8 8.7 Rural 14,636 1.5 27,941 2.5 90.9 N.A. N.A. 45,212 3.3 61.8 56.3 3.5 Other (foreign invested, etc.) 11,741 1.2 16,306 1.5 38.9 33.7 3.9 27,877 2.0 71.0 66.4 2.7 TOTAL 971,647 100.0 1.119,426 100.0 15.2 8.8 5.9 1,381,299 100.0 23.4 17.7 4.8 of which, Rural Total 172,138 17.7 234,677 21.0 36.3 N.A. N.A. 321,812 23.3 37.1 33.2 3.0 88/87 88/87 IWPLIE) 89/88 89/88 IMPLIED 1988 % OF TOTAL % GROWTH % GROWTH DEFLATOR 1989 % OF TOTAL % OROWTH % GROVTH DEFLATOR (nominal) (nominal) (real) (%) (nominal) (nominal) (real) (%) State-Owned Enterprises 1,035,128 56.8 25.5 12.6 11.4 1,234,291 56.1 19.2 3.9 14.8 Collective-Owned Enterprises 658,749 36.1 37.8 28.2 7.5 785,805 35.7 19.3 10.5 8.0 Township Enterpri"s 184,669 10.1 43.8 35.2 6.4 219,384 10.0 18.8 10.2 7.8 Village Enterprises 170,363 9.3 46.2 36.9 6.8 211,768 9.6 24.3 17.9 5.4 Joint City-Town Enterprises 3,927 0.2 29.6 21.3 6.8 5,003 0.2 27.4 19.8 6.3 Joint Rural Enterprises 40,011 2.2 39.8 30.9 6.8 44,627 2.0 11.5 0.7 10.8 Other Collectives 1/ 259,779 14.3 28.9 18.4 8.8 305,023 13.9 17.4 7.2 9.5 Individual Enterprises .79,049 4.3 57.3 47.3 6.8 105,766 4.8 33.8 23.8 8.1 Urban 6,848 0.4 36.2 27.6 6.8 8,970 0.4 31.0 22.4 7.0 Rural 72,200 4.0 59.7 49.5 6.8 96,796 4.4 34.1 23.9 8.2 Other (foreign invested, etc.) 49,532 2.7 77.7 61.5 10.0 75,844 3.4 53.1 42.7 7.3 TOTAL 1,822,458 100.0 31.9 20.8 9.2 2,201.706 100.0 20.8 8.5 11.3 of which, Rural Total 471,170 25.9 46.4 37.3 6.6 577.578 26.2 22.6 14.3 7.2 1/ CalculIstd by differences. A"umd to comprise urban col lctives.  major shares in the output of metal products (36.2 percent of the national total), furniture (34.6 percent) and plastic products (31.6 percent). The largest TVCE subsectors in terms of absolute output and employment at the beginning of the 7th Plan period were building materials (17.8 percent of all TVCE output and 28.0 percent of employment) and textiles (12.9 percent and 8.5 percent, respectively), followed by machine building (11.1 percent and 8.9 percent), metal products (6.7 percent and 5.8 percent) and food processing (6.6 percent and 5.7 percent). 1.10 Over 1986-88, industrial TVCEs outgrew national averages in all but two of the 40 identified subsectors. The largest absolute gains in employment and output came in traditional TVCE subsectors (textiles, building materials and machine building), but TVCEs also advanced into more technologically sophisticated areas. Among the most rapid absolute (and relative) growth rates in both employment and output (nominal and also an estimated real series 9/) were those in electronic and telecommunications equipment and chemical products. The third subsector with particularly fast growth was coking, gas and coal-related products, possibly reflecting the market-oriented TVCEs' response to China's tight energy markets during the period. Regional and Provincial Aspects 1.11 The previous World Bank-CASS analysis found that the growth of rural nonagricultural sectors up to 1984 had been fastest in China's more advanced provinces, especially those along the coast, and that both the pace of devel- opment and the relative importance of these sectors tended to diminish as one moved inland. Advantages of the more advanced provinces can include superior domestic and international communications and contacts, better endowments of human capital and a stronger local financial base. While more remote areas have generally been able to develop basic industries, such as building materials for local markets, and to capitalize to some degree on the process- ing of specialized foodstuffs and other primary products, development of a broader-based manufacturing sector has proved more of a challenge. The earlier work also pointed to "fiscal predation" as the cause of a possible vicious cycle, in which excessive local government revenue demands on firms, sometimes even in excess of profits earned, could suppress the possibility of takeoff in backward communities. There are, however, offsetting success stories from less-developed areas (such as Wenzhou and Jieshou) which have restrained their fiscal demands and successfully encouraged the development of "below-village" firms. In detecting an apparent reversal of the earlier ranking of economic regions by growth rates during 1984/85, the Bank-CASS team speculated that this might reflect the impact of more vigorous private RE development in backward regions.10/ 1.12 Analysis of post-1985 developments appears, however, to show that, at least during periods of relatively unconstrained expansion such as 1986-88, both rural industrial and total nonagricultural output growth have still proceeded, in overall terms, faster in the more advanced regions. This is 9/ For details of methodology and results see Ody (1991 a). 10/ See especially Wang Tuoyu in Byrd and Lin (eds.).  -8- shown in Table 1.2, which contrasts the experience of Region A, comprising developed provinces mostly on the coast, Region B, largely the immediate hinterland, and Region C, the more remote interior areas. A similar regional pattern was found for employment growth in rural industry. By contrast, the relative regional growth rates were reversed in 1989, which--like 1984/85--was a period of considerable economic austerity, and this suggests that the more developed and commercialized regions may as a rule be more vulnerable to slow- downs in market growth. Table 1.2: REGIONAL DIFFERENTIALS Region A Region B Region C Total Developed Developing Undeveloped Regional share (%) of 1988 national 56.0 35.5 8.5 100.0 rural total output value (RTOV) Regional share (%) of 1988 national 67.4 28.7 3.9 100.0 nonagricultural RTOV Regional share (%) of 1988 national 72.5 24.8 2.7 100.0 industrial RTOV Share of nonagricultural output 64.0 42.9 24.9 53.2 in total regional output Share of industrial output in 49.4 26.6 12.1 38.1 in total regional output 1986-88 (nominal) % growth in 77.1 54.0 51.8 65.9 total RTOV 1986-88 growth in nonagricul- 95.5 77.7 58.6 88.3 tural RTOV 1986-88 growth in industrial 106.2 89.1 78.0 100.8 RTOV Source: China Statistical Yearbook (various dates). 1.13 Beneath the regional differences, however, there are further varia- tions in performance between different provinces in the same region. Within Region A, for example, the original pioneers of RE development, Zhejiang and Jiangsu, showed below-average rural industrial growth in both 1985/86 and 1986-88, while provinces such as Guangdong and Shandong significantly outper- formed the average. Similar variations can be found in the other regions. A full understanding would require more detailed work: an element of catching-  - 9 - up may be involved, especially in Shandong. But the differences may also reflect variations in the local policy environment. Guangdong, in particular, more than any other province, has become noted for a policy climate featuring openness to international contacts, private sector development and labor immigration.11/ Both it and Shandong also appear to have maintained significantly lower effective tax rates on TVCEs than Jiangsu and Zhejiang (para. 1.26). Exports 1.14 The late 1980s confirmed the potential of China's RE sector to penetrate export markets on a major scale. As shown in Table 1.3, the dollar value of estimated TVE exports increased tenfold over 1985-90, and the sector's share in total exports rose from 4.4 percent to an estimated 21.0 percent (though the earlier year's figures may miss some indirect RE exports routed through SOEs). Although special export promotion schemes for REs were announced towards the end of 1988, it seems likely that the overall process of decentralization of trade responsibility, together with improved firm-level incentives and the maturing of the RE sector, may have played a more important part in stimulating growth (para. 2.11). Table 1.3: TOWNSHIP AND VILLAGE ENTERPRISE EXPORTS As share Export-oriented Exports of total enterprises ($ billion) (%) (number) 1985 1.20 4.4 8,500 1986 2.67 8.6 9,000-11,000 1987 4.35 11.0 18,000-20,000 1988 8.03 16.9 n.a. 1989 10.00 19.1 n.a. 1990 13.00 (est.) 21.0 56,000 Source: For exports 1986-88: Township and Village Enterprise Yearbook, 1989; for other data, Chinese press reports (various dates): see Ody (1990 a) for details. 1.15 RE exports appear to originate primarily from a fairly small number of firms--the "export-oriented enterprises" shown in Table 2.3 increased from 0.5 percent of all TVCEs in 1985 to perhaps 3.5 percent in 1990 (though officials stress that these estimates do not necessarily provide comprehensive coverage of all exporting firms). Constraints to exporting cited by rural firms interviewed recently in Jiangsu included both difficulties in meeting li/ See Vogel (1989).  - 10 - international quality standards and adverse prices relative to the domestic market, in part reflecting the monopsonistic buying power of local foreign trade corporations (larger state firms increasingly enjoy "direct export rights," permitting them to by-pass the trade corporations).12/ RE exports are also concentrated geographically: in 1987, the predominantly coastal Region A provided 83 percent of the reported total, Region B 14.6 percent and Region C 2.4 percent. Sectoral analysis indicates (Table 1.4) that RE exports still originate primarily from relatively labor-intensive manufacturing activ- ities, and secondarily from processing of primary raw materials. As in domes- tic production, however, the electrical and electronics group has shown espe- cially rapid trade growth. Table 1.4: COMPOSITION AND GROWTH OF TVE EXPORTS Composition by Average Growth Value, 1988 1986-88 (Z) (Z) Chemicals 5.25 16.31 Electrical 3.83 19.36 Minerals 4.57 12.04 Foodstuffs 9.16 13.63 Traditional products 1.63 10.77 Animal husbandry 4.05 15.64 Textiles 15.53 11.19 Clothing 12.63 17.00 Arts and crafts 13.51 12.78 Other 29.87 15.89 Total 100.00 14.15 Source: Township and Village Enterprise Yearbook, 1989. Investment. Credit and Factor Supply 1.16 Research into smaller firms in other developing countries has gener- ally found them to be less capital-intensive and more labor-intensive than larger firms and there has been extensive discussion of how far these contrasts reflect possible distortions in factor markets which raise the cost of capital to smaller firms and/or the cost of labor to the larger enter- prises.13/ In China, as will be shown below, similar contrasts can be drawn between State-owned firms and the nonstate sector. Institutional fea- tures suggest significant segmentation of factor markets between the two sec- tors: with credit to SOEs largely allocated through a national credit plan 12/ See Montes-Negret. 13/ See e.g., Little et al.  - 11 - (often at concessional rates), and loans for rural firms mostly generated from incremental rural deposits (and generally paying higher, more market-oriented interest rates). Similarly, in the labor market, many graduates of higher education institutions are assigned directly to employment in the state sector. Wage determination also differs markedly between the two sectors, with state sector employees paid largely on national scales and nonstate workers usually on locally-agreed payment by results systems. In addition, markets for many materials and services continue to operate on a multiple-tier basis, with lower prices applying to supplies included within the Plan allocation to state firms. Further examples could be provided of technology support services (e.g., research institutes and information networks) created primarily to support SOEs under central ministerial control. Beyond the apparent dualism of factor markets between the state and nonstate sectors, it should also be appreciated that considerable geographical fragmentation of factor markets applies within the rural economy itself. Rural deposits generated within the banking system can generally be expected to be loaned out within the same jurisdiction - bank loans to rural enterprises are thus highly concentrated geographically in the advanced Region A (over 70 percent of the national total, compared to 25 percent for Region B and less than 4 percent in Region C), and it will be argued that rural labor markets are similarly frag- mented. 1.17 The ability of the RE sector to absorb labor on a much larger scale than the SOEs (para. 1.8) in part reflects the substantially lower capital intensity of the nonstate firms. Published data on the stock of fixed assets, based on historical values, show that fixed assets per worker in industrial SOEs in 1988 were approximately Y 20,800, compared to Y 3,240 in TVCEs (including a small proportion of service establishments). Allowing for working capital, total capital per worker would be Y 26,860 in SOEs and Y 6,385 in TVCEs. A recent research program, which prepared deflated (con- stant 1980 price) capital stock estimates for industrial SOEs and for urban and township collectives over 1980-88, found that, while the gap in fixed assets per worker narrowed from its initial ratio of 5.7:1, it remained as high as 4:1 at the end of the period.14/ 1.18 Rural firms accounted for a very low proportion of total investment in the economy during the 7th Plan years. Their growth over the first part of the period was nonetheless accompanied by a moderate increase in that share.15/ Total rural collective investment in fixed assets increased from 8.1 percent of total national fixed asset formation in 1986 to 10.2 percent in 1988. The share estimated to be financed by bank loans varied between 32.1 percent and 38.0 percent over 1986-88, while the portion self-financed by firms declined from 49.0 percent in 1986 to 41.1 percent in 1988, so that enterprises appear to have had greater recourse to other external sources of funding (these may include some community resources, as well as loans from nonbanking sources, bonds, "shares," funds raised from the workforce, accounts 14/ Outputs include: Jefferson (1988) and Jefferson, Rawski and Yuxin Zheng (1990). 15/ This trend was, however, to be at least temporarily interrupted during the selective credit squeeze of 1989 (Section III).  - 12 - payable, etc.). If working capital loans are also allowed for, rural firms clearly succeeded in expanding their access to the banking sector, albeit modestly, over the 1986-88 period. Loans to REs as reported by the two prin- cipal lenders to the sector--the Agricultural Bank of China (ABC) and the rural credit cooperatives (RCCs)--rose from Y 55.4 billion in December 1986 (7.3 percent of all loans by the domestic banking system) to Y 86.4 billion in December 1988 (8.2 percent of all loans). While it is sometimes suggested that TVCEs have in general borrowed excessively from the banks and are over- leveraged, this claim does not receive much support from the aggregate data. In 1988, the average TVCE ratio of bank loans to net fixed assets plus working capital was just 26.4 percent and, although this is somewhat above levels estimated for the SOEs, it seems low by most international standards (it should, though, be noted that most bank lending in China, as in many other developing countries, is extremely short term in nature, and this factor can expose firms with apparently low gearing to considerable financial pressure during periods of tight credit such as 1989). 1.19 The earlier World Bank-CASS research found rural labor markets to be highly fragmented geographically, with labor still relatively immobile between provinces, localities and communities (except for temporary migrant labor, mostly in low-skill activities). In Wuxi, labor was allocated to community enterprises by the township authorities, who discouraged inflows of workers from outside. While there was less administrative intervention elsewhere, various social and institutional frictions still inhibited the free flow of labor. In three of the four counties studied, rural industrial workers were overwhelmingly local in origin: between 79.7 percent and 94.0 percent from the same township and between 92.4 percent and 97.0 percent from the same county. Only Nanhai county, in Guangdong, was more open, with 15.3 percent of workers originating outside the county.16/ It is probably significant that Guangdong has a relatively high proportion of individual and private firms in its RE sector, as there are indications that such enterprises may be more open than collectives to recruiting outsiders. To the extent that this contrast in hiring practices continues to hold, it could have significant implications for employment and poverty alleviation policy in China (as extreme poverty is by now highly concentrated geographically, and greater internal migration one of the few potential escape routes), suggesting the importance of an environment conducive to the growth of the private and individual sector 17/ (it is, of course, possible to argue that, if the collectives are subject to sufficiently keen competition from private firms, they may themselves adjust their employ- ment policies, but this strengthens rather than invalidates the policy impli- cation). 1.20 Any generalizations about REs' access to qualified personnel (or, more broadly, to technological information and improved plant and equipment) must be tempered by an awareness of the very wide range of variation between conditions in different regions. Much of the early strength of the TVCEs in southern Jiangsu and northern Zhejiang may be attributed to links formed with the advanced domestic industrial sector in the greater Shanghai area. Corre- 16/ Byrd and Gelb. 17t See Ody (1991 b).  - 13 - spondingly, the rapid development from a later start of Guangdong province has owed much to its international connections, which other coastal areas such as Fujian and Shandong now seek to emulate. By contrast, conditions in more dis- tant interior provinces are sometimes characterized by considerable isolation from wider developments in technology and marketing and overall shortages of relevant skills. In the aggregate, in any case, the RE sector has clearly experienced difficulty in recruiting more highly trained and experienced staff in managerial, financial and technical fields. An investigation of 1.52 mil- lion rural enterprises in 1986 found only 635,000 "qualified technicians" (0.7 percent of the total workforce, compared to 8.8 percent for SOEs),18/ while the State Science and Technology Commission (SSTC) estimated 19/ that less than 1 percent of TVE personnel had completed college, 15 percent upper secondary school and 42 percent lower secondary school (the latter now a minimum requirement for SOE employment). Though these observations are now somewhat dated, there is reason to believe that the picture may not have changed markedly since.20/ Partly these findings reflect the constraints to geographical mobility alluded to above. Furthermore, as noted earlier, graduates from colleges and from Secondary Technical Schools and Skilled Worker Schools have traditionally been assigned directly to the SOE sec- tor.21/ In addition, state employees benefit from near-absolute job security and a wide range of side benefits (heavily subsidized tied urban housing, noncontributory pension programs, etc.) which have tended to discour- age them from moving to a rural collective sector generally offering fewer such inducements. Personnel already retired from SOEs are said to have con- stituted a fertile recruiting ground for REs, and a variety of public programs (including the Spark program, see para. 2.9) have encouraged shorter-term advisory or consulting assignments for SOE technicians to rural firms. Labor Costs and Competitiveness 1.21 Rural wages in China have generally been below levels in urban areas (and state firms). The World Bank-CASS study found, however, that in the two more advanced communities studied (Wuxi and Nanhai) average monthly incomes in rural industry had surpassed those in the "local" state enterprises (the low- est level of SOE). A large part of the explanation was that RE workers were far more likely to be paid by performance than those in SOEs: 92 percent of RE workers were on piece rates but only 6 percent in the SOEs. Several con- tributors, however, expressed concern that RE wage increases might erode the sector's competitiveness.22/ Subsequent developments do not to date seem to have borne out this fear. Over 1986-88, the average wage bill per worker 18/ China Daily, November 3, 1988. 19/ World Bank (1990 e). 20/ In 1989, the proportion of both college and high school/technical graduates in total new hiring by the urban collective sector was less than 10 percent of that in SOEs (Ministry of Labor). 21/ World Bank (1987 b). 22/ See Du Haiyan and Meng Xin in Byrd and Lin (eds) op. cit.  - 14 - in TVCEs increased by 36.6 percent, compared to 39.1 percent for state indus- trial units, and even the provinces with the highest average TVCE wage bill have remained below the national average for SOEs. Indeed, the average share of wages in TVCEs' total expenses declined over 1986-88 from 18.6 percent to 13.5 percent.23/ A more comprehensive comparison of labor costs between the state and nonstate sectors would also need to take account of the wide differences in side benefits received by their workers. Rural inhabitants provide their own housing, while rents paid by urban residents have on average covered only about 4 percent of full cost recovery. Subsidies on food grain and vegetable oil--available to urban but generally not to rural inhabitants-- were recently estimated to be equivalent to Y 215 per person (about 18 percent of average urban income). Permanent SOE workers make no contribution to their own pensions (set at 75 percent of the standard wage)--employers' contribu- tions in new pension pooling systems range from 19.0 to 25.5 percent of the wage bill--while rural residents are generally expected to fund their own contributions to voluntary provident funds: medical coverage is also more generous in the state sector.24/ One recent study suggested that cash wages accounted for only 62.0 percent of urban workers' incomes in 1988, but even this estimate almost certainly understates seriously the true economic costs of both housing and food subsidies.25/ Profits and Taxes 1.22 The seventh Plan period saw the continuation of a longer-term down- ward trend in reported TVCE profitability that has, to date, persisted over all phases of the business cycle (Table 1.5). Although part of the explana- tion is provided by a steady increase in tax rates (para. 1.23), the trend is also evident in pretax indicators. In retrospect, the extremely high levels of the early reform years seem to have represented an artificial situation, with the sudden opening up of new business opportunities--some long closed off by administrative sanctions--temporarily permitting supranormal profits to early entrants. Heightened competitive pressures, which have substantially lowered all the profitability indicators shown in Table 1.5, are apparently increasing the pressure on RE managers to maximize the efficiency of their operations.26/ In addition, they may be reducing the previously high ability of rural firms to finance their expansion from retained earnings and increasing the importance of access to external sources of funds (para. 1.18).27/ 1.23 In the early reform years, REs benefitted from a variety of tax exemptions, and their effective tax rates were generally low both in absolute 23/ For details, see Ody (1991 a). 24/ World Bank (1990 a,d; 1991 d). 25/ Wang Yukun (1990). 26/ It is also very likely that increasing tax rates have resulted in enter- prises under-reporting their profits, thus exaggerating the trend. 27/ For provincial variations in profitability, see Ody (1991 a).  - 15 - Table 1.5: SELECTED TVCE PROFITABILITY INDICATORS (Percentages) 1981 1983 1985 1986 1988 1989 Gross profit on sales 21.9 19.0 15.3 13.4 11.7 10.6 Net profit on sales 16.8 12.7 9.4 7.2 6.1 5.0 Gross profit on fixed assets /a 39.2 37.1 37.3 31.5 31.3 26.7 Net profit on fixed assets 30.0 24.8 22.8 17.0 16.4 12.5 Gross profit on (fixed assets plus working capital) 25.5 23.9 20.9 17.4 15.9 13.5 Net profit on (fixed assets plus working capital) 19.6 16.0 12.8 9.4 8.3 6.3 Gross profit on (fixed assets plus working capital minus bank loans) n.a. n.a. 26.3 22.8 20.7 17.4 Net profit on (fixed assets plus working capital minus bank loans) n.a. n.a. 16.1 10.5 10.8 8.2 /a Fixed assets at original value. Source: China Statistical Yearbook (various dates). terms and by comparison to other sectors. Over the course of the 1980s, how- ever, such exemptions were progressively phased out. Between 1981 and 1986, the share of TVCEs' pretax profits paid as various state taxes, as reported by the State Statistical Bureau, rose from 23.3 percent to 46.0 percent--and by 1988 stood at 47.7 percent, comparable to the average 49.7 percent paid by state firms. The combination of gradually rising rates and a booming tax base among TVCEs, on the one hand, with relatively stagnant revenues from most other sources (partly reflecting the low short-term elasticity of the prevail- ing revenue-contracting system) has resulted in TVCE tax payments rising sub- stantially faster than public revenues as a whole. Between 1986 and 1988, for example, TVCE taxes rose in total by 72.2 percent in nominal terms (from Y 13.7 billion to Y 23.7 billion), compared to only 16.3 percent for total public revenues. This resulted in an increase of TVCEs' share of the total from 6.0 percent to 9.0 percent (while 1989 saw a further rise to 10.0 per- cent). 1.24 State regulations provide for progressive corporate income tax rates on both collectives and "small" SOEs, rising from 10 percent to 55 percent-- against a flat rate of 55 percent on larger SOEs. By contrast, TVCEs are expected to repay loans before tax, while SOEs may still deduct these repay-  - 16 - ments from their tax bill (though the intention is to phase such deductions out). In practice, the present administration of the tax system is reported to allow considerable scope for negotiation over effective tax rates. The published statistics for TVCEs classify a surprisingly low proportion of their tax payments as income taxes (in 1989, 22.4 percent); the balance would com- prise a variety of indirect taxes and other levies.28/ 1.25 In addition to formal taxation, researchers have found that REs may be subject to a variety of ad hoc levies by local agencies. A survey con- ducted in Hunan province reportedly discovered over 60 industrial administra- tions and organs demanding funds from REs and over 100 types of expenses being collected from them (it is not known how far such charges are included in the reported state taxes).29/ Research has also shown that taxes on REs (toge- ther with remittances of post-tax profits from TVCEs) can assume the dominant position in the overall finance of the lower levels of government.30/ In part, this reflects the difficulty for these levels to levy significant explicit taxes on staple agriculture--already subject to heavy implicit taxa- tion through the compulsory procurement system: indeed, especially in more advanced regions, localities levy significant charges on REs for the specific purpose of subsidizing local farmers. As noted, the earlier World Bank-CASS study team coined the term "fiscal predation" to refer to excessive demands by local governments on rural firms: in one of the more backward counties ana- lyzed (Shangrao), local enterprises had been required to make transfers even beyond the extent of their profits, and development had been essentially choked off. 1.26 A recent analysis of provincial differences in effective tax rates revealed a remarkably wide range of variation.31/ In 1986, for example, the average proportion of TVCE profits paid in taxes ranged at provincial level from 22.5 percent to 64.4 percent. Perhaps surprisingly, though, the higher rates were not in general found in the less-developed areas (as the concept of fiscal predation would predict). With the exception of relatively high tax levels in inland Sichuan, the heaviest demands were consistently in the central coastal provinces of Shanghai, Jiangsu and Zhejiang: by 1989, when the national TVCE average was 53.2 percent, firms in the latter two prov- inces were paying 74.9 percent and 77.9 percent respectively (20 percentage points higher than SOEs in the same provinces). More work would be needed to confirm the reliability and comparability of these data,32/ but it may be significant that other coastal provinces, such as Shandong and Guangdong, which have achieved higher growth rates over recent years(para. 1.13), have 28/ See e.g., Oi (1990). 