China - Strategies for reducing poverty in the 1990s

I' Report No. 10409-CHA China Strategies for Reducing Poverty in the 1 990s June 29, 1992 Environment, Human Resources and Urban Development Operations Division China and Mongolia Department East Asia Region FOR OFFICIAL USE ONLY MICROFICHE COPY Report No. 10409-CHA Type: (SEC) PIAZZA, A./ X84067 / A7 041/ AE2AG 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 otherwis. be disclosed without World Bank authorization.  OFFICIAL EXCHANGE RATE Currency Unit = Yuan (Y) Effective March 31, 1992: US$ = Y 5.46 Y = US$ 0.18 Up to November 29, 1990: US$ = Y 4.72 Y US$ 0.21 Up to December 15, 1989: US$ = Y 3.72 Y = US$ 0.27 CONVERSIONS 1 jin 0.50 kg 1 kg = 2.00 jin 1 mu = 0.0667 ha 1 ha = 15.00 mu ACRONYMS ABC - Agricultural Bank of China CASS - Chinese Academy of Social Sciences COE - Collectively-Owned Enterprise ICBC - Industrial and Commercial Bank of China IMR - Infant Mortality Rate (first year) LEMO - Labor Export Management Office LGEDPA - Leading Group for the Economic Development of Poor Areas MOA - Ministry of Agriculture MCA - Ministry of Civil Affairs MOF - Ministry of Finance MOL - Ministry of Labor MOPH - Ministry of Public Health PADO - Poor Area Development Office PBC - People's Bank of China POE - Privately-Owned Enterprise RCRD - Research Center for Rural Development RE - Rural Enterprise SEdC - State Education Commission SOE - State-Owned Enterprise SPC - State Planning Commission SSB - State Statistical Bureau STC - State Science and Technology Commission Y - Yuan (Rennminbi)  FOR OFFICIAL USE ONLY Preface This report is based on the findings of a mission which visited Beijing, Yunnan, Guangxi and Gansu during April and May, 1991. Mission members were Alan Piazza (EA2AG, mission leader), Wang Yuan (EA2CH, public finance), Elaine Chan (EA2CO, labor and rural enterprise), Lee Travers (consultant, urban poverty and social services), Paula Harrell (consultant, poverty programs and southwest China case study), Elisabeth Croll (consultant, gender issues and village studies), and Suresh Tendulkar (consultant, social services and poverty measurement). The mission worked closely with the Leading Group for Poor Area Development (Leading Group), was accompanied by Messrs. Gao Hongbin (Deputy Director of the Leading Group) and Zhang Yiming throughout its stay in China, and was hosted by provincial and local level staff of the Leading Group in Yunnan, Guangxi and Gansu. The excellent support afforded the mission at all levels of the Leading Group system is gratefully acknowledged. In addition, the mission met with staff of the central, provincial and local level offices of the Ministry of Civil Affairs, Ministry of Agriculture, Ministry of Public Health, State Planning Commission, State Education Commission, Minorities Commission, All China Women's Federation, and State Statistical Bureau. The mission also benefitted from interaction with UNICEF, WFP, UNDP and Ford Foundation staff. The report's analysis and recommendations are scheduled to be discussed at an international conference on "Poverty Issues in China" to be held in Beijing during late-1992. UNDP financial support for the proposed conference is gratefully acknowledged. Judith Banister (consultant, demographics), Bert Keidel (consultant, macroeconomics), and John Quinley (consultant, public health) each completed desk studies in support of the report. Mr. Yao Xianbix. provided valuable research assistance. The mission also wishes to acknowledge the extensive review and comments on the report provided by Lyn Squire, Paul Glewwe, Terrice Bassler, Jeffrey Taylor and Martin Ravallion (peer reviewers) and Robert Parker (UNICEF). Helena Ribe, Samuel Lieberman, Barbara Herz, Zafer Ecevit, Shahid Yusuf, Joseph Goldberg, Paul Cadario, Richard Bumgarner, Ivy Chang, Caroline Jen, Himelda Martinez, Tony Ody, and Christine Wallich (Bank) also gave valuable suggestions and comments. This document has a restricted distribution and may be used by tecipients only in the performance of their official duties. Its contents may not otherwise be discloseu without World Bank authorization.  CHINA STRATGOIRS FOR iNDUCING POVERTY IN TM 1990s Table of Cont-nto UmCUV sgumay .. ..............*....***.. ......... ..... v-xiv 1. *POVERTY ALLEVIATION AND NACROECONONIC TRENDS IN TME 19803 ... 1 A. Absolute Poverty ..........,........................ 