China - Rural industry : overview, issues, and prospects

Report No. 7267-CHA China Rural Industry: Overview, Issues, and Prospects March 1 1989 China Country Department Asia Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  OPP,*, - # FOR OFFICIAL USE ONLY CHINA RURAL INDUSTRY: OVERVIEW, ISSUES, AND PROSPECTS Table of Contents Page No. SUMMARY AND CONCLUSIONS ............................................... i I. BACKGROUND AND RECENT DEVELOPMENTS ............................. 1 China's Changing Strategy for Industrial Development. 1 Evolving Patterns of Rural Industrialization. 3 Institutional Framework. 4 Different 'Models' of Local TVE Development. 7 Recent Growth Performance .11 Overall Dimensions of China's TVE Sector .14 Conduct of Business and Market Interactions .19 Industrial Structure ....................... .................. 21 The Question of Relative Efficiency .24 II. MACROECONOMIC ISSUES AND INTERSECTORAL RELATIONSHIPS .26 Macroeconomic Impact .26 Competition and Integration with State Industry .29 Relationship with Agriculture .30 Urbanization and Migration .33 Foreign Economic Relations .34 1II. LEGAL AND REGULATORY FRAMEWORK .36 Legal Framework for TVP Business Activities. .36 Regulatory Environment .38 Treatment of Private Enterprises .40 Discriminatory Treatment vis-a-vis the State Sector .41 This report, prepared by the World Bank, is based on a collaborative research project (RPO 674-05) between the World Bank and the Institute of Economics of the Chinese Academy of Social Sciences and other Chinese research institu- tions. The project involved in-depth fieldwork in four counties, each repre- senting a different level and pattern of rural industrial development in China. As part of the project, survey data on a sample of 122 rural indus- trial firms, 1,174 employes, and 67 rural townships were collected and analyzed. Aggregate statistics at county, provincial, and national levels also were utilized. The main research output of this project is a book entitled China's Rural Industry: Structure, Development, and Reform (in process), from which many observations and ideas, contributed by Chinese and World Bank researchers, have been incorporated in this report. Two fieldwork missions occurred as part of this project (in June-July 1986 and February- March 1987), for which the World Bank team consisted of the following persons: William Byrd (mission leader), Alan Gelb, Josephine Woo, Jan Svejnar (con- sultant), and Christine Wong (consultant). The Chinese research team for this project consisted of Lin Qingsong (team leader), He Jiacheng, Luo Xiaopeng, Du Haiyan, Du He, Gu Xiulin, Li Hanlin, Meng Xin, Song Lina, Wang Hansheng, Wang Tuoyu, Wang Xiaolu, Wu Quhui, Xu Xinxin, Yang Xiaodong, and Yang Xiaodong. Research assistance was providr l vy 7hu Ning, Hong Wei, and Liang Xuefeng. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - ii - Page No. IV. FINANCIAL ISSUES ............................................... 42 Current Situation and Problems ................................. 42 Enhancing Capital Mobility ..................................... 43 Enterprise and Community Risk Management ....................... 47 Capital Structure of Larger PEs ................................ 48 The Future Role and Structure of the Banking System ............ 49 V. INDUSTRIAL STRUCTURE, LOCATION PATTERNS, AND SUPPORT SERVICES .................................................... 51 Enterprise Size and Economies of Scale ......................... 51 Community Industrial Structure and Subsectoral Composition ..... -3 Technology ..................................................... 54 Industrial Location and Spatial Patterns ....................... 55 Industrial Support Services .................................... 56 VI. LABOR AND HUMAN RESOURCES ...................................... 59 Current Wage Practices and Labor Relations ..................... 59 Workers' Role and Protection ................................... 60 Wage/Incentive Systems and Pay Differentiation ................. 61 Development of Rural Nonagricultural Labor Markets ............. 62 Mobility and Nurture of Human Resources ........................ 63 VII. PROBLEMS OF BACKWARD AREAS ..................................... 66 Ameliorating Fiscal Predation .................................. 68 Human Resource Development in Backward Areas ................... 69 Greater Population Mobility and Rural Industrialization ........ 70 VIII. PROSPECTS FOR THE FUTURE: INSTITUTIONAL AND POLlCY ISSUES ...... 72 Fiscal Incentives Related to TVE Development ................... 73 Future Role of Community Governments ........................... 74 Community Orientation of CEs ................................... 75 Role and Importance of Private Enterprises ..................... 75 Diversified Ownership in the TVE Sector ........................ 76 Legal and Regulatory Environment ............................... 77 Competition and Market Structures .............................. 77 Government Industrial Policies ................................. 79 The Faciliteting Role of Government ............................ 79  - iii TABLES IN TEXT 1.1 TVE Ownership Structure in Four Counties, 1985 .................. 8 1.2 Growth Performance of TVE Sector, 1978-87 ....................... 12 1.3 Growth and Ownership Structure of TVE Gross Industrial Output Value, 1978-87 ........................................ 13 1.4 Composition of Rural Gross Output Value, 1978-87 .14 1.5 Structure of Rural Labor Force, 1978-87 .15 1.6 Ownership Structure of Chinese Industry, 1971-871 .17 1.7 Subsectoral Composition of GVIO, 1982 and 1985 .18 1.8 Capital Structure of Large TVEs ................................. 23 2.1 Bank Credit to TVCEs, 1980-87 ................................... 27 2.2 Use of TVCE Profits to Support Agriculture, 1978-87 ............. 32 5.1 Average Sise of TVCEs, 1980-87 .................................. 52 7.1 Economic Indicators ic rour Counties, 1985 ...................... 67 7.2 Economic Indicators !n Two Townships ............................ 67  - iv - GLOSSARY TVE - Township and village enterprise (including all kinds of rural non- agricultural firms, both private and community-owned) CE - Community enterprise - TVE owned by a community government (town- ship, village, or producrion team) PE - Private enterprise TYCE - Township- or village-owned CE TE - Township eiterprise VE - Village enterprise CG - Community government (township, village, or production team) TG - Township government TIC - Township industrial corporation ABC - Agricultural Bank of China RCC - Rural Credit Cooperative PRS - Production responsibility system in agriculture SE - State enterprise CPE - Centrally planned economy GVIO - Gross value of industrial output  - i - SUMMARY AND CONCLUSIONS China's TVE Sector 1. The last decade has seen a tremendt,us growth of rural indtustrial and other nonagricultural activities in China, accompanied by large increases in rural personal incomes and major shifts in the structure of the rural labor force (see Chapter I). Rural GVIO increased at over 231 p.a. in real terms in 1978-86, while employment in rural nonagricultural activities (industry, construction, transport, and commerce) jumped by 1242 (or by over 30 million people) during the same period.l/ Millions of new firms have been established, new activities and technologies have become operative, and a sizable group of industrial entrepreneurs, managers, technicians, and skilled workers has emerged. China's TVE sector has become both an integral component of the rural economy and the most dynamic, rapidly growing part of Chinese industry. 2. Though China had achieved a certain degree of rural industrialization prior to the late 1970s, this is dwarfed by the burst of growth and develop- ment which has occurred since then. The Government's reversal of previous ideological and political prohibitions against large-scale involvement in nonagricultural activities by the rural population, starting in 1978, played a crucial facilitating role.2/ The production responsibility system (PRS) in agriculture (implemented in 1980-83) also was e- remely important because it transformed hidden surplus labor under the old commune system into visible surplus labor, which generated great pressure to create new nonagricultural employment opportunities. In the more developed areas, the PRS prevented local labor supply conditions from constraining TVE development for a while (permitting annual growth of TVE GVIO of 30-100% for two or more years in a row). Rising rural incomes and sharp increases in credit provided to TVEs by the banking system, along with TVEs' internal accumulation (generated in part by concessional tax treatment) were able to finance the rapid expansion of the sector. Perhaps most important, the release of human energies and talents in 1/ This figure is based on China's official labor force statistics, which appear to underestimate the number of people involved in part-time or seasonal off-farm actJvities and those employed by very small firms. The total number of nonagricultural TVE employees including part-time and seasonal workers rose from 22 million in 1978 to 77 million in 1986, an increase of 55 million. 2/ The prereform emphasis on "division of labor' between urban and rural areas, with the former engaging in industry and other nonagricultural activities and the latter forced to specialize in agriculture (with emphasis on grain production), created an institutional and policy environment generally inimical to rapid TVE development. Only the few areas where local government eubverted national policies (like southern Jiangsu Province) achieved a reasonable degree of success in rural industrialization before the late 19709.  - ii - China's rural population brought on by rural reforms was extraordinary, with its most dramatic manifestations in the burgeoning TVE sector. In this context, the progressive liberalization of Government policy toward rural private enterprises (PEs) was most helpful. 3. Dramatic changes in rural policies and institutions unleashed the TVE sector on the supply side, but the demand side also played a critical role. The state sector had left numerous "vacuums", where it did not meet existing demand, providing an opening for TVEs. Rising personal incomes in both urban and rural areas generated strong new demand for many consumer goods (as well as for building materials for rural housing). Enterprise reforms, which allowed firms to procure inputs outside the plan, also generated considerable demand for many goods that could be produced by TV'Es. High investment demand throughout the economy meant more opportunittes for TVEs. In responding to demand from all these sources, the TVE sector generally seems to have been quiCKer and more flexible than state enterprises (SEs). 4. The institutional structure of China's TVE sector is intimately linked with the three-level system of rural community governance structures in place since the late 1950s (the township, the village, and the production team). Chinese rural communities are characterized by fixed memberships (due to population immobility), various forms of income-sharing within the commu- nity, and a considerable degree of financial self-reliance in external rela- tionships. Community governments (CGs) at the township level functior- in many ways like "mini-states", and they also serve as profit-oriented business entities supervising community-owned TVEs (CEs). In the prereform period, all TVEs were owned by CGs (mainly townships and villages); this "traditional" form of ownership survives and remains entrenched in some parts of the country where it has achieved great successes. Since the early 1980s, private forms of TVE ownership have emerged, ranging from individual proprietorships to partnerships to a tiny handful of genuinely "cooperative" firms and to joint vent-,res among different types of entities. There is thus considerable diver- sity in ownership structure, along with great variation across localities in the relative importance of different types of enterprises as well as in policies adopted by local governments and CGs. 5. TVEs are characterized by great dynamism (rapid growth, ability to take advantage of opportunities, and flexible adjustment), competitiveness vis-a-vis the state sector, "hard budget constraints" (at the community level if not always at the enterprise level), diversity in institutional arrange- ments and development models, outward orientation in terms of product markets (with respect to the locality the producer operates in), and strong local community ties (through community ownership as well as other linkages). Many characteristics of the TVE sector appear to have been conducive to its rapid growth and development since the late 1970s. The outward orientation of TVEs toward product markets seems to have been crucial in this respect, along with relatively "hard" budget constraints for CGs. The interaction with CG fiscal incentives and local employment needs generated a potent stimulus for nonagri- cultural development in virtually all localities and communities (though rural industrialization did not succeed everywhere).  - iii - 6. Despite their impressive recent growth performan-ce, TVEs face numerous problems. Though these have not prevented the TVE sector from growing extremely rapidly so far, they may become a more serious hindrance in the future. 7. As in the Chinese economy as a whole, factor mobility in the TVF sector is limited. This is a source of rigidity which may not be so important if output, investment, and employment are all growing very fast. But as growth slows, and as diffusion and adaptation of relativelv advanced techno- iuies and practices become more important determinants of TVE perrormance, immobility could become an increasing handicap. A major cause of factor immo- bility, as well as a problem area in its own right, is the weak legal frame- work governing TVE activities. Though much codification has taken place in recent years, enforcement of laws and adjudication of cases are still problematic. 8. The TVE sector is now facing increasingly severe competiti- and market saturation in many industries. Though in part merely reflecting market adjustment, this does result in numerous investment failures and financial problems. Problems of backward, inefficient technology and poor product quality plague TVEs in many localities. In some industries, TVEs' ability to compete in changing circumstances may be hindered by continuing small firm sizes--due mainly to the strong community orientation of enterprises combined with factor immobility. Still another weakness is the inadequate network of support services for rural industrialization. This mirrors the neglect of the service sector in the Chinese economy as a whole, but it could well be a greater problem for the TVE sector since self-provision is less feasible for small firms. Scattered, dispersed locstion patterns for TVEs may be a significant source of static inefficiency at present, and may become an increasingly severe problem in the future. Like the limitations on firm size, these are largely the result of community orientation, which dictates that the bulk of TVEs are located in their home communities, and factor immobility. 9. Though the role of %Gs has been highly beneficial in some regions where TVE development has prospered, i.s many of the more backward areas CGs have been a hindrance. Given their own tight resource constraints and the lack of other sources of funds to tap, CGs in backward areas may make exces- sive levies on their CEs, which the latter are for.ced to financd through addi- tional borrowing from local banks or drawing on their depreciation funds. This makes it harder for TVE development to get off the ground. Due to uneven endowments, particularly of human resources, and factor immmobility, spatial patterns of TVE development have been highly uneven. To some extent this is a natural consequence of rapid growth in the TVE sector cverall, but spatial imbalances are more oE a problem than they might be in many other countries because the population of the poorer localities is not permitted to leave them en masse. Finally, remaining ambiguities and discrimination in the treatment of private firms may still be a significant problem.  -iv - The Need for an Improved Legal and Regulatory Framework 10. TVEs suffer from the inadequate 'egal framework that characterizes the Chinese economy as a whole, and from an unstable, sometimes discriminatory regulatory system (Chapter III). Though a great deal of business and economic legislation has been promulgated during the past several years (including a contract law, a bankruptcy law for SEs, a state enterprise law, and others), enforcement and adjudication still appear to be somewhat uneven, with apparent biases against "outsiders" in legal cases handled by local economic courts (which are organs of county governments). There are also indications that TVEs, and within the TVE sector PEs, are sometimes discriminated against by the legal and regulatory systems. The community orientation of TVEs may in large part be due to the inadequate legal framework and poor treatment of outsiders, rather than to any inherent "community consciousness" in China's rural population. 11. Though additional legislation is needed in some spheres, further codification will not solve problems of inadequate enforcement and uneven adjudication. In this respect, one option would be to move toward enforcement and adjudication mechanisms that cut across localities and communities, at least for legal disputes between economic entities from diZferent communities and localities. Traditional methods might for some time to come continue to be used in disputes within communities. To improve the regulatory environment of the TVE sector, greater separation of the regulatory and ownership functions of government agencies would be desirable (indeed, this is a criti- cal aspect of Chinese economic reforms as a whole). As in the case of the legal framework, it would make sense for regulatory functions to be carried out at a level of government higher than that of enterprise ownership (e.g. at the provincial level for regulation of the TVE sector). In addition, greater simplicity and transparency in regulatory processes is needed. 12. A better legal and regulatory framework may require new government institutions of certain kinds, like court systems that cut across localities. Alternatively, it may be possible to strengthen and improve existing institu- tions like the Industrial and Commercial Administration Bureaus so that they can provide adequate recourse in cross-community and cross-locality disputees. Evenhanded treatment of economic disputes that cut across localities and regions is an important part of the Government's efforts to foster free inter- locality trade and a unified national market. Finally, the temptation to establish a ministry or similar government agency to provide administrative support and supervision for the TVE sector should be avoided. One of the main reasons behind the successful performance of TVEs may well be their lack of powerful "mothers-in-law", so to create such an entity would be counterproduc- tive. Ownership and Institutional Structure 13. China's TVE sector is characterized by diverse, evolving patterns of ownership combined with a cor.tinuing strong community orientation (see Chapter I). The "traditional" system from the prereform period, based on ownership and supervision of TVEs by township and village governments, has been highly successful in a few of the better-off areas like southern Jiangsu Province.  - v - Where entrenched, it appears to have considerable staying power. On the other hand, in many parts of the country wholesale privatization uf the traditional CE sector has occurred in various forms (contracting and leasing of firms to individuals, outright sales, abandonment of failing CEs, etc.). In these areas (often the poorer, more backward ones), due to difficult conditions the traditional system never got well-established in the prereform period, and existing CE assets were often divided or sold off when the PRS was implemented. New enterprise creation and incremental growth come largely from the private sector. In still other parts of the country like southern Guangdong Province, a diverse, multi-level ownership pattern has emerged, with numerous joint ventures among different entities adding to the diversity. 14. Diversity of ownership in the TVE sector has more advantages than disadvantages, and there is no need to impose uniformity (Chapter VIII). However, strengthening and clarification of the property rights of owners (whatever kinds of entities these may be) with respect to their firms would be desirable. Evenhanded treatment of different forms of ownership by the legal and regulatory systems also is needed (Chapter III). In particular, any remaining discriminatton against PEs should be reduced to the extent possible. At the same time, the ownership rights of communities in CEs that are under- going various forms of privatization should be recognized and appropriately recompensed. 15. Fiscal incentives have played a major role in promoting TVE indus- trial growth. While these linKages will likely weaken over time, as part of the separation of regulatory from fiscal functions in CGs, it is important to ensure that CGs and their officials continue to have adequate incentives to promote TVE development. In this context, the uniform nationwide pay scale for township leaders may be counterproductive given the great differences in average per-capita income levels across regions and localities (in the most developed rural areas, a township leadet's base pay is lower than that of an ordinary TVE worker). To enhance their incentives, abolition of the uniform nationwide pay scale for township leaders could be considered, along with development of appropriate performance-related bonus systems geared to local conditions and needs. Capital Mobility 3/ 16. Immobility of capital leads to inefficiency in resource allocation, but more important it hinders the rapid growth of successful TVEs and thereby the attainment of larger firm sizes where this is desirable. There is a need to facilitate movements of capital to the investment opportunities with highest returns, within localities. But capital flows across provinces and regions raise complex issues because of the continuing immobility of other factors. For example, large outflows of funds from the backward areas to earn higher returns in more developed localities would further exacerbato the problems of the former. 3/ Options for enhancing capital mobi'.ity in the TVE sector are discussed in a-re detail in Chapter IV.  - vi - 17. In addition to encouraging greater capital flows within the rural banking system, one possible way to enhance local capital mobility and improve the efficiency of investment would be through emergence of new financial institutions in rural areas. These could improve the provision of financial services to TVEs and ,enerate competition with the existing state-owned (but at lower levels largely community-oriented) banking system. To realize the benefits from greater institutional diversity, many of the new institutions should cut across community/locality lines (up to now, most of them appear to be community-based). Some financial institutions could be structured as joint ventures among different CGs and local governments. Cross-community joint ventures among TVEs are another possible means of concentrating scattered rural capital and directing it to high-return uses. 18. Still another possible option for enhancing capital mobility is for well-performing rural communities to issue financial instruments to attract capital and spread risk. These "income bonds" could involve a combination of fixed and variable, profit-related returns and could be marketed through the banking system. Secondary trading could emerge gradually over time, and income bonds could eventually evolve into more equity-like instruments. Since there is no international experience and little past Chinese experience to provide guidance on how well the income bond concept would work in practice, and moreover massive, possibly destabilizing flows of capital across regions and provinces should be avoided in the short run, experimentation with income bonds probably should start on a small scale, in a few of the more advanced areas. Capitalization of Large Private Enterprises 19. Many iarge PEs depend heavily on bank credit, and property rights with respect to firm equity are still vague and ambiguous (founders' original contributions are typically "repaid" once the enterprise is financially viable). While private property rights need to be further strengthened and clarified, these firms' capital structure could be improved by allowing banks to make nonvoting equity investments in larger PEs (Chapter IV) This would require a change in orientation and practices on the part of the rural banking system, and it wouJ4 also require careful monitoring and regulation of emerging linvestmer banking" practices. The increased risk from equity participation shoul be manageable because this would occur only in the case of larger, rapidly growing, and successful PEs. Bank equity participation would allow communities to reap a larger portion of the returns from the activities of PEs that they now finance through credit, enhance PEs' ac:ess to bank funds, and partially socialize their ownership in a way that should preserve the incentives and autonomy of private proprietors. The Need for More Concentrated Location Patterns 20. The strong community orientation of China's TVE sector has resulted in highly dispersed location patterns, with factories dotting the countryside in many areas (see Chapter V). As a result, infrastructure is costly or is inadequately provided; it is harder for a dense network of services to emerge because of scattered clients; advantages of "economies of agglomeration" are  - vii - largely lost; transport costs are high; wasteful patterns of land use are pervasive, in particular occupation and pollution of high-yield agricultural land; and more generally, industrial pollution is spread widely in the countryside, with consequent difficulties and high costs for pollution control. Since the root cause of dispersed location patterns is the community orientation of the TVE sector, any solutions to this problem must address the firm-community nexus. But a more efficient location structure is urgently needed in the short tun and cannot await gradual weakening of community orien- tation over time. 21. One possilble option for promoting more concentrated location patterns is for local government authorities at county or higher levels to estRblish 'industrial parks" at which all types of TVEs could set up operations. These would not be under the jurisdiction of any particular rural community, and would offer (on a fully costed basis) infrastructure, public utilities, and services for TVEs which locate there. International experience indicates that there are numerous pitfalls to be avoided if the industrial park concept is to work well; most such schemes in other countries have not been very successful. Thus industrial parks need to be well thought-out and carefully designed, and in the Chinese context particular care would need to be taken in determining policies toward movement of people into them. Moreover, the attractions of locating in industrial parks should if possible be complemented by measures to discourage location of firms in the old dispersed manner. Appropriate land- use and land-conversion fees, cost-based utility charges, and meaningful pollution control standards and pollution fees might help encourage relocation to industrial parks. Support Services for TVEs 22. The network of services supporting rural industrialization (commercial. financial, 4-ansport, and technical, among others) is for the most part inadequate (Chapte- V). This reflects similar problems in Chinese industry as a whole and in the et-tire economic system, but the problems for TVEs are exacerbated by (1) their relative lack of access to services provided by state institutions and (2) the greater need of small firms like TVEs for specialized services (since it is herder for them to engage in self- provision). The focus on gross value of industrial and agricultural output as a performance indicator may have resulted in relative neglect of opportunities in the service sector by CGs. Price distottions also may have been important, and the community orientation of the TVE sector and inadequate legal framework most likely have played a role. 23. Part of the solution is to improve the access of TVEs to existing urban, state-oriented service institutions, which is already beginning to occur as the latter become more profit-oriented. This does not mean that large government programs for provision of support services to TVEs should be established. On the one hand, TVEs should pay the full cost of any services they obtain from government institutions, while on the other, these agencies should not take advantage of their monopoly position or administrative power to gouge the TVE sector. The critical need is for development of industrial support services within the TVE sector itself, on a profit-oriented, commer-  - viii - cially viable basis. This is not only essential for the healthy future development of rural industry, but it can also be an important source of growth and employment in its own right. Many services are most efficiently provided by small firms, so they represent a great opportunity for the TVE sector. 24. Among the measures which might help encourage rapid development of TVE services are (a) greater freedom for professionals and skilled workers to leave their firms and set up their own service businesses;4/ (b) greater concentration in TVE location patterns, which would also concentrate demand for services; (c) improvements in the legal and regulatory environment for PEs (since the relative importance of services is greater for them); and (4) shifting the evaluation of CG and local government leaders by higher levels from a narrow focus on production to a broader assessment of local economic development, with an appropriate emphasis on services. The Role of Workers 25. Lower labor costs and better labor motivation are among the most important sources of TVEs' competitive advantages vis-a-vis SEs (Chapter VI). Though TVE workers play virtually no role in firm-level decisionmaking, they are members of the rural communities to which most sizeable TVEs belong. Moreover, nearly universal use of performance-based wage systems (piece rates, firm-level workpoints, bonuses, year-end "dividends", etc.), combined with greater ease of firing poor performers, provide potent incentives for hard work and long hours. Smooth labor relations in the TVE sector are facilitated by its community orientation and by the fact that most workers are "first generation" and have spent at most a few years outside of agriculture. 26. As the community orientation of the TVE sector weakens over time, as in at least some areas more and more workers are hired from outside the commu- nity (often on a short-term basis), and as the TVE labor force becomes more aware of its group interests, this largely beneficial pattern may change. Carefully designed institutional and policy measures will be needed to prevent a possible deterioration in labor relations from becoming a drag on the effi- ciency and growth of the TVE sector in the future. Regardless of how the role of workers and labor relations evolve in the future, basic legislation delineating acceptable labor practices in the TVE sector would be desirable. This should not be overly restrictive, but it would be helpful both in protecting workers and in protecting employers (especially private proprietors) from arbitrary charges of "exploitation". These regulations would have to be widely perceived as appropriate and fair, and they would need to be backed up by meaningful sanctions and an enforcement mechanism. 4/ This could also have some negative effects, like reduced identification with firms by these personnel, weaker incentives for TVEs to provide training for their employees, and sharply rising wage levels as a result of competition among enterprises for highly-qualified personnel. Nevertheless, the benefits most probably outweigh these costs.  - ix - Improving Prospects for Backward Areas 5/ 27. The degree of spatial inequality in rural incomes, economic develop- ment, and industrialization is considered a problem by many in China. This is caused mainly by the combination of highly uneven endowments of human and other resources and immobility of population and other factors of production. Problems of backward areas are exacerbated by the "fiscal predation" many CGs engage in, which sucks resources away from local TVEs and alienates the banking system. Redistribution through the national fiscal system is signifi- cant but does not address the fundamental causes of uneven development. More- over, "aid" from higher levels of government is often wasted or mainly goes to finance local government and CG payrolls. 28. The counterproductive role of CGs in many backward areas needs to be changed. A two-pronged approach can be envisioned, including major reductions in the size of local and community governments in backward areas and, after careful evaluation, financing some key public and social services through the state budget (rather than with CG extrabudgetary funds as at present). These measures would reduce the need for fiscal predation by easing the constraints on CGs in backward areas. Certain other actions at the local level also may help ameliorate fiscal predation, but these are unlikely to work well unless the resource squeeze on CGs is reduced. 29. Augmenting the human resource base in backward areas could help stimulate rural industrialization in them. This can be furthered by training and other measures to improve the quality of the indigenous human resource base (though a serious "brain drain" may arise), as well as by establishment of a regulatory and policy environment conducive to an inflow of these resources from elsewhere. Attracting outside entrepreneurs, who can then bring in the other skills needed to organize production and marketing success- fully, is crucial. Though international experience suggests that it is very difficult to promote regional development in backward areas where it has not occurred spontaneously, the strongest possible efforts have to made in this direction (in ways that are not costly to the Government), as long as popula- tion and labor remain largely immobile. How3ver, over the longer term, considerable exit of population from the backward areas may be the only means of pulling up their average income levels and bringing labor force into line with local resource base. Labor Markets and Mobility 30. One of the most difficult issues for the future faced by the TVE sector (and indeed by the Chinese economy as a whole) is the extent to which open labor markets should emerge and there should be freer mobility of labor and population (see Chapter VI). Mobility of 'high-quality human resources" (entrepreneurs, managers, engineers, marketing personnel, technically skilled workers, etc.) is a highly effective means of diffusing technology, best management and other practices, and industrial experier.Ze throughout the economy. Thus the costs of immobility of this kind of labor are very high. 5/ See Chapter VII for more details.  i~~~~~~~~~~~~~~ On the other hand, large-scale population movements generate substantial short- and medium-term social costs, particularly if they involve excessive migration into large cities. 31. Chinese reality may dictate sul)stantial restrictions against massive permanent movements of population and ordinary labor into major urban centers or from poor to rich rural areas, at least in the short run. But large-scale "temporary" movements of workers (t-'thout their families) are already occur- ring, in some ways reminiscent of the "guest-worker" phenomenon in Europe. This at least temporarily provides employment and income to people from back- ward areas. It also helps check the rise in labor costs in the more advanced areas, and provides on-the-job training and experience, which augments the human capital of backward areas if the temporary laborers return home. The long-term impl_cations are still unclear, however, particularly whether these migrants will return home or eventually will view themselves as permanent residents of the localities they work in. Even if mobility of population end ordinary labor remain restricted, measures to increase the mobility of high- quality human resources may improve efficiency and growth prospects in the TVE sector. The Evolving Role of TVEs in the Chinese Economy 32. The future role and share of the TVE sector in the Chinese economy depends on many hard-to-predict factors. But from all indications, its impor- tance will continue to grow over the short and medium term. TVEs represent the only hope for achieving the Government's ambitious goals for shifting a large part of the rural population from agriculture to nonagricultural activi- ties over the next several decades, and for transforming the countryside through "urbanization in small towns". Great hope is also invested in the TVE sector as a future generator of exports and export earnings. More generally, competition from the TVE sector may well serve as a potent source of pressure to stimulate the state-owned industrial sector to reform itself and improve efficiency. 33. The TVE sector will need to further evolve and develop if it is to continue to be a highly dynamic "leading sector' in the economy, pErticularly if urban reforms are successful in drastically improving the efficiency and competitiveness of the state sector (see Chapter VIII). The barriers against reasonably large firm size in the TVE sector will have to be overcome, but this can only occur gradually as institutional, administrative, and other constraints are removed over time. As part of this process, the community orientation of larger TVEs will need to weaken, and oGs should gradually sepa- rate their local government/fiscal functions from their corporate business development role. Diversified ownership patterns may be appropriate for larger TVEs. Closer and more organic forms of integration with state industry may become more and more desirable, even necessary, for smaller TVEs. The TVE sector will likely continue to be characterized by a great deal of local variation and, for some time to come, experimentation. 34. The future development of the TVE sector does not call for massive Governaent support, financial or otherwise, or for large new goverument agen-  - xi - cies, major, costly new programs, or detailed administrative regulation and promotion. Nevertheless, the role of government will be crucial in some respects. More important than any positive measures to support the TVE sector (though some of these may be called for) is the need to avoid policies and restrictions which do major harm to TVEs, based on the mistaken notion that TVEs are sources of problems elsewhere in the economy.61 The TVE sector should not be made a scapegoat for the problems of Chinese agriculture, serious as these may be; fundamental solutions must lie in agricultural policies and reforms. Similarly, TVEs should not be held back because competition from them is rmaking life harder for many SEs; obviously it is necessary to move toward a "level playing field" on which enterprises of all types can compete with each other on an equal basis, but this should be accomplished by freeing restrictions on other sectors rather than by tighter restrictions on TVEs. Finally, the TVE sector is not the main source of macroeconomic instability in the Chinese economy; hence it should not be disproportionately affected by any credit squeezes or other restrictive macroeconomic policies to deal with "overheating" of the aconomy. The long- term prospects for the TVE sector are bright, as long as it is allowed to continue to develop, evolve, and adapt to the changing economic and institutional structure of the Chinese economy. 6/ Macroeconomic issues and intersectoral relationships involving the TVE sector are discussed in detail in Chapter II.  - 1 - I. BACKGROUND AND RECENT DEVELOPMENTS 1.1 This chapter will first briefly review China's overall industrial development strategy as it has evolved during the past several decades, as well as changing patterns of rural industrialization. Then it will look at the institutional framework within which the TVE sector functions, particu- larly the organizational structure of rural communities. The impressive recent growth performance of the sector will be described, and the ove-all dimensions of the TVE sector in terms of output, employment, income, and other related aspects will be outlined. The conduct of business in TVEs and the characteristics of markets they face will also be discussed. Finally, the chapter will look at salient features of the industrial structure of the TVE sector and at its changing role in Chinese industry as a whole. China's Changing Strategy for Industrial Development 1.2 Prereform industrialization. Between the late 1950s and the late 1970s, a distinctive pattern of industrialization took shape in China, based on an evolving industrial strategy and a unique combination of institutional, demographic, political, and ideological constraints. Chinese industry on the eve of reforms was characterized by problems and weaknesses that in many respects resembled those of other centrally planned economies (CPEs). But there were some important differences, which in part explain subsequent developments and in particular the extraordinarily rapid growth of rural industry since the late 1970s. 1.3 China's prereform industrial policies and industrialization strategy very much followed the Soviet model. Industrialization relied on a very high investment rate (often over 30Z of GDP), with most industrial investment financed through the government budget. Heavy industry, particularly basic goods like steel, was emphasized at the expense of light industry and consumer goods production. Production, resource allocation, and investment were all subject to planning and administrative direction, exercised through a hierar- chical system of industrial administration based on ministries and bureaus, and backed up by state ownership of all sizeable industrial firms. Enter- prises had little autonomy in business decisionmaking, and all of their profits were turned over to the government. 1.4 But industrialization in China departed from the Soviet model in important respects even before the late 1970s, and evolved its own distinct characteristics. These differences were related to China's initially very low level of development, its huge rural population still at a low level of sub- sistence and relying primarily on agriculture, transport and other bottlenecks which fragmented the domestic market, and elements of Maoist ideology and economic strdtegy. Among the most striking deviations from the Soviet model were (a) elimination of material incentives (no bonuses for workers, no profit retention by enterprises); (b) a much cruder and less comprehensive planning apparatus; (c) a territorially decentralized system of industrial administra- tion, with even large enterprises nominally "reporting to" provinces and municipalities rather than the center; (d) similar partial administrative decentralization of planning and control over goods allocation; (e) a tendency toward provincial and even local self-sufficiency and autarchy; (f) immobility  - 2 - of population, enforced through the household registration system and ration- ing of basic consumer goods in urban areas; and (g) a clearcut "division of labor' between urban and rural areas, with the latter largely prevented from engaging in nonagricultural activities, particularly industry. 1.5 There was also an important regional dimension in China's prereform incustrialization strategy. Each province was supposed to develop its own 'complete industrial system". Massive investments in industrial development were made by the central government in some of the interior provinces. This was motivated partly by military considerations, partly by the ideology of provincial and local "self-sufficiency" in industry. High-cost and often highly capital-intensive state-owned industrial firms were nurtured in the cities of interior provinces. This generated a more balanced pattern of industrial growth across provinces, but at very high cost. 1.6 These features greatly affected the pattern of industrialization in China. In addition to vertical insulation between ministries, there was a great deal of horizontal insulation between provinces and localities. The separation between city and countryside brought about a sharp polarization between the urban and rural population. The former were guaranteed employ- ment, for the most part in the state sector, and were entitled to a comprehen- sive set of subsidies and benefits, whereas the latter were not, had lost the bulk of their productive assets (land and implements), and were largely prohibited from engaging in nonagricultural activities. This resulted in a dualistic economic structure, with urban areas engaging in capital-intensive industrialization while hidden surplus agricultural labor accumulated in the rural areas with very little outlet. Overall, inefficiency and waste were pervasive, and growth was generated primarily by the high investment rate rather than by improvements in factor productivity. 1.7 Changes since the late 1970s. China's industrial landscape has changed dramatically as a result of reforms, but the legacy from the prereform period is still evident in many spheres. Bonuses for w%.rkers and profit retention for enterprises were reinstated relatively early (the former in 1978 and the latter, for most large enterprises, by 1980). Markets have become increasingly important in the distribution of industrial goods. In contrast with the past predominance of state ownership, a variety of different owner- ship forms (including collective and community ownership, partnerships, indi- vidual proprietorships, and joint ventures among different types of firms) now flourish. But perhaps most important from the perspective of this report, China's rural population is now allowed, even encouraged to move into nonagri- cultural activities, as a means of soaking up surplus labor from agriculture and augmenting personal incomes. As a result of this fundamental change in orientation, there has been an explosion of rural industrial development. Industry in the countryside has now reached the point where it is an important determinant of the overall pace of China's industrial development and interacts in complex ways with urban industry. 1.8 A strong element of continuity is the immobility of China's popula- tion and labor force, despite some loosening in the system at the margins. Thus there is still considerable separation between administratively designa- ted urban and rural areas, even though many of the latter are industrializing and urbanizing (without the traditional urban privileges and subsidies) at a  -3- rapid pace. Territorial decentralization of the industrial administration system has been carried even further than in the prereform period, without completely freeing enterprises from local control. Except in the private sector (where the bulk of firms are still relatively small) every Chinese industrial enterprise still reports to a government supervisory body, usually an industrial bureau. Finally, "investment hunger" continues to pervade the system, and the aggregate investment rate has climbed to near 40Z in the last two years. 1.9 It is this combination of change and continuity which has generated China's current industrial structure, strategy, and policies. Unlike what has occurred in some developing countries, massive migration tc the large cities and consequent social problems have been avoided. The burst of rural indus- trialization has been at least in part the result of administrative rigidities which prevent the population from leaving rural areas. The continuing insula- tion between many facets of the urban and rural economies also partly explains why the inefficient urban/state industrial sector has only slowly been "awakening" in response to competition from rural industry. Despite nearly a decade of reforms, state enterprises (SEs) still face great obstacles in trying to improve efficiency and instill dynamism in their activities. Evolving Patterns of Rural Industrialization 1.10 Rural industrialization in China before the late 19709 occurred in an administrative straitjacket, albeit one which was loosened on occasion. The pressures of surplus agricultural labor and the desires of the country's leaders to promote rapid industrialization led to several "waves" of rural industrial development. One occurred as part of the Great Leap Forward move- ment of 1958 and was a disastrous failure. The bulk of rural factories and workshops created at that time were shut down in the early 1960s. A second wave of rural industrialization occurred as part of the campaign to promote agricultural mechanization launched in 1970. It resulted in rapid rural industrial growth, but only in some parts of the country, like southern Jiangsu Province, and in the peripheries of large cities. Though rural areas in principle were only allowed to develop industries directly supporting agri- culture (e.g. agricultural machinery and fertilizer), some of them moved intc more profitable activities, ranging from machine building to subcontracting parts and products for urban factories. This prereform rural industrial development laid the foundation for a subsequent "take-off' in the lower Yangtse valley (Shanghai environs, southern Jiangsu, northern Zhejiang), it the Pearl River delta, and in a few other rural areas with conditions condu- cive to industrialization (proximity to urban markets, technology, and infor- mation, a rich agricultural base, a core of entrepreneurs and skilled workers, etc.). In the prereform period, rural industrialization occurred entirely on the basis of community government (CG) ownership of firms. 1.11 Since the late 1970s, China's rural industrialization strategy and development patterns have undergone a remarkable transformation, much greater than the corresponding changes in the urban industrial sector. Ideological and political prohibitions against rural nonagricultural activities were  - 4 - erased starting in 1978.1/ Moreover, private industrial and other nonagricul- tural activities are now permitted, even encouraged. Most recently, in 1987 limits on the size of private firms (a ceiling of seven non-household employees) were formally lifted. As a result, new patterns and "models" of rural industrialization have arisen. The production responsibility system in agriculture (PRS) transformed hidden surplus labor under the old commune system into visible surplus labor, which generated great pressure to create nonagricultural employment opportunities and may have facilitated the emergence of new, private forms of ownership. The PRS also eased local labor supply constraints in some of the most rapidly growing areas. Rural industry has thrived by taking advantage of the increased scope for market activities and the slower response of the state sector to new opportunities. It also has benefitted from tax advantages and, since 1984, enhanced access to credit from the state banking system. But the underlying institutional structure retains many features from the prereform period. Institutional Framework 1.12 China's rural economy since the late 1950s has been organized under a framework of rural community governance structures, linked to higher levels of government primarily through the Communist Party hierarchy. Recent reforms have changed this pattern somewhat, by restoring the township as the lowest level of government in rural areas and allowing much greater scope for eco- nomic activities and asset accumulation by private individuals, but neverthe- less, rural community organizations retain their cohesiveness and a major role in the local economy. Rural communities in China exist at three levkls: (a) the township (formerly the commune), which is now the lowest level in China's government hierarchy and has an articulated government structure--the typical township has 15-30,000 people; (b) the village (formerly the brigade), which is not a separate level of government but has governmental functions and a community structure--villages generally have a population of 1-2,000; and (c) the production team (sometimes referred to as 'villagers' smsll group"), which is purely a community structure and has lost most of its administrative functions as a result of the PRS--a team on average has about 30 households and 150 people. There are substantial variations in size and economic power among communities in different parts of the country.2/ I/ Though applied unevenly and never fully enforced in some parts of the country, the prereform emphasis on "division of labor" between urban and rural areas, with the former engaging in industry and other nonagricultural activities, the latter forced to specialize itl agriculture (with emphasis on grain production), created an institutional and policy environment generally inimical to rapid TVE development. Only the few areas where local government subverted national u?olicies (like southern Jiangsu Province) achieved a reasonable degree of success in rural industrialization before the late 1970s. 2/ For example, in Nanhai County, Guangdong Province, the average township has over 50,000 people and a gross industrial output value of more than Y 35 million. In Shangrao County, Jiangxi Province, by contrast, the average township has only 20,000 people, and gross industrial output value averages only Y 393,000.  -5- 1.13 The most important characteristic of China's rural communities is their largely fixed membership (population), due to the household registration system which has prevented any substantial rural-urban or rural-rural m.gra- tion for three decades. Most people born into a rural community stay there for their entire lives, and their children also will belong to the same community. Immobility of population implies immobility of human resources, though "temporary" migration to cities and other rural areas to seek jobs is increasingly common. 1.14 Another key feature is the combination of income-sharing within the community and a considerable degree of financial 'self-reliance' in relations with other communities and higher levels of government. Income-sharing has occurred in various ways over time and in different parts of the country. Before the PRS, the workpoint system in production teams assured relatively equal distribution of income within them. Allocation of jobs in enterprises owned by townships and villages (TVCEs) also has been a means of balancing family incomes in areas where TVE development occurred relatively early. Even CG decisions on establishment and location of new enterprises have had income- sharing considerations in mind. Finally, the equal distribution of agricul- tural land when the PRS was implemented contributed to a relatively equal distribution of income within most rural communities. To some extent income- sharing has been weakened by the rapid growth of private economic activities and the abolition of the commrune system. 1.15 On the other hand, the degree of self-reliance of rural communities, particularly at the township level, has if anything increased since the late 1970s. Fiscal revenue-sharing systems have gradually been extended down to the township level, and the township has been established as a separate level of government administration and public finance in 1984-86. Most townships depend heavily on extrabudgetary revenues--primarily from TVEs, in the form of profit remittances to the CG and various "management fees" 3/ --to meet public expenditure needs. Even for poorer, more backward communities which receive fiscal subsidies, incremental financial resources at the margin depend largely on the efforts of the CG and its subordinate economic entities. Reforms in the local banking system have made lending ceilings in a township to a large 3/ Management fees do not represent payment for services provided; rather, they are often ad-hoc levies by CGs, which occur ot;tside the budget. TVEs also generate the lion's share of budgetary tax revenues in the more industrially developed rural communities.  extent dependent on deposit generation within tne township.4/ Given immobility of human resources, rural communities must largely rely on the "human capital" of their own members. 1.16 CGs have complex, often conflicting roles and responsibilities. In the first place, they function essentially as "mini-states", providing a whole range of public and social services and (especially at the township level) setting policies and regulations within their jurisdictions. The instruments available to CGs include establishment and ownership of CEs, decisions on CG expenditures, a considerable degree of influence over the lending decisions of the local banking system in many areas, encouragement and support of different types of private enterprises (PEs), etc. CGs cannot use certain policy instruments normally available to states, however. They cannot engage directly in deficit financing: their expenditures cannot exceed available funds (including transfers from higher levels of government and from TVEs and the portion of locally collected taxes allocated to them). Hence CGs face a relatively 'hard" budget constraint. CGs cannot effectively impose barriers to trade in goods, so they cannot meaningfully "protect" their enterprises from outride competition. In any case, rural communities, even at the town- ship level, are too tiny markets to be of much use to local TVEs in most industries. 1.17 In addition to being local governments, CGs are profit-oriented economic entities akin to holding companies, investment corporations, or even headquarters of loosely controlled multidivisional corporations. CGs tend to be intimately involved in major investment decisions for CEs, decisions to establish or dismantle firms, changes in product lines, appointment of enter- prise management, managerial compensation (sometimes wages as well), and bonuses. CGs also can play a crucial role in risk absorption for subordinate CEs, investment financinp 'both directly and through arranging other sources of funding like bank loais), and more generally by representing their CEs in an often hostile administrative environment. 1.18 There are serious conflicts inherent in this dual role of CGt, which will be explored i:i subsequent chapters. But the combination of strong fiscal and employment incentives for CGs, their business development role, and the market environment for outputs and material inputs TVEs function in (see below) has generated an extremely powerful 'push" for TVE development. 1.19 TVEs consist of all nonstate, nonagricultural urdertakings in rural areas, as well as some agricultural activities organized as (nonstate) firms rather than on a household basis. They fall into a number of main categories. Traditional TVCEs are owned and controlled by township or village authorities. Firms at the production team level often were not independent accounting units and instead put their proceeds into the "pot" of collective distribution to team members. Many team firms were privatized during or in the aftermath of the PRS. Some township enterprises (TEs) and especially village enterprises 4/ Though there are large gaps between the two (positive or negative) in many localities, incremental loans to a large extent depend on incremental deposit generation.  (VEs) have been "contracted" or leased to individual management.5/ PEs include both individual firms and partnerships owned and managed by several people. Many of them are based on family or kinship ties. Close personal relationships with CG officials are often necessary for sizable PEs to allow them to navigate the complicated administrative environment. Only a few TVEs are owned cooperatively by all or most of their workers.6/ Finally, there are an increasing number of joint ventures between different forms of ownership, within the TVE sector as well as including urban state and collective enter- prises. Joint ventureq with foreign partners are still very rare, but in places like Guangdong Province, many TVEs have cooperative relationships with foreign firms short of equity participation. 1.20 China's rural communities have been profoundly affected by the PRS, which gave peasant households partial rights to land which had previously been vested in the production teams and greatly reduced direct supervision of agri- cultural production by CGs. Moreover, it abolished the system of community workpoints and collective distribution of income, reverting to primarily household-based personal income generation and distribution. The communes were abolished and replaced by townships [xiarig], now the lowest level of government in rural areas. Another important policy change already mentioned was the relaxation of previous restrictions against large-scale expansion of nonagricultural activities by rural communities. Similarly, removal of restrictions against household nonagricultural activities and, more recently, against sizable PEs stimulated a great spurt of TVE development. Finally, related to the PRS was a crucial cnange in wage systems. In the past many ; TVCEs paid workers' wages directly to the production teams or which workers were members--workers received workpoints and participated in year-end collec- tive income distribution in their teams. This greatly dilvted incentives, since any fruits of harder work were shared not only with other workers in the enterprise but also with all the members of the production team. Even in TVEs where workers were paid directly, fixed time wages were the norm. With the PRS, TVEs shifted to direct payment of workers, and moreover, performance- based pay in a variety of forms became the norm (enterprise-based workpoints, year-end dividends, profit-related bonuses, and above all, piece-rate wages). This had a great impact in improving incentives and efficiency. Different 'Models" of Local TVE Development 1.21 China's TVE sector shows a remarkable degree of local and regional institutional diversity, despite the basically uniform overall framework. This is the result of both exogenous and endogenous factors. The former include the traditional resource base (most importantly in terms of human capital and agricultural land); proximity to markets and to urban industry; pre-1949 industrial and commercial traditions; delivery requirements for grain 5/ These contracts typically involve remittance targets or sharing rates for the enterprises, with extra profits being kept by the contractor. 6/ But community ownership has some of the characteristics of cooperative ownership in small communities where the employees of the main enterprise make up a significant proportion of the community population  - 8 - and other crop& under the agricultural planning system; and prereform ruras industrialization. On the other hand, local authorities and CG leaders have sometimes influenced the path and pace of ruLal industrialization as well as the local institutional structure. One example is the decision of Wuxi County authorities in the early 1970s to support continued development of TVEs, despite the prevailing national trend against this at the time.7/ Authorities in Nanhai and other counties of southern Guangdong Province permitted and protected "undergrotnd" private firms even in the prereform period. In numerous other localities, authorities did not promote development of TVEs until very late, handicapping rural industrialization. 1.22 A key dimension of variation is ownership. In the four counties where extensive field research was conducted, striking differences were encountered. Table 1.1 shows shares of different types of TVEs in GVIO in these four counties in 1985. The share of PEs ranged from a low of 3Z in Wuxi County to a high of 51Z in Jieshoi County. The relative weights of TEs and VEs also varied considerably, with the share of TEs nearly three times that of VEs in Jieshou and close to twice as high in Shangrao but roughly equal to each other in Wuxi. Only in Nanhai County did team-run firms account for a substantial part of total industrial output. Table 1.1: TVE OWNERSHIP STRUCTURE IN FOUR COUNTIES, 1985 (I of total gross industrial output value of TVEs) Wuxi Nanhai /a Jieshou Shargrao China Total (Y million) 3,705 ..421 127 31 175,008 TEs lb 48 43 36 43 45 VEs 47 31 13 22 38 Team firms 2 16 - - 9 PEs 3 10 51 35 8 /a Shares in gross revenues of industrial TVEs. /b Including town-run firms. Source: Fieldwork information; Table 1.3. 7/ Jiangsu Province as a whole several times successfully ignored national campaigns to shut down and "readjust' TVs, which gave it a commanding lead in rural industrialization on the eve of reforms in the late 19709.  