29/ Furusawa (1990). 30/ Byrd and Gelb. 31/ Ody (1991 a). 32/ Jiangsu officials have reportedly questioned the accuracy of the State Statistical Bureau data for their province; it is possible that, again, some underreporting of profits may be involved (World Bank (1991 c)).  - 17 - apparently succeeded in holding effective RE tax rates within the 40-50 per- cent range. Efficiency and Performance 1.27 As shown above, the RE sector has since 1986 made impressive contri- butions to the growth of output, employment, exports and public revenue, both in absolute terms and relative to other sectors of the economy, and in spite of enjoying comparatively restricted access to formal credit and highly trained personnel. At the same time, various aspects of the sector's develop- ment have attracted domestic criticism, especially during the rectification program launched in late-1988, and the balance of the section will review the controversies and the available evidence. 1.28 Economic Efficiency. To date, only limited quantitative results are available on the economic efficiency of the RE sector. Studies undertaken within the World Bank-CASS program found that the sample firms had achieved significant increases in total factor productivity over time and these econometric findings were supplemented by results showing that almost two- thirds of the firms had achieved moderate or extensive innovation as measured by the number of new products introduced over a five-year period. An examina- tion of the determinants of employment levels suggested that enterprises had shifted away from an emphasis on employment creation per se towards a scale of workforce closely reflecting the marginal productivity of labor. At an aggre- gate level, there was some evidence of suboptimal economies of scale, but it was emphasized that meaningful results could only be achieved through more disaggregated analysis based on homogeneous subsectors.33/ 1.29 A more recent study 34/ has compared the performance of the state industrial sector with that of collectives at township level and above over 1980-88 (i.e., including both township and urban collectives but excluding village firms). As would be expected, given the wide differences in capital intensity between the two sectors (para. 1.17), it was found that the nonstate firms achieved substantially higher capital productivity than the SOEs, while the state firms' labor productivity was higher. Among the most striking results, however, were those for changes in productivity over time. The average productivity of labor in the collective sector had increased by over 12 percent per annum over the period (state sector 5.2 percent), and that of capital by 4.5 percent (SOEs, 2.1 percent). The estimated annual growth in collectives' total factor productivity was 3.5 percent over 1980-84 and 5.9 percent over 1984-88. These results must be treated with some caution, as possibly subject to the upward bias in real output measures discussed above (para. 1.7). Such effects, however, are likely to be more serious at the vil- lage level and below than for the urban and township firms covered by the study. To date, almost no work has been undertaken in China at the subsec- toral level of detail needed to provide accurate results on comparative effi- ciency and scale economies, and this would appear to be an important priority for further research. A direct comparison of township enterprises and local 33/ Svejnar, Byrd and Zhu in Byrd and Lin (eds). 34/ Jefferson et al.  - 18 - state firms in the silk industry in Nanhai, however, found that the township firms showed the higher labor productivity, almost certainly reflecting the greater prevalence of payment by results.35/ 1.30 Rural firms' performance has been challenged additionally on grounds of poor product quality. More specific allegations include not only obsolete designs but also fraudulent attempts to pass off inferior products as "famous name brands". A legislative framework has been put in place to deal with legitimate concerns of fraud, and to compensate those who suffer actual physical or financial damage due to failure to meet recognized standards (though enforcement is presumed still to be uneven). There is little doubt that many rural firms could benefit from additional positive support in rais- ing their levels of technology and product quality--as noted earlier, diffi- culty in meeting international standards was cited by rural firms in Jiangsu as a major constraint to exporting. However, quality problems are by no means exclusive to the RE sector. An investigation of 800 firms by the State Eco- nomic Commission to assess the share of output meeting state standards found that, while large SOEs (90.0 percent) outperformed township enterprises (78.3 percent), the weakest performance (72.8 percent) was by small to medium SOEs.361 1.31 A further perception is that rural firms are unusually wasteful in their use of energy. However, the main evidence offered consists of stylized comparisons for a few key energy-intensive sectors (e.g., steel, ammonia, glass, cement, bricks and coke) between supposedly representative unit con- sumption rates in "small plants" and those considered attainable in "large" or "modern" plants.37/ There does not appear to be any body of survey-based work for China which can identify actual energy consumption in plants of dif- ferent scale or ownership (although the State Planning Commission (SPC) plans to launch such work). Chinese plants in aggregate still fail to achieve energy standards observed internationally (though the energy intensity of GNP has been declining in recent years). Most industrial subsectors noted for heavy energy use, however (including practically all steel and ammonia and most glass), are concentrated in the state sector: the major exceptions being brick-making (where REs account for four fifths of national output), and cement and coke (in each of which rural firms supply about one quarter of the total). Anecdotal evidence suggests wide variation in the technologies applied in these subsectors. In brick-making, for example, many village pro- ducers still employ traditional, energy-inefficient batch kilns, but more advanced REs have introduced improvements such as rotating kilns and "internal firing" (which involves incorporating coal in the clay and can save one third of energy use). Similarly, traditional beehive coke ovens (both energy-inef- ficient and polluting) have been largely phased out in the more populous coas- tal regions and now operate primarily in major inland coal-producing regions (e.g., Shanxi province). Continued survival of such inefficient facilities 35/ Song Lina in Byrd and Lin (eds). 36/ China Daily, September 4, 1987. 37/ See estimates for 1980 in World Bank (1985) and for 1987 in World Bank (1990 c).  - 19 - probably largely reflects the severe bottlenecks still applying to domestic freight transport. 1.32 Compared to SOEs, nonstate enterprises generally face considerably stronger financial incentives to economize on energy consumption, as the non- state firms' requirements are purchased almost entirely outside Plan allocations, at prices that in some cases probably exceed long-run marginal supply cost. One 1987 report 38/ estimated market-to-plan price differ- entials for electric power of between 3.4:1 and 4.6:1, while market prices for coal can range between twice and four times Plan prices. Given their hard budget constraints and high degree of market orientation, REs must be expected to respond positively to such incentives. There is some evidence that deficiencies and lack of competition in the domestic capital goods sector constitute some of the main constraints to better energy efficiency: a 1986 review of the cement machinery manufacturing sector, for example, found that even the best suppliers lagged behind international standards in technology, product quality and operational efficiency, and most of the sector's output continued to comprise energy-inefficient plants.39/ While official policies in recent years have leaned towards greater emphasis on administr- ative mechanisms to promote improved energy efficiency among end-users, the strength of the financial incentives already provided, at least to nonstate firms, suggests that public policy towards this sector might appropriately focus more on provision of improved information on technological alternatives and on promoting modernization and greater competitiveness within the capital goods sector. 1.33 Relationship with Other Sectors. Internal controversy over the relationship between rural enterprises on the one side and state firms and the agricultural sector on the other to a large degree reflects the currently incomplete nature of market-oriented reforms, primarily in the pricing of industrial goods and raw materials and staple agricultural products. Competition between REs and SOEs, both in product markets and for supplies of factors and raw materials, has certainly been unequal, though current advan- tages are not all one-sided. As shown above (para. 1.23), earlier tax concessions favoring REs over SOEs seem by now, on average, to have been largely eroded. SOEs have enjoyed more favorable access to budgetary finance, bank loans, long-distance (especially rail) transport, energy supplies and trained personnel. Transport, energy and raw materials, in particular, have been publicly provided through the Plan, usually at prices far below market- clearing levels. On the other hand, SOEs' output prices have also often been held down quite stringently by administrative controls. The higher market prices obtained by REs for their output have at times permitted them to outbid SOEs for less effectively controlled raw material inputs. These conflicts have also sometimes had a provincial overlay, as when silk cocoon-producing regions, seeking greater downstream integration, clashed during 1988 with traditional coastal processing centers over supplies. Such conflicts have also in part reflected an implicit cross-jurisdictional competition for tax revenues, generally collected disproportionately at later processing stages. 38/ Far Eastern Economic Review, June 4, 1987. 39/ World Bank (1987 a).  - 20 - The authorities attempted under stabilization and rectification to "tilt" material supplies towards the SOEs, though there is little reliable inform- ation on the effectiveness of these efforts. As noted, such opportunities for "unfair competition" are clearly endemic to the present "two-tier" price sys- tem, and provide part of the rationale for swiftly moving to single-tier mar- ket-determined pricing mechanisms. The other area where SOEs' competitiveness is artificially constrained concerns the extensive welfare benefits they cur- rently provide to their workforce, including the cost of unfunded pensions paid to their retirees. There is widespread official recognition of the need for further policy reforms to alleviate the burden of such costs on individual SOEs. 1.34 The relationship between REs and agriculture is another in which blame appears misplaced. It is true that labor has flowed from farming to REs, but this is fundamental to relieving the pressure of the agricultural population on the land and permitting farm labor productivity to rise. REs' demand for labor, capital and land is not in general the result of any artifi- cial stimulus favoring the nonagricultural firms. To the extent that finan- cial returns to labor are as a rule higher in nonagricultural occupations than in farming, much of the explanation should be sought in the continued use of compulsory procurement of staple crops as an implicit means of taxing agricul- ture (recent World Bank research indicates that, in 1987, the mixed average procurement prices for rice and wheat yielded daily returns to farm labor that were, respectively, 25.7 percent and 45.5 percent of the levels farmers would have received if international prices had applied).40/ The focus of future reform efforts thus needs to be on improving the attractiveness of agriculture rather than seeking artificially to constrain RE development--recent measures to raise prices of staple products and to commence development of more market- oriented distribution networks represent promising moves in the right direc- tion. 1.35 Environmental Aspects. There is widespread anecdotal evidence of increased rural water pollution (and, to a lesser degree, atmospheric contami- nation) resulting from rural industrialization, but the dimensions of the problem are not well documented. One recent survey indicated that rural firms accounted for only 10.7 percent of wastewater discharges, 9.4 percent of air emissions and 11 percent of solid wastes--well below their 25 percent share of industrial output--though this in part reflects REs' greater concentration in light industry and their limited presence in the chemical and smelting indus- tries which are the leading sectors for wastewater discharges.41/ Within the rural sector, pulp and paper has been identified as the leading source of wastewater emissions (28 percent of the total), though electroplating, coke, sulfur and building materials are also seen as problematic. A survey by the Chinese Academy of Sciences confirmed the widespread belief that rural pollu- tion has become particularly severe in the area of concentrated development in southern Jiangsu and around Shanghai.42/ Even in this region, however, the 40/ World Bank (1990 d). 41/ Quoted in World Bank (1991 a). 42/ China Daily, October 13, 1988.  - 21 - proportion of REs which are serious offenders appears to be comparatively small: a comprehensive survey of Jiangsu's REs during 1989 identified just 3.8 percent of the firms as "serious" polluters (and a further 28.5 percent as "light" polluters).431 1.36 Few direct comparisons exist between the environmental impact of state and nonstate firms within the same sector. A study of wastewater dis- charges per yuan of output between TVEs and "all industry" (in food processing, textiles and pulp and paper in five municipalities) found that, while TVEs caused more pollution (sometimes substantially more) in all sectors in three of the cities, in the other two the rural firms consistently had emission levels below "all industry.""/ Questions can be raised about the comparability of the results, and better data are clearly needed [the Ministry of Agriculture (MOA) and the National Environmental Protection Agency (NEPA) are currently collaborating on a major survey of RE pollution control], but the above results suggest significant local variation in regulatory efforts. Future policy development will need to take account of several elements generic to rural industrialization: (a) the small size of the typical firm and the large enterprise population inevitably raise the administrative cost of regulatory efforts (it is reported that less than 30 percent of the pollu- tion levies due in many rural areas is actually collected); and (b) the close links, particularly between collectives and lower-level administrations, can raise issues of conflict of interest in the aggressive enforcement of environ- mental regulations. These factors suggest the need to concentrate regulatory efforts on priority plants, subsectors and localities. They also support the recent policy decision by NEPA to vest enforcement responsibility at one administrative level above the ownership of the relevant enterprise. II. REFORMS AND EXPERIMENTS 45/ 2.1 In early 1988, central authorities drew up a medium-term reform agenda designed to clarify and improve the operating climate for rural enter- prises. Underlying the agenda were two main priorities. First was the need to provide a clearer legal framework for the establishment, capitalization, operation and dissolution of nonstate enterprises of different ownership forms. The most obvious anomaly concerned the status of privately owned enterprises with more than the prescribed limit of eight employees, whose legitimacy then owed much to local discretion (para. 1.5). However, the legal framework was weak more generally for recognizing multiple ownership claims to an enterprise, thus leaving unclear, for example, the rights of workers who invested in a "collective" enterprise, and possibly helping to inhibit the development of firms across jurisdictional boundaries. 43/ Xinhua December 5, 1990. 44/ World Bank (1991 a). 45/ In addition to cited sources, this section draws on interviews in China over the period 1988-91, connected with the preparation of World Bank (1990 e; 1991 b).  - 22 - 2.2 The second basic theme was the desire to improve further REs' access to official services and programs in such areas as credit, staff training, technological support, information and export promotion, where, as indicated in Section I, government efforts have in the past generally been largely con- centrated on SOEs. In support of these two broad aims, the proposed reform program envisaged a combination of initiatives at the national level with trials of additional measures in three "experimental zones." The overall program, as set out in a Statement of Reform Policies communicated to the World Bank,46/ was projected to include: (a) adoption of a private enterprise law designed to encourage and guide such firms' development, ensure their equal status and treatment vis-&-vis tax and credit, and promote fair competition between firms of different ownership types; (b) adoption of a law on rural collective enterprises to clarify their property relationships, and promotion of management contracting and joint-stock organization; (c) accelerated implementation of an ongoing effort (the Spark program) designed to upgrade REs' level of technology and staff training; (d) programs to facilitate RE exports, including foreign trade system reforms; and (e) financial reforms, including expanded financial services to REs, adoption of bankruptcy regulations and other measures to promote capital mobility. 2.3 The three experimental zones comprised Wenzhou in southern Zhejiang, already noted for its promotion of a largely private model of RE development (para. 1.5); Fuyang in northern Anhui, an area with a relatively mixed owner- ship pattern, selected in part as a model for other poor labor-surplus regions; and Zhoucun, a comparatively highly industrialized area of Shandong province, whose RE sector was-primarily collective in ownership. It was anti- cipated that a number of the measures on the national agenda would first be applied on a trial basis in the experimental zones: e.g. joint-stock enter- prises primarily in Wenzhou and Zhoucun, bankruptcy regulations specifically in Fuyang. Additional experiments would address the clarification of relations between REs and government agencies, the further development of support services and the systematization of arrangements for the temporary export of labor to more advanced regions. In each zone, an Experimental Zone Office (EZO) was established to coordinate the local program. Review of Implementation 2.4 Private Enterprise Regulations. China's 1982 Constitution recog- nized the "individual economy of ... working people" as a complement to the public economy and offered state protection to its lawful rights and inter- ests, but individual firms remained officially subject to employment ceilings. 46/ World Bank (1988).  - 23 - In April 1988, a constitutional amendment provided equivalent legitimacy to "the private sector." Subsequent regulations, adopted in June 1988, differen- tiate "private" from "individual" firms on the basis of employment exceeding the previous limit of eight persons. Private enterprises may be either sole proprietorships, partnerships or limited liability companies: the latter are restricted to no more than thirty investors (though exceptions may be granted) and their stock may not be offered to the public or disposed of without the consent of a majority of the other investors. Bankrupt enterprises are to liquidate their assets and pay their creditors: further measures on bank- ruptcy, to be formulated separately, have not in fact been issued.47/ Pri- vate enterprises enjoy the right to hire and fire workers, to enter into con- tracts, etc. They may also enter into Sino-foreign equity joint ventures, processing contracts or compensation trade (points that had earlier been the focus of considerable controversy). Reflecting concern over ad hoc local levies, the regulations underline private firms' right to refuse demands that they provide financial or material resources or manpower, except as provided by state laws and regulations. Stipulations on profit retention and directors' remuneration apparently reflect the authorities' concern that entrepreneurs might adopt a short-term, high-consumption approach: no less than 50 percent of post-tax profits are to be retained as a production development fund, which may be used to expand production, repay loans, make good losses or invest in other enterprises. Overall, the 1988 regulations should in principle offer many existing enterprises with more than eight work- ers the chance to reregister and thus emerge from their previously gray legal status, and also provide a legally-sanctioned means to organize firms with multiple individual investors. The normal statistical series do not yet dif- ferentiate between private and individual firms, but it is reported that 225,000 private enterprises existed by the end of 1988, half of them sole proprietorships, with total employment of 3.6 million people (on average, twice the size authorized for individual firms).48/ 2.5 Rural Collective Regulations took two years longer to be approved (June 1990), a possible indicator of the controversies surrounding RE develop- ment during 1988-90. Ownership rights to TVCEs are defined as residing in the entire peasants' collective of the township or village, and are to be exer- cised either through the local representative congress or through the collec- tive economic organization representing all peasants. Collectives may "attract investments from shareholders on the condition that the nature of the enterprises does not change." This appears to imply that such shareholders would not enjoy a formal share in the control of the enterprise. As with the private firms, reference is made to liquidation in the event of bankruptcy, but no more detailed regulations have appeared. There is a similar stress on the right to reject unwarranted financial demands: there are also stipula- tions on the allocation of post-tax profits, with a minimum of 60 percent to be retained by the enterprise and used primarily for development purposes (expansion and modernization)--a minority may be used for welfare or bonuses. The retention provisions may be interpreted as an attempt to combat predation 471 Although trial bankruptcy regulations for REs were drafted in Fuyang, they were not officially adopted. 48/ Gold in Davis and Vogel (eds.) (1990).  - 24 - by local governments or management committees on collective enterprises. By contrast to the private enterprise regulations, and perhaps in part because of policy shifts in the intervening two years, the collective regulations place much greater emphasis on the need for all levels of government to strengthen "guidance, management, supervision, coordination and service" to enterprises, including the formulation of plans for enterprise development, and the need for enterprises, on their part, to comply with state industrial policy. Par- allel regulations issued in May 1990 encourage formal management contracting by rural collectives, and provide guidance on inclusion of quantified targets, the linking of remuneration to performance, the bidding process and the rights and responsibilities of both parties. 2.6 Joint Stock Organizations. Although no definitive national regula- tions on joint stock organizations have to date appeared, experiments with "shareholding cooperative enterprises" have been ongoing in the experimental zones. In Wenzhou, approximately 22,000 shareholding enterprises have report- edly been formed since the mid-1980s, under a series of local regulations. The most common (and frequently very successful) model in this region seems to involve a relatively small group of individual investors, who expect to play an active day-to-day role in management and who would probably not normally wish to dilute their control by issuing equity to the general public (as such, the Wenzhou model may provide a portent for the limited liability companies likely to be organized under the national private enterprise regulations). In Zhoucun, experience is being gained primarily with the conversion of collective enterprises into a shareholding form: of 811 shareholding enterprises recorded in 1990, 48 were former township collectives and 249 former village collectives (the balance were joint-household firms, but these appear to have been much smaller in size). The EZO has been providing assist- ance with "property rights clarification" to aid firms in the allocation of shares, helping to undertake a historical examination of different parties' inputs to the firm together with prorating of past retained earnings. By contrast to Wenzhou, share-ownership in Zhoucun remains heavily concentrated in the public sector: in 1989, 94.2 percent of shares were held either by state bodies, township or village collectives or other corporations or legal persons, the balance being owned by employees or other individuals (local rules currently prescribe a maximum 30 percent shareholding for employees). 2.7 After a period when the central authorities appeared ambivalent on the extension of joint-stock principles beyond the zones, it is understood that during May 1991 guidelines were issued elsewhere in the country author- izing continued experimentation with both enterprise leasing and shareholding systems (the latter especially for newly established enterprise groups and cooperative enterprises).49/ The experience of the zones suggests that a number of issues are likely to arise. Both responsibility for administrative registration and supervision and corporate income tax schedules differ between private and collective firms. In the latter case, private firms pay a flat rate of 35 percent, while collectives face a progressive schedule rising from 10 percent to 55 percent: the tax status of joint-stock companies appears to remain undefined by the center. In seeking to help "regularize" the function- ing of shareholding firms, the zones have also identified weaknesses in 49/ Information from Nanjing Business School (1991).  - 25 - accounting systems (and limited experience in the valuation of assets) as major bottlenecks: nongovernmental accounting associations are starting to develop, but their resources may as yet be inadequate to meet potential demand. 2.8 To date, the zones can offer only limited experience in the develop- ment of a formal market for the sale of enterprise securities (a field in which, more broadly, few recent precedents exist within China). In Wenzhou, the EZO has promoted the establishment of an Equity Transfer Consulting Ser- vice, which became operational in October 1990. The service is prepared to act as a broker for the seller, reportedly requiring disclosure of financial statements and charging a 1.5 percent service fee, and is in principle chartered to handle transfers from one shareholder to another, the issue of new equities or the issue of "redemption shares." The latter, which are essentially one-year, fixed-interest bonds, accounted for all of the Y 2 million worth of transactions (on behalf of five firms) handled by the service in its first three months of operation. Further extension of share- holding concepts could provide an appropriate mechanism for organizing larger private firms, worker-owned cooperatives and joint ventures across jurisdic- tional lines, though the Zhoucun experiment leaves open the issue of whether existing collectives are likely to be transformed into a more mixed form of ownership (local governments and management committees will probably resist ceding majority control at least of profitable firms). A survey conducted by the Zhoucun EZO found shareholding firms to have above-average profitability and productivity but, failing baseline data, the direction of causality is unclear. 