1 S. Impact of Macroeconomic Trends ..... ................. 5 -Bconomic Reform and Growth .. . . . ...... .. .. . . . . 5 -Prices and Income .... .. ........ *... . . . .. 7 -Employment Levels and Demographic Change ..... 9 -Regional Growth Trends ................. ....... . .... 12 -Public Finance and Redistributive Transfers ...... 16 -Macroeconomic Trends and Poverty Alleviation Work in the 1990s ................................... .. 19 2. INCIDENCE AND CORRELAES OF ABSOLUT5 POVERTY .........o....... 21 A. Incidence of Absolute Poverty ... ........... ... 22 -Income Levels ... .....* . ... .... * 22 -Urban Poverty ...... ......... ............... 26 -Rural Poverty ................... ............. 32 D. Correlates of Absolute Poverty . ..... .......... ....... 41 -Natural Resources .............................*.... 42 -Infrastructure . ... ...... ..... *.... . 42 -Minorities . . ...... 0****.......... ... .. . .... 43 -Involuntary Resettlement ...*.. .... .*............ 45 -Gender .. o... .. o..... .... .. .. . . ............oo 45 3. * WOY.T AND PRODUCrIVITY ......* ... ....*.*.. . . . ... 49 A. Demographic Change, Employment Levels and Labor Mobility * 49 -Demographic Change and Labor Supply .......... 50 -Labor Market Development .. ........ ...... 52 -Expected Labor Market Trends ..................... 54 -Labor Mobility and Poverty Alleviation ........... 57 3. Rural Enterprise ..................... ...... 60 -overview ............................... ......... 60 -National Trends ...... 62 -Rural Enterprise Development in the Poor Areas ... 65 C. Agriculture and Natural Resources ................... 66 -Employment in AgrJculture ........................ 69 -The Loess Plateau ............ 7................... 72  4. SOC1AL SERVICES IN TME POOR AREAS ....... . . . . . . . . . . . ............. 81 A. Financing Social Services in the Poor Areas . ... 81 B. Education so ....... .*..*. .*.................. 82 -The Poor's Accose to Education .................. 83 -Returns to Education so ....... . 84 -Education Financing and services ......s.......... 85 -Intervention Strategies 2....... so...... 87 C. Health . 90 -Maternal and Child Health ....................... 90 -Disease Morbidity ..... . ......... ... ........ 93 -Health Financing and Services .................... 96 -Costs of Improving Health Services and Strategies in the Poor Areas .........*... ............... 101 D. Relief . ............................... * *.*..*........*...**. 107 -IntroductLon ..................................................... 107 -Rural Relief .. ...... ........ .. ......... . .........*.. 109 -Financing Rural Relief: An Expanded Role for the Minietry of Civil Affairs ........................ 112 5. POVERTY ALLBVIATION ?OLICIES, PROGRAMS AND INSTITUTIONS ..... 115 A. Poverty Alleviation DurLng the 7th Five Year Plan ... 115 -strategy ....................................................... 115 -Institutions ........... ................. so ... 116 -Defining the Poor . ............. ................. . 117 B. Development Assistance Programs * .................* 121 -Domestic Programs . *.. 121 -International Assistance .. ..... ........ 125 C. 8th FLve Year Plan: Reaching the Chronic Absolute Poor ............................... 127 -strategy and Funding ..... . ............. ....... 127 -Monitoring and Targeting ......... .... 0........ 128 -Institutional Issues ............................. 129 REFERENCES .......... *...... .............. 131 Annexes 1. Incldence of Absolute Poverty .........137 -National o ..oo.o . o. ..o ...137 -Provincial oo.o....... ... 00. ... 0 . 100... 141 -Conclusion .oo... o.. *oo... 0..o. 141 -Appendix 1: Rural Income Reporting Systems .... 152 mAPs 1. 1989 Incidence of Rural Poverty (SSB Data) ....156 2. 1989 Incidence of Rural Poverty (MOA Data) .......... 157  Tables in the Main Report 1.1 Selected Indicatocs of Social Development: China and Other CountrLes .............................................. 1 1.2 Macroeconomic Indicators and Estimated Incidence of Absolute Poverty ..... 00........*..................... 4 1.3 Provincial Average Annual Real Income Growth (%) ....... 14 1.4 Real Provincial Average Annual Budget Balances ......... 18 2.1 Distribution of Annual Per Capita Income, Poverty Lines, and Incidence of Absolute Poverty .......... ......... 23 2.2 Provincial Indicators of Social Development ............ 37 2.3 MOA County Per capita Income Data .................... 