1.23 Another important but more subtle dimension of variation is the degree to which lower-level communities are under hierarchical control by townships. In Nanhai, villages and teams have been relatively independent, whereas in Wuxi, the township government (TG) rules with a stronge- hand. The degree of administrative control over wages, the degree to which workers receive profit-related bonuses, the methods of appointment of TVE managers, labor allocation, and a host of other practices can differ greatly across regions, localities, and even communities within a locality. There is also considerable variation in policies followed by local and community authori- ties. In southern Guangdong, some counties promote large enterprises owned by townships, others village and team firms, others PEs, still others outside or foreign entrepreneurs. Wuxi has an administrative climate which makes it virtually impossible for large PEs to emerge. Jieshou County authorities personally sponsor some of the more successful private entrepreneurs, support- ing them with bank loans and administrative help. 1.24 The most important dimension of variation is in levels of development and industrialization (see Chapter VII). The richest and most industrialized counties like Wuxi and Nanhai already have in large measure achieved the equivalent of "middle-income status", whereas the most backward areas, an order of magnitude poorer and often hundreds or thousands of zimes less industrialized, are nowhere near the "take-off" stage. The former face issues of how to organize, finance, and furtl.er develop the TVE sector most efficiently, while the latter are striving merely to get nonagricultur3l development started. Local diversity in the TVE sector can be illustrated by a brief look at the four counties that were subjected to intensive fieldwork. 1.25 Wuxi County in Jiangsu Province for a long time has been the most industrialized rural county in China, even though it is by no means the most urbanized. Wuxi is also the quintessential example of the "traditional" TVE system based on TVCEs, integrated under the supervision of TGs. This is especially evident in wage determination and labor allocation. Industry in Wuxi is highly efficient in the Chinese context; TVCEs are relatively large; many of them use relatively advanced technologv; and they compete effectively with state industry. Industrial PEs are severely hampered by administrative restrictions. Neither labor nor land markets have developed, though large amounts of financial capital have been brought in from other parts of the country by local banks to support local TVE development, and many TEs or town- ship industrial corporations (TICs) have issued bonds to employees ar.d local residents. Labor is not encouraged to flow in from other parts of the country. 1.26 Nanhai County in Guangdong Province is also highly industrialized, but more urbanized than Wuxi. Its TVE sector is "open" to the outside world in terms of foreign trade and business relationships (most notably extensive processing arrangements between Nanhai TVEs and Hong Kong firms). Nanhai has a diversified structure of rural communities, with much greater independence at the lower levels (villages and teams) than is the case in Wuxi. This has spawned a variegated pattern of TVE ownership, in which TVCEs are important but by no means dominant, and production team enterprises, PEs, and various kinds of firms "contracted" to private management, "Joint ventures", and even  - 10 - 'cooperative' entities have emerged. There is a relatively open labor market, which has stimulated a large influx of migrant workers from poorer areas out- side the county. As a result, wage controls are largely ineffective. Capital flows from outside appear to be less important, mainly because the county is well-endowed with local bank savings deposits. 1.27 Jieshou County in Fuyang Prefecture of Anhui Province represents a more-or-less "average" level of TVE development for China as a whole, but its policies toward the TVE sector are unusual. It is much less industrially developed than either Wuxi or Nanhai, and its average rur-al per-capita income in 1985 was somewhat below the national average. The most distinctive features of Jieshou are a lively market in land and the emergence of a subttantial ntumber of large PEs, under the personal sponsorship of county and CG authorities. The county has encouraged relocation of PEs to Jieshou Town (the county seat), where infrastructure and support services are more readily available. PEs in Jieshots Town are vulnerable to "creeping collectivization" in various forms, however. In the rural areas ottside the county seat the level of industrialization remains extremely low, and most CEs have either gone out of business or have been "privatized" in one way or another. "Specialized villages" engaged in handicraft production have emerged in Jieshou and neighboring counties as a new mode of rural nonagricultural development. 1.28 Shangrao County in Jiangxi Province represents a level of rural industrialization and TVE development significantly below the national average. Even though the county is endowed with substantial mineral resources, rural incomes depend primarily on agriculture, industrial firms are small and mostly unprofitable, and industrial development has not yet been put on a self-sustaining basis, despite an infusion of capital from the central government. The ownership and management of TVCEs is in the 'traditional' mode, but unlike in Wuxi this has not been successful. TVEs are plagued by problems of excessive levies by CGs to meet public expenditure needs, wide- spread overmanning, poor management, lack of contact with markets outside the locality, etc. 1.29 These four counties to a certain extent exemplify forr different "models" of TVE development in China. One other important "model" is that of Wenzhou Prefecture in Zhejiang Province. A poor and remote area, Wenzhou has achieved great success in TVE development by permitting almost completely free development of PEs; a thriving "informal" financial market, based largely on private financial institutions; and extensive commercial relationships with distant parts of China. Wenzhou is different from Jieshou because the development of PEs may be somewhat less related to personal support by CG or local government officials, and because of the much greater emphasis on private service sector development, especially in finance. Like Nanhai, Wenzhou has a well-functioning labor market, with large flows of workers from poorer, mountainous parts of the prefecture into the booming tcrwns. It also has a market in land. 1.30 The diversity of the TVE sector means that reforms and policy measures which work well in some areas may be unsuccessful in others. For  example, the turn to PE-based industrialization in many poorer areas is based on a recognition that the traditional CE system (epitomized by Wuxi) cannot be successful in the more backward areas, for a variety of reasons. It is impossible at present to assert unequivocally that any one model of local TVE development is better than the rest under all conditions and circumstances. Over time, the relative advantages and disadvantages of different models will be tested during the process of market competition, and the TVE sector as a whole may gravitate toward one or more of the most successful and efficient models. But this is a natural process which should not be rushed by premature Government decisions and administrative directives. Recent Growth Performance 1.31 The growth of the TVE sector since the late 1970s and particularly since 1983 has been spectacular, whether measured in terms of output, employ- ment, assets, or profits (see Table 1.2). Nominal gross output value of all TVEs increased at an average annual rate of 25% in 1978-87. GVIO rose by over 262 p.a. in nominal terms during the same period. Real growth rates are hard to calculate due to lack of accurate price deflators, but even with generous assumptions about inflation (which was low until the mid-1980s), real output growth has been in the 202 p.a. range. The nonagricultural labor force in rural areas (industry, censtruction, transportation, and commerce) grew by 13? p.a. in 1978-86, for a total increase of 124% (over 30 million people).8/ TVCEs' total wage bill, assets, profits, and tax payments all increased rapidly in line with the growth of output and employment. Profits have risen much more slowly than other indicators, however, suggesting that profitability declined in the TVCE sector. Average wages in TVCEs rose by nearly 132 p.a. in 1978-87, considerably faster than the rate of inflation.9/ Though the number of TVCE firms has increased only very slowly, new enterprise creation in the PE sector has been tremendous in recent years.10/ 8/ These calculations are based on China's official rural labor force statistics, compiled by the State Statistical Bureau, which appear to seriously underestimate the number of people involved in part-time or seasonal off-farm activities and those employed by very small firms. According to figures compiled by the Ministry of Agriculture's TVE Bureau, the total number of nonagricultural TVE employees (including part-time and seasonal workers) rose from 22 million in 1978 to 77 million in 1986, an increase of 55 million. 9/ Average wages in TVCEs rose far more rapidly than in SEs in 1978-83 (12X p.a. versus 6% p.a.), but this pattern was reversed in 1983-86, when TVCE wages increased by 141 p.a. and SE wages by 162. 10/ The number of PEs reportedly rose from 4.2 million at the end of 1984 to 10.4 million at the end of 1985, 13.4 million at the end of 1986, and 15.9 million at the end of 1987.  - 12 - Table 1.2: GROWTH PERFORMANCE OF TVE SECTOR, 1978-87 (X p.a.) 1978-68 1988-87 1984 1986 1986 1987 1976-87 Output *lural nonagricultural output value (nominal) t 16.6 36.4 84.9 46.8 80.2 84.8 25.0 Industrial output valuo (nominal) 19. 41.6 56.2 45.7 82.1 84.4 26.4 Employment Nonigricultural labor force 6.2 19.9 38.6 20.1 16.1 8.1 12.1 Rural lndustrlal labor force 3.1 16.9 20.0 18.2 14.6 5.0 7.4 Number of Firms MrTVs -2.5 4.1 22.6 -4.9 -8.8 4.0 0.4 Industrial TVCEs -1.8 6.8 21.1 -5.1 2.7 10.6 2.2 Financlol Indicators (TVCEs only) Fixed assets 15 16.7 26.7 20.9 80.5 26.2 29.6 20.6 Profits /c 7.4 14.1 18.4 81.6 -2.1 16.0 10.8 Taxes L/ 21.8 80.0 34.4 87.2 26.9 22.0 25.3 Bank loan. 84.7 55.9 102.6 40.8 47.0 41.4 48.8 Orose revenues 17.8 88.8 86.6 44.1 21.7 82.0 24.2 Employment 2.7 9.8 18.9 7.9 5.8 7.1 5.8 Wage bill 15.2 24.9 86.1 26.0 17.9 20.8 19.4 Average wage 12.1 18.7 14.5 16.7 11.4 12.8 12.9 Averago wage, state units 6.1 15.6 19.5 17.8 16.6 9.8 10.2 /a Oross output value of rural nonagricultural materiale-producing *sctors (industry, construction, transpor', and commerce). b Value of fixed assets at original cost, without subtracting depreciation. Gross profits before payment of income tax (but net of Indirect taxes). tc Includes both direct and Indirect taxes. Sourcoe: Statistical Yearbook of China 1981, 1986, 1986, 1987; China Rural Statistical Yarbook, 1985, 198s, 1987. 1.32 Growth in China's TVE sector has been highly uneven over time (Table 1.2). There was a sharp acceleration in 1983-87 as compared with 1978-83, with GVIO rising by 202 p.a. in the earlier period and 42Z p.a. in the later period. Growth in 1984 and especially in 1985 was substantially higher than in 1986 and 1987. TVE growth has also been unbalanced geographically, with some areas, particularly more advanced coastal regions, experiencing more rapid growth than backward, less industrialized interior provinces. Between 1980 and 1985, TVE nominal GVIO rose by 4152 in Zhejiang  - 13 - Province, by only 372 in Inner Mongolia. In eight provinces the increase was greater than 2502, in seven it was below 120Z. Many though not all of the fast-growers were more industrialized from the beginning, and a number of backward provinces in terms of rural industrialization slid further into relative backwardness. Provincial growth rates mask great differences across counties within provinces. 1.33 Table 1.3 provides a more detailed picture of industrial output growth for TVEs under different forms of ownership. TVCE GVIO has grown at 23Z p.a. in 1978-87 and was especially high in 1984-85. GVIO of PEs (including production team firms) has risen sharply, starting from a iiny base in 1980. Some of this may reflect incorporation of economic activities which previously escaped the statistical network, but most of it represents real growth. Differential growth rates have led to substantial changes in the composition of TVE industrial output. The shares of TEs and VEs have declined, while that of PEs jumped from negligible levels in the late 19708 to 21? in 1987. Table 1.3: GROWTH AND OWNERSHIP STRUCTURE OF TVE GROSS INDUSTRIAL OUTPUT VALUE, 1978-87 (Z p.a. and Z of total) 1978-83 1983-87 1984 1985 1986 1987 1978-87 Annual Growth Rates TVCEs 18.4 34.7 36.8 40.9 27.6 34.1 23.1 PEs /a 54.6 95.7 328.7 68.0 50.0 35.6 63.5 TVE sector 19.7 41.5 55.2 45.7 32.1 34.4 26.4 Shares in Total (1978) (1983) Township firms 56.8 53.5 45.9 43.7 42.3 42.0 Village firms 40.9 40.2 36.7 36.2 34.9 35.9 Team firms } n.a./b n.a.tb 2.0 } Partnerships 12.3 6.3 8.1 10.5 8.2 123.1 Individual firms } } 9.4 9.7 12.7 /a Including production team firms. lb Many production-team firms were "privatized' in the early 1980s. In 1984 and 1985, they seem to have been included in the category of partnerships. Since 1987 no PTEs have been reported in the statistics. Sources: Statistical Yearbook of China, 1981, 1984, 1985, 1986; China Rural Statistical Yearbook, 1985, 1986, 1987; and China Economic Yearbook, 1981, 1985.  - X Jt - Overall Dimensions of China's TVE Sector 1.34 Given the rapid growth documented above, the overall dimensions of the TVE sector have been changing substantially. For example, measured in terms of real GVIO, China's industrial TVE sector was roughly four times as large in 1986 as it was in 1980. Thus any attempt to describe the size and dimensions of the TVE sector can only be a snapshot of a rapidly changing situation, which may quickly become out-of-date. 1.35 Table 1.4 shows the dramatic changes in the composition of gross output value in China's rural areas. The share of crop cultivation dropped from somewhat over half in 1980 to less than one-third in 1986; the share of other agricultural activities remained more-or-less constant; and industry rose from less than one-fifth to nearly equal the share of crop cultivation in 1986 and surpass it in 1987. Construction, transport, and commerce rose modestly as a share of the total, indicating there is still much potential for future gains. Table 1.4: COMPOSITION OF RURAL GROSS OUTPUT VALUE, 1978-87 1978 1980 1983 1984 1985 1986 1987 Rural total gross social product /a 203.8 279.2 412.4 506.8 634.0 755.4 943.2 Shares in Total Crop cultivation 52.6 49.3 47.1 43.3 36.0 33.1 30.1 Forestry, livestock, fishery, and sidelines 16.0 19.5 19.6 20.1 21.1 20.0 19.5 Industry 19.4 19.5 20.0 22.9 27.6 31.5 34.8 Construction, transport, and conmerce 12.0 11.7 13.3 13.7 15.3 15.4 15.6 /a In current prices. Includes gross output value of the five "materials- producing" sectors (agriculture, industry, construction, transport, and commerce), excluding nonmaterial services. Sources: China Rural Statistical Yes..book, 1986; and Statistical Yearbook of China, 1987, 1988.  - 15 - 1.36 Shifts in the structure of the rural labor force have also been significant, as is shown in Table 1.5. Cultivation declined sharply in 1984 and 1985, but much of the slack was taken up by agricultural sidelines, which more than doubled their share. The proportion of the rural labor force engaged in industrial activities rose only modestly (from 6Z in 1978 to 8Z in 1986). Construction, transport, and commerce more than quintupled their share, however, rising from 12 in 1978 to 6Z in 1986. Overall, the movement of labor out of agriculture in rural areas has only begun. Nearly four-fifths of the rural labor force is still engaged primarily in agriculture and sideline activities. Table 1.5: STRUCTURE OF RURAL LABOR FORCE, 1978-87 (million persons and Z of total) 1978 1980 1983 1984 1985 1986 1987 Total Rural Labor Force /a 306.4 318.4 346.9 359.7 370.7 379.9 390.0 Shares in Total Agriculture (cultivation) 83.8 82.8 81.5 70.9 67.3 } } 80.2 79.2 Forestry, livestock, side- } lines and fishery /b 5.9 6.4 6.4 13.3 14.6 } Industry 5.7 6.1 5.8 6.7 7.4 8.3 8.5 Construction, transport, and commerce /c 1.2 1.6 2.5 4.4 5.5 6.1 6.7 Other /d 3.4 3.1 3.8 4.7 5.2 5.4 5.6 /a At year-end. Does not include labor force of state farms. /b The industrial portion of sidelines labor force has been included under industry. /c Commerce includes storage, catering, and other related personal services. /d Including real entate administration, public utilities, residential services, consultancy, public health, education, sports, social services, scie:utific research, technical services, banking and insurance, govern- ment, Party, etc. Sources: China Rural Statistical Yearbook, 1985, 1986; Statistical Yearbook of China, 1987, 1988.  - 16 - 1.37 The TVE sector has become increasingly important as a generator of fiscal revenue ir. rural areas and a major provider of personal incomes for rural inhabitants. Total tax payments by TVEs in 1985 (including both profit tax and indirect taxes) were nearly Y 14 billion, accounting for only 7.7? of total national budgetary revenue but the lion's share of tax collected in rural areas (the stagnant agricultural tax generated only Y 4 billion). Tax payments by TVEs have been rising sharply as earlier exemptions are gradually phased out, though "special treatment" remains common and PEs in particular can evade much of their tax liability. The TVE sector probably is still some- what more lightly taxed than the state sector (though the difference is reduced if extrabudgetary remittances of TVEs are factored in), but on the other hand its tax burden is much heavier than that of agriculture. 1.38 In 1985, wage payments and distributions to community members by TVCEs totalled Y 31 billion. This represented Y 40-50 per member of the rural population, or about 15Z of average per-capita rural income. PE wages add significantly to these figures. But nevertheless, the share of the TVE sectcr in personal income generation is far less important than its role in produc- tion. This is partly due to much lower net/gross output value ratios in industry and especially commerce than in agriculture. Moreover, most TVCE profits go into investment and thus affect personal incomes only indirectly. 1.39 Extrabudgetary remittances of TVEs to CGs are very important, especially in the more developed areas. In 1985 TVCEs turned over Y 6.8 bil- lion in profits to their supervisory CGs, not including management fees which may have totalled as much as Y 2 billion or even more. These figures compare with total tax payments by TVCEs of Y 10.9 billion. More than half of TVCE remittances to CGs were ploughed back into reinvestment in them, but substan- tial amounts were used by CGs for supporting agricultural production, for social services, for covering the CG's payroll, and a considerable amount (Y 620 million) was distributed to community members. 1.40 The TVE sector has become the most dynamic part of Chinese industry, an important actor in most subsectors and dominant in a few. Its share in total national GVIO has risen sharply, from only 3Z in 1971 (when the movement to promote commune- and brigade-run enterprises was getting underway) to 9Z on the eve of reforms in 1978 and then, rising very rapidly starting in 1984, to over 23? in 1987 (Table 1.6). The state sector's share has declined consider- ably (from over four-fifths in 1975 to three-quarters in 1980, five-eighths in 1986, and under 60% in 1987), while that of urban nonstate enterprises has risen moderately. All parts of the TVE sector have seen their shares in Chinese industry as a whole increase, but PEs have shown the most spectacular growth, rising from negligible levels in the late 1970s to account for nearly 5Z of China's total GVIO in 1986.  - 17 - Table 1.6: OWNERSHIP STRUCTURE OF CHINESE INDUSTRY, 1971-87 (X of total Industrial output value) /a 1971 1976 1978 1980 1981 1982 1983 1984Tb 198 1986 1987 state 86.9 81.2 77.6 76.1 74.8 78.8 72.6 67.6 75.9 62.8 69.7 Urban collective 10.9 13.7 13.7 14.4 14.1 14.3 14.4 15.9 15.9 165. 14.6 Urban IndividualsC - - - 0.0 0.0 0.1 0.1 0.2 0.8 0.4 0.6 Urban other - - - 0.6 0.6 0.7 0.8 1.1 1.2 1.6 2.0 Rural nonstat. 3.2 6.1 8.7 10.0 11.0 11.2 12.1 16.2 17.7 20.8 28.1 Of which: Township 1.6 2.6 4.8 6.4 5.9 6.0 6.8 7.7 7.8 8.8 9.8 Village ) ) } } } } 6.0 6.8 7.6 8.4 1.6) 2.6 8.9 4.6 5.1 6.2 6.8 Below village } } } } } } } 1.6 8.1 4.5 6.4 Of which: Individual fires 1.6 2.6 8.8 /a Percentage. for 1971-80 are bascd on Industrial output value figures In 1970 constant accounting prices, for 1981-88 in 1980 constant prices, and for 1L84-87 In current prices. Since these figures are based on data compiled by the Stat. Statistical Bureau, they show slightly lower TVP shares than Ministry of Agriculture statistics would Imply. /b In the absence of current price data for ownership forms other than rural nonstat. industry, the shares of different urban ownership forms were calculated on the basis of their constant-price shares. There may be some inconsistencie between 1984 date and earlier and later data, which are from different sources. /c Includes partnerships. Sources: Statistical Yearbook of China, 1988, 1986, 1986, 1987, 1988; China Rural Statistical Yearbook, 195, 1986, 1987; Statistical MaterialseonCR'is Industrial Mae rla i, 1949-84, 1986. 1.41 TVEs are active in virtually all the broad subsectors of Chinese industry, as can be seen from Table 1.7 (which due to data limitations pro- vides information on TVCEs only). Only in one of the 14 major subsectors (petroleum) is the share of TVEs truly negligible. The most important TVCE industrial subsectors are machinery, construction materials, textiles, chemicals, and food (in that order), which together account for nearly three-  - 18 - quarters of TVCE gross industrial output value. Among these major subsectors, however, only the share of construction materials is far higher than it is for industry as a whole--otherwise shares tend to be no higher than or moderately lower than in industry as a whole. The composition of TVCE industrial. output has remained reasonably stable over time, with gradual rather than abrupt changes in relative shares. Table 1.7: SUBSECTORAL COMPOSITION OF GVIO, 1982 AND 1985 (Z of total) 1982 1985 Industry Industry as a whole TVCEs as a whole TVCEs Metallurgy 8.7 2.4 8.0 3.4 Electric power 3.7 0.7 3.3 0.4 Coal and coke 3.0 5.5 2.5 3.8 Petroleum 5.2 0.1 4.5 0.1 Chemicals 11.8 8.2 11.2 8.4 Machine building 22.0 24.8 26.9 25.5 Construction materials 4.0 20.7 4.2 18.9 Forestry 2.0 2.5 1.6 3.0 Food 13.5 8.3 11.5 7.9 Textiles 15.5 11.6 15.3 12.5 Garments 2.5 3.8 2.4 3.7 Leather 1.0 1.3 0.9 1.6 Paper 3.6 1.6 1.3 2.4 Cultural/educational n.a. 2.6 2.6 3.0 Other 3.5 5.9 3.8 5.4 Source: Statistical Yearbook of China, 1983 and 1986. 1.42 TVE sector shares irn the output of particular industrial products vary greatly, but again, the coverage is rather broad. TVEs produce nearly 80Z of all bricks in China, but perhaps more surprising, they are responsible for almost 20Z of total national cement production, three-eighths of national output of silk textiles, nearly a quarter of paper qnd cardboard, around 30Z of phosphate fertilizer, and 262 of coal. The share of metal-cutting machine tools is under 42, but TVEs are increasingly important in electronics (where  - 19 - they accounted for 6Z of total national output value in 1986), consumer dur- ables of many kinds, and even some more sophisticated industries. Only certain basic industries using process techniques with substantial economies of scale (e.g. petroleum refining) are impervious to entry by TVEs. 1.43 All in all, the TVE sector has established its importance in a number of crucial spheres of the Chinese economy. It has been responsible for the bulk of incremental output and income in rural areas, and it represents the only hope for moving rural labor into productive employment outside of agriculture, particularly as long as restrictions against permanent migration into large cities are kept in place. Without TVE development, much of the efficiency gains from agricultural reforms could not be realized. At the same time, the TVE sector is increasingly important in Chinese industry as a whole, both in terms of meeting rising demand for the whole range of industrial goods and as a competitor/partner to state industry. Conduct of Business and Market Interactions 1.44 There is ample quantitative as well as anecdotal evidence that TVEs function primarily in a market environment for their outputs and nonfactor inputs. The share of government mandatory or "guidance" plans in the total output of most TVEs is minuscule, usually zero. Moreover, in most cases where planning is reported to account for a substantial share of total enterprise output, plans are either merely reference targets or more in the nature of market-based procurement orders by state agencies. On the input side, TVEs rely primarily on the market mechanism. the only major exception is electric power, but even here, the market is playing an increasingly important role at the margin. Unlike many SEs, the bulk of TVEs buy and sell goods for money rather than through barter-like exchanges. 1.45 A striking feature of the TVE sector is the outward product market orientation of firms with respect to their home communities. This is only natural given the relatively small size of rural communities and their limited "home markets". CGs have little or no ability to 'protect" local TVEs from outside competition anyway. Outward orientation is most striking in the more industrialized localitiess in Wuxi County in 1984-85, for example, only 42 of the gross sales of all industrial TVCEs occurred within the home township of the producer. Eighty-two percent of a sample of firms in Wuxi sold less than 202 of their output within the county in 1985, while large majorities of similar samples of enterprises in all four counties subjected to intensive fieldwork sell at least 40% of their output outside the home province. For many of the larger, more successful TVEs, the relevant sphere of sales activity is the national market. The combination of outward orientation and inability to rely on CG protectionism in large part explains the dynamism and competitiveness of the TVE sector. 1.46 In contrast, the TVE sector is heavily community-oriented with respect to factor markets and investment and location decisions. TVEs of all forms of ownership (including PEs) must obtain key factors of production through the good offices of CGs. This is especially true of land and capital from the banking system, but in some areas CEs also get labor allocations .rom  - 20 - CGs. Moreover, numerous different CG permits and approvals are needed to start a TVE business and stay in operation, whether CE or PE. Virtually all of China's TVEs are rooted in particular rural communities, obtain factor inputs from within them, and do not consider relocation (particularly outside the county) as an attractive option. 1.47 TVEs to a surprisingly large extent do their own procurement and marketing. This is due to the relatively "thin" commercial network in rural China, as well as TVEs' relative lack of access to state plan distribution channels and their need to penetrate new, often hostile markets. The relatively large purchasing and marketing staffs of TVEs and nigh marketing expenses are undoubtedly essential in the environment in which TVEs currently operate. Similarly, TVEs cannot rely on a well-developed network of other industrial support services and instead often must engage in self-provision or do without them (see Chapter V). 1.48 The markets for industrial goods that TVEs operate in were just emerging in China in the late 1970s, so they have been to a large extent shaped by TVEs. The state sector left many gaps of unfilled demand, and the sharp rise in urban and rural personal incomes starting in the late 19709 generated additional demand which SEs were often initially not well-placed to meet. There were many opportunities to earn extraordinarily high profits at the outset, and hence early entrants (mostly from the richer areas which had already undergone TVE development to a certain degree) benefited greatly, and could grow rapidly through internal accumulation. 1.49 However, as competition became .aore severe due to widespread entry into the industries concerned, prices and profit margins tended to fall sharply, squeezing the later entrants. in markets where entry by TVEs is relatively easy, the typical pattern has been erosion of initially high profits as widespread new entry occurred (which sometimes overshot the appro- priate long-run industry capacity). Of course new opportunities have been generated by the growth of the TVE sector itself, but these often require better skills and more advanced technology. 1.50 There is considerable variation across industries and products in the market structure and nature of competition faced by TVEs. For a few indus- trial goods, markets are essentially nationwide in scope and competitors include SEs as well as TVEs. Specialized machinery produced by many firms in Wltxi falls into this category, and Wuxi TVEs have often emerged successful ';om competlsion wIth the state sector. In some industries, the competitive advantages of TVEs are so great that most SEs have already exited or are in the process of doing so (usually by changing their product lines); competition may be quite severe, but it is entirely within the TVE sector. In still other industries, transport costs or other restrictions may result in segmented markets along local or provincial lines. Even so, there may be considerable competition among TVEs and with SEs within provinces and localities. There are at most only a few TVE industries c.iaracterized by very weak or non- existent competition.  - 21 - 1.51 Overall, TVEs face a relatively "hard" and generally highly competi- tive market environment. This puts great pressure on many of them to perform well and is relatively unforgiving of mistakes, undoubtedly contributing in a major way to the efficiency, dynamism, and flexibility of the TVE sector. Unlike SEs, TVEs have no "captiver markets for their products. They are also less constrained by price controls, making it less likely that they will face chronic sellers' markets for their outputs. TVE product markets have been highly unstable, which is only natural given their recent emergence and administrative and economic "shocks" which have occurred at times.ll/ TVEs have shown impressive ability to adjust successfully in the face of adverse circumstances, through measures ranging from changes in product lines to downward adjustments in wages, replacement of management, major restructuring, privatization, and to a lesser extent et the extreme, going out of business.12/ 1.52 There are some problems, however. Many TVEs produce shoddy, poor- quality goods using backward techniques which waste materials or energy and often generate considerable pollution as well. Sometimes these shortcomings reflect the weakness of competition from the state sector (combined with limi- tations on the size of the bulk of TVEs). More generally, the "overheating" of the Chinese economy generates an overall environment of high demand, in which firms and production techniques (both inside and outside of the TVE sector) which otherwise would fall by the wayside can survive and even prosper. Further improvements in the functioning of markets for industrial goods in China, a more competitive state sector, and some cooling off in the macroeconomic demand environment would all help ameliorate these problems, though they might also result in a slowdown in the TVE growth. Industrial Structure 1.53 The average TVE is a relatively small concern, and most rural communities are still largely nonindustrialized. The average industrial TE in 1985, for example, had 56 employees and a gross output value of a little over Y 300,000, while VEs were only about one-third as large. Many rural townships have only a handful of TVEs and GVIO of less than Y 1 million, generating profits of only Y 100,000 or so--perhaps Y 5 per member of the community on average. Although average size of firms and number of firms per community are much larger in the more advanced areas like southern Jiangsu and Guangdong Provinces, there are very few large TVEs and virtually none with annual GVIO of over Y100 million. This is related to the fact that all TVEs are closely tied to rural communities, even the most economically developed of which have limited resources, and factors of production cannot be freely acquired from elsewhere. 