2.9 RE Support Services. The Spark Program, launched by the State Sci- ence and Technology Commission (SSTC) in 1985/86, is an initiative to involve the technology sector in support of RE development.50/ Implemented nation- wide through SSTC's lower-level counterpart Science and Technology Commissions (STCs), the Program undertakes: (a) enterprise modernization subprojects for REs, which introduce improved technology and management methods with high potential for replication, and are designed and implemented with assistance from technical partners such as research institutes or larger firms; (b) focused training programs ("Spark training"), in technical and managerial areas, for RE staff; and (c) reorientation of the technology information net- work of SSTC's affiliate, the Institute of Science and Technology Information of China (ISTIC), to meet business information needs of REs and other small enterprises. The Program also helps selected localities with support for preparation of comprehensive development strategies. An original goal of developing improved capital goods ("complete sets of equipment") suitable for RE adoption does not, however, appear to have been institutionalized in a systematic manner. The Program has already approved some 14,000 enterprise modernization subprojects for financing, and officials estimate that Spark training courses may have reached 4 million people. Although formal informa- tion/evaluation systems appear weak, early indications, based on inspection visits by World Bank specialists, are that most enterprise subprojects are well chosen and efficiently implemented, and that Spark training is meeting needs not adequately addressed by other existing programs (employers typically 50/ World Bank (1990 e).  - 26 - pay at least half the cost of such training). Although Spark was initially financed largely by the central government, such funding fell off fairly rap- idly and at this stage Beijing's support primarily takes the form of policy advice to lower levels of government which, together with the banking system, now provide the overwhelming majority of Program resources. While the Program was able to maintain its operations during the 1988-90 stabilization period, the overall shortage of credit appears to have obstructed earlier hopes of acceleration. As Spark's direct support can reach only a very small propor- tion of the RE population, greater focus appears indicated on programs of diffusion. In addition, SSTC itself recognizes that to date most support has been concentrated on TVCEs (plus county collectives) and has urged greater efforts to support lower-level firms. 2.10 Other attempts to upgrade support services to REs have been under- taken on a trial basis, especially in the Fuyang zone. Among the more impor- tant was the establishment of an organized market in raw materials to serve REs. Lacking access to Plan allocations, rural firms have had to invest sub- stantial resources in locating supplies through alternative channels, and it was anticipated that a permanent market would improve the efficiency and transparency of dealings. The market, established in Fuyang in late 1988, deals primarily in cement, steel, timber and vehicles: there are reportedly 30 traders handling cement alone. In 1990, over 20,000 tons of cement and over 7,000 tons of steel passed through the market. The EZO itself credits the market with narrowing the differentials between state and free-market prices (for steel, the ratio shrank from 2:1 to 1.15:1), though part of this reduction may be due to more depressed demand resulting from the macroeconomic climate: it is also reported that the market became subject to price regula- tion under the stabilization program. The national Bureau of Township and Village Enterprises (BTVEs), an organ of MOA, is considering establishment of raw materials supply services for REs in other parts of the country. The Fuyang EZO additionally played a facilitating role in the establishment of a financially self-sustaining TVE training center providing both pre-service and in-service training in business-oriented subjects. 2.11 Export Promotion. As shown earlier (para. 1.14), expansion in RE exports over the late 1980s was dramatic. Key measures in stimulating enthusiasm for exporting included regulations in 1984 and 1985 to formalize arrangements for "retention rights," under which exporting enterprises and local governments receive a proportion of their export earnings in the form of rights to purchase hard currency, and the opening to domestic enterprises, in 1988, of the Foreign Exchange Adjustment Centers (FEACs), at which surplus foreign exchange may be converted to renminbi at a floating exchange rate. A World Bank team found no evidence of discrimination against REs in retention rights or access to the FEACs in Jiangsu in 1990,51/ though the reiteration of their rights in a central policy statement in 1988 (para. 2.12) suggests there may have been earlier difficulties in some places. Despite this pro- gress, disincentives to exporting still remained: REs interviewed in Jiangsu in 1990 (para. 1.15) still complained about monopsonistic pricing practices by local Foreign Trade Corporations (FTCs), and few REs could expect to reach the export volume needed to qualify for direct export rights (a standard subject 51/ Montes-Negret; Panagariya.  - 27 - to local variation, although exports of $3.0 million is reported to be a typi- cal threshold).52/ 2.12 A central package of export incentives targeted at REs was announced in November 1988.53/ Some points essentially involved the reiteration of existing policies, presumably in the face of local contraventions: thus, enterprises' right to their share of foreign exchange retentions was defended against unauthorized demands by localities and departments, as was their right to sell such foreign currency through the FEACs. Many of the other incentives left scope for local interpretation. Both access to export credits at the same interest rates as SOEs, and the possible reduction or exemption of customs duties and tax on the import of technology and advanced equipment, were directed at "export-oriented" enterprises. This term has apparently not been explicitly defined at the central level, although officials state that common practice would grant the designation to firms which either export at least 60 percent of their output or achieve an absolute export volume of over $1.0 million per annum.54/ The package also included provisions to stream- line REs' access to imported inputs and components required for export produc- tion, by substituting customs verification for the normal import license requirement, but again it appears that such arrangements will be subject to case-by-case approvals. No information is available on the extent to which localities have in practice extended these incentives to REs. 2.13 Financial Sector Development. As seen earlier, the period prior to adoption of the Statement of Reform Policies was one in which REs' access to credit through formal banking channels was gradually expanded (para. 1.18). It also witnessed considerable institutional diversification in the financial sector, including the development of small nonstate deposit-taking, loan-mak- ing bodies in both urban and rural areas. Urban credit cooperatives (UCCs) grew particularly rapidly between 1986 and 1989, when their assets rose from Y 3.2 billion to Y 28.4 billion, and have emerged as significant lenders to collectives and private firms (in 1989, 65 percent of loans outstanding were to collectives and 9 percent to individual proprietors). A similar role is performed in rural areas by various types of nongovernmental financial associ- ations (NGFAs), sometimes referred to as rural financial service units (RFSUs). No national data are available for this sector, though in Wenzhou, a leader in their development, there were by 1990 some 42 such units, with com- bined deposits of over Y 300 million. 2.14 The stabilization measures implemented from late 1988 interrupted the growth in REs' share of lending by the formal banking sector (Section III). In addition, the central authorities announced in 1988 a suspension in approvals of new NGFAs (though this appears to have been aimed primarily at Trust and Investment Corporations formed by local governments, and it is not clear how strictly or for how long it may have applied to other types of NGFAs). Review of a number of NGFAs in Wenzhou and Fuyang zones suggests that 52/ Interview with BTVEs officials (1991). 53/ China Economic News, November 1988. 54/ Interview with BTVEs officials (1991).  - 28 - such units can play a useful role in improving competition for deposits and meeting demand for small loans which might not be of interest to the special- ized state banks. It also, however, confirms that arrangements for prudential supervision of nonstate banking institutions require considerable strengthen- ing: rules on concentration of lending to individual borrowers, for example, do not appear to be adequately understood or enforced, which can leave at least some NGFAs susceptible to pressure from local authorities or corporate shareholders to indulge in risky lending practices. 2.15 Additional Experiments. Fuyang zone included a number of additional components of wider interest. Under one, designed to address reported bureau- cratic obstacles to enterprise formation, a one-stop shop approach was adopted to facilitate the registration of new enterprises, with representatives of all the relevant agencies assembled together in a specific location on a weekly basis to give on-the-spot approval: following a trial period, the approach is now being extended throughout the prefecture. Further experiments designed to eliminate predatory behavior by official bodies addressed the rationalization of road tolls and market fees: the EZO helped establish that, in each case, only a single agency would be authorized to levy such charges, and the prefec- ture then published widely the types and levels of fees permitted--this was followed by rectification of such practices in 300 markets and 200 road check- points, and is reported to have resulted in a sharp increase in the volume of transactions in local free markets. This program is being extended throughout Anhui province and has also received central recognition. 2.16 The Fuyang program also included the establishment of five indus- trial zones, intended to promote a more rational spatial pattern of develop- ment, compared to the frequently scattered location of REs, and thus permit more efficient shared provision of common services, including effluent treat- ment. As of early 1991, some 61 enterprises were reportedly operational in the five zones, and a further 20 under establishment. The zones appear to have succeeded in establishing a model that is largely self-managing and self- financing: in Taihe zone, for example, which was launched in 1987 and currently has 13 factories, new enterprises pay Y 13,000 per mu (Y 195,000 per hectare) for land use rights and infrastructure connections, of which Y 6,000 is for compensation to the former occupants and Y 7,000 covers power supply, water, effluent treatment and roads. The relevant funds are handled by a management committee, representing the enterprises in the zone. It is, however, striking that, with the exception of one county-level firm, all the enterprises in the zone are essentially private in nature (joint-household or shareholding enterprises). As might be predicted, the strong community ties of the collective firms make it far less likely that they will locate outside their home territory. 2.17 Of special interest to other poor "labor surplus" regions is the program for labor export management launched in Linquan county within Fuyang zone. As noted earlier, geographical mobility of labor has generally been low in China since 1949, which has inhibited progress in raising living standards in areas with adverse ratios between population and resource base. Barriers to mobility eased from the mid-1980s, when a sizeable "floating population" became established on an informal basis in many cities, though attempts were made during the stabilization and rectification campaign of 1988-90 to reen- force residency regulations. In Linquan, a Labor Export Management Office,  - 29 - established in 1988, identifies employers with a demand for additional workers in more advanced, labor-deficit areas (typically the Northeast, Shanghai, Jiangsu or Beijing) and negotiates a three-year contract defining the level of educational qualification or other test the candidates must meet, together with details of pay scales, housing conditions, medical and accident insurance and any on-the-job training to be provided. In turn, the Office uses its contacts with local farmers' associations to recruit young men and women interested in the openings, and provides them with any necessary predeparture training. The Office contends that its own involvement as a signatory to the contract provides the migrants much stronger protection and security than if they had sought work in other regions as individuals. As of 1990, it had processed 23,000 new migrants, and had also set up files on an additional 44,600 workers who had left the county before the organized scheme commenced, for whom it was seeking to negotiate similar protection. Press reports indi- cate interest on the part of the Ministry of Labor in replicating organized temporary migration schemes.55/ The broader issues raised by possible ero- sion of the segregation of rural and urban labor markets have, however, yet to be adequately addressed by public policy.56/ III. STABILIZATION. RECTIFICATION AND REVIVAL (1988-90) 57/ 3.1 Objectives. From September 1988, the central government adopted a range of policies to stabilize the overheated macroeconomic situation, includ- ing measures to curb the growth of money and credit, to cut directly public investment, to postpone previously contemplated price increases and to encourage savings through indexation of deposits. The credit squeeze was implemented largely through quantitative ceilings, including, in late 1988, the introduction of mandatory ceilings for total lending by specialized banks to REs, reportedly for the first time. Early 1989 saw ceilings introduced for total lending by the rural credit cooperatives (RCCs), reinforced by higher mandatory RCC redeposits.58/ In early 1989, it was reported that fixed asset loans to REs would be cut by half 59/ and that total bank credit to REs would be frozen, in nominal terms, at its 1988 level.60/ 3.2 Beyond selective credit controls, central statements between late 1988 and the end of 1989 gave increasing emphasis to the use of administrative measures to "rectify" the RE sector. The Fifth Plenum of the Thirteenth Central Committee in November 1989, while noting that TVEs had become a major 55/ China Daily, May 10, 1991 and June 21, 1991. 