40 3.1 Labor Supply and Employment Trends, 1978-2000 .... ...... 52 3.2 Rural Enterprise Development, 1978-90 .........I.. 62 3.3 Provincial Rural Enterprise Employment, 1980-90 64 3.4 Estimated Agricultural Labor Requirements, 1978, 1984, 1989 and 2000 ......................... 71 4.1 Rural Disaster and Poverty Relief -- Numbers Assisted and Funding ............... <. 113 5.1 PADO (1985) and MOA (1989) Poor Counties and SSB (1989) Poverty Incidence k Y'.rcvince ..................... 119 Figures in the Main Report 2.1 Rural Income ar.d Poverty ..................... 33 Boxes in the Main Report 2.1 The Women's Federation and Poverty Alleviation ......... 48 3.1 Jingtai Phase II -- Laudable Objectives, Questionable Economics ......................................... .... 80 4.1 "Self-Relief Through Development" Enterprises .......... 110 Working Papers (available upon request) 1. Demographic Aspects of Poverty 2. A Program to Improve Health Services in the Poor Area& 3. Economic Analysis of the Loess Plateal Land Rehabilitation Program 4. Poverty in Southwest China 5. Rural Poverty: Characteristics, Causes and Policies in Baise Prefecture  -In tdence and Determinants of absolute Poverty (i) Incidence of Absolute Poverty. Broad participation in strong rural economic growth brought about a tremndous reduction in absolute povetcy in China durlng 1978-85. The number of absolute poor are estLmated to have declined from roughly 270 mllion ln 1978 to 100 million ln 1985, or from about one-thlrd to less than one-tenth of the total population. However no furzner reductlons Ln poverty were achLeved during the second half of the 1980e -- the proportion of total population living ln absolute poverty remained roughly constant at about 9%. The stagnation of poverty durlng 1985- 90, whlch contrasts the strong overall economlc growth of those years, is conslitent wLth the modest obuerved lncrease of rural Lncome disparltv. The poverty gap also Lncreased slLghtly durlng the second half of the ' -- lndlcatlng a modest deepenLng of poverty -- but remalned extremely 'e ;y developlng country standards. The summary poverty assessment fLguruu shown below document Chlna's impressive overall achlovements ln poverty allevlation sLnce the LnLtLation of economlc reform ln 1978: Total Populatlon (million) 963 1059 1143 Urban 172 (17.9%) 251 (23.7%) 302 (26.4%) Rural 790 (82.1%) 808 (76.3%) 841 (73.6%) Average Per CapLta Income (1978 Yuan) Urban -- 557 685 Rural 134 324 319 Poverty Line (Yuan/year) Urban -- 215 319 Rural 98 190 275 Ineidence of Poverty (million) \A Total 270 (28.0%) 97 (9.2%) 98 (8.6%) Urban 10 (4.4%) 1 (0.4%) 1 (0.4%) Rural 260 (33.0%) 96 (11.9%) 97 (11.5%) GLnL CoeffLcient Rural -- 0.28 0.29 Poverty Gap (Index) \k Rural -- 2.1 2.5 _________________________ \a SensLtlvlty analyses shows that these trends ln the LncLdence of poverty are robust to al: #easonabl- changes in the assumptLons underlylng the estimated poverty lines (see paras 2.05 and 2.06). \k The poverty gap li a measure of the additional lncome that would be necessary to brLng a poor person up to the poverty line. Summed across the poor populatlon and normallzed agalnst the poverty llne and total populatlon size, the poverty gap lndex' can be used to assess the relatlve depth of poverty among s*bpopulatLons, across time, or lnternatlonally.  vi (ii) Correlates of Poverty. Absolute poverty in China is almost entirely restricted to resource constrained remote uplaid areas -- it is extremely limited among the registered urban population and is not a problem of landless laborers. The government has effectively limited urban poverty through a system of full employment, linked to narrow ranges of earned income, crupled with heavy subsidization and price contrcls on basic foodstuffs and :ousing and near universal access to education and health care. The incidence )f urban poverty, which was already contained to very low levels by the late- 1970s, is estimated to have declined to less than 1% of the registered urban population by 1984 and to have remained at about that level thereafter. This optimistic portrayal, however, fails to reflect the status of the unregistered urban population. Those people, who number at least ten million, formally hold rural residence but work in urban areas without the job, food, health, and other rights normally accorded urban people. Most earn a decent