1.54 The more industrialized rural communities tend to have a mixture of 'specialized' and "conglomerate" industrial structures. There is usually some specialization in a broad industrial subsector or two (for instance machine building in Wuxi County or textiles in Xiqiao Town of Nanhai County), but ll/ For example, the tight national credit policy imposed in early 1986 had an immediate and severe impact on TVEs. 12/ In NanhRi County in ;986, for example, about 4Z of CEs and about 15Z of PEs went out of business.  - 22 - typically there is also some spreading of firms in a community across indus- tries. Such diversification serves as a mears by which communities can spread risks from market fluctuations, in the absence of financial instruments for this purpose. 1.55 There is great regional variation in industrial structure, just as there is in firm size, financial performance, and TVE development itself. The more backward areas tend to have a high share of resource-based industries and "transport-protected" activitiea (like construction materials), along with basic food processing. Manufacturing in general, and more 3pecifically the dynamic industries with potential to stimulate rapid output growth and labor absorption, tend to be concentrated in the more advanced coastal regions. Despite their increasingly higher wages, these areas still offer a better administrative and economic environment for industrial development. 1.56 Turning to financial structture, industrial TVEs on the whole are much more highly leveraged than SEs. This partly reflects their recent rapid growth and relative lack of reliance on state and CG grants of funds. The capital structure of a sample of 200 larger TVEs (nearly all of them TEs) is presented in Table 1.8. CG contributions are very important in providing "seed" capital for CEs at time of founding (contributing close to 30Z of the total), but their share declines sharply as successful firms expand. The two most important sources of funds for established, growing CEs are internal accumulation and bank loans, though arrears to other firms have grown considerably in recent years (partly reflecting the tight credit situation). Capital from individuals (whether the firms' own workers or other community residents) accounts for only a small part of the total.  - 23 - Table 1.8: CAPITAL STRUCTURE OF LARGE TVES (2 of total) At End- End- End-June founding /a 1984 1985 1986 Government grants and loans 4.0 1.8 1.6 1.7 CG contributions /b 29.4 6.3 6.0 4.4 Enterprise's own funds /b - 41.0 40.7 43.2 Bank loans 38.2 28.9 24.1 23.4 Capital from other firms Ic 13.0 14.8 19.1 19.8 Capital from individuals td 6.5 3.5 4.6 4.0 Other 8.9 3.7 3.9 3.5 /a This represents different points in time for different firms. /b The share of CG contributions is somewhat understated and that of internal accumulation is overstated because capital provided by the township gov- ernment on a nonrepayable basis is included in the latter rather than in the former. However, given rapid growth of the firms concerned, the orig- inal contribution of CGs would account for a much smaller share of total capital than at founding. /c Mostly payables (possibly including some to government agencies), but also including small amounts of explicit loans and "investments" from other enterprises. Id Including bonds purchased by workers (jizi), wages payable to workers, and capital provided by individuals outside the firm. Source: Zhou Qiren, 'The Macroeconomic Effects of Debt Management--A Policy- Oriented Analytical Report on an Investigation of 200 Large Township and Village Industrial Enterprises in 10 Provinces," Draft Report, May 1987, pp. 11, 20. 1.57 PEs rely primarily on household capital (from founders, friends, relatives, workers getting jobs, etc.) when they first get started. Like CEs, if successful in their business activities they can rely to a large extent on internal accumulation to finance expansion. But beyond a certain point, many larger PEs exhaust these two sources of capital and try to get loans from Rural Credit Cooperatives (RCCs) or (less commonly) Agricultural Bank of China (ABC) offices. Some of them may also turn to new nonbank financial inter- mediaries which are emerging in certain parts of the country. Overall, balance sheets of relatively large and successful PEs might not look too different from those of CEs. 1.58 Declining profitability in the TVCE sector is evident from Table 1.2, though profit rates on sales and assets are still considerably higher on average than in state industry. This trend reflects (a) exceedingly high profits initially; (b) increasing competition in many TVE markets, which has  - 24 - put downward pressure on prtces and squeezed profit margins (especially since prices of many raw materials have been rising); (c) rising average capital intensity of TVE production (though even now, it is still far below that in state industry); (d) disproportionately rapid rises in prices of investment goods, which overstate the increase in capital stock and hence the decline in profitability as well; end (e) possibly concealment of profits for tax evasion purposes, as previous exemi1ptions are phased out and the effective profit tax rate rises. 1.59 Poor financial performance is a problem for some TVEs, and there are certain inhibitions against exit by failing TVCEs, though not as strong as in the case of SEs. The need to maintain employment is obviously important in localities where there is still substantial surplus labor. In the more developed areas where surplus labor may not be a problem, local banks generally try to force CGs to take responsibility for bank debts of defunct CEs. As a result, there is no incentive to shut down an enterprise unless it has negative cash flow, since CEs cannot reduce their debt obligations by going bank.upt. In 1985, 41,000 TEs (9.8Z of the total number) made losses totalling Y 600 million, while a smaller number of VEs (24,000, or 2.1Z of the total number) made losses totalling Y 250 million. Though much lower than the correspcnding figures for SEs, these losses are nevertheless substantial. Going downward through the ownership spectrum of the TVE sector, exit becomes easier and hence there are fewer chronic loss-making firms in existence. PEs simply fold, effectively goirg bankrupt, in the face of severe financial problems. The Question of Relative Efficiency 1.60 It is extremely difficult to assess the degree to which the TVE sector is more or less efficient than the state sector, in terms of production costs and factor usage. Finding comparable firms in the two sectors with comparable raw data is a major exercise in itself, but ascertaining the reasons behind observed differences in performance would be even harder, given the numerous distortions in the economy and a generally weak accounting system for the TVE sector. Input costs for TVEs may be higher than those of SEs, for example, not because the former are less efficient but because the latter have access to inputs at lower controlled prices through the plan. TVEs may also be able to inflate their costs to a greater extent than SEs, in the interest of tax avoidance. On the other hand, many SEs are still forced to sell a large part of their output through the plan, at plan prices, which would make them appear less profitable than TVEs. Similarly, the social expenditures of TVEs (on workers' welfare, pensions, housing, etc.) are much lower than those of SEs, which would allow them to incur lower costs and earn higher profits than SEs even if production efficiency happens to be the same. In view of these considerations, any comparison of costs and profitability between TVEs and the SEs must be undertaken with great caution. 1.61 At the aggregate level, TVEs use somewhat more labor and much less capital than SEs per unit of output value. Hence any comparison of factor productivity which gives a substantial weight to capital would indicate that the TVE sector is more "productive". But these kinds of estimates are biased  - 25 - by distortions on both the input and output sides, and in any case, there is variation across products and industries. It would not be surprising if by and large, the TVE sector is more efficient than the state sector in the industries where it already accounts for a large share of total output. On the other hand, "sellers' markets" and "two-tier" plan/market pricing systems may allow high-cost TVEs to survive in some industries like cement and steel- rolling. 1.62 The question of relative efficiency is perhaps not as important as that of market orientation, dynamism, and competitiveness in the TVE sector. To the extent that these characteristics are present, they will move the sector in the direction of lower costs and greater efficiency, regardless of precisely where it happens to find itself at any particular point in time. Thus the crucial issues concern the functioning of markets faced by TVEs, the degree to which TVEs can flexibly respond to market opportunities, their adjustment to changing market conditions, and their overall competitiveness. In these respects, the TVE sector so far seems for the most part to have done much better than the state sector, though it will need to continue to evolve and develop in order to maintain a competitive edge in the future.  - 26 - II. MACROECONOMIC ISSUES AND INTERSECTORAL RELATIONSHIPS 2.1 This chapter looks at the broader macroeconomic and intersectoral issues facing the TVE sector, whereas those that follow will concentrate on issues more "internal" to the TVE sector itself. The increasing importance of the TVE sector in China's rural economy, in industry, and in the economy as a whole has given rise to questions in China about its impact on macroeconomic stability. Has the TVE sector contributed in a major way to demand pressures and "overheating' of the Chinese economy? If so, how should demand pressures from TVEs, for bank credit and other resources, be restrained? The TVE sector's interactions with the state sector and with Chinese agriculture also raise certain policy issues of concern to Chinese authorities. Are TVEs "inappropriately' competing resources away from state industry on the one hand or from agriculture on the other? Are financial flows from TVEs to agricul- ture a heavy burden on the former, and how effectively are these funds being used? Still another concern is the role of the TVE sector in foreign trade. The dynamism and flexibility of TVEs make them poten-ially successful exporters, but to expand exports they need contacts with foreign businessmen, access to foreign exchange resources, better market information, etc. The nature and implications of urbanization generated by TVE development form another crucial set of policy questions. Related to this are the types and magnitude of rural labor and population flows in the future, both rural-rural and rural-urban (discussed in Chapter V). A fundamental issue underlying all of the others is the future role of the TVE sector in the Chinese economy. Is TVE development just a transitional phase, or will TVEs still be prominent in the Chinese rural economy and in Chinese industry 20 years from now? A comprehensive assessment of this last issue must encompass the analysis of 'internal" TVE topics in subsequent chapters, however. Macroeconomic Impact 2.2 Due to its rapidly increasing size and importance, the TVE sector naturally has become a significant source of demand for bank credit, raw materials, energy, intermediate inputs, and other resources in the Chinese economy. But its role in causing the 'overheating' of the economy as a whole in 1984-85 and subsequently should not be overrated. In the first place, the share of national credit and other resources taken by TVEs has not been large in relation to the increasing size of the sector (Table 2.1). The share of TVCEs in total national loans outstanding was very low in the early 1980s: even after more than doubling to 52 in 1986, it is still quite modest in relation to the share of TVCEs in China's gross social product (122 in 1985, 142 by 1987). The share of TVCEs in incremental credit creation was only 7-81 even in the boom period of TVE development in 1984-85. Finally, it appears that the :ratio of bank loans outstanding to GVIO is much lower for TVCEs than it is for SEs (222 versus 382 in 1986, according to rough estimates).13/ Thus overall, the TVE sector's role in credit inflation, a primary source of macro- economic instability in the Chinese economy, is modest. 13/ Because of much lower capital-output ratios in the TVCE sector than in state industry, bank credit may well be more important in relation to total sources of capital for TVCEs than for SEs.  - 27 - Table 2.1: BANK CREDIT TO TVCEs, 1980-87 (Loans outstanding at year-end, Y billion, and Z) Total national bank and rural Share Share of TVCEs Loans to credit coop- of loans in incremental TVCEs erative loans to TVCEs credit creation 1980 5.600 249.594 2.2 - 1981 7.000 286.105 2.4 3.8 1982 8.200 317.342 2.6 3.8 1983 9.770 359.479 2.7 3.7 1984 19.796 477.409 4.1 8.5 1985 (1)/a 27.780 576.580 4.8 8.1 1985 (2)/a 27.780 630.547 4.4 1986 /a 40.825 815.873 5.0 7.0 1987 57.713 980.37 5.9 10.3 /a Before 1985, aggregate loan statistics do not include credit by the People's Construction Bank of China. The 1985 (2) data and 1986/87 infor- mation do include the Construction Bank. Loans to TVCEs presumably include credit from all state banks as well as from rural credit cooperatives. Sources: Statistical Yearbook of China, 1981, 1984, 1986, 1987, 1988; China Statistical Summary, 1987; China's Finance, various issues. 2.3 Caveats and qualifications slightly weaken but by no means nullify the above conclusion. Loans to PEs have not been factored in because of the difficulty of separating them from loans to individual peasant households (for agricultural, sideline, or consumption purposes). Though they have risen sharply, these still comprise a tiny portion of total credit. In 1986 total loans to individual peasants were Y25.8 billion, only 3X of total national loans outstanding, and only part of this was loans to nonagricultural PEs. Some loans to TVEs may have been hidden in various agricultural loan cate- gories, but the amounts must be small in relation to total credit. Some credit to SEs also has been transferred (voluntarily) by the enterprises concerned to TVEs through subcontracting and other arrangements. Finally, after tighter national credit policies were imposed starting in 1985, TVEs resorted to short-term bond issues to finance their continued expansion. Purchasers of these bonds may have drawn down their bank deposits to pay for them, possibly enlarging aggregate demand.  - 28 - 2.4 Overall, TVEs clearly have not taken a disproportionate share of national credit or government financial resources. Most CGs face a relatively "hard budget constraint', which appears to be effectively communicated to their subordinate CEs, and PEs do not face problems of a "soft budget con- straint" in the first place. Hence TVE demand for economic resources largely reflects legitimate economic and business concerns, and any distortions in this demand for the most part are due to distortions in the market and other "signals" TVEs respond to. If the state sector generates very high demand for certain kinds of goods, which the TVE sector then steps in and produces, this may exacerbate problems of overall macro control. But the source of the problem is not the TVE sector, which is merely responding to perceived effec- tive demand, and solutions should lie in restraining demand by the state sector. 2.5 More generally, while naturally a growing source of demand for inter- mediate and capital goods, the TVE sector has generated a tremendous increase in supply of goods demanded by SEs as well as by the urban and rural popula- tion. This has been done with overall capital-output ratios much lower than in the state sector. Thus it is misleading to look only at the demand side- -on balance the TVE sector has if anything contributed to macroeconomic stability through huge increases in supplies of goods that are in demand. For example, it has greatly expanded production of building materials, which help satisfy investment demand by all sectors of the Chinese economy. 2.6 There are two factors which may inflate TVE demand for resources, however. One is the enormous pressure CGs are under to increase nonagricul- tural employment, which may cause them and subordinate CEs to demand large amounts of investment funds and take imprudent risks. This could be a problem particularly in the backward areas with more surplus agricultural labor and a weaker base of TVEs, entrepreneurs, market opportunities, etc. But the macro impact should not be exaggerated: leaving surplus labor in agriculture saps the vitality of that sector, with adverse consequences. A second factor which may inflate TVE demand is the ambiguous relationship between CGs and local banks and credit cooperatives. In some of the more developed areas, symbiotic ties have evolved: CGs recognize that bank funds represent the community's own accumulation and should not be wasted, and there is a clear assignment of responsibility for bad debts of CEs to their CG owners. But in many other localities, particularly backward areas, CGs may use the local banking system as a means of obtaining "deficit financing", weakening their hard budget constraints and opening the way for inefficient use of funds. This may fuel demand for credit above what is economically appropriate. 2.7 The immediate policy question is whether additional quantitative credit controls should be applied to TVEs beyond those implied in the linkage between deposits and loans at the township level.14/ The latter, along with redeposit requirements, imposes some limitations on total credit in rural 14/ Under current Chinese banking practice, lower-level bank branches are limited in the amount of credit they can provide to their clients by the amount of deposits they generate locally--if they can attract additional deposits, they can grant a similar amount of additional loans (less required redeposits).  - 29 - communities. It is doubtful whether further quantitative limits on credit to TVEs should be instituted, unless the national macroeconomic situation requires severe stringency, and other parts of the economy (most notably the state sector) are being subjected to equally severe credit controls. 2.8 TVE demand for other resources (e.g. raw materials and energy) follows a broadly similar pattern to credit. It would not appear to be "excessive" in economic terms, though certain distortions in the system may make it higher than otherwise or may give TVEs an advantage in bidding for resources against SEs. The key point is that the remedies to these problems, to the extent that they are serious, lie in removing or ameliorating the distortions that give rise to them, not in direct restrictions on TVE activi- ties or growth. The TVE sector is the most dynamic and market-oriented part of Chinese industry, and it has been highly successful so far in generating employment opportunities for China's vast rural labor force. Hence direct administrative restrictions against TVEs would most likely harm the overall efficiency of the Chinese economy, and in particular they might weaken compe- titive pressures on SEs to improve efficiency. Competition and Integration with State Industry 2.9 Closely related to the macroeconomic impact of the TVE sector is its interactions with state industry. These fall into the following main catego- ries: (a) competition in product markets; (b) competition for material inputs and energy; (c) competition for factors of production, particularly human resources; (d) subcontracting by SEs to TVEs; (e) joint ventures, licensing of SE brand names, etc.; and (f) relations between TVEs and state distribution agencies. One central issue is whether and to what extent there is "unfair' competition, which prevents SEs from fully making use of their econol.lic advan- tages (like large scale and more advanced technology), thereby allowing exces- sive expansion of TVEs. Another key question concerns the integration of the TVE sector into Chinese industry as a whole. 2.10 Precisely assessing the relative advantages of TVEs and SEs in a complex, fragmented administrative and regulatory system still riddled with direct controls is extremely difficult (see Chapter I). TVEs may gain considerable benefits from their greater freedom from price controls, ability to spend large amounts on marketing and procurement, and tax advantages, but on the other hand SEs have access to material inputs through the plan, to state marketing channels, to advanced technology and foreign exchange, etc. TVEs have lower labor costs than SEP. not so much due to lower wages but because SEs must provide a whole set of mandated welfare benefits to their employees (see Chapter VI). The situation may be radically different in different industries and markets. For example, the "two-tier" system in the iron and steel industry allows high-cost small producers (many of which are local state-owned enterprises rather than TVEs) to survive, but they might vanish if a unitary pricing system is implemented and large complexes are permitted to compete freely. On the other hand, in the textile industry TVEs will remain and if anything grow in importance. Increasingly, these outcomes will be determined by economic factors rather than by administrative decisions.  - 30 - 2.11 A "level playing field" needs to be established for competition among all kinds of industrial firms (see Chapter III), but this will be a long-term task. In the interim, the temptation to impose direct restrictions against TVEs to ease competitive pressures on the state sector should be avoided. There are many industries, particularly those in which labor costs are a dominant factor, where TVEs could come to play an increasingly important role. Though adjustments by the state sector will be painful, they will lead to improvements in aggregate efficiency. In other industries, where economies of scale are significant, most TVEs may eventually exit, except perhaps for a few of the largest ones, if they are allowed to grow to competitive size. 2.12 The most pervasive form of cooperation between SEs and TVEs is sub- contracting of parts production, assembly, ancillary activities, etc. to TVEs by SEs. This is usually motivated by the need of the latter to get access to certain factors of production which they cannot obtain easily otherwise, most notably land and labor, and by their desire to take advantage of lower labor costs in TVEs. Sometimes TVEs merely obtain (for a price) SE brand names to help sell their products. TVEs can even serve as a dumping ground for old equipment of SEs. The benefits to TVEs from these arrangements include access to more advanced technology and equipment, money, distribution channels, brand names, (sometimes) low-priced inputs through the plan, etc. Whereas in the past many subcontracting and other arrangements were based at least in part on noneconomic motivations and administrative decisions by local governments, they are now increasingly entered into voluntarily by all participants to serve their own interests. 2.13 There are two key points concerning this kind of cooperation: (a) in the more advanced rural areas, it is no longer essential to the TVE sector--in Wuxi County, the share of TVE industrial output consisting of such subcontrac- ting declined from 702 in 1978 to only 21. in 1985; and (b) like "outsourcing" arrangements between firms in the United States and newly industrializing countries, it may lead to partial "de-industrialization" of the state sector and transfer of manufacturing capabilities to TVEs. 2.14 Overall, a policy and regulatory environment conducive to greater integration between state industry and the TVE sector needs to be built up over time. But such integration cannot be "forced", and its form should not be prejudged. Given the competitive environment many TVEs function in and their demonstrated adaptability, it might well turn out that some of them grow to the point where they are more-or-less on an equal footing with SEs in the same industry.l5/ TVEs can also play a positive role in stimulating efficiency improvements in SEs through competition. Relationship with Agriculture 2.15 The TVE sector is closely linked to Chinese agriculture in a number of ways. Though some of the linkages have been weakened by reforms since the late 1980s, in other respects ties are now closer than ever. The TVE sector 15/ In order to achieve this, these firms would have to transcend the limits of their rural community origins. See Chapter VIII.  - 31 - relies on much the same resource base as agriculture--the same land, the same human communities, the same banking institutions and other sources of invest- ment funds. Hence TVEs naturally compete with agriculture for these resources. Agriculture and TVEs are the two most important means of employ- ment and personal income generation in rural areas. Finally, there are impor- tant financial flows between the two sectors. Past TVE development relied to a considerable extent on investment funds from agriculture. But financial flows now also go in the opposite direction: TVEs provide subsidies to agricultural production and for agricultural investment projects. 2.16 It has been suggested that competition between TVEs and agriculture for resources is harming the long-run development potential of the latter. Funds of the rural banking system are increasingly being used in rural industry rather than in agriculture. Able-bodied and skilled personnel in rural areas with dynamic TVE sectors are reportedly shunning agriculture in favor of TVE jobs. The rapid growth of the PE sector also may have drained people from agriculture. TVEs are occupying large amounts of good agricul- tural land in the more developed areas, since the profitability of land in TVE activities is much higher than in agriculture. 2.17 It is difficult to assess the magnitude of this problem and the severity of its adverse impact on agriculture.16/ But regardless, administra- tive restrictions against the TVE sector or diversion of additional funds from TVEs to agriculture are not the answer. The problems of Chinese agriculture relate to continued underpricing of some farm products; underpricing and inefficient allocation of key agricultural inputs like chemical fertilizer; the need for new methods of financing investments in agricultural infrastruc- ture; a land tenure system which does not facilitate consolidation of farm plots large enough to support highly profitable farming; and other factors. The fundamental solutions to these problems must be found in agricultural policies and institutional reforms; the TVE sector should not be made a scape- goat for the problems of Chinese agriculture. Of course, solutions to China's present agricultural problems may not come immediately, and during the interim some administrative measures to prevent excessive and irreversible encroach- ment of agricultural resources (particularly land) by the TVE sector may be appropriate. These should not distract attention from or serve as a substi- tute for needed institutional and policy reforms in agriculture, however. 2.18 Funds flow from TVEs to the agricultural sector through a number of channels. A portion of TVCE profit remittances to township and village autho- rities is used by them to purchase agricultural equipment, finance agri- cultural investment projects, and aid poorer communities. These transfers to agriculture have been declining both in absolute amounts and as a share of total TVCE profits (Table 2.2). In 1978 they totalled Y 2.6 billion, 30Z of net profits of TVCEs, whereas in 1986 they were only Y 700 million, 42 of TVCE profits. During this period the ratio of TVCE net profits used for reinvest- ment in TVCE development (including both retained and remitted profits) to profits used for agricultural support increased from 1.17 to 11.6. Thus this 16/ After all, one of the main goals of rural nonagricultural development is to absorb 'surplus" labor from agriculture.  - 32 - flow of funds from TVEs to agriculture would not eppear too burdensome at present. There are other flows, however, whose magnitude is uncertain but which may well be increasing. Most important is spending by township and village authorities on various kinds of subsidies for agricultural production, also financed largely from TVCE profit remittances. Table 2.2: uSE OF TVCE PROFITS TO SUPPORT AORICULTURE, 1976-87 (Y billIon and %) 1978 1980 1981 1982 1988 1984 1985 1986 1987 Total Net Profits of TVCEs 8.81 11.84 11.28 11.66 11.78 12.87 17.18 16.10 16.78 Shares in Total Used for *id to agricul- tur/ 30.0 19.2 15.1 12.4 11. 6 7.8 6.1 4.8 4.6 Of whTeh: Purchase of equipment 18.1 7.7 6.2 4.6 3.6 n.a. na. n.s na. Field construction 13.4 7.9 7.1 6.2 6.1 n.s. n.. n.o n a. Aid to poor brigades 8.6 8.6 1.8 1.6 1.9 na. n a. na. n. . Used for investment in TVCE sector 86 1 39.7 8. 1 41.2 48.0 47.4 46.4 49.8 58.2 Ratio of TVCE investment to agricultural support (X) 117.0 207.0 262.9 881.9 878.4 919.0 902.8 1,159.4 1,161.7 /a Apparently Includes profits turned over to township and village authoritioo by TVCE* and then used by the latter for expenditures to support agriculture. Sources: Statistical Yearbook of China, 1987, 1988; China Rural Statistical Yearbook, 1986, 1986, 1987; China Agricultur- Yoerbook, 1988, 1984, 1986, and 1986. 2.19 Administratively directed flows of funds from the TVE sector to agri- culture are not the most appropriate response to the development needs of the latter. There is some evidence that "aid-to-agriculture" funds are used in an inefficient and wasteful manner. More fundamentally, since they compensate for underpricing of agricultural products, these flows effectively constitute a "tax" on the TVE sector (and hence on rural communities), which helps make agricultural goods available to the urban population at low, subsidized prices. Similarly, restrictions against TVE activities can be viewed as a measure to force rural areas to maintain agricultural production for the benefit of the cities. A much better approach would be to improve the incen- tives for agricultural production and the allocation of agricultural factors of production (e.g. land) and key inputs like chemical fertilizer.  - 33 - Urbanization and Migration 2.20 Rural industrialization in China has been accompanied by urbanization in small towns, mostly traditional market towns and township or county seats. In many areas agglomerations of people and nonagricultural activities have sprung up almost overnight. But many TVE factories are scattered throug iout the countryside, away from towns. A major reason behind urbanization in small towns and previously rural areas is the restrictions against migration into larger cities, which have been relaxed to some extent but still greatly hinder permanent moves and moves by families. But even within administratively demarcated rural areas, emerging patterns of industrial location and urbaniza- tion may be excessively dispersed. This is due to the strong community orien- tation of TVEs and entrepreneurs. Even at the village level, the bulk of TVEs (including PEs) are located in the communities to which they belong. Since a typical county has several hundred villages, the potential for overly dispersed patterns of industrial location is serious. 2.21 Dispersed factories and small-town urbanization carry high economic costs. Infrastructure is scattered and inadequate; transport costs are higher; excessive amounts of high-quality agricultural land may be diverted to industrial uses; and industrial pollution is spread widely in the countryside, with consequent difficulties and high costs for pollution control. Over the longer term, as rural industrialization proceeds, the costs of not moving to a more concentrated pattern of urbanization and industrial locationi will likely become progressively higher. Thus even if restrictions against permanent, family-based migration into large cities remain in place, there is consider- able scope and need for greater concentration of population and factory loca- tions within counties and, in the more industrialized areas, townships. This will require incentives and possibly a zoning framework which encourages sizable TVEs to locate in towns, even if these are outside the home community. Such policies and reforms might best be carried out by counties or higher levels of government setting aside land for 'industrial parks", not under the jurisdiction of any particular rural community, and providing economic incen- tives for TVEs and entrepreneurs to locate there (see Ghapter V). 