56/ See Ody (1991 b). 57/ A more detailed examination of this period is provided in Ody (1991 a). 58/ World Bank (1990 b). 59/ Xinhua, January 21, 1989. 60/ World Bank (1990 f).  - 30 - pillar of the rural economy and an important component of the national economy, pointed to such problems as "poor product quality, backward manage- ment, low economic benefits and competition for raw materials and energy supply with large- and medium-sized backbone enterprises owned by the state" as creating the imperative for "readjustment, consolidation, transformation and improvement". Future development of TVEs should: (a) draw on farm and other raw materials available locally but avoid competition with large indus- tries for materials and energy; (b) exploit opportunities for labor-intensive activities, traditional technologies, and export development; and (c) coordin- ate with and serve the needs of large industry. TVEs "characterized by high consumption and poor product quality, having serious pollution problems, com- peting with large enterprises for raw materials and energy supply and yielding very poor economic benefits" should be "closed, stopped, merged or converted into making other products." 3.3 Implementation. A clear assessment of the impact of the stabiliza- tion and rectification program on the RE sector is complicated by the partial and contradictory nature of the evidence available to date. In general, press reports, particularly during 1989, portray a more critical situation than is apparent from the official statistics so far published for the period. The squeeze on bank credit was at its tightest during 1989, especially in the middle two quarters: ABC, the main bank lending to REs, reportedly cut loans to the sector by about 15 percent in nominal terms during the first half of the year,61/ and by August, an MOA spokesman was forecasting that 30 per- cent of all REs might go bankrupt during the year due to shortages of raw materials and credit.62/ In many cases, however, rural firms were able to mitigate the effects of the squeeze on bank lending by locating alternative sources of finance--borrowing from their employees, private lenders or the RCCs (which were subject to overall but not sectoral lending ceilings): REs rose from 45.3 percent of total RCC lending to 49.0 percent over 1989, and this--together with an easing of ABC's position over the second half--enabled total ABC-plus-RCC lending to REs to increase by 14.9 percent (only moderately below total credit expansion during the year). 3.4 The austerity measures nonetheless had a dramatic impact on RE output. This is only partially discernible from the annual indicators, which show, for example, estimated real RE industrial growth declining from 37.3 percent in 1988 to 14.3 percent in 1989 (and nominal results for construction and commerce implying a significant real decline). A clearer picture of the cumulative impact of the program during 1989 is given by the monthly data for real industrial output (for firms at village level and above). As shown in Chart 3.1, while industrial TVCEs' growth over the previous year remained above that of other ownership categories (SOEs and the broader collective sector including urban firms), the rural firms' growth declined more sharply -- from close to 40 percent at the start of the austerity program to zero a year later. 61/ China Daily, August 10, 1989, reported that ABC had invested Y 6 billion less in REs than in the previous year. 62/ China Daily, August 3, 1989.  - 31 - Chart 3.1: GROWTH OF GROSS INDUSTRIAL OUTPUT 1987-90 Percentage Change Over Same Month of Previous Year 40 40 _ _.__ _ 30 - -__4_ _------- ----- ------ 20 - -lo 10 , ,I, 1 1 14 I I 1 t I 1 1 1 ,1 1 | 1i I 1 4 7 10 1 4 7 10 1 4 7 10 1 4 7 1 1987 1 1988 I 1989 I 1990 1 State-Owned Collv-tiw* TVCE 3.5 Considerable press publicity has been given to reports of widespread closures of enterprises (and ensuing losses of jobs), and these are areas where the evidence remains confused. The highest figure for closures of rural enterprises during 1989 is an estimate of 3 million (or about 16 percent) first issued by MOA in January 1990.63/ Other sources, by contrast, have suggested that perhaps 600,000-800,000 rural firms closed as a result of the stabilization program.64/ The published official data for December 1989 appear to suggest a less dramatic outcome than the MOA estimates, as the net decline in the number of REs was a comparatively modest 195,000 (from 18,881,600 to 18,686,300). It is, however, more difficult to assess how much gross turnover of firms may underlie these relatively small net changes. Micro-level studies from earlier periods suggest that firm closure is compara- tively rare for township collectives, whose sponsoring governments will gener- ally attempt to ensure their survival. Closure is somewhat less rare for 63/ The 3 million figure has been reprinted on several occasions (e.g., China Daily, January 6, 1990; Xinhua, August 27, 1990; and China Daily, April 12, 1991), but its definition has not remained entirely consistent. Originally described as the number of firms "closed in accordance with the central government's policy to adjust rural enterprises," it was more recently cited as the number of firms that had either "closed, stopped production, merged with other enterprises or changed the products they manufacture." 64/ For example, South China Morning Post, March 7, 1990; Oxford Analytica, May 22, 1990; and information provided by Nanjing Business School (1991).  - 32 - village firms, and a fairly common phenomenon for cooperative or individual firms. In Nanhai county, Guangdong, for example, World Bank-CASS researchers found that 4 percent of collective (mostly village) firms closed during 1986, but 15 percent of private businesses. A business failure rate of around 15- 16 percent during a difficult year would not necessarily be considered abnor- mal for small firms in an international context (the US Small Business Admin- istration has estimated that, of the firms entering its data base between 1976 and 1978, 23.7 percent had closed within two years). On the other hand, assuming the official data to be accurate for the number of enterprises opera- tional at the end of year, three million closures would imply the offsetting birth of 2.8 million new enterprises--and while this is not much above the net increase in firms in 1987 (2.3 million), it seems less likely that such a birth rate would be sustained during a period of austerity.65/ 3.6 As a minimum, it seems clear that the program brought to a temporary standstill the extraordinary growth of the previous few years, which had been adding 2 million or more new enterprises and 7 or 8 million new employees to the sector each year--even the most favorable series show some falling back of employment. The reports of more widespread closures and job losses are more difficult to evaluate. Given the relative flexibility and low capital- intensity of many REs, it is very likely that a large proportion of the reported closures may have been temporary in nature. It is also possible that lower level administrations, under pressure to show results from rectification, may have reported more "closures" than actually occurred. It seems, in any case, doubtful that a high proportion of such closures as did take place was the result of direct administrative intervention, as opposed to overall credit and demand conditions. The center could only expect to achieve such a program with the active cooperation of local authorities, and both earlier research and contemporary reports stress the efforts made by lower- level governments to defend "their" enterprises. If elements like pollution, energy consumption or competition with SOEs had been governing factors, these might have been expected to show up in heavy losses among the larger indus- trial REs (i.e., TVCEs), but the data show no such pattern--nor does an exami- nation of results by industrial subsector provide much evidence of unusual losses among firms in areas noted for pollution or high energy con- sumption.66/ The view that most of the firms which closed fell victim to the general economic situation and credit controls is lent some credence by a breakdown of the much smaller number of closures in 1990, which indicates that 4 percent were closed for excessive energy consumption and 2 percent for breaking pollution regulations--reasons for the other closures are not fully accounted for but many appear to have reflected cash-flow or material supply 65/ There are similar contradictions in the area of RE employment, where some unofficial estimates have reported job losses as high as 8 or even 15 million, but the official data show a net decline of 1.8 million jobs in 1989 (1.9 percent) and a further fall of 1.0 million in 1990. See, e.g., Zweig. 66/ For details, see Ody (1991 a).  - 33 - problems.67/ It may well be that the recession-induced "shakeout" of non- state firms over the period achieved some improvement in average levels of efficiency; it should however be noted that SOEs, whose closure is almost inconceivable under prevailing circumstances, were not exposed to any compara- ble economic discipline. 3.7 Beginnings of Revival. In late 1989, the emphasis of central state- ments on the RE sector was still placed heavily on the need to control perceived negative aspects of RE development: indeed the Fifth Plenum decision from November (para. 3.2) was the definitive statement of this view. In the same month, the official news agency reported that no new REs would be permitted except for those with an export orientation or which processed agri- cultural or nonstaple materials, while ABC stated that 1990 would be a second year with no increase in RE lending.68/ In the early months of 1990, how- ever, the policy climate began to change. Considerable publicity was given to a speech in January by a prominent member of the Party's old guard stressing that rural industrialization was Party policy and a "strategic task for invig- orating the rural economy."69/ Shortly afterwards, the Premier visited Jiangsu, where he was quoted as saying that "insufficient stress has been laid" on rural industry and that its significance must again be recog- nized.70/ This was followed in March by a similar admission on behalf of SPC that "we ourselves are not sure we have fully understood the importance of rural industries."71/ At the same time, SPC reversed the earlier ABC posi- tion on lending during 1990 by announcing that bank loans to rural industry would increase over the 1989 level, and also announced an "instructive, not mandatory" target of 15 percent real growth for the RE sector in 1990 (com- pared to a 6 percent target for the industrial sector overall). 3.8 The annual government report to the National People's Congress in March 1990, makes an instructive contrast with the November 1989 decision of the Fifth Plenum. While the need for "readjustment, overhaul, transformation and upgrading" was repeated, the earlier negative emphasis was missing. Instead, the sector's contribution to developing the rural economy, expanding employment and raising living standards was recognized, and all levels of government were called on to help REs overcome their current difficulties. Local authorities were to "encourage the establishment of collectively owned village and township enterprises." Credit policies towards REs were in fact significantly eased over 1990. Combined RE lending by ABC and the RCCs over the first ten months (at Y 110 billion) was reported as 19.8 percent above the same period of the previous year. The bulk of increased lending came from the 67/ China Daily, April 12, 1991. See also Naughton, who quotes MOA sources from 1991 to the effect that about 100,000 REs were actually "shut down by the government" during 1989/90. 68/ Xinhua, November 7, 1989; China Daily, November 13, 1989. 69/ See Zweig. 70/ Nongmin Ribao, March 15, 1990. 71/ China Daily, March 19, 1990.  - 34 - RCCs (whose loans to REs rose by nearly 18 percent in the first six months alone),-but ABC lending also showed an increase (approximately 9 percent over the year as a whole). Output indicators also indicate that 1990 witnessed a revival of RE growth: rural collective industrial output rose by about 12.5 percent in real terms over the year as a whole, and by 24 percent on a December-to-December basis. 3.9 Subsequent statements indicate recognition that, with state-owned firms still showing sluggish growth in output and employment, strong develop- ment by the nonstate sector will be needed if overall targets are to be met. The 1991 Annual Plan, for example, projected overall industrial output to grow by 6 percent, but with SOEs contributing just 3-4 percent and TVEs 10-12 percent growth. MOA has. forecast a rebound of RE employment by about 1.6 million during 1991, and indicated that public financing of RE fixed asset investment would rise by 15 percent over the year. Finally, guidelines set for RE development over the next five-year plan (1991-95) call for an average growth rate of 11 percent (10 percent for industry), the absorption of 18 million additional workers (a 20 percent increase) and a near doubling of RE exports to $25 billion (19 percent per annum growth). The most recently available data suggest that at least some of these targets--even if ambitious by international standards--may prove unduly conservative under Chinese conditions: TVCE industrial output showed real annual growth of 26.3 percent over the first eleven months of 1991 (compared to 13.5 percent for the industrial sector as a whole, and 9.0 percent for SOEs). IV. CONCLUSIONS 4.1 It is, as yet, too early to state with confidence that the shift detected above in the tone of central pronouncements on the RE sector since the early months of 1990 represents a permanent change in official thinking: that, most likely, will only become apparent the next time the Chinese economy comes under severe demand pressure. But perhaps the strongest reason for believing that something more than a cyclical reaction has occurred is the fact that central leaders now talk in different terms not only about REs, but also about the SOE sector. Recent discussions of state firms no longer attempt to divert blame for their continued problems to "unfair competition" from nonstate enterprises, but instead show a new willingness to confront the need to seek solutions through extensive reforms and restructuring within the SOE sector itself.72/ A second piece of evidence is the emphasis placed within the Outline of the Eighth Five-Year Plan (1991-95) on the need for rapid future growth in the services sector (both to offset past underdevelopment of services and as a means of employment generation): services provision is an area in which the nonstate sector (especially private 72/ Among numerous recent speeches stressing the centrality of (state) enter- prise reform, see e.g., the Finance Minister's Budget Speech to the National People's Congress (March 26, 1991).  - 35 - and individual firms) has during the 1980s made a major contribution in both rural and urban areas (para. 1.8).73/ 4.2 Previous World Bank discussions on policy guidelines towards China's nonstate sector have emphasized the distortions which result from treating enterprises differently based on their ownership classification, and stressed the arguments for seeking a "level playing field" between firms.74/ The present paper has pointed to a number of areas in which discrimination in practice continues to apply, e.g., between SOEs and REs. In taxation poli- cies, for example, recent experiments suggest that the authorities intend to move in the direction of standardized tax rates between firms of different ownership types (not only SOEs and REs, but also joint ventures), and this is to be welcomed. For many raw materials and services (e.g., power, rail freight), achieving less distorted allocation must be fundamentally linked to further progress with price reform, to narrow and eventually eliminate the current wide differentials between plan and market prices, and replace remain- ing vestiges of the materials allocation system by efficiently functioning competitive markets. In the financial sector, too, there seems little doubt that effective segmentation of markets has resulted in widely differing returns to capital in the state and nonstate sectors.75/ In view of the dangers of undue influence by lower-level authorities over local bank branches, however, improving nonstate firms' access to formal credit needs to be part of a broader pattern of change in financial markets, in which greater autonomy for lenders is tempered by enhanced competition and the more effective exercise of prudential supervision.761 4.3 Several dimensions are relevant to upgrading the access of the RE sector to improved technology. Measures have already been adopted to induce domestic research and development institutes to seek a higher proportion of their financing from contracts with industry, and these should continue to be pursued.77/ Similarly, there is substantial scope for further relaxation of import restrictions which effectively shield the domestic capital goods sector from international competition, to the detriment of consuming industries. Among the more difficult challenges, however, will be the facili- tation of greater domestic mobility of highly trained and skilled specialist staff, which has long been far below international norms. In this respect, it is particularly important that ongoing reforms in e.g., social security and housing provision be designed to weaken the current ties which bind employees 73/ For a more detailed discussion of issues in the rural services sectors, see Ody (1991 a). 74/ See e.g., World Bank (1989). 75/ See Jefferson op. cit., which detects some evidence for a moderate con- vergence between returns in the two sectors over the 1980s. 76/ World Bank (1990 b). 77/ World Bank (1990 e).  -36 - to a single enterprise--for example, by guaranteeing the full portability of pension rights.78/ 4.4 Realism suggests, however, that changes in skilled labor markets may only occur quite gradually, and this indicates the importance of measures to improve human capital development within the existing rural population. Experiments such as the Puyang TVE Training Center (para. 2.10) and the Spark Training Program (para. 2.9) indicate that there is considerable willingness to pay, both among students' families and employers, for relevant pre-service and in-service training, and suggest greater efforts to mobilize such resources. At the crucial secondary level, however, controversy continues over the relative merits of a general junior secondary education (followed by on-the-job training) versus more specialized--and substantially more costly vocational secondary schooling: World Bank reports have suggested that inter- national experience does not support the current official emphasis on expand- ing vocational schools,79/ and this is an area in which it would be highly useful to survey directly the preferences of the potential employers. 4.5 More controversial still is the future mobility within China of unskilled workers. In many other economies, the possibility of migration has been a crucial escape valve for those raised in impoverished regions. For much of the post-1949 period in China, however, such mobility was actively discouraged by official policy, which specifically prohibited the migration of rural inhabitants to urban areas. The broader questions raised by the possible easing of such prohibitions go beyond the scope of the present paper- -and are to be addressed by forthcoming pieces of World Bank sector work dealing, respectively, with poverty alleviation strategies and the functioning of urban labor markets. Here it may suffice to reiterate that the nonstate sector in general has already demonstrated itself to be the leading generator of employment in the Chinese economy (in both industry and services). To the extent that job creation possibilities within currently disadvantaged areas are likely to remain constrained (by limited natural resources, unfavorable location, weak infrastructure, etc.), greater internal mobility of labor must receive serious consideration as at least part of the solution to remaining absolute poverty: on current indications, the more that can be done, at both local and higher levels, to facilitate the growth of the private and individual economy, the more likely are potential migrants to find employment (para. 1.19). 4.6 This paper has taken, on the whole, a favorable view of the growth of the rural, nonstate sector in China, and encouraged the removal of artifi- cial controls to its further development. It has also, however, sought to face some of the genuinely worrying aspects of RE development. One of the most obvious concerns is the environmental aspect of rural industrialization, where the need must be recognized for a more determined effort--not only during crisis periods--to build a permanent institutionalized basis for rural 78/ World Bank (1990 a). 79/ World Bank (1987 b).  - 37 - pollution control.80/ Rural firms' environmental record relates closely to the intimate relationship which exists between the TVCEs and local govern- ments, which, it has been suggested, can lead to conflicts of interest between regulatory and ownership functions, and which may also impose limits on the ability of TVCEs to reap potential economies of scale to the full. While no ready international parallels exist to suggest the possible future direction of the TVCEs' evolution, there is much experience to indicate the importance of clearly defined property rights, objectively protected in law, to stable long-run development. Strengthening enterprise law and clarifying property rights in China might help some existing TVCEs to outgrow their present community orientation, as well as allowing a richer range of inter- relationships between the state and nonstate sectors (including possible cross-sectoral mergers, or takeovers of less successful enterprises by more dynamic firms). If combined with broader measures of deregulation within urban markets8l/--reduction of entry and exit barriers, liberalization and development of markets in land, labor and finance--it might also extend to China's cities more of the dynamism which, in the reform years to date, has been so striking a feature of development in the rural economy. 80/ A related concern is the occupational health and safety of RE workers, reviewed in greater detail in Ody (1991 a). 81/ See Annex 2 for a brief outline of China's urban nonstate sector.  - 38 - BIBLIOGRAPHY Byrd, W. and A. Gelb (1990). "Township, Village and Private Industry in China's Economic Reform," PRE Working Paper, WPS 406, The World Bank. Byrd, W. and Lin Qingsong (ed.) (1990). China's Rural IndustrV: Structure. Development, and Reform, Oxford. Cortes, M., A. Berry and A. Ishaq, (1987). Success in Small and Medium-Scale Enterprises, Oxford. Furusawa, Kenji, (1990). "Rural Enterprises under Reevaluation," China Newsletter (JETRO), September/October. Gold, T.B., and Jean G. Oi, (1988) in Chinese Economic Policy, edited by Bruce Reynolds, Paragon. Gold, T.B., (1990). "Urban Private Business and Social Change" in Davis D. and E.F. Vogel (ed.), Chinese Society on the Eve of Tiananmen, Harvard. Ho, S.P.S. (1980). "Small-Scale Enterprises in Korea and Taiwan," World Bank Staff Working Paper No. 384. Jefferson, G.H., (1988). "Productive and Allocative Efficiency Within Chinese Industry," April (processed). Jefferson, G.H., T.G. Rawski and Yuxin Zheng (1990). "Growth, Efficiency, and Convergence in China's State and Collective Industry," July (processed). Lau, L.J. (ed.) (1990). Models of Development, ICS, San Francisco. Liedholm, C. and D. Mead, (1987). "Small-Scale Industries in Developing Countries: Empirical Evidence and Policy Implications," Michigan State University International Development Paper No. 9. Little, I.M.D., D. Mazumdar and J.M. Page. (1987). Small Manufacturing Enterprises, Oxford. Montes-Negret, Fernando, (1990). "Rural Industrialization in Jiangsu," October (processed), The World Bank. Naughton, B. (1992). "The Chinese Economy: On the Road to Recovery?" in Joseph, W.A. (ed.), China Briefing. 1991, Westview. Nolan, P. and Dong Fureng (eds.) (1990). Market Forces in China: Competition and Small Business--the Wenzhou Debate, Zed Books, London and New Jersey. Ody, Anthony J. (1991a). "China: Rural Enterprise, Rural Industry, 1986-90," January 4 (processed), The World Bank.  - 39 - Ody, Anthony J. (1991b). "Overcoming Poverty in China: the Role of the Nonstate Sector," Transition, Volume 2, No. 3, March, The World Bank. Oi, Jean G., (1990). "Economic Management and Rural Government," April (processed). Panagariya, Arvind. (1991). "Unraveling the Mysteries of China's Foreign Trade Regime: A View from Jiangsu Province," Policy Research Working Papers, WPS 801, November, The World Bank. Vogel, E.F. (1989). One Step Ahead in China: Guangdong Under Reform, Harvard. The World Bank, (1985). "China: The Energy Sector," The World Bank. The World Bank, (1987a). "China: Cement Machinery Manufacturing Subsector Study," Report No. 6551-CHA, February 5, The World Bank. The World Bank, (1987b). "Technical/Vocational Education for China's Development," Report No. 6789-CHA, August 7, The World Bank. The World Bank, (1988). "China: Rural Sector Adjustment Loan," President's Report, P-4801-CHA, June 3, The World Bank. The World Bank (1989). "China--Rural Industry: Overview, Issues, and Prospects," Report No. 7267-CHA, March 1, The World Bank. The World Bank, (1990a). "China: Reforming Social Security in a Socialist Economy," Report No. 8047-CHA, June 25, The World Bank. The World Bank, (199Ob). "China: Financial Sector Review," Report No. 8415- CHA, June 29, The World Bank. The World Bank, (1990c). "China: Efficiency and Environmental Impact of Coal Use," Report No. 8915-CHA, September 21, The World Bank. The World Bank, (199Od). "China: Managing an Agricultural Transition," Report No. 8652-CHA, October 17, The World Bank. The World Bank, (1990e). "China: Rural Industrial Technology (Spark) Project: Staff Appraisal Report," Report No. 7484-CHA, October 22, The World Bank. The World Bank, (1990f). "China: Between Plan and Market," The World Bank. The World Bank, (1991a). "China: Environmental Strategy Paper," Report No. 9669-CHA, The World Bank. The World Bank, (1991b). "China: Rural Sector Adjustment Loan and Credit," Project Completion Report, June 10, The World Bank. The World Bank, (1991c). "China: Economic Development in Jiangsu Province," Report No. 9354-CHA, June 25, The World Bank.  - 40 - The World Bank, (199td). "China: Urban Housing Reform," Report No. 9222-CHA, June 26, The World Bank. Wang, Yukun, (1990). "The Size of Housing Subsidies in China," paper prepared for the Housing Reforms in Socialist Economies Policy and Research Seminar, June 2-13. Wong, Christine, (1988). "Interpreting Rural Industrial Growth in the Post- Mao Period," Modern China, Vol. 14, No. 1, January. Zweig, D., (1991). "Rural Industry: Constraining the Leading Growth Sector in China's Economy," in China's Economic Dilemmas in the 19909, Joint Economic Committee, Congress of the United States.  STRUCTULE OF RURAL ENTEPRISE SECTOR - 1988 CaRCEc, FWoD INDUSTRY AGRICULTURE CONSTkUCTION TRANSPORT 4D O*lE TOTAL CT 473.418 bln 010) 1 26.97 bin products for 0.24 in enter- 0.42 Uin jk export prises (0.22 aln)/b snt.rprisY 4 T(WSRIP 14.93 g in lb 24.90 sin L total 3 xpeud4 liure COLLECTIVES workers workers 1 184.67 bin GVO 1 236.65 bUn La p 312r t6 bin cL Y 197.19 al) Lb 1.00 snl enterprises rovenup (0.97 Uin) 123.65 bin la 35.07 vila worker* 0.23 uin *nterprises 0.07 iin enterprises 0.06 enterprises 0.23 sin enterprises taes3; 35.07 inl verkers (33.37 in) 2.50 uin workers 8535 %in workars 1.00 &In warkers 1.82 sin vorkers V 2592 bin Is Y 322.67 bin revene Y 11.77 bln revenue 1 45.77 bin revenua * 7.39 bin revene 1 353.3 bin revenue CY 378.21 bin (0VO) (Y 12.38 bin VO) (1 58.21 bin GVO) (Y 7.65 bin VO) (C 17.03 bin (1O0) 1 S4.12 bin la 0.73 sin Lb 1.17 Uin la wages enterprise enterprises 1 158.43 bInuA VILLAE 18.42 sin Lb 24.04 rs La orig. valu fed aset COLLECTVES workgaer wrkers 1 154.05 bin La 1 170.36 bin lb V 186.57 bli la circulating fliEds GVO l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~revoenu 1 73.24 bin Is bonk loans Uy difforencs, 0.69 ala enter- 17.29 sn enterprises JOIW RWAL prises ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~46.51 lsn workers 1 40.01 bin C00 by differecesee By diff,reaes 46 Sn e r Not included in 2 d s d30 bf Gin)VO 21.96 sin urkerks enterprise 0.89 uin anterprise 3.67 sln enterpriece 6.00 sna enterprirs (23.66 sin) statistics V 7.30 blu ta e Ct 121.08 bin 6VO) 6.30 sli yorkers 5.4 Suin workers 12.41 esn workers 5.79 alu enter- (T 26.67 bin CVOI CT 40.80 bin 0W) (T 39.75 bin 0O9) T 29.08 bln poet-t profits RRAL IDIYI L prices 1 72.20 bla CV 1 42.23 bin s 1 51.44 bin erig. Valu fixed assets 6.23 sin enterprises I0 sin zterprisoo 7.74 sin enterprises 0.96 nla enterpriss 3.73 aIn enterprises 85.45 a" workers (7.40 ain--sus of above) 0.23 sin enterprises 14.23 sin workers 57.03 sin yorkers 14.85 sin yorkers 6.94 sIn workers 1 649.57 bin 0VO 2.50 join workers I 69.30 bin 0V0 CY 701.76 bin) 1 452.94 bin 1C0 (1 478.12 bin) la 1 92.77 bin CV0 1 47.33 bin 0V0 CY 3.935 binl) P TOTAL RUAL (Y 471.17 bin--sum of sbove) Y 11.53 bin 0VO CY 89.53 bin) JR ( 43.44 binl) is C 57.68 bin 1 31.03 bin taxe WITERPRSES (I 4998.2 bin) CY 12.38 bl) (C 84.8 bln CY 48.45 binj Rural social labor I 53.00 bin post-tan profits Rural ociai labor fores, lural social labor Rural social lobor Rural socil labor force (comerce. food 34.13 sin Ls force not ccparsbl forces 15.26 isn /a forces 6.07 sin /a snd sarketina only)s 1 96.34 bin wages 6.57 *'- JR 1 209.97 bin orig. value fied assets t Dats a1so av ilable broken down by provinc Data a1lo available broke do"n by (40) branebs of industry. curce9 Chin Statistioel Yearbook 1989, China Statistics Abstract 1990, and World Rank staff calculations. Figures in parenthesis represent alternative estimates svailable within the official publications. >4  -42 - ANNEX 2 The Urban Nonstate Sector: An Outline 1. Although the present paper has focussed on nonstate firms in admin- istratively rural locations, it should be recognized that a substantial non- state sector also exists within China's designated urban economy. As in the rural areas, the most basic distinction is between the collective and indiv- idual/private subsectors. The origins of the urban collective subsector have been traced 1/ to three distinct lines of development: (a) the original socialization, commencing in the early 1950., of pre- 1949 private and household activities, primarily in commerce and handicraft (i.e , nonpower-using) industries; (b) the "neighborhood industry" movement, dating from about 1958, under which urban residents--often primarily married women with limited alternative employment possibilities--took advantage of policy shifts to promote new small-scale, labor-intensive development at the local level (this subsector, however, declined sharply following policy reversals after 1960); and (c) the more recent (post-1978) promotion of job creation schemes for unemployed urban young people, often the children of SOE employees who could not be absorbed by the SOE sector itself. In many cases the "collective" enterprises set up in this period have been closely affiliated with sponsoring SOEs, frequently--for example--undertak- ing subcontracting of more labor-intensive operations. 2. Unlike their rural counterparts, many urban collectives have come under the supervision of official subsectoral bodies (i.e., industry-specific bureaus, corresponding to central ministries), and as such have enjoyed more limited scope for exercising their own initiative, e.g., in choice of product mix.2/ An important distinction in this respect exists between the so- called "big" collectives (in effect often enterprise arouws) which are directly supervised by the relevant specialized bureau at the city level (as such, acquiring a status in some ways almost indistinguishable from that of the smaller SOEs), and the "small" collectives, which are generally supervised by more broadly-based (but lower-level) district or neighborhood management bureaus. Though the official statistics cannot be broken down between the two groups, it is on record that the "big" collectives account for the majority of 1/ Lockett, M. (1986)."Small Business and Socialism in Urban China," Development and Change, Vol. 17. 2/ Wong, C. (1986). "Ownership and Control in Chinese Industry," in China's Economy Looks Towards the Year 2000, Joint Economic Coanittee, Congress of the United States.  - 43 - ANNEX 2 urban collective output, while the "small" firms contribute proportionately more to employment.3/ 3. Urban collectives make a significant overall contribution to the national economy. In 1989 they employed a total of 35.0 million staff and workers (about one third as many as the SOEs) of whom 52.7 percent were in industry, 22.5 percent in commerce and related fields and 10.2 percent in construction. In the industrial sector, urban collectives accounted for approximately one seventh of national gross output. Some of the data for urban collectives suggest that they occupy an intermediate position between SOEs and TVCEs: for example, in their factor proportions (in 1988, fixed assets per worker in urban industrial collectives were approximately Y 10,000, compared to about Y 20,000 for industrial SOEs and Y 3,200 for TVCEs).4/ The recent growth performance of the urban collectives, however, suggests rather closer similarities to the state enterprise sector. Their employment grew at only about 2 percent per annum over 1985-88 (before falling slightly in 1989), while the decline in the subsector's share of total industrial out- put (from 15.9 percent in 1985 to 13.9 percent in 1989) was, in proportionate terms, almost identical to that experienced by the SOEs.51 4. The post-1978 period has also witnessed the emergence of a signifi- cant individual/private subsector in urban areas. A series of official policy statements and regulations, commencing in 1978, relaxed earlier restrictions on such businesses (permitting them, for example, to use power-driven machin- ery and hired labor), while at the same time stressing that eligibility to found such firms was to be restricted to those possessing urban registration and lacking other employment: the intended target groups included the growing numbers of unemployed school leavers and such socially marginal individuals as ex-prisoners. Participants in the subsector fall largely outside the normal range of social service provision (though many may benefit indirectly through relatives with public sector jobs). At least between the mid-1980s and the onset of the rectification campaign in late 1988, however, the sector became sufficiently attractive (with average earnings exceeding those of most public sector employees) 6/ for its originally intended participants to be joined, albeit in fairly small numbers, by personnel who had given up government or SOE jobs and by recent college graduates. More significant in quantitative terms, though, were large numbers of migrants from rural areas, who took advantage of the period's more relaxed enforcement of residency requirements: these came to constitute the so-called "floating population," and were almost entirely unrecognized by official services. 3/ Lockett, op. cit. 4/ Wong op. cit. also found wages and benefits in urban collectives to be set at lower levels than in SOEs. 5/ China Statistical Yearbook (various dates). 6/ Gold, T.B. (1990) "Urban Private Business and Social Change" in Davis, D. and E.F. Vogel (eds.) Chinese Society on the Eve of Tiananmen. Cambridge (Mass.): Harvard.  - 44 - ANNEX 2 5. Unlike the situation in the rural areas, the urban individuallpriv- ate subsector seems to have made little headway within the industrial sector, which continues to be dominated by SOEs and the official collectives: the share of urban individual firms in total industrial output inched up only from 0.3 percent in 1985 to 0.4 percent in 1989. Official data on "urban individual laborers" show them to be heavily concentrated (almost 70 percent) in commerce and catering. The absolute magnitudes shown by these published statistics, which record growth from 150,000 such workers in 1978 to 6.6 million in 1988, are widely believed to reflect substantial under- statement--it is assumed that not all of those involved, especially among the migrants, will have taken out official registration. But more focussed surveys also confirm the subsector's concentration in functions such as commerce, catering, services, repairs and handicrafts, where capital require- ments are generally low. 6. The urban individual/private subsector has visibly succeeded in responding to demand for services and products that were previously not effectively supplied. At the same time, the security enjoyed by such enter- prises has sometimes appeared precarious, and potentially vulnerable to periodic shifts of emphasis in official policy. One area of controversy is the subsector's perceived propensity for tax evasion. Accurate assessment of tax liability is a difficult issue in dealing with small private businesses in almost any country, but in China--with its long egalitarian tradition-- official statements at times appear to reflect not merely a determination to collect what is due but a more broad-based questioning of the legitimacy of the incomes enjoyed by some individual entrepreneurs.7/ Migrants to the cities are doubly vulnerable to possible disapproval by local officials, in view of their sometimes grey residential status. Controls over migration were officially relaxed somewhat during 1988, to facilitate the movement of those prepared to assume personal responsibility for their own food supply. Certain migrants may be eligible to obtain a "temporary residence card" or, for a longer stay, a "lodging card" (iuzhuzheng).8/ However, in times of economic downturn, the security enjoyed even by those holding such registration may be incomplete, and the austerity period which commenced in late 1988 reportedly witnessed a variety of pressures by city authorities to return migrants to the countryside. It is not clear whether such campaigns in fact achieved a sign- ificant aggregate reduction in the floating population--the official data on individual laborers showed only a modest fall from 6.6 million in 1988 to 6.5 million in 1989, though as noted earlier the series probably undercounts migrants. It does, however, seem apparent that such efforts must constitute a source of instability for the subsector, and presumably encourage a short-term focus among many of its participants. 7/ On the general issue of taxation of private firms, see Chinese Economic Studies, Fall 1987. For examples of official statements on the subject see China Daily, July 30, 1989 and August 10, 1989. 8/ Linge, G.J.R. and D.K. Forbes (eds.) China's Spatial Economy, Hong Kong: Oxford University Press (1990). See especially chapter 7 and Appendix 1.  Distributors of World Bank Publications ARGENTINA FINLAND KUWAIT SOTH Caile Hired, SRL AkinSminIb KJokauppa SORUTHkNi wvo Fa AFRICA,ii TWAN Gc Guenes Pr. Box 128 P.O. Box S65 Oxrd UIdvuity Piu FlIdda 165,4th FloorOfc 453/465 HdlinrtdID SmtAhic 1333 8uenne Alco SPLWOID MALAYSIA PJ. 1141 Univ ity of Malya Coopepratie CapeTownECD AUSTRAL'iA, PAPUA NEW GUINEA, Bookihop, Umiled F%l,ASOLOMONlSLANDS, FRANCE PO. Bso 1127.jaln PFat Earu VANUATU, AND WESTERN SAMOA WoeId Bank Publications KOAla Lumzpur Interationals Subscription Sornie DA.Boe0&j,uma1se 66. avnus d'ira 648WhitnoeRoad 7511676Parl eMCO P.O. Box 4109S tham 3132 INOTECCashal Victorid GERMANY Aprdo Po1e-860 2D24 UNO-Varlag Aafd oW246 AUSTRIA Poppedarfer A SD60TlpMadoD.. SPAIN Genrld md Co. D-530D Enr I MOuOC t dlmPrU ikens, SA Grabre 31 MRCOCastelo 37 A-ll Win GREECE3 Sod cEdM en Muccine 28a Madrid AEME 12 nue Maurt. Bd. d'AnrE C.asablancs Liber. Internadenal AEDSn3 RAHLAY 24, Ippodsmou-Street Pl" PI-ia nbaaIlADS Bahrain Research and cozwwway Affiens-11635 ~~~~~~~~~~~~~~~~~~Consadl de Cant 391 ABrsinRrdtaid Con y Aihue-11635 NETHERLANDS C nedBaeCntj3 Ad oBding iNd. DeLindeboon/tnOr-Pubnaties Se dBu(lding mo SRI LANKA AND TrE MALD S P.. o 2 ? Lllai aUPed7ra SantHou74B3ok5 o L M pan a m s, A e T o w C2 llZ 7 So7 E DAaH gl P .O . M A L 244S MRSm Tovn 317Gata 1 NEW ZEALAND 1D, Sir Odao.paI A. BANGLADESH H~~~uadnlaCiy EioIUbn., mad Infone,aticn Samic Gardnder Mawatha BANGLADESH 2utmb