living, but many dc not and all are extremely vulnerable if struck by illness, accident, or other calamity. By convention, the Chinese statistical system continues to treat these people as rural residents, therefore those among them in poverty are counted as rural poor, leaving the total estimates of poverty unchanged but shifting the apparent sectoral composition. (iii; Rural poverty is estimated to have declined from about one-third of the rural population in 1978 to roughly one-tenth by 1985. Since virtually A.l all of China's rural population received land use rights as part of the implementation of the production responsibility system during the early-1980s, there are few if any landless laborers. Instead, the majority of the rural poor are now concentrated in resource deficient areas, and comprise entire c -.anities located mostly in upland sections of the interior provinces of nor .hern, northwestern and southwestern China. Although these poor have land u rights, in most cases the land itself is of such low quality that it is not possible to achieve subsistence levels of crop production. Consequently, most poor are net consumers of grain and other subsistence foods, and are negatively affected by price increases for these products. The poorest households are typically those further disadvantaged by high dependency ratios, ill health and other difficulties. Minority peoples are known to represent a highly disproportionate share of the rural poor. Available evidence does not suggest that women are overrepresented among the poor, though poverty does certainly exacerbate society-wide problems of lower rates of female participation in education, higher relative female infant mortality rates, and higher rates of maternal mortality. (iv) The educational and health status of China's remaining absolute poor is deplorable. At least 50% of poor boys and, particularly in some minority areas, nearly 100% of poor girls do not attend school and will not achieve literacy. Infant and maternal mortality rates in very poor counties - - which exceed 10% and 0.3% respectively -- are at least 50% to 100% greater than the national average, and are much greater yet in the poorest townships and villages. Incidence of several infectious and endemic diseases, including tuberculosis and iodine deficiency disorders, is concentrated in poor and remote areas. Roughly 50% of children in households at or below the absolute poverty line are at least mildly malnourished (stunted), and iron, vitamin A,  vii and other micronutrient deficiencies remain a severe problem among the poor. As many as 90% of poor children suffer chronic helminthic infection. (v) Determinants _of Trends in Poverty. Broad participation in reform- driven agriculture sector growth played the key role in the tremendous two- thirds reduction in absolute poverty during 1978-84. Rural per capita income grew at an average annual rate of 15% in real terms during this periods and increased a total of more than 130%. The fail.re to achieve further reductions in poverty during the second half of the 19809, despite modest agricultural growth and very strong industrial growth, is more difficult to explain. A number of macroeconomic developmento stymied efforts to reduce poverty during 1985-90: (i) sharply increased prices for grain and other subsistence goods adversely affected the real incomes of the majority of the rural poor; (ii) rapid growth of the working age population exceeded the expansion of employment opportunities, contributing to a worsening of rural underemployment; and (iii) economic growth was greater in the higher income coastal provinces than in the lower income inland northwestern and southwestern provinces. In addition, fiscal decentralization passed an increasing share of the costs of rural social and relief services to local governments. While successful in better off areas, this fiscal reform put pressure on the limited revenues of poor area governments. In the absence of adequate levels of appropriately earmarked funding from higher levels of government, poor area local governments have been unable to support either adequate social services or economic growth. (vi) The key determinant of the stagnation of poverty during 1985-90, however, was the absence of meaningful levels of agricultural growth and rural enterprise development in the upland areas. Many of the rural poor in 1978 resided in less