2.22 At the broader level and from a long-term perspective, the policy of restricting migration into large cities and controlling city sizes can be questioned. In the first place, it is increasingly ineffective, and large numbers of "temporary" migrants are already living in large cities, taking on various menial and construction jobs. In any case, larger cities per se are not necessarily inefficient, as long as the population and firms located in them are paying the higher costs associated with large city size.17/ There are many economic activities with high returns which need the benefits of big- city location. Thus the crucial priority is more appropriate pricing and costing of infrastructure and public services in large cities, so that the economic activities for which large-city location is economically appropriate will indeed be located in them and other activities will be established in or move to other areas. Such reforms are a prerequisite for any substantial relaxation of migration controls. Even if there is considerable growth of 17/ See World Bank, Zheiiang: Challenges of Rapid Urbanization (World Bank Sector Report 6612-CHA) for a detailed analysis of urbanization in China's cities.  - 34 - population in China's large cities, a large part of future urbanization will necessarily have to occur in smaller towns. Foreign Economic Relations 2.23 There is great variation in the degree of access of different parts of the TVE sector to foreign partners, markets, funds, technology, and equip- ment. In Nanhai County, TVEs are free to engage in all kinds of processing and compensation trade agreements with foreign partners. Because of the area's large exports to Hong Kong and high remittance incomes, TVEs also have reasonably good access to foreign exchange. Most imported equipment and tech- nology comes to Nanhai TVEs through compensation trade. In Wuxi, by contrast, TVEs are not permitted to deal directly with foreign p'irtners, despite their large direct and indirect exports. Access to foreign exchange and foreign capital is highly restricted. In Jieshou, contacts with foreign businesses and even direct foreign investment are permitted, but activity has been limited so far due to lack of traditional connections. Inexperience in deal- ing with foreign businessmen has led to problems. in Shangrao County, there is not one piece of imported equipment installed in any enterprise (including state and urban collective firms as well as TVEs). The county's exports (primarily handicrafts) go through Shanghai. or Jiangxi provincial foreign trade authorities; producers have no direct contact with international markets and no access to the foreign exchange proceeds from their exports. Overall, the TVE sector is already a major actor on the export scene, generating annual exports of over US$5 billion, more than one-eighth of total export value. 2.24 The TVE sector's access to foreign economic contacts should be opened up. This undoubtedly would benefit areas like Wuxi greatly. Small firms often make excellent exporters because of their quick response and flexibi- lity. Many TVEs potentially could be competitive in export markets, but except in southern Guangdong and a few other places, this potential has not been realized. As part of reforms in China's foreign exchange system, TVEs should be permitted to participate in foreign exchange markets and buy foreign exchange at market prices to import equipment and technology, and to have access to the foreign exchange proceeds of their exports. TVEs should be permitted to have direct contacts with foreign businessmen and enter into various kinds of cooperative arrangements w'.th them, as is now possible in southern Guangdong. For small firms to be successful exporters, a strong network of trade-related services is necessary. TVEs should be given access to the existing network of state trading agencies (on a fee-for-service basis). But liberalization of and greater competition in the state foreign trade system also is needed. Trading companies and other kinds of enterprises providing trade-related services should spring up over time withiin the TVE sector itself. 2.25 There are a host of other mundane yet important changes which would help facilitate TVE exports. Greater standardization of product specifica- tions for many export goods and better quality control (including testing systems) are two examples. TVE access to export financing on an equal basis with other types of firms also would be very helpful. There may be great  - 35 - potential for subcontracting of export production tasks to TVEs by state foreign trade corporations as well as state-owned industrial firms, which is already occurring to some extent, and this should be further encouraged. At the broader level, the Government's exchange rate and related policies will, to a large extent, determine the competitiveness of many TVE exports and potential exports. 2.26 A final question concerns what actions could be taken to give back- ward areas more access to beneficial foreign contacts. At the very least, they should not be given lesser access to internationai markets than the more advanced coastal areas (which appears to be the common pattern at present), because this would only further increase the handicaps they already face. But it is only to be expected that given their lack of traditional ties and weaker industrial base, the more backward areas will not be able to take advantage of opportunities presented by opening up to foreign economic relations as quickly or as easily as the developed coastal regions. Development of markets for trade-related services could be very helpful for the backward areas, as would an inflow of export-oriented entrepreneurs from more developed localities (see Chapter VII).  - 36 - III. LEGAL AND REGULATORY FRAMEWORK Legal Framework for TVE Business Activities 3.1 As in the Chinese economy as a whole, the legal framework governing TVE business activities is weak and rudimentary. Contracts to a considerable extent cannot be enforced through the legal system, which leads to a number of problems. The ineffectiveness of contracts reduces economic efficiency and forces reliance on other, more personalized and customary means of doing busi- ness, which may have higher transactions costs. The.ce is considerable diver- sity in local practices and hence uncertainty for enterprises. 3.2 Given the close ties between TVEs and the rural cammunities they are based in, "outsider" firms or individuals are at a major disadvantage.l/ This inhibits factor mobility and mobility of human resources, because people are reluctant to set up businesses and production facilities outside their home communities. The community rootedness and commuiiity orientation of the TVE sector may be based primarily on the inadequate legal framework, rather than on strong "community consciousness" on the part of China's peasantry. 3.3 The ir.adequate legal framework, combined with the strong community orientation of the TVE sector, may also partly explain the relatively low share of services in output. Delivery of goods outside the home community is generally fairly straightforward--the value created by the producer is embo- died in the goods and moves with them to the place of purchase. Delivery of services outside the home community, on the other hand, often requires actions by employees and other factors of production in the locality where the service is "consumed". Thus the inhibitions against inter-community mobility of enterprises and factors also serve to restrict the growth of those services which are most economically provided on a multi-community basis. 3.4 Contract law also may not adequately support the rights of TVEs in dealings with SEs. TVEs have found it hard to enforce contracts with SEs when changing business conditions caused the latter to renege on them, and conse- quently sometimes have suffered heavy economic losses without remedy.2/ The unenforceability of many business contracts w'th SEs has forced TVEs into practices that guard against losses from breaking contracts, even though this may be more costly and less efficient for all concerned. 1/ In dealings with outsiders. there is identification between local TVEs and county governments and legal institutions. In this situation, criteria for determining where a case is to be adjudicated are obviously important, but these appear to be somewhat vague. Suits in principle are to be brought where the violation occurred or contract was signed, but if it is "inconvenient" for a defendant to travel to that location, they must be adjudicated in the defendant's home jurisdiction. 2/ SEs voice similar complaints about TVEs, which only further supports the assertion that the legal framework is highly inadequate.  -37 - 3.5 Strengthening contract law and particularly the enforceability of contracts is an urgent task. One option would be to follow the general prin- ciple that economic regulation of business entities under "local" ownership should be handled at higher levels of government so as to avoid conflicts between ownership and regulatory roles of government. This could be achieved through a system of courts and some kind of enforcement mechanism for dealing with business disputes that cut across communities and localities, probably at the provincial level.31 Such a system would provide recourse to "outsiders", while allowing disputes within communities and counties to continue to be handled by local economic courts or traditional methods. It would also provide recourse for TVEs in contract and other business disputes with SEs under local government ownership, and indeed for SEs in disputes with TVEs. 3.6 An alternative which does not involve establishment of entirely new institutions would be to strengthen existing mechanisms like the contract arbitration committees of the Industrial and Commercial Administration Bureaus.41 To become effective and credible means for enforcement of contracts, these committees or other existing institutional mechanisms would need to be "de-localized'. This might be a very difficult task, given the multitude of administrative, political, financial, and personal ties that bind local branches of the Industrial and Commercial Administration Bureaus as well as other government agencies to community and county authorities. Neverthe- less, if these difficulties can be overcome, use of existing institutions might well be less costly and more effective than creating entirely new ones. 3.7 Another important part of the legal framework is the rules for estab- lishing and operating enterprises of different types (known as 'company law" in other countries). China is in the process of drafting and promulgating enterprise laws (including one for TVEs), but there are already numerous registration and other regulations in place. Generally the procedural requirements for establishing TVEs appear not too onerous, but for PEs they vary considerably at the local level--in some places like Wuxi, local regula- tions are used to prevent the emergence of sizable PEs, whereas in other areas they are encouraged, or at least tolerated. Very small firms, as in other countries, are often set up without going through registration procedures anyway. There is also a great deal of variation in the degree to which 3/ Since much TVE business transcends provinces, a central government judicial body and enforcement mechanism for such disputes might also be necessary. This would be a key part of efforts by the central government to ensure "free trade" within China and prevent damaging 'internal protectionism". 4/ The Industrial and Commercial Administration Bureaus (nominally under the supervision of the General Bureau of Industrial and Commercial Administration in Beijing) are generdlly responsible for registration of firms, monitoring compliance with business laws, registration of certain contracts, etc. Like many other Chinese government agencies, they are to a considerable extent localized in terms of orientation, personnel and personnel decisions, and even actions.  - 38 - iocalities permit, support, or welcome the establishment of firms by outsid- ers. Some private entrepreneurs from Fujian Province have been allowed to set up plants in Nanhai County, for example. This is exceptional, however, and in many localities allowing outsiders (particularly from outside the county) to establish firms would he unthinkable. 3.8 The legal framework for establishing and operating enterprises to the extent possible should be neutral with respect to different types of firms. Substantially different regulations for CEs and PEs would not be appropriate, because this would facilitate discriminatory treatment and would adversely affect the objective of achieving a "level playing field" for competition between them. In principle, "company law' should be similar for TVEs, SEs, and urban collective firms, though this may be hard to achieve immediately, and in any case smaller TVEs have a much more rudimentary organizational structure than SEs or larger CEs.5/ Nevertheless, a broadly similar legal framework as between CEs and PEs should be possible. 3.9 Another crucial part of the legal framework is the termination of the existence of financially unviable firms and the adjudication of claims on them. A bankruptcy law might facilitate exit or rehabilitation of loss-making firms, but the specifics and even the concept of bankruptcy in the TVE sector raise complicated issues. For example, what happens if a community's CE sec- tor as a whole cannot meet payment obligations, since bankruptcy for a fixed- membership rural community is not realistic or meaningful? Another danger iL that CGs might have an incentive to squeeze their CEs for funds, forcing the latter to borrow excessively from local banks, and then allow them to go bankrupt, writing off what in effect are CG obligations (see Chapter IV). 3.10 Though new legislation is required in company law and bankruptcy law and some strengthening of contract law and other business laws may be desira- ble, the crucial difficulties will increasingly lie in enforcement and judi- cial treatment rather than in law-making. These are especially essential in legal matters that involve parties from different localities. Within communi- ties, laws may differentially affect TVEs with different forms of ownership. Given the likely continuing internal cohesion in China's rural communities and the powers of CGs, especially at the township level, ensuring evenhanded legal treatment within communities probably has to be a gradual process. Hence the highest priority in the short run should be given to improved handling of cases which cross community and county boundaries. Regulatory Environment 3.11 TVE regulatory issues are closely related to the legal framework. Indeed, the distinction between the two is blurred as a result of the weakness of the latter. The problems already mentioned of uncertainty, lack of even- handedness, bias against outsiders and consequent inhibitions against factor mobility, etc. all arise in much the same way. In addition, certain other 5/ One possible approach would be to have a single "basic" company law which applies to all kinds of enterprises, along with different supplementary laws for SEs, urban collective firms, CEs, and PEs.  - 39 - aspects also are very important, like administrative approval processes for production of specific goods, environmental protection regulations, treatment of mineral resources, etc. "Management" of rural marketplaces by local gov- ernment agencies often leaves much to be desired, and in many rural areas there are problems with excessive collections of road tolls by CGs and other organizations. More generally, predatory behavior on the part of government regulatory agencies and individual officials constitutes a severe hindrance against the smooth and efficient functioning of markets TVEs operate in. 3.12 The regulatory environment faced by TVEs is generally much weaker than that in place for SEs, due to the small size of TVEs and also their non- state ownership, as well as the weakness of economic regulation by the central and provincial governments in rural areas. Attempts have been made from time to time to restrict entry by TVEs and locally-owned small SEs into lucrative lines of business like cigarettes and textiles, or into heavy industries with presumed economies of scale, like iron and steel. But these have generally not been very successful due to lack of enforcement. 3.13 The solution to all these problems is not to build up an extensive government hierarchy to provide administrative support and a better regulatory environment for TVEs. On the contrary, the lack of "mothers-in-law" for TVEs outside the local community may be a major factor in their success. What is needed instead is a more evenhanded and "neutral" regulatory system, which gets away from the biases inherent in the present modes of intervention by the government administrative hierarchy. Greater separation of the regulatory and ownership functions of government agencies would be desirable in this regard. As in the case of the legal framework, a key principle to follow is that the design and implementation of economic regulations should be divorced from the exercise of ownership functions with respect to SEs or CEs. It would make sense for regulatory functions to be carried out at a level of government higher than the level of enterprise ownership (e.g. the provincial level for regulatory policies and enforcement with respect to the TVE sector). Detailed regulation of numerous, mostly small firms is costly and ineffective anyway, and use of "macro" levers can be more efficient. 3.14 In addition to greater evenhandedness, more simplicity and transpa- rency are urgently needed in government regulation of TVEs and their activi- ties. This would help check predatory behavior and corruption by government regulatory agencies and improve the functioning of markets in rural areas. Some localities are experimenting with reforms designed to "clean up" market management, fees levied on traders, collection of road tolls, and approval and registration procedures for TVEs. Consolidation of fees, specified in a precise manner and with only one agency empowered to collect them, appears to be an effective measure. Similarly, "one-stop procedures", with all the agencies brought together to a single place to process registration applica- tions by TVEs and other administrative matters, have been successful in cutting down red tape and the potential for corruption. These kinds of very basic reforms can be implemented rapidly at the national level, once effective and viable "models" have emerged from local experiments.  - 40 - 3.15 Improved regulation and a strengthened Government regulatory appara- tus are urgently needed in the area of )ollution control. Rural industrial pollution has increased sharply with the rapid growth of the TVE sector. Less effective pollution controls in rural areas appear to have encouraged some transfers of polluting activities from urban SEs to rural TVEs. The proximity of many industrial TVEs to rich agricultural land and rural water supplies heightens the impact of their pollution. Policy measures and institutional mechanisms to curb pollution in China's rural areas, whatever their precise nature, should be imposed in a transparent and nondiscriminatory manner on different types of firms. In Wuxi County, local authorities used pollution controls as a means of obstructing the emergence of industrial PEs, in order to maintain and even strengthen the dominant position of TVCEs. Treatment of Private Enterprises 3.16 The fundamental issue in this area is the need to strengthen, clarify, and systematize legal ownership rights with respect to PEs. Ambi- guity concerning the ideological desirability and legality of large PEs has generated numerous distortions and has opened the door for a variety of unhealthy personal relationships between PEs and local CG and county officials. Private entrepreneurs, unsure of their position, may divert resources to consumption or housing or otherwise run their enterprises in a way that doesn't maximize their efficiency and growth. In the absence of appropriate actions, these problems and distortions could get worse as the TVE sector continues to grow and the number of substantial PEs increases. 3.17 The central government's political and ideological blessing to PEs with more than seven employees, given in 1987, was a major step forward. PEs in principle are no longer limited in terms of the size they can attain and the number of employees they can hire. But central government pronouncements alone may not resolve all the ambiguities and problems. The political accep- tance of sizeable PEs should be backed up with concrete laws and regulatory measures to ensure that they receive fair treatment. Issues of how to treat net worth of PEs, transferability, inheritance, roles and rights of workers, etc. still need to sorted out, and clear guidelines issued. 3.18 Discrimination against PEs is still practised in some areas. Wuxi County, for example, directly restricts the formation of industrial PEs and their entry into different industries, through a host of regulations (energy conservation, land use, environmental protection, etc.). Competition for personnel between the PE and CE sectors is generally the most severe source of conflicts in Wuxi and elsewhere. As part of the legalization of PEs and the strengthening and clarification of private property rights, outright discrimi- nation against PEs by local governments should be prohibited. There is a large "grey area", however: county governments and CGs have many instruments and positive rewards as well as negative sanctions which they can use to promote or restrain different parts of the TVE sector. It is probably inevitable that differential treatment will continue in some localities, which makes a clear and enforced policy on basic property rights of PEs all the more important.  - 41 - ! Discriminatory Treatment vis-a-vis the State Sector 3.19 A major concern is whether TVEs suffer from discriminatory treatment as compared with the state sector. This would only be natural because many government industrial regulatory agencies also have hierarchical ("quasi- ownership") ties with SEs. Such discrimination would be damaging to the TVE sector and to China's efforts to instItute a "level playing field" on which all kinds of enterprises can compete on an equal basis. 3.20 By and large, there is little if any discrimination against local CEs by county governments. These usually have a strong interest in promoting local TVE development, and moreover the number of county-run SEs and their output are often small in relation to the local TVE sector. At higher levels of the government regulatory apparatus, policies harmful to TVEs sometimes are instituted. Restrictions against entry by TVEs into certain fields have some- times been imposed, but not very effectively. The whole system of mandatory planning treats the TVE sector differently from the state sector, by giving the latter access to low-cost inputs through the plan. National credit restrictions imposed in 1986 may have had a disproportionate effect on the TVE sector. TVEs may suffer from disadvantages in their commercial relationships with SEs, in particular biases against them in the legal and financial enforcement mechanism. On the other hand, TVEs in the past received conces- sional tax treatment, paying profit tax at much lower effective rates than most SEs. This is gradually being ended, as TVE tax rates are raised to con- form to the 8-level progressive income tax system for collective enterprises, and the "adjustment tax" levied on SEs is reduced. However, more lax account- ing systems for TVEs may permit them to continue to avoid high taxes. 3.21 Obviously, different forms of discrimination against or in favor of the TVE sector should be eliminated, so that competition among different types of industrial enterprises can occur on an equal footing. There is an uneasy balance between different biases now in the system, however; to remove one form of discrimination while maintaining others might not be very helpful. For example, sharply raising tax rates for TVEs to bring them into line with those for SEs might be counterproductive, in the absence of comparable mea- sures to reduce and eventually eliminate the competitive advantages SEs gain from access to low-priced inputs, energy, and bank credit. As a general principle, movement toward a more neutral and balanced environment for compe- tition among different types of industrial enterprises should come about through loosening of restrictions affecting more highly controlled firms (SEs) rather than through greater restrictions imposed on firms now subject to looser government controls (TVEs).  - 42 - IV. FINANCIAL ISSUES Current Situation and Problems 4.1 Financial capital is a critical resource for TVEs. The community- based institutional structure in China's rural areas and the community- orientedness of both TVEs and local financial institutions are in some ways beneficial in this respect. The CG can serve as "guarantor" of bank loans to its subordinate CEs, enhancing their ability to obtain credit, providing informational advantages to banks and borrowers, and reducing risk to the former. Sin,ce bank credit funds for TVEs in the more developed rural areas are based entirely on deposit mobilization within the township, CGs and community banking institutions--local Agricultural Bank of China (ABC) offices and Rural Credit Cooperatives (RCCs)--recognize that bank loans represent "community capital" and hence may be more careful in using it. Though the banking system has gained a measure of independence from CGs in recent years, local l)anks are still subject to considerable influence from local governments in lending decisions.6/ But the relationship between the two is largely symbiotic in many of the more developed rural areas, where CG involvement in lending decisions is in line with CG assumption of most of the risk of nonre- payment. 4.2 The same basic institutional structure and incentive pattern often works very poorly itl the backward areas. CGs do not have adequate financial resources to absorb the risk from credit-financed investment projects. As a result, risk is borne (involuntarily) by the banking system, in the form of delayed repayments.7/ CGs sometimes even use CEs as a means of obtaining "deficit financing" for local public expenditures (and CG payrolls) indirectly from the banking system. CEs are required to turn fixed amounts of "profits" over to CGs, regardless of their actual performance, and to meet these targets CEs borrow from the bank, consume their capital (drawing on depreciation allowances), or both. Under these circumstances, banks become very conserva- tive and fearful of lending to CEs, and as they gain more independence they may reduce such lending. 4.3 Spatial flows of capital across communities and localities in China's rural economy are still limited. Outside the banking and fiscal systems such flows are very small. Redistribution through these systems, considerable in the past, may now be increasingly limited. Immobility of capital can lead to inefficient resource allocation. Combined with considerations of risk (see 6/ ABC is nominally structured as a nationwide vertical hierarchy under ABC headquarters in Beijing. RCCs, though nominally independent cooperatives, were in fact part of the ABC system and subject to supervision by ABC. However, local bank branches and RCCs tend to be very responsive to local government and CG demands. 7/ Outright default is very rare in the case of CEs, since banks do not have procedures for writing off bad debts (let alone for creating reserves against anticipated losses). Thus the common practice is simply to continue carrying unrepaid (and unrepayable) loans on the books.  - 43 - below), it may also limit the maximum size TVEs are able to achieve, prevent- ing them from taking advantage of economies of scale and slowing down the growth of the most efficient, well-managed firms. 4.4 Investment resources for TVEs to a large extent must come from the financial resources of the local community. This need for financial "self- reliance' can provide beneficial incentives for efficient allocation and use of funds within the community, but it may result in wasteful and inefficient patterns of investment across communities. The issue is not so much large- scale movement of financial capital across provinces and regions, which raise complicated issues if other factors like labor are immobile, but rather con- centration of funds for efficient TVE investment within localities. 4.5 TVEs on the whole are rather highly leveraged in comparison to SEs (see Table 1.8). This is due partly to their recent emergence and rapid growth, and also to their inability to draw on government cEpital to any sig- nificant extent. In a situation of extremely rapid growth mnd great opportu- nities, a high degree of leverage may be beneficial. But 's TVE output markets become more saturated and competition intensifies, heavy reliance on bank debt may become more of a burden for the TVE sector. Many TVEs already are or may become financially unviable, unable to repay their large bank debts. More generally, risk management issues are coming to the fore, at the community level as well as at the enterprise level. 4.6 In terms of capital structure, many larger PEs have no equity base other than internal accumulation, since it is typical practice for PEs to "repay' their founders' initial investments as soon as they are on a sound financial footing. Thus how to augment (indeed, create) an equity base for these firms is an important question. For CEs as well, risk-absorbing finan- cial instruments need to be devised. 4.7 The future institutional structure, ownership, and relationship to CGs of the rural banking system is another key issue. In some areas the present system has worked reasonably well, despite low interest rates, lack of competition, and administrative intervention in lending decisions. In others, the banking system is not stimulating local development and carries a portfo- lio laden with bad loans. Among the key issues for national policy are: (1) decentralization of the "ownership' of the banking system; and (2) how to introduce competition into the rural banking system. Enhancing Capital Mobility 4.8 Greater capital mobility, at least within localities like counties, could help improve efficiency of resource allocation in China's TVE sector. More efficient, successful enterprises and communities would attract more capital, allowing them to grow more rapidly and thereby improving aggregate efficiency. Enhanced capital mobility could also serve as a means of imposing tighter financial discipline on CGs, preventing them from using the TVE sector and banking system to "soften" their hard budget coz;straints. Greater capital mobility should start at the "local" level--between townships in a county, counties in a prefecture or province--to avoid large flows from poor to more  - 44 - developed regions, which in the absence of labor and population outflows could adversely affect development prospects in the former. 4.9 The banking system has been and is likely o remain the most impor- tant channel for capital flows in rural aras. Reforms in the rural banking system could increase the mobility of bank credit funds. Higher interest rates on discretionary interbank lending and borrowing across localities might well elicit larger capital flows. Restructuring of the banking system so that branches are not exactly contiguous with community and county boundaries also might enhance capital mobility (see the section below on the banking system). But banking reforms may be hard to implement and make effective in the absence of meaningful competition for the existing state banking system. 4.10 Reform of the existing banking system would be facilitated by pressure from competition. Establlshment of new commercially-oriented finan- cial institutions will also directly improve provision of financial services to TVEs. A considerable amount of institutional diversification is already occurring in China's rural financial sector, but most new entities appear to be community-oriented and community-based. To strengthen competition and enhance capital mobility, many of the new financial intermediaries created in the future should cut across community and locality boundaries and should not be owned or controlled by any particular CG. Cross-community financial insti- tutions could be structured as joint ventures, for example including several CGs and possibly national- or provincial-level financial entities. 4.11 Another way to enhance local capital mobility in the TVE sector is to establish a regulatory and administrative environment conducive to cross- community joint ventures. The community orientation of the TVE sector and the inadequate legal/regulatory framework create substantial barriers against such joint ventures. For instance, each partner would want the joint venture to be located within its own jurisdiction. libor hiring might also be a difficu- issue. One way of encouraging cross-community joint ventures is for locai governments to establish "industrial parks" at which joint ventures could set up operations (see Chapter V). These areas would not be under the jurisdic- tion of any particular rural community: hence they would provide a "neutral" location for joint ventures involving several communities. 4.12 Still another approach would be to bAild on the existing practice of issuing short-term (1-2 year maturity) ands with profit-related variable interest rates by some CEs and township industrial corporations (TICs) in the more advanced areas like Wuxi. These bond issues (called jizi) have uncertain legitimacy and an uneasy relationship with the banking system, which fears losing some of its deposit base. Moreover, up to now they have been issued to workers and to local community residents--few if any bond issues have crossed township boundaries. Finally, the linkage of returns to profits often contains a large discretionary element. 4.13 TICs or county industrial authorities in the more developed localities could issue variable-interest bonds (hereafter referred to as "income bonds"), which would be purchased primarily by entities and indivi- duals outside the community. Such bonds would carry a fixed minimum nominal  - 45 - interest rate, but they would also entitle their purchaser to a share of TE or TVCE net after-tax profits in a township (if issued by a township) or in several townships (if issued by a county). Thus they would spread some of the variability in profits to investors, while at the same time assuring them of a certain minimum nominal return. Initially, bonds would be of fairly short maturity (say, three years), but maturities could be lengthened as people become accustomed to the new instruments. TICs or county entities rather than individual firms would be the appropriate bond issuers. TVEs are far too small; moreover, CEs in some ways are more like subsidiary parts of a conglomerate than like independent firms. With perhaps only few exceptions, lower-level communities also would not be appropriate bond issuers. Only TICs have a sufficient industrial base and adequate standing to be credible to outside investors. Purchasers of income bonds could include other CGs, individuals, firms, and financial institutions. They would be marketed on a provincial or local basis (but outside the community) and sold through banks, which would also serve as guarantor and collect a service fee. Banks would market income bonds of a number of TICs jointly, with a joint prospectus allowing would-be investors to compare past economic performance and terms of issue. There would be provisions for advance redemption, but at a penalty which would discourage doing this and which would give banks that redeem bonds an attractive return for holding them to maturity. 4.14 Pricing of these bonds would be based on the fixed minimum interest rate and a set share of profits. The fixed interest rate could be mandated by the government and set at the same level for all issuers. It should be signi- ficantly below the interest rate for time deposits of similar maturity at banks, so there is indeed some risk to investors, though the maximum loss (in terms of alternative opportunities) is limited. A relatively low minimum fixed interest rate would mean that only profitable issuers could sell their bonds (since investors would not be attracted if there would likely be no profits to share in as well). Thus poorly performing communities would be automatically excluded from the bond "market". Given interest rates of over 8Z p.a. on three-year time deposits at banks, a fixed nominal interest rate of 4-5 p.a. might be the most appropriate level to meet the twin needs of assuring investors some minimum return and providing sufficient variability and a large profit-related component. The share of profits carried by each bond would then be set to provide an expected total rate of return attractive to investors. This would have to be considerably above the "safe" interest yield on bank saving deposits. If the bank interest rate is 8Z and the fixed interest rate on the income bonds is 4Z, an expected total yield of 122 or even as much as 15Z might be necessary to attract investors.8/ 8/ As an example, assume that the fixed interest rate is 4Z p.a. and the desired total rate of return is 12Z. The issuing entity expects that its net after-tax profits will be (for simplicity) Y1 million in each of the three years the bond is outstanding. A holder of a YlOO bond should receive Y8 in profit share each year in order to get a 12Z p.a. total return, so the bond's profit share should be 8/1,000,000 or 0.000008Z. The calculations become more complicated if profits are expected to change over time, but the basic principle is the same.  - 46 - 4.15 Presuming the bonds are issued in series, an issuer that does unexpectedly well, with sound financial performance, could lower the profit shares for subsequent bond issues. On the other hand, entities who could not sell their bonds in the first round or whose performance is worse than expected would have to raise profit shares subsequently in order to stay in the market. Profit shares perhaps could be lowered gradually over time as incone bonds gain acceptability and the required "premium" over the deposit interest rate declines. The total value of bonds that could be issued by a TIC would be subject to a ceiling, perhaps specified in relation to its profits or assets. This would provide a measure of protection to investors, but more important is the "natural" protection offered by the fact that unprofitable entities would be unable to market bonds. 4.16 Income bonds would lend themselves toward gradual transformation into more "equity-like" instruments. This could be accomplished by reductions in the fixed interest payment (eventually to a zero nominal rate), which might require increases in profit shares, or at least smaller reductions than otherwise.9/ Once suitable trading arrangements are formulated, resale of bonds on secondary markets could occur. Another ingredient in this transfor- mation would be lengthening maturities, perhaps first to five years and later to 10 years or longer. In the end, income bonds would become something akin to equity shares, but without voting rights or claims to the assets of the issuer. 4.17 Income bonds could be expected to have some important benefits in the more developed TVE areas. First, they would strengthen the capital structure of TICs issuing them and promote the evolution of TICs into true holding companies, with less direct linkage to CGs. Second, they would stimulate greater capital mobility on the basis if profitability, generating competition among firms and CGs for capital and providing a counterweight against intru- sion of efficiency-reducing nonbusiness objectives. Third, they would serve as an indicator of performance for banks in their lending decisions. A TIC unable to generate profits and hence unable to sell bonds, for example, could expect less sympathetic treatment from banks in decisions on loan applications of subordinate CEs. Fourth, they offer the rural banking system an opportu- nity to assume a more diversified role, and to focus on profits rather than only total lending (the primary indicator of bank performance in the past). Fifth, they would require and hence promote more accurate and open financial accounting and reporting in the TVE sector. Sixth, they would serve as an evolutionary bridge to development of equity instruments and capital markets, without the instability that characterizes the small capital markets of many developing countries, or that would likely be the result of an overly rapid move to capital markets by the Chinese. Since there is no international experience and little past Chinese experience to provide guidance on how well the income bond concept would actually work out in practice, experimentation should start on a small scale, in a few of the more advanced areas. 9/ A lower minimum interest rate would require a higher profit share just to reach the previous expected total return, and moreover the greater variability in returns might well cause investors to demand a higher expected total return as compensation for the greater risk.  - 47 Enterprise and Community Risk Management 4.18 Risk management undoubtedly has played a role in the success of the traditional CE sector in the more developed areas. TICs pool and absorb the business risk of subordinate TEs, which gives firms facing a downturn a tem- porary respite to recover, change product lines, etc. The main instrument for TIC risk management has been ad-hoc variation in the financial flows between TEs and TICs. This also serves as a means of pooling investment funds within townships for major projects and new enterprise creation. CGs manage their business risk through their decisions on the number of firms to operate and the degree of diversification of product lines. Even so. business downturns have an inordinately large impact on CG finances and operations. Inadequate risk management exacerbates the problems of backward areas. CGs there do not have the discretionary funds to absorb enterprise risk, so much of it is borne involuntarily by the banking system, with harmful consequences. 4.19 There are three main issues concerning risk management: (1) how to pool and absorb risk in the backward areas; (2) how to manage risk at the community level, in developed and backward areas; and (3) how to deal with failing CEs. In the more backward areas, wholesale assumption of CEs' risks by CGs is not appropriate, unlike in the developed areas where CEs generate substantial amounts of discretionary funds for CGs. De facto risk absorption by the banking system is not desirable either, as noted above. The financial strength of TICs is insufficient for them to issue risk-spreading financial instruments. This is an issue that needs further thinking--attractive options are not immediately apparent. 4.20 At the comm.aity level, CGs need to diversify their risks, beyond what is feasible through enterprise creation and diversification of product lines. Beyond a certain point, both carry considerable economic costs; more- over, the degree to which CGs in different localities and regions can diver- sify risks in this way varies greatly. The need for new means of risk spreading applies both to backward areas, where economic returns of firms may be low, and to developed areas, where enterprises are often relatively large and products highly specialized. One way to allow CGs to further diversify risk is through purchase of income bonds issued by other TICs (mentioned above) or other high-return financial instruments. CGs would then have the choice between investing in these financial instruments and undertaking physi- cal investments in enterprise development, and could reduce their risk by carrying a diversified "portfolio' of physical and financial assets. 4.21 The question of how to deal with failing firms primarily concerns CEs and large PEs. Small PEs enter and exit the scene frequently and apparently quite easily. In more developed areas, the CG commonly takes responsibility for repayment of debts of defunct firms under its ownership, usually by assigning them to other local CEs. This inhibits CGs from closing down money- losing firms, as long as their cash flow can service at least some of their bank debt. But it also gives CGs strong incentives to supervise their enter- prises in such a way that they are able to service their debts. In backward areas, CGs generally do not view themselves as responsible for bad debts of subordinate CEs. Banks in these localities are often reluctant to call in loans of money-losing CEs, in the absence of any reserves against bad loans or accounting procedures for dealing with them.  - 48 - 4.22 Carefully designed bankruptcy regulations would be useful in adjudi- cating claims on the assets of defunct PEs and clarifying priority among claimants, and they could play a similar role in facilitating exit by nonvia- ble CEs. But careful thought needs to be given to their precise terms, and in particular the claims of banks should not be weakened vis-a-vis implicit or explicit CG claims. Though keeping failing firms in business is often waste- ful, in the more successful TVE regions this usually does not go on for a very long period of time, as the enterprises are forced to make drastic adjustments like developing a completely new line of products if necessary. On the other hand, the danger of CG misuse of overly generous CE bankruptcy regulations is great in many of the backward areas. Capital Structure of Large PEs 4.23 As has already been mentioned, large PEs have a problematic capital structure. Many of them depend heavily on bank credit, and in addition the more successful among them have built up considerable assets generated by internal accumulation. But founders are usually "repaid" their seed capital from enterprise profits as soon as the firm is financially viable, and the property rights of the owner(s) with respect to internally generated funds, like the ownership of the firm itself, are still vague and incomplete. Both risk spreading and clarification of ownership/property rights would be facili- tated by broadening the equity base of large and successful PEs. At the same time, clarifying and strengthening the property rights of private owners of PEs may be problematic if such firms depend largely on bank loans, since these represent "the community's capital" in a certain sense. This is particularly true because interest rates are low. 4.24 Ways need to be found to give large PEs sufficient access to capital and other local community resources, in a manner which does not affect CG incentives. One possible solution is for banks to provide some capital to such firms in the form of nonvoting equity (or "ownership") shares rather than bank loans. This would broaden the equity base of large PEs and allow clari- fication of their ownership. Banks could capture a significant part of the profits of these firms in a transparent, legitimate manner; the access of large PEs to bank funds would probably be enhanced; and the ownership of these firms would be partly 'socialized', in a manner that should not be threatening to private entrepreneurs and would not interfere with their autonomy in managing their firms. This option obviously would require a major change in orientation and practices on the part of the rural banking system, and it would also require careful monitoring and regulation of "investment banking" practices. The banking system would be making (probably nontradable) equity investments and hence would increase its risk, but since the firms concerned are by definition large, successful, and profitable, this should be manage- able. Banks could continue providing funds to these PEs in the form of loans as well. 4.25 Direct investments in larger PEs by CGs and other local community entities could be handled in the same way. For example, TICs, as part of their evolution into investment companies, could provide some equity financing  - 49 - to larger PEs in the community. Given present Chinese conditions, to ensure that the private owner/entrepreneurs maintain control and are not threatened by the changed ownership structure, such equity participation should be on a nonvoting basis. 4.26 There are numerous other options for structuring flows of community capital resources to PEs, expanding their equity base, and strengthening and clarifying ownership rights. The key requirement is to ensure that a consid- erable part of the return to capital provided by the community does accrue to the community (which may be best accomplished through equity participation), while incentives of private entrepreneurs are not dampened and their autonomy is not threatened (which suggests nonvoting arrangements). The Future Role and Structure of the Banking System 4.27 An issue of great importance is how to structure the rural banking system. The:..e are many pitfalls to be avoided, and few if any transparently obvious and feasible options. Putting the local banking system under CGs (e.g. townships) would probably be harmful, making it easier for the latter to directly or indirectly (through CEs) obtain "deficit financing" from banks and thereb- soften their budget constraints. This danger is especially serious in the niore backward areas where these problems are already present. 4.28 One possibility would be to decentralize the ownership of the banking system, but only to one or two levels of government above the level to which clients belong. For example, county-level ABC branches could be the main lenders to TVCEs, while township-level financial institutions like the RCCs could lend to PEs. To some extent this resembles the current division of labor between ABC and RCCs.10/ More generally, any restructuring of the ownership of the banking system in rural areas should if possible avoid creating separate, independent banks that are exactly contiguous with CG or county jurisdictions. Simply making the township ABC offices, perhaps combined with the RCCs, into independent township banks would not be a good idea, because they easily could be coopted by township governments. 4.29 At present, there is virtually no competition between ABC township offices and RCCs.ll/ Some competition has been introduced by the attempts of the urban Industrial and Commercial Bank of China to penetrate more industrialized rural areas, which seems to have been beneficial in improving deposit services. On the lending side, competition gives borrowers more options, and by the same token it may weaken CG influence over allocation of 10/ However, ABC township offices in many respects are highly responsive to township authorities rather than to the county government or ABC branch. 11/ Though RCCs are nominally independent cooperative institutions, in practice they are under ABC. Even in counties where they have been nominally separated from ABC, there is a clear division of labor (ABC lends to TEs, RCCs to lower-level CEs and PEs), which severely limits competition in lending.  - 50 - the community's capital resources. Competition should result in improved provision of banking services to TVEs, but it will be difficult to generate a great deal of local competition among existing rural financial institutions. 4.30 Institutional diversification should be an important part of reforms in the rural banking system, and indeed it is already occurring to some extent, through a proliferation of community "trust funds", creation of some "investment companies" and similar entities, and emergence of private finan- cial institutions in some areas like Wenzhou. This trend should be encou- raged. though any sizable financial institutions (especially those which take deposits) should be subject to appropriate regulatiGn. Up to now, most insti- tutional diversification has occurred at the community level, which suggests that there may be systemic obstacles against financial institutions whose business areas cut across community and locality lines, and which are not closely tied to any particular township or county. Nevertheless, future institutional diversification in the rural financial system should proceed in this direction to the extent possible.  - 51 - V. INDUSTRIAL STRUCTURE, LOCATION PATTERNS, AND SUPPORT SERVICES 5.1 China's industrial TVE sector is characterized by for the most part relatively small, community-oriented firms; a surprisingly broad range of industrial subsectors and products; a mix of simple, labor-intensive produc- tion techniques and more modern technologies; a dispersed spatial patteri of industrial location; and widespread lack of industrial support services (an inadequate independent commercial system, limited financial services, lack of technical services, etc.). Some of these features could become major sources of inefficiency and severe constraints on future TVE growth and development. Enterprise Size and Economies of Scale 5.2 Despite rapid development of the TVE sector and extremely fast growth of GVIO in recent years, the average size of firms has not increased very much. As can be seen from Table 5.1, the average number of employees of industrial TVCEs increased at an average annual rate of only 5Z between 1980 and 1986. GVIO and assets per firm increased at a faster pace, but these figures are specified in nominal terms and there has been significant infla- tion in the past several years, especially in construction costs and building materials prices. TVCEs on average tend to be no smaller than small-scale manufacturing firms in other developing countries, and indeed many of them fall into the category of medium-sized firms.l/ However, the problem is not so much absolute firm sizes but rather the obstacles that may be preventing more efficient and successful TVEs from growing and eventually entering the ranks or large enterprises. 5.3 In many of the industries where TVEs account for a significant share of output, economies of scale are not important enough to be a determining factor. But in other industries they may be substantial, and TVEs are probably below the minimum optimal scale. One example is cement, another steel-rolling. In industries like these, small firm size may be a significant barrier against efficiency gains. But more important than any static ineffi- ciencies are the dynamic costs of limitations on firm size in the TVE sector. The more efficient, successful, and rapidly growing enterprises at some point are subjected to a slowdown in growth (or may be split up into smaller firms as they develop new products), which could adversely affect the dynamism and efficiency of the TVE sector overall, as well as its competitiveness vis-a-vis the state sector. 1/ Generally firms are considered "very small" if they have fewer than ten employees, "small" if they have 10-50 employees, and "medium-sized" if they have 50-100 employees. See Little, Ian M.D. etal, Small Manufacturing Enterprises: A Comparative Analysis of India and Other Countries (New Yorks Oxford University Press, 1987), p. 8. By these standards, many TVEs would be considered medium-sized or even "large".  - 52 - Table 5.1: AVERAGE SIZE OF TVCEs, 1980-87 (number of persons and Y '000) All TVCEs Industrial TVCes Employ- Gross Fixed Total Employ- Gross industrial ment revenues /a assets /b assets /c ment ou.put valur! /a 1980 21 42 23 31 26 71 1983 24 69 35 47 29 102 1984 23 77 35 51 28 116 1985 26 116 48 75 33 170 1986 29 147 62 100 35 214 1987 30 186 78 133 n.a. n.a. /a In current prices. /b Undepreciated value of fixed asses at original purchase prices. /c Net value of fixed assets (after accumulated depreciation is subtracted) plus "quota" circulating assets (inventories of inputs and outputs and goods in process). Sources: Statistical Yearbook of China, 1981, 1985, 1987, 1988; China Rural Statistical Yearbook, 1985, 1986, 1987. 5.4 To some extent slow growth of firm size may reflect spread of TVE development to more backward areas where firms are much smaller on average. But average firm size even in Wuxi County has not increased very much. The average industrial TVCE in Wuxi had 57 employees in 1980, 64 in 1985, and 76 in 1986. Another possible reason for continuing small firm size may be an effort to diversify industrial structure at the community level, accomplished by establishing new firms in different lines of business. A large number of firms and a "conglomerate" industrial structure may help reduce community financial risk. Frequent changes in products to take advantage of shifting market conditions may also have been a factor preventing rapid increases in firm size. Another reason may be that the period of unfettered TVE develop- ment has been relatively short. Communittes may also desire to keep their firms small to facilitate management. PEs have been discouraged from reaching large size by the vagueness and weakness of private property rights. But perhaps the most important reason for the absence of large enterprises in the TVE sector is their community orientation, combined with risk considerations and lack of mobility of key factors of production across communities and loca- lities. This limits the pool of resources available to any TVE.  - 53 - 5.5 If the main barrier against larger size is indeed the community orientation of firms, limited community resource bases, and immobility of key factors of production across communities, the short- and medium-term remedy is to find ways in which successful and rapidly growing TVEs can draw in resources, in the first instance financial resources, from outside the commu- nity. Income bond issues, cross-community joint ventures, and greater mobi- lity of capital in the financial system, options suggected in Chapter IV, might facilitate growth of relatively large TVEs. Creation of 'industrial parks" where TVEs could locate their activities and which are not under the jurisdiction of any particular rural community (see below) could encourage emergence of TVEs which draw on resources from more than one community. Over the longer term, transcendence of their rural community origins by some TVEs may occur (see Chapter VIII). 5.6 Administrative restrictions against entry into certain lines of activity by TVEs or against 'inefficient" small firms are not the answer to possible problems with suboptimal firm size. TVEs function in a market envi- ronment, so those that are truly inefficient will be weeded out as the func- tioning of China's industrial product markets improves. The pressure of rural surplus labor will stimulate a great deal of TVE development, but the sector's ability to absorb this labor efficiently will be weakened if there are restrictions against entry into particular industrial activities. Finally and perhaps most important, administrative restrictions invariably turn out to be costly and inefficient, and more often than not they fail to achieve their objectives. Thus fundamental solutions lie in removing or weakening the barriers against larger firm size in the TVE sector, not administrative measures to exclude it from particular industries. Community Industrial Structure and Subsectoral Composition 5.7 Some of the methods to enhance capital mobility and allow greater diversification of community risk through use of financial instruments should also permit greater specialization in community industrial structure (and . rger firm sizes) where this is economically appropriate. In this context, ti.e question of whether particular communities and localities or indeed the TVE sector as a whole have overly diversified industrial structures would become moot--not an issue for specific policy measures but rather one that will be settled gradually over time as communities adjust their strategies in response to the reduced need for industrial diversity from a community risk management perspective. 5.8 An even broader issue is the degree of concentration of TVE activity in industry. The most advanced TVE industrial centers like Wuxi are probably overindustrialized, as well as being under-urbanized (Wuxi has over 4,000 substantial industrial TVEs). One of the reasons for increasingly fierce competition in the TVE sector is that each locality and community is trying to engage in manufacturing operations, regardless of whether this is economically appropriate given the local resource base, market situation, etc. In part this is the result of population immobility--to absorb surplus labor, locali- ties have to do something, given that outmigration is so limited. Price distortions also may play a role, making industry overly attractive compared  -54 - to nonagricultural services or specialized agriculture. Many services are closely related to the level of local industrial development, however, and there would be little or no demand for them in the absence of the latter. Hence a large part of the problem may be underdevelopment of services in the highly industrialized rural areas, not in nonindustrialized localities. Rather than continuing the proliferation of industrial plants which dot the countryside, places like Wuxi should be expanding industry-related services. Technology 5.9 Technology in the broad sense is considered a crucial bottleneck in TVE industrialization efforts, and the shortage of technical personnel is perceived to be one of the most severe constraints on TVE development. The flow of technological information is still somewhat restricted, and there is no real "market" for the host of technology-related services needed in a modern industrial economy. Technology problems vary with the level of TVE development in different regions and localities. In some of the more backward areas, there is no nucleus of technicians and only a weak base of entrepre- neurs. Combined with lack of market contacts and ignorance of technological and market conditions, this often results in costly failures and use of inappropriate technologies. The problem is not so much one of obtaining tech- nology per se, but rather inability to appropriately choose and adapt techno- logy suited to local conditions, and even to manage production and keep equip- ment running. In the more advanced areas, the existing technological base is muci stronger. There is a core of technologically proficient people, and usually established contacts with urbani SEs and research institutes or other providers of technological information and services. In some areas like southern Guangdong, there is access to foreign equipment and technology. Adaptation of new technology and management of production hence tends to be less difficult. What hinders the more advanced TVE regions is systemic restrictiens, for example against acquisition of technology from abroad, as well as financial resource and foreign exchange constraints. 5.10 This suggests that solutions to technology problems may be somewhat different as between backward and more developed areas. In the former, human resource development and augmentation of humar capital is the mcst urgent requirement (see Chapters VI and VII). Easing resource constraints or systemic restrictions (e.g. against foreign business contacts) may not be very helpful in the absence of the human resources needed to choose, adapt, and manage technology. In the developed areas, such policies are likely to be more effective. In places like southern Jiangsu Province, just giving TVEs greater access to the foreign exchange they are already generating through exports may be highly beneficial. One of the main reasons southern Guangdong has been able to largely "catch up" with southern Jiangsu in terms of TVE industrial and technological development, despite a weaker prereform base, is its ability to rely on foreign technology, along with needed management and technical personnel from Hong Kong.  - 55 - Industrial Location and Spatial Patterns 5.11 Related to the community orientation and small average size of TVEs is the highly dispersed location pattern of TVE plants. Though rural indus- trial development has been highly uneven as between different parts of the country, within many of the most industrialized rural localities factories dot the countryside. Virtually all CEs are located in their home community, as are most PEs. The costs of this dispersed location pattern are substantial, though further research would be necessary to quantify them with precision: (1) infrastructure either is very costly or simply is not invested in, and there is under-provision of 'urban-type" public services for TVEs; (2) the advantages of 'economies of agglomeration" may be largely lost; (3) transport costs are higher; (4) it is much harder for a dense network of industrial support services for TVEs to emerge, since potential customers are scattered; and (5) wasteful patterns of land utilization becanme pervasive, in particular cccupation and pollution of high-yield agricultural land by industrial plants. Though pollution problems in the industrial TVE sector are significant in their own right, they are exacerbated by the dispersed location pattern. 5.12 Since the basic cause of dispersed location patterns is the community orientation of the TVE sector and the near-immobility of firms and factors of production across communities, any solutions to the problem must be aimed at the firm-community nexus. Though over the long term the community orientation particularly of larger TVEs may weaken (see Chapter VIII), this will be a gradual process which may not even affect the bulk of smaller firms. Thus ways must be found in the short run to achieve more concentrated location patterns for TVEs without a severance of firm-community ties, through economic incen- tives rather than administrative decrees. 5.13 One option is for local government authorities at county or perhaps provincial level to establish "industrial parks" at which CEs, PEs, and cross- community joint ventures could set up operations. These areas would not be under the jurisdiction of any particular rural community; hence they would provide a "neutral" location for all types of TVEs.2/ Infrastructure invest- ments would be undertaken by the sponsoring government of the industrial park, and public services would be provided to enterprises which locate there, priced on a full cost-recovery basis. Sites and buildings could be rented by enterprises, or firms could be permitted to acquire land and construct their own plants. If necessary, enterprises locating in the new industrial parks could be given certain tax advantages at the beginning (reflecting the external economies and social benefits of more centralized location), but in general, an advantageous site and availability of services (including, perhaps most importantly, electric power) should be sufficient to attract firms. Another important inducement might be credit availability. 5.14 The industrial park concept would have to be carefully designed to work well; most such schemes in other countries have not been very successful. 2/ Obviously, for this to work the agency administering the industrial park would have to be perceived as neutral, without strong interests of its own that would affect its management of the territory.  - 56 - International experience suggests that industrial parks wi11 not be successful if they are located far from cities and try to attract small-scale entrepre- neurs away from the urban environment they tend to prefer. In this context, it should be recognized that the goals of industrial parks for the TVE sector would be quite different from those of industrial parks in most other develop- ing countries. In China, a primary objective would be to promote more concen- trated location patterns for the TVE sector, not to move small-scale industry out of large cities. Hence if systemic obstacles can be overcome, industrial parks may have a greater chance of success in China's TVE sector than has typically been the case in other countries. 5.15 A number of issues specifically related to China would need to be resolved. For example, tax policy toward enterprises that locate in the industrial parks would have to be carefully balanced between covering the costs of provision of government services to these firms (which calls for certain taxes to be levied and collected on the spot) and maintaining CEs' fiscal ties with their home communities (which would call for tax collection at the home community). Another issue relates to employment--since providing jobs for local community residents is one of the main motivations for estab- lishing CEs in labor-surplus areas, they would be unwilling to relocate if this means severing employment linkagee. This is one reason why limiting the scope (at least initially) to counties makes sense. Flows of labor and popu- lation into the industrial parks should be facilitated for those who have gainful employment there (as well as for private entrepreneurs who want to set up their own firms). Ways also need to be found to discourage setting up new firms in a dispersed manner, based on the real economic costs involved in this kind of location pattern. Appropriate land-use and land-conversion fees and public utility charges, as well as pollution fees and pollution control stan- dards, might help encourage communities to locate their firms in industrial parks. Still another issue relates to the primary reliance on gross indus- trial and agricultural output value as a performance indicator for communities and CGs. This would have to be modified for CGs to be willing to let "their" CEs locate in industrial parks (perhaps by allowing them to count the output value of these firms in their communities' GVIO). 5.16 Certain Chinese counties are moving toward something akin to the industrial park concept. Jieshou County encourages private entrepreneurs to relocate their activities (and residences) to the county town, through a variety of inducements including the possibility of changing household regis- tration from rural to 'long-term urban". Other counties are putting large amounts of resources into building county seats of their own, and encouraging TVEs to locate there. But these latter efforts are to some extent vitiated by their primary objective of establishing an administrative center for the county. Finally, several counties in Fuyang Prefecture (where Jieshou County is located) have created "industrial development zones", which in some respects come close to the industrial park concept elaborated above. Industrial Support Services 5.17 The network of services supporting TVE industrialization is for the most part inadequate. The underdeveloped commercial system serving TVEs  - 57 _ forces them to do their own marketing, spending large sums to sell their products as well as to procure inputs. Financial services also appear not to have kept up with the rapid growth of the TVE sector. Transport services for TVEs may also be insufficient, though the proliferation of individually owned rural trucking companies must have eased bottlenecks somewhat. The inadequacy of technical assistance, market information networks, consulting, and "extension-like' services to TVEs may be the most severe zonstraint emanating from the service sector. The underdevelopment of services in the TVE sector mirrors the same problem in Chinese industry as a whole and in the entire economic system. But the problems for TVEs are exacerbated by (1) TVEs' relative lack of access to services provided by state institutions and (2) the greater need of small firms like TVEs for specialized services, since it is harder for them to engage in self-provision. 5.18 Some of the main causes of the weakness of the TVE service sector have already been noted. These include the dispersion of customers (indus- trial TVEs), possible price distortions, remnants of the prereform ideological disdain for "nonmaterial" services in general, a narrow focus on the part of CG leaders and county authorities on development of manufacturing industry, etc. Another major factor most probably has been TVEs' relative lack of access to urban/state service networks. The community-oriented institutional structure in China's rural areas may well inhibit the flow of services across communities much more than the flow of goods, generating a bias in favor of goods production and against services. As reforms and the rapid growth of the TVE sector continue, the need for industrial support services may become even more acute, and continuing inadequacy in the service network could become an increasing liability for TVEs. 5.19 Part of the solution is to improve the ability of TVEs to tap existing urban state-oriented service institutions. This is already beginning to happen, as the latter become more profit-oriented and freed from restric- tions against taking fees for services provided. But this trend needs to be encouraged, and competition needs to be promoted, so that monopolistic state- owned service institutions will not overcharge TVEs. "Bridge' service organizations also need to be formed, which TVEs can turn to, rather than being forced into costly direct searches in (often distant) urban areas. To some extent this is already occurring in Wuxi County, which has made major efforts to improve human capital, technology, and related services for TVCEs. More important than access to the state service network is development of industrial support services within the TVE sector. This is not only essential for the healthy development of the sector in the future, but it also has important positive benefits. The employment generation potential of services is very great, and since many of the services concerned are most efficiently provided by small firms, they represent a great opportunity for the TVE sector. 5.20 A number of measures .o encourage rapid development of TVE services can be envisaged. Freedom for CE professionals and skilled workers to leave  - 58 - their firms and go into business for themselves is crucial.3/ Greater concen- tration in location patterns will increase the effective demand for various support services to TVEs, which in turn should elicit a supply response. There may also be scope for industrial support services to be mobilized through the voluntary TVE industrial associations that are now forming in many rural areas. Since the relative importance of the PE sector is greater in services than in manufacturing, improvements in the legal and regulatory environment for PEs could also help stimulate more rapid growth of services. CG cadres could be evaluated as much for their progress developing services as for achievements in local industrial and agricultural development, which would h3lp get rid of any remaining ideological bias against service activities. The key role of government is to create an administrative and regulatory environment conducive to the development of TVE support services on a profit- oriented basis, to a large extent provided by specialized TVEs themselves. Except in the case of services which carry substantial externalities, like education and training, governments should not get involved in direct provi- sion or financial subsidization of TVE services. Finally, there is a need for further research on the obstacles which hinder development of TVE service activities. 3/ This would likely encourage a proliferation of new "spin-off" firms and service activities. It could also have some negative effects, like disruptions in activities, reduced identification with their firms by these personnel, disincentives for TVEs to provide training for their employees, and sharply rising wage levels as a result of competition among enterprises for highly-qualified personnel. Nevertheless, the benefits most probably outweigh these costs.  - 59 - VI. LABOR AND HUMAN RESOURCES Current Wage Practices and Labor Relations 6.1 There is abundant evidence that TVE workers are not treated purely as a variable input in a "capitalist' production process. The bulk of TVEs (both CEs and PEs) use wage systems based on individual and enterprise performance. Job security is not as airtight as in SEs, but it is substantial, and turnover is low. TVE workers appear to have implicit claims on the returns generated by their firms. There is virtually no formal worker ownership or participa- tion in enterprise decisionmaking, but most CE workers are also members of the rural community that "owns" their employer. 6.2 One of the main sources of TVEs' competitive advantage vis-a-vis SEs is their lower labor costs. Differences in nominal wage rates are often not very great. But the mandated benefits package for SE workers (subsidized housing, pension, medical care, etc.) is very expensive, and there are few benefits which TVEs are legally required to pay their workers. Nor are there anything like government-set minimum wage laws for TVEs. Overstaffing in SEs also contributes to their higher labor costs. Wages of TVE employees are much more closely tied to individual and enterprise performance (growth of output and/or profits) than are those of most SE workers. Piece rates are extremely common, and many CEs pay some profit- or output-related year-end bonus. There are powerful incentives to put more effort into work and to work long hours. Labor relations in TVEs seem relatively good, particularly considering the hard work, long hours, difficult working conditions, and often relatively low wages. Superior labor motivation and effort are probably even more important than lower wage costs in explaining the performance of the TVE sector. 6.3 The community orientation of TVEs makes labor relations very different from the typical pattern in capitalist firms or in Chinese SEs, and in particular workers may be much more willing to accept piece rate wages.4/ Managers and owners (i.e. the CGs) are probably less willing or able to use piece rates to "exploit" workers (e.g. by increasing work norms) than in capitalist firms. The ubiquity of piece rates also is facilitated by the simple, easily quantified nature of many tasks in TVEs. The rapid growth of the TVE sector and increasing capital intensity may have allowed many enter- prises to be generous in setting piece rates and bonuses. Most TVE workers have spent only a few years outside of agriculture, and the overwhelming majority are first-generation. The drastic improvement in their living stan- dards accompanying the shift from agricultural to nonagricultural activities must make these workers more receptive to performance-based wages and willing to put in long hours (to which they are accustomed from farming). The possi- bility of being forced to return to agriculture may inhibit excessive concern about wages, wage systems, or working conditions. 4/ Piece rates are now rare in industrialized market economies, due to conflicts over work norms and competition among workers which threatens worker solidarity. Experiments with piece rates in Chinese SEs in the early 1980s were for the most part unsuccessful, for similar reasons.  - 60 - 6.4 Some of these factors may change over time, leading to more difficult labor relations in the TVE sector and consequent deterioration in worker moti- vation and efficiency. To avoid this, it is necessary to carefully consider different options for wage systems, ownership, worker participation (in finan- cial and decisionmaking spheres), firm-CG relationships, etc., which will help maintain a relatively efficient incentive environment for TVE workers. The issues are complex, and the danger of costly wrong choices is great, as can be seen from the experience of nonstate, noncapitalist firms in most parts of the world. Workers' Role and Protection 6.5 As the community orientation of CEs as well as PEs weakens, as in at least some areas they hire more workers from outside the community and/or on a short-term basis (which is already occurring in places like southern Guangdong), and as TVE workers become more conscious of their own interests, pressures may arise for greater and more formalized worker participation. Moreover, the lack of effective legislation protecting workers from abuses, or of unions to represent workers, which may not have been too harmful in a system of community-oriented enterprises where the bulk of employees were community members, may become less and less acceptable in the future. 6.6 Over the long run, the main alternative to unionization on a national, provincial, or industrial basis may well be some form of worker participation, at the enterprise or possibly community level.5/ This could help preserve the advantages of the present, community-oriented system even as firms become more autonomous, but well-known pitfalls cooperative and labor- managed firms have run into in other countries should be avoided. A system without any significant worker participation but with trade unions, on the other hand, could encourage a trend toward "Taylorist" management practices and confrontational labor relations, which might have considerable costs in terms of worker motivation, effort, and efficiency. 5/ It is only to be expected that organizations will emerge to promote workers' interests and represent them in bargaining with employers. The real question concerns hcw these entities will be structured and whether they will be enterprise-based (involving some form of worker participation) or more broadly organized.  - 61 - 6.7 Regardless of whether China attempts to introduce some form of worker participation or goes the trade union route, some basic legislation delineating acceptable labor practices is needed.6/ This should not be overly restrictive, but it would be helpful, both in protecting workers and in protecting employers from arbitrary charges of "exploitation". Legislation is especially necessary if there is a relatively free labor market, with neither worker participation nor effective unions. These regulations would have to be widely perceived as appropriate and fair, and they would need to be backed up by meaningful sanctions and an enforcement mechanism. Wage/Incentive Systems and Pay Differentiation 6.8 Choices on ownership/participation have obvious implications for TVE wage systems. Even in the absence of worker participation, performance- related wages, s0 successful in the TVE sector in recent years, obviously should be continued. The precise form of such wage systems depends greatly on the characteristics of the industry, product, and production process con- cerned. Over time, as complexity increases in many firms, there may need to be a shift from piece rate payments based largely on individual physical output to year-end, profit-related bonuses, variable pay based on subjective evaluations, etc. Rewards could become linked to future as well as current profits, to further strengthen worker incentives. 6.9 A number of issues concerning wage levels and differentials are evident. One of the most serious is the difficulty of attracting managers, technical personnel, salesmen, and other highly skilled persons to backward areas, given limited pay and income differentials within communities. For example, if the allowable difference between "high" pay and average pay within a community is 3 to 1, and the differential between averages in advanced and backward areas is 6 to 1, it would be necessary to pay highly qualified people (generally from the more developed areas) at least 18 times the local average pay level in order to attract them to the poorer, backward areas. Such dif- ferentials would meet social resistance. Among the options for dealing with this problem are bringing in entrepreneurs as owners of firms with large residual incomes (but whose fixed base pay may not be too out of line with local standards); providing major noncash benefits like housing to pad incomes; or treating such personnel as "temporaries", whose pay is off the local scale but who cannot build up a permanent stake in the community. 6/ At present, some surprising labor practices are being used in China's rural areas. One is labor contracting (under which an individual contracts to obtain labor for a given job and is paid an agreed total sum, and he himself decides how much to pay the workers). Bringing in workers from poorer regions to do hard work that local residents are no longer willing to do, often with lower wages and benefits, also is increasingly common. Wiien it occurs on a large scale, this can adversely affect both incomes and employment opportunities of local community members. In Nanhai County in 1986, an inflow of workers from poorer rural areas and hiring of these people by local TVEs apparently forced some local residents to return to agriculture.  - 62 - 6.10 In the past, worker-manager differentials within enterprises were very low. Community attitudes and pressures tend to limit such gaps, though these limits have eroded in recent years, and differentials have widened. The rapid development of private firms in many parts of the country has given CE managers the option of moving to the private sector, which has sometimes forced CGs to allow them to earn higher rewards. There may be conflicts between managerial incentives, narrowly construed, and community and worker interests and "solidarity". For example, in Wuxi County managerial salaries are linked to workers' average pay at the enterprise level. This results in more harmony between workers' and managers' financial interests, but on the other hand it may encourage excessive wage increases. In Nanhai County there is no such linkage, and directors of CEs, responding to profit incentives, are actively recruiting low-cost labor from outside the community. Development of Rural Nonagricultural Labor Markets 6.11 The TVE sector is characterized by highly segmented labor markets. Though there is an increasing flow of "temporary' migrant labor from poorer rural areas to the more developed ones, so far this has not been sufficient to cause any significant equalization in TVE wage rates in different regions. 'Two-tier" labor and wage systems are emerging in the developed areas which permit inflows of workers from outside. Wages of "outsiders" may or may not be somewhat lower than those of community residents, but the former are not eligible for benefits or for profit-related "dividends". Moreover, outsiders tend to move into the most difficult, onerous, and dangerous tasks, like brick-making and construction work. This emerging "guest worker-like" system has some important advantages. It provides considerable amounts of employment and extra income to people from backward areas. By reducing surplus labor in the backward areas, the outflow of temporary migrant labor may also help raise incomes of the remaining residents. On the other hand, the influx of low-cost labor into the more advanced and rapidly growing areas helps check the rise in local labor costs, with possible benefits for capital accumulation and labor- intensive industrial growth. On-the-job training and experience gained by the temporary laborers can augment the human capital of backward areas if they return to their homes and set up small businesses. But large-scale temporary labor movements could lead to social strains in the future, and the long-run benefits for the backward areas which supply temporary migrant labor may be limited. In areas which inhibit the inflow of outsiders, like Wuxi County, labor shortages and tendencies toward capital intensification may result. 6.12 One of the most critical institutional and policy questions faced by China's rural economy is whether and to what extent well-functioning rural labor markets should be allowed to develop, along with consequent labor and population flows. Mobility of labor, especially of highly skilled and talen- ted people, is one of the most important means by which new and improved tech- nologies and practices are diffused throughout the economy. Movements of labor may also permit more flexible adjustment to changing economic conditions by firms as well as by workers. But there are counterexamples like Japan, with its near-guaranteed lifetime employment system for a large segment of the labor force. What seems more important than large-scale movements of ordinary labor is mobility of "high-quality human resources" (entrepreneurs, managers,  - 63 _ engineers and technicians, procurement and marketing agents, and other skilled persons). This was gravely hindered in China's prereform system, and there has been little improvement in this respect in the urban/state sector since the late 19709. TVEs in the more developed areas increasingly have been able to attract high-quality himan resources from the urban/state sector in recent years, by offering very high wages and benefits (often more than twice what these people were earning previously). Retirees from SEs and other state institutions are another fertile source of human resources for the TVE sector. Nevertheless, there are still shortages, and within the TVE sector CE managers and professionals often are prevented from leaving their firms by various sanctions available to CGs (e.g. threatening to refuse to let family members and relatives work in the local CE sector if a person leaves his job). 6.13 This all suggests that measures to enhance the mobility of high- quality human resources are desirable and urgently needed (some options are discussed in the following section of this chapter). Large-scale mobility of ordinary unskilled labor is a much more problematic issue. Though in princi- ple freer flows of labor across communities P.nd localities would be desirable, it is unrealistic to expect the emergence of anything approaching national or even regional rural labor markets in the near future. The social costs of large-scale, "permanent" movement of peasants and their families across local- ities would be very great; unlike tempora:.y migrants (who invariably do not bring their families with them) permanent migration would meet with resistance from community residents in the richer, more developed areas, since this would dilute the per-capita resource base in the latter. 6.14 Policies toward labor mobility and development of rural labor markets must be coordinated with those toward mobilit:y of other factors of production. For example, it may be more. efficient to improve overall efficiency through movement of capital (Chapter IV) and entrepreneurs (see below) from the more developed, high-wage areas to backward, low-wage regions than through large- scale movement of ordinary labor in the opposite direction.7/ In the absence of such labor migration, free flows of capital from backward to developed regions could be counterproductive. Greater labor mobility and development of rural labor markets should also be linked to a more liberalized attitude toward rural-urban migration. The bulk of the shift of China's vast rural labor force from agricultural to nonagricultural activities will have to occur through "local" rural-urban migration and local urbanization rather than as a result of large-scale movements of labor and population from one rural area to another. Mobility and Nurture of Human Resources 6.15 Restricted m( 'lity of high-quality human resources has an adverse effect on TVE sector eiiiciency and performance at the micro level, as noted above. It also may contribute in a major way to the problems of backward areas and to uneven spatial patterns of rural nonagricultural development (see Chapter VII). Imbalances in endowments of high-quality human resources are 7/ This would require major improvements in the policy and regulatory environment for "outside" investors in backward areas (see Chapter VII).  - 64 - almost certainly a main cause of the severe gap in economic development and industrialization between successful and backward areas. The very immobility of conventional factors of production means that backward areas have plenty of labor and at least some capital, and often natural resources of various kinds as well. But the entrepreneurial, managerial, and marketing abilities needed to organize these factors of production into profitable businesses as well as the necessary networks of contacts with potential customers and suppliers are missing. Under these circumstances, increasing the mobility of conventional factors might well exacerbate gaps if entrepreneurs and other high-quality human resources remain relatively immobile. The fact that in recent years an advanced TVE area like Wuxi has developed its textile industry in a major way (from a zero base in 1978) while a poorer low-wage locality like Shangrao has no textile plants suggests the depth of the problem. 6.16 If immobility of high-quality human resources is the major source of uneven rural economic and industrial development, at least in one respect the problem is somewhat manageable: the sheer volume of resources that need to be made more mobile is relatively small, and large population movements may not be necessary, at least in the short run. Moreover, selective investments in 'human capital" may have high pay-offs for the backward areas. The problem can be attacked in two ways: (1) measures to encourage greater flows of human resources to areas where they can earn the highest returns. and (2) policies and investments to nurture "indigenous" human resources its the backward areas. 6.17 The first of these options runs into obstacles because of the whole nexus of rural community connections that makes for immobility in the first place. Moreover, not just temporary mobility of high-quality human resources is desired. In the case of entrepreneurs at least, what is important is where they set up firms, and major institutional and policy changes will be required before the strong preferences of entrepreneurs for locating in their home com- munities will be swayed. Increasing mobility of entrepreneurs will not neces- -arily cause them to flow toward backward areas, at least initially. The ease of doing business in the more developed areas and the financial rewards that can be earned may cause entrepreneurs to gravitate toward them for some time to come. Implicit limits on the margin by which 'high" personal incomes can exceed the local average make it even more difficult to attract entrepreneurs to backward areas. 6.18 Despite these and other obstacles, it may be possible for greater entrepreneurial mobility to generate enterprise development and industrializa- tion in the more backward areas. The more developed areas are increasingly -^rowded' with entrepreneurs, so it is becoming more difficult to earn extra- ordinarily high returns. On the other hand, the lack of indigenous entrepre- neurial resources in some of the backward areas means that many market niches remain unfilled and opportunities for entrepreneurial profits are greater. To encourage movements in this direction, backward areas need to provide a rela- tively good administrative environment, with a minimum of red tape and govern- ment intervention; they should also ensure that entreprenrurs can earn returns which are competitive with those that can be earned in the more developed  - 65 - areas;8/ and they should facilitate movement of entrepreneurs' families and their establishment of longstanding residence. 6.19 The second option requires resources and is not a short-term measure. but it may be effective. A great deal can be gained by sending out managers and entrepreneurs from the backward areas to the developed regions, not to do menial, low-skill tasks but to gain useful experience through apprenticeships and various forms of on-the-job training. Formal education and training also can play a role, but to avoid exacerbating the "brain drain" problem that many poorer localities already suffer from, people receiving such education or training should be required to return to their home communities to work for some minimum length of time. Finally, the entrepreneurial and other high- quality human resources that the backward areas already have are often not well-utilized. Thus a top priority is to create an administrative and regulatory environment thet is conducive to indigenous entrepreneurs. 8/ This is probably best done by letting outside enterpreneurs set up firms which are essentially private in their ownership and autonomous in their management (though the local CG perhaps could participate in their capitalization on a nonvoting basis).  - 66 _ VII. PROBLEMS OF BACKWARD AREAS 7.1 The degree of spatial inequality in rural incomes, economic develop- ment, and industrialization is striking and, to some extent, worrisome. The magnitude of the problem can be illustrated w4th some local data shown in Tables 7.1 and 7.2. Table 7.1 provides basic information on a per-capita basis for the four counties that were subjected to intensive fieldwork inves- tigation. They by no means span the full range of variation in China, but even so, the gap is rather large. Average per-capita income in Nanhai County is 3.6 times that in Jieshou County; GVIO per capita in Wuxi is an astonishing 95 times that in Shangrao, and agricultural output per capita is nearly 1302 higher. Wages paid by TVCEs per capita in Wuxi are 28 times those in Jieshou, gross profits of TVCEs per capita in Wuxi 128 times those in Shangrao. County statistics mask even greater differentials between rural communities. A highly developed, industrialized township in Wuxi County is compared with a backward township in Shangrao County in Table 7.2. Not only is the gap in per-capita income larger (nearly 7 to 1), but differences in industrialization are astronomical. GVIO per capita in the developed township is close to 700 times the figure in the backward township. These differences in income levels and rural industrialization translate into huge gaps in local financial resource mobilization capacity (about 1%,000 times). 7.2 Such unbalanced rural and TVE development patterns are caused mainly by the combination of highly uneven endowments of human and other resources and immobility of population and factors of production. The problems of back- ward areas are exacerbated by their lack of entrepreneurs, managers, and contacts with markets; by predatory fiscal practices on the part of many CGs toward CEs; and by greater difficulties their TVEs now face in getting a foot- hold in increasingly crowded and competitive markets. Fiscal redistribution is substantial but does not address the fundamental causes of uneven develop- ment. Moreover, "aid" from higher levels of government is often inefficiently utilized, and the widespread use of interest subsidies in backward areas encourages waste of capital. Massive state investments in urban industrial development in the cities of interior provinces had virtually no beneficial impact on the backward rural areas. 7.3 The extremely rapid growth of the TVE sector in many parts of the country in recent years has highlighted the problems of the backward areas which have not fully participated in such growth. Emergence of a certain degree of relative inequality is only to be expected under such circumstances. The TVE sector itself is not the source of rural inequality in China, and it would be unrealistic to try to reduce rural inequality solely through reforms and policy changes affecting the TVE sector alone. Nevertheless, to the extent that it is possible, more rapid growth and better performance of TVEs in backward areas can play a significant role in redressing the most severe manifestations of rural spatial inequality.  - 67 - Table 7.1: ECONOMIC INDICATORS IN FOUR COUNTIES, 1985 (Y per capita of agricultural population) La Wuxi Nanhai Jieshou Shangrao China Average per-capita income of rural population 754 1,029 285 322 398 Rural GVIO 4,656 2,313 244 49 207 GVAO 604 610 210 266 429 TVCE gross profits 512 208 9 4 24 TVCE wage bill 364 276 13 23 36 /a For China as a whole, the figures are per member of the rural population. The agricultural population is an administratively defined category used at local levels, which includes people engaged in nonagricultural activities but not receiving grain rations or other subsidies. Sources: Information from fieldwork; Statistical Yearbook of China, 1986. Table 7.2: ECONOHIC INDICATORS FOR TWO TOWNSHIPS (Y per capita) Township A, Township B, Ratio Wuxi County, 1985 Shangrao County, 1986 A/B (X) Average per-capita income of rural population 907 131 692 GVAO 399 197 203 GVIO 11,268 17 663 Sales revenues of TEs 5,846/a 14 418 Local revenue mobili- zation /b 1,268 9 14,089 Expenditure 522 18 2,900 la Industrial TEs only. /b Includes both budgetary and nonbudgetary revenues, but not additional budgetary funds provided by the county for 'quota' budget expenditures. Sourres Tnformattnn from fieldwork.  - 68 - 7.4 The main issues concerning backward rural areas include: (1) the need to eliminate "fiscal predation" by CGs; (2) the need to improve the adminis- trative and regulatory climate in these areas, especially to get rid of exces- sive CG intervention in the operations and personnel decisions of TVEs; and (3) the need to take positive actions to augment and improve human resources. Given the seriousness and complexity of their problems, major efforts will be required to stimulate more rapid TVE growth and development in the backward areas. Mere liberalization of controls and elimination of restrictions against TVEs will be helpful but may not be sufficient. Ameliorating Fiscal Predation 7.5 "Fiscal predation" occurs when CGs' regular revenues are insufficient to cover their largely fixed exp%--.itures on public and social services and on the CG payroll. In this situatiku which generally arises only in the more brckward areas) CGs are forced to levy funds from their subordinate CEs, regardless of whether or not the latter earn profits. CEs in turn obtain the money they need to meet these remittance requirements by "eating" their capi- tal stock (i.e. using depreciation funds), borrowing from local banks, or both. This severely strains bank-CG relationships and harms the financial viability and development prospects of the CE sector (see Chapter IV). 7.6 Specific measures to deal with the problem of fiscal predation have been implemented at the local level (in parts of Shangrao County) with some degree of success. These involve a combination of election of factory directors by workers, "contracts" between elected factory directors and the CG, and a credible pledge by the latter not to levy more than a certain pro- portion of enterprise profits. Another part of this strategy is mandatory retirement (sometimes with a 'golden handshake") for incompetent CG officials and enterprise managers. But these solutions may not work everywhere, and they have been most successful where community resource constraints were at least temporarily eased by inflows of funds of one kind or another. 7.7 There are some nationwide policy changes which could help reduce the need for fiscal predation, but they represent a sharp departure from past practices and may have substantial resource requirements. The fiscal problems of backward townships can be viewed from two different perspectives: (1) In these areas, government and its activities are too big (in particular, the number of people on the CG payroll is too large) in relation to the resources of the community. (2) One way China has been successful in meeting basic needs of the rural population is by placing much of the financing burden on grassroots rural community institutions, regardless of their ability to pay. 7.8 Thus the problem of fiscal predation needs to be attacked from two angles. In the first place, backward areas cannot afford and do not need the full government administrative structure and the numbers of CG employees that may be necessary in the more advanced areas where economic activity is so much greater. The organizational structure of CGs in the backward areas should be simplified and the number of CG personnel reduced to the point where it is  - 69 - more in accord with local community resources. While the process of retrench- ment may be a difficult and gradual one, the objective should be clear. The large numbers of CG cadres nearing retirement age in many localities present an opportunity for reductions through attrition which should not be wasted. To encourage CGs to reduce their staffs and cut overhead expenditures, such actions could be made a condition for expanded budgetary allocations for social expenditures (see below). 7.9 There are various public and social services now being financed from nonbudgetary rural community resources which are important in meeting the basic needs of the Chinese people and in developing the "human capitaln that is so necessary for local economic development. These should not be adversely affected by CG retrenchment in backward areas. A good example is the finan- cing of rural primary education (a large part of which occurs outside the budget). If these kinds of expenditures get neglected in the more backward areas, inequality in resource endowments, levels of living, and development prospects could become even worse. 7.10 A reassessment of the public expenditure responsibilities of CGs would be desirable. After a careful evaluation, those which are found to be truly important in furthering national social objectives could be incorporated partially or wholely into the state budget. This would require additional budgetary resources, but the pay-off could well be much higher than that from the large amounts of government resources which now go into financing or subsidizing interest charges for poorly-chosen investment projects in backward areas. The budgetary costs could be held down, if necessary, by permitting budgetary financing of the expenditures concerned only in localities that meet specified criteria concerning poverty and backwardness, which would also give the measure a more decidedly "redistributional" tinge. 7.11 Eliminating fiscal predation will greatly help in the task of reducing CG intervention in the operations of CEs. Educating CG officials on the benefits of releasing controls also could be efficacious, particularly if there are examples at hand where such reforms he-'e stimulated TVE development and general prosperity. Human Resource Development in Backward Areas 7.12 Chinese rural communities have been largely dependent on the local human and material resource base to stimulate TVE development. Those that have failed to industrialize successfully tend to lack needed human resources, due to a variety of historical, geographical, economic, and systemic reasons. As was noted in Chapter VI, high-quality human resources in backward areas can be augmented and improved both by encouraging the inflow of these resources from the developed localities and by investments in "human capital" in the backward areas themselves. Though international experience provides ample ground for pessimism on the prospects for success of such measures, there are really no attractive alternatives short of a substantial outflow of population from backward areas (discussed in the following section).  - 70 - 7.13 Any program for high-level human resource development in the backward areas would have to be carefully designed and to the extent possible market- oriented. Sending local people out for on-the-job training and appren- ticeships in managerial work in the more developed areas may be an effective option. Training programs within the backward areas should be geared toward actual needs rather than narrow technical specialties. Finally, measures would need to be taken to ensure that human talent remains in the backward areas and contributes to TVE development there. The fact that Shangrao County has a high-quality basic education system whose graduates each year include a considerable number who pass their examinations, go on to universities, and then never return to the county should give pause for thought. 7.14 To encourage inflows of high-quality human resources from other areas, the most important tasks are to create an administrative and regulatory environment conducive to development of profitable enterprises and to offer potential returns that are competitive with those in the more developed areas. The best way to achieve the latter may be to allow outside entrepreneurs to set up PEs in the locality, possibly with nonvoting equity participation by the local CG. These entrepreneurs will then have a strong incentive to identify and bring in the other human resources needed to run a profitable firm (managers, technicians, marketing personnel, etc.). Greater Population Mobility and Rural Industrialization 7.15 The options and measures discussed above, as well as other possible reforms, could have some impact on poverty and backwardness in China's under- developed rural areas. But experience in other countries w'th government- mandated 'regional development programs" has been decidedly mixed. Over the longer term, outward flows of population from some of the more backward areas may be the most effective means of alleviating poverty and raising incomes. This would mark a sharp departure from the present system of fixed-membership communities in rural areas. The weakening of this system might be only a gradual process, to minimize the possibly large social costs involved. Though resistance agelnst moving out by the populations of poorer localities may not be very great, it could still be a factor, and moreover local CGs and county authorities would probably be opposed because it would weaken their power base. Perhaps more important, there would be strong resistance from the localities absorbing such migration. Though immigration of temporary, low- paid workers for the most part doing menial tasks may be accepted by the local population, this would not be true of permanent immigration, since it would mean sharing community resources (land and local nonagricultural job opportu- nities) among a larger pool of "members". Local governments might resist permanent settlement of immigrants because of the added responsibilities and financial burdens involved. Despite these obstacles, outward movement of populatlon from backward areas deserves consideration as an option and as part of a long-term national urbanization strategy. It is probably much more feasible if it occurs as part of widespread urbanization, in which large numbers of people from richer and poorer areas alike are leaving their home communities.  - 71 - 7.16 Even with substantial human resource development and elimination of fiscal predation, it is unrealistic to expect every nonindustrialized county in China to be able to follow the high-industrialization path of Wuxi or Nanhai. Though there may be some room for development of other nonagricul- tural activities, many of these do not have great potential in the absence of industrialization. Many of the localities concerned may well be suited pri- marily for various types of agriculture only. Development of nonagricultural activites to absorb surplus labor from agriculture would still be the best means of raising incomes in these areas, but the nonagricultural activities concerned would emerge primarily elsewhere, requiring substantial outmigration (which would leave the remaining local population, still primaril.y engaged in agriculture, better off).  - 72 - VIII. PROSPECTS FOR THE FUTURE: INSTITUTIONAL AND POLICY ISSUES 8.1 The future role and share of the TVE sector in the Chinese economy depends on many hard-to.-predict factors. But from all indications, its impor- tance will continue to grow over the short and medium-term. The TVE sector will likely maintain double-digit growth for the next several years, at a pace considerably faster than that of state industry. TVEs represent the only hope for achieving the Government's ambitious goals for shifting a large part of the rural population from agriculture to nonagricultural activities over the next several decades. TVE development is also to provide the foundation for 'urbanization in small towns". Great hope is invested in the TVE sector as a generator of exports and export earnings. The potential role of competition from TVEs in stimulating the state-owned industrial sector to reform itself and improve efficiency should not be underrated. 8.2 But the TVE sector will need to further evolve and develop if it is tn continue to be a highly dynamic "leading sector" in the economy, particu- larly if urban reforms are successful in improving the efficiency and competi- tiveness of SEs. The barriers against reasonably large firm size will have to be overcome, so that more efficient and successful TVEs will be able to grow beyond community-based limits. At the same time, closer and mot. organic forms of integration with the state sector may become increasingly necessary for many of the smaller firms. The costs of the present highly dispersed location pattern of TVEs will likely become a progressively severe competitive disadvantage; hence there needs to be evolution toward greater concentration. TVEs will also have to improve technology and product quality, cut wastage of material inputs and energy, and engage in a host of related efficiency improvements in the increasingly competitive business environment which is emerging. Competition within the TVE sector is already severe in many industries and will only become more heated over time. A reformed, more dynamic, and efficient state sector may also put increasing competitive pressure on TVEs in the future. 8.3 One set of issues of crucial importance for the future development of China's TVE sector concerns Government policies. Policy liberalization has played a major role in allowing the TVE sector to grow so rapidly during the past decade. Government policies may be even more important in encouraging or hindering needed transitions by the TVE sector in the future. The legal and regulatory framework for TVE activities has already been discussed (Chapter III). The future evolution of industrial markets will be crucial and will depend to a considerable extent on Government policies. Government industrial strategies and policies may also have an important bearing on future TVE development. These policy issues will be examined in the second half of this chapter. 8.4 Institutional aspects lie at the heart of both many of the problems and much of the needed future evolution of the TVE sector. They have also played a major role in its successful performance to date. The most striking characteristic of the TVE sector is its strong community orientation, yet it is possible to foresee a weakening of this in the future, with major  - 73 - consequences. In this context, the freure role of CGs at different levels is a crucial question. A more immediate issue concerns the strong fiscal incen- tives for TVE development and possible distortions these may induce. Owner- ship is a key part of the institutional structure, particularly since owner- ship patterns vary so greatly across different regions and localities (see Chapter I). One question concerns the future evolution of the ownership form of CEs. Another issue is the scope and potential for further development of PEs. There are also more specific issues and ptions, related to the identity and roles of entities exercising ownership righits (CGs, TICs, banks, outside entities, workers, community members, etc.). Institutional issues will be taken up in the first part of this chapter. Fiscal Incentives Related to TVE Development 8.5 The need to generate revenues to meet local public expenditure requirements has been a potent incentive for CGs to actively promote TVE development. It is much easier for CG officials to levy funds from TVEs, especially CEs, than from agricultural activities or directly from peasant households. Rural communities with successful, profitable TVEs also tend to have a better CG financial situation and adequate funds for public services, social services, and investment in needed infrastructure, public facilities, and further TVE development. On the other hand, localities that have not been able to develop TVEs seem to be caught in a vicious cycle of fiscal predation (see Chapter VII). 8.6 Though fiscal incentives have played an important role in the recent rapid growth of China's TVE sector, problematic aspects of these fiscal linkages also are readily apparent, particularly in the backward areas. They carry the danger of increasing spatial inequality in provision of various rural social services, which in turn could result in further entrenchment of uneven rural development and industrialization patterns. The fiscal incen- tives of CGs related to TVE development may be in a certain sense "excessive". In the backward areas, excessive reliance of CGs on CE remittances is obvi- ously harmful and adversely affects TVE development. But even in the more developed areas, the fiscal impact of the TVE sector on the local community is so great that it may lead to undesirable buffeting of government expenditures, uneven provision of some government services over time, and possibly inappro- priate uses of TVE profits. 8.7 Some separation of the fiscal/governmental and business development roles of CGs may be desirable. Though this might weaken fiscal incentives for TVE development, other incentives like employment generation would still be quite strong, and enterprise profit incentives might well be strengthened. The beginnings of such a separation can be seen in the TICs of Wuxi. While still under overall supervision by the TG, the TIC there does serve as the intermediary between TEs ar.d the TG, and it effectively pools resources of subordinate enterprises. The business development role of TICs might be further strengthened by allowing them to issue income bonds (see Chapter IV) and through other measures. TICs could move further in the direction of becoming profit-oriented holding companies, owning and supervising local TEs  - 74 - and mobilizing funds for them through bond issues, investment in joint ven- tures, etc. The link between the TIC and the TG could gradually evolve into a more arms-length relationship. 8.8 Overall reforms in China's public finance system may profoundly affect the fiscal incentives and linkages related to local TVE development. For example, the trend toward giving tax collection responsibilities and incentives to lower and lower levels of government (culminating with the establishment of townships as a new level of public finance in many areas) may be reversed in the future. If the tax collection functions of township governments are taken away and put in the hands of higher levels of govern- ment, this may greatly weaken the former's fiscal incentives to promote TVE development. It thereby might facilitate the separation of CG corporate and fiscal roles. Future ROLe of Community Governments 8.9 CGs have played a crucial role in the development of the TVE sector so far, which undoubtedly will continue for some time into the future. But there are already signs of friction and change. CGs have not been effective in stimulating TVE development everywhere. The traditional CE model of rural industrialization is now only one of several; other models have different implications for the role and importance of CGs. Though PEs are in many ways still part of the communal economic and social structure and often have close informal ties with CG leaders, these are uneasy, unstable relationships which may gradually erode over time if the property rights of PEs are strengthened and clarified. Greater population and labor mobility will affect CGs in ways that are hard to predict but will tend to weaken their authority. The increasing independence of the local banking system and possible future reforms in its structure may affect the ability of CGs to allocate community savings. The market situation faced by TVEs is becoming increasingly competi- tive, which makes the task of CGs in the areas where TVEs have not yet devel- oped even more difficult. More generally, economic development will likely erode the communal nature of Chinese rural society over time. 8.10 It is hard to forecast how rapidly the community structure in China's rural areas will evolve and change. In 1980, nobody would have predicted the universal implementation of the PRS, the demise of the commune system, and, in many areas, the virtual elimination of production teams as significant commu- nity entities, all of which actually occurred during the next few years. Thus there is at least the possibility of further rapid institutional change in China's rural areas. On the other hand, there are elements of stability, particularly in the more developed localities where CGs have built up dynamic and relatively efficient CE sectors. 8.11 As part of the separation between the corporate and fiscal roles of CGs, mentioned above, TICs may begin to transcend the limits of their commu- nity orientation. This could occur first in their obtaining resources from outside the community, mainly capital (e.g. through bond issues), but later such transcendence may become broader and may lead to changes in orientation. T!Cs may eventually become relatively autonomous, linked to CGs by ownership relations but without the many noneconomic ties which exist at present.  . 75 - 8.12 CGs would then focus on their more narrowly defined "governmental" functions, relying on shared local revenues and/or grants from higher levels of government, as well as possibly some continued remittances from TICs. This evolution should help improve the provision of public and social services in rural areas. More generally, CGs in many parts of the country will have to shift their orientation away from a narrow focus on rural industrialization as the primary means to promote economic development, generate employment, and solve fiscal problems. It is simply impossible for every township in China to attain the level of industrialization achieved by Wuxi and Nanhai. Some of them may develop specialized agriculture, while others may be able to carve out a niche in nonindustrial activities outside of agriculture. Many of them may lose population eventually, as urbanization elsewhere attracts labor from poorer regions and restrictions against migration erode (see Chapter VII), but it will require a major change in the mind-set of CG leaders to accept this. Community Orientation of CEs 8.13 Like TICs, some CEs will need eventually to transcend the limits of their rural community roots if they are to attain large firm sizes and fully exploit economic opportunities. This will require some difficult transitions, but if it does not occur the CE sector will be constrained in its ability to compete with SEs and may run into severe problems if the latter become more competitive. Under these circumstances, CEs would also be in danger of being left behind by PEs, which might find it easier (in an appropriate policy and regulatory environment) to transcend their community ties. Transcendence for CEs could occur as part of the change of orientation of TICs; another transi- tional possibility is for some large CEs to be structured as joint ventures among several neighboring communities. But eventually, the successful CEs that are able to become large firms will probably be structured much as large corporations elsewhere--as joint stock companies in which the "founding" CG may still be the largest shareholdei but no longer dominates. Such firms would no longer be narrowly construed as TVEs, though they might derive much of their efficiency and competitiveness from their TVE antecedents. Role and Importance of Private Enterprises 8.14 This issue will become increasingly important as the PE sector continues to grow rapidly, more and more relatively large PEs emerge, and they become more autonomous vis-a-vis CGs. Though PEs are still largely community- oriented despite their lack of formal community ownership ties, this could easily change over time, in which case PEs might become a vehicle for weaken- ing the community orientation of the TVE sector in general. 8.15 A key issue is the different kinds of interactions and competition between PEs and CEs. As was mentioned earlier, competition for labor, and in particular for "professionals", has been one of the key sources of friction where PEs have been permitted to develop rapidly, and served as the main justification for a harsh clampdown against PEs in Wuxi County. Over time, continuing growth of the PE sector may sap the vitality of CEs by attracting  - 76 - away many of their best people, or alternatively will force CEs to pay high wages to these personnel (as is already occurring in Nanhai), leading to larger pay differentials within firms. 8.16 Another important linkage between the two parts of the TVE sector is various forms of 'privatization' of CEs. These usually involve some type of contract-based autonomous management of CEs by individuals or small groups, with a large element of residual claimant status with respect to profits. Such methods often seem to work well in the short run, permitting greater autonomy of management while still maintaining links to the community and access to community resources. But there needs to be a clearer specification of property rights in these firms to avoid future problems. Outright sale of CEs to individuals is becoming more widespread; if this trend continues, it could lead to substantial further reductions in the share of the CE sector. 8.17 Still another kind of linkage is joint ventures or other forms of "attachment" by PEs to local CGs. These are often motivated by the need for private entrepreneurs who have gained a certain degree of wealth and visibil- ity to obtain "administrative protection' from the local community, as well as access to community resources. The potential for abuse and personal corrup- tion in these arrangements is great. Many of them might turn out to be unnec- essary if the legal framework and regulatory environment for PEs is improved. 8.18 In the absence of heavy-handed restrictions, the PE sector is likely to continue to grow rapidly and further increase its share in the rural non- agricultural economy. This will have important implications for CEs and CGs. But there will probably continue to be a great deal of regional and local diversity. Diversified .wnership in the TVE Sector 8.19 As part of the weakening of CG-CE ties, and as part of the integra- tion of sizeable PEs into the rural economy, diversification of ownership of larger TVEs would appear desirable. This should not be considered as a goal in itself, but rather it should be permitted to occur as enterprises pursue different channels for obtaining capital needed for expansion; as firms and CGs strive to spread and reduce the financial risks they face; and as inves- tors seek high returns. Among the different rural institutions and persons who could be appropriate 'owners' of TVEs are the following: (1) individual proprietors and partners; (2) CGs at all levels, both inside and outside the community in which a firm is located; (3) ABC branches and RCCs (with limited exposure and only in the case of large and relatively profitable TVEs); (4) enterprise workers (also to a limited extent); (5) community members not employed by the firm; and (6) individual investors from outside the community. 8.20 Though the TVE sector would benefit greatly from a clearer specifica- tion of ownership and property rights, free trading of ownership claims might not be possible for a long period to come (though trading of the income bonds described in Chapter IV could become feasible). Most enterprises are too small for this, and it might not be appropriate for trading of shares in TICs to occur, since these really represent their communities, and the latter cannot be liquidated.  - 77 - 8.21 In terms of the precise exercise of ownership functions and how these are to be structured, something akin to the board of directors concept might be appropriate for larger TVEs, though there are many other workable options. Most if not all owners would be on the board, given their small numbers, so it would really function more like an owners' council. In the case of PEs, partial ownership by banks and other entities could be on a nor.voting basis, as suggested earlier, to help maintain autonomy of the private proprietors. Legal and tegulatory Environment 8.22 Desirable improvements in the legal and regulatory framework for TVEs parallel those needed in the rest of the economy (Chapter III). In the future the focus will gradually shift from codification to meaningful adjudication and enforcement, where the system still has a long way to go until it becomes supportive of autonomous, business-oriented enterprises in a well-functioning competitive market environment. 8.23 The importance of a nondiscriminatory legal and regulatory environ- ment for the TVE sector and a "level playing field' for competition within the TVE sector as well as between TVEs and other types of firms has been stressed. Progress in achieving this goal will probably be only gradual. Nevertheless, it is important to keep the ultimate objective in mind, and to ensure that specific policy measures in the short run go in the right direction even if they leave something to be desired from a long-term perspective. 8.24 In this context, the separation of regulatory and ownership functions in the state-owned industrial sector should prove helpful. At present some government agencies with regulatory responsibilities for industry as a whole or for entire industrial subsectors simultaneously have close hierarchical ties with SEs, which may bias their regulatory actions. Similarly, a gravita- tion of government regulatory functions toward higher levels, particularly provincial and central governments, would facilitate the separation of owner- ship and regulatory roles at lower levels. 8.25 A problem that pervades the legal and regulatory apparatus, with profound adverse consequences, is the bias against outsiders (see Chapter III). This has to be addressed in some way, even though fundamental improve- ments in the situation may come only gradually. In business disputes, outsiders need to have some recourse to reasonably impartial legal institu- I tions, so that they can be conf.dent about engaging in commerce and other activities outside of their home localities. Whether existing government agencies can be improved to the point whiere they can provide such recourse, or whether new institutions with much weaker local ties need to be created, is a question which requires further study. Competition and Market Structures 8.26 Government policies toward markets and competition can have a major impact on the future evolution of the TVE sector, since it is the most market- oriented part of Chinese industry. Measures which improve the functioning of  - 78 - markets and promote the emergence of more unified regional and national markets will have a beneficial effect on the TVE sector overall. Tt.is is true even though in the short run and in some industries and localities, broader and better-functioning markets may cause considerable suffeting for particular TVEs, their employees, and their communities. It is only through flexible adjustment to increasingly competitive markets that the TVE sector as a whole can prosper over the medium- and long-term future. 8.27 The TVE sector at present benefits very little if at all from the various kinds of "internal protectionism" that provincial and local govern- ments in China sometimes engage in. Thus effective enforcement of the central government's repeatedly enunciated stand against such protectionism will cause little if any harm to TVEs even