remote and less hilly areas, where increased application of fertilizer, better seed and other modern inputs could bring about rapid productivity gains, and so were able to participate in the rapid agricultural growth of 1978-84. However, the quick reductions of poverty through agricultural growth were largely exhausted by end-1984. Most of the residual poor have remained trapped in more remote upland areas where agricultural productivity gains have proven far more problematic. Measured on a per capita basis, output of grain and subsistence foods in such areas failed to sustain any significant increase during the 1980s. Although the agriculture sector did expand in real terms during 198S-89, increased output of nongrain crops and animal and aquatic products, products which few of the poor either produce or consume in significant quantities, accounted for more than all of the modest growth which did occur. By comparison, the annual growth rate of per capita production of grain and oilseeds -- which are the subsistence crops of most immediate importance to the poor -- declined from 2.8% and 14.5% respectively during 1978-84 to -0.4% and -1.6% during 1985-89. (vii) Rural enterprises, an important source of employment in the rural economy as a whole, have developed very slowly in poor areas. In the early 1980s such employment actually decreased in poor areas as the commune system was dismantled and workers could no longer be paid in work points instead of cash. Employment rose only after a 1984 policy initiative supporting private rural enterprise opened opportunities for small, family firms particularly  viii suited to the small market niches available in poor aiaao. Despite private enterprise gains, by 1990 only 4% of the rural labor force in China's 120 pooreot counties had found employment in rural enterprises, in contrast to the 22.1% finding such employment in the nation as a whole. Not only are fewer employed, but rural enterprise wages and profits are lower in poor areas than the rest of the country. Rural enterprise growth will continue to be concentrated in the wealthier areas with inexpensive access to national and international markets. If the poor are to sAhare in that growth, It will be largely throubn migration to take advantage of employment opportunities where they occur. (viii) Government's Poverty Alleviation Program. The central and provincial governments sustained their strong commitment to poverty alleviation efforts throughout the 1978-90 economic reform period. While continuing pre-existing rural social and relief services, the poverty alleviation strategy adopted during the Seventh Five Year Plan introduced a new emphasis on economic development programs in the poor areas. The Ministry of Civil Affairs provides disaster relief and income maintenance support, and coordinates the distribution of relief grain with the Ministry of Commerce's Grain Bureau. The State Education Commission and the Ministry of Public Health administer some special programs to improve the education and health status of the poor. The Agricultural Bank of China and several other banks offer subsidized loans for poor area development through a variety of funds administered by provincial bank branches and their networks of county and lower level banks. The Regional Office of the State Planning Commission administers a Food-for-Work Program which assists with the building of roads and riverine transport, drinking water systems, irrigation works and other capital construction in poor areas. In addition, each of 27 central ministries and agencies has its own special poor area project and every province has i*s own specially-funded programs. The Leading Group for the Economic Development of Poor Areas (LGEDPA) was established in 1986 in part to provide greater coherence to these many poverty alleviation initiatives and, in particular, to expedite economic development in the poor areas. Since its establishment, LGDPA has emerged as the principal advocate of China'a rural poor and iA now the key agency responsible for coordinating the nation's more than rY billion in annual funding for poverty alleviation programs. -Strategies for the 1990. (ix) Achieving further reductions in the number of absolute poor will prove a more difficult challenge than in the early-1980s. However, the transition and significant reduction of absolute poverty, from large numbers of poor spread widely across the countryside in the late-1970s to pockets of poverty In remote resource deficient areas by the mid-1980s, makes it easier and fiscally lese burdensome to target increased development and social services -sfsiatance to the absolute poor. Most importantly, in order to guarantee a minimum safety net while improving the productivity of the poor over the longer term, revitalized social services must be integrated with improved agriculture and rural enterprise development programs. In order to realize the poverty alleviation potential of the expected reduction of rural  ix underemployment during the 1990o, it is also essential that all opportunities for expanding out-migration of surplus labor from thn poor areas be explored. (x) Economic Reform and Growth. Steady adoption of a range of economic reforms is expected to sustain overall growth of the Chinese economy at about 7% per annum during the 1990. (World Bank, 1992b). While this is undoubtedly essential to further reductions in poverty, the analysiG of this report shows that growth alone would not be sufficient to achieve desired levels o9 poverty reduction. Reducing poverty requires that economic growth be coupled with expanded and new social service and rural development programs directed to the poor. Fortunately, the additional funding demands for these programs come at a time when planned fiscal and price reforms promise to cut the central government deficit and, over the medium term, free resources that will allow government to match with a financi. commitment its moral commitment to poverty alleviation. Price refoy J cluding the expected decontrol of distorted raw material and agric. :..al prices and, in particular, the further liberalization of grain prices, need not adversely affect the real incomes or welfare of either the urban or rural poor. Though the reform of grain prices in recent years was associated with some erosion of transfers to the rural poor, it is unlikely that completing grain price reform would further erode such transfers since the poor now pay the equivalent of market prices for any grain they need purchase (in excess of relief grain provided to them by government relief agencies at no cost). Refozz of state-owned entezprise, another priority for the 1990. with important impLications for the poor, could be associated with some increase in unemployment in cities where there is a large concentration of state enterprises. However, if should be possible to maintain the welfare of the urban population at current levels without incurring additional social welfare cc3ts by (i) shifting welfare functions from urban enterprises to specialized government agencies and (ii) making adequate provision, in advance, for unemployment insurance. On balance, the compintion of the economic reform ag;enda should prove consistent with. and even hslp finance, the realization of poverty alleviation objectives. (xi) Poverty Alleviation under the 8th Five Year Plan. Two key poverty alleviation initiatives introduced under the government's Eighth Five Year Plan (1991-95) are (i) the extension and strengthening of assistance to the poorest of the poor residing in the worst physicai environments and (ii) the integration of production, education, health, family planning and transport programs into comprehensive local intervention packages. The first initiative will be supported by individual programs tailored to the special needs of minorities, communities residing in remote, high altitude, and karst areas, aru.i other disadvantaged groups. Other initiatives include (i) an additional Y5u%O million in annual support for the roughly 200 counties newly included to -.he national roster of poor counties, (ii) expansion of China's Food-for-Work uogqran, with the bulk of incremental assistance directed to construction of 6erraced agricultural lands, and (iii) increased support for agricultural eytension and training end farmers' marketing systems in the poor areas. (xii) Demoaraohic Trends and Labor Mobility. Pronounced demographic changes are expected to enhance employment and income prospects during the 1990s. Driven by fertility trends of the 19608 and 70s, annual growth of the  a total working age population is projected to decline by more than half from 3.1% during 1978-90 to 1.3% during 1990-95 and then to 1.1% during 1995-2000. Employment growth is also expected to drop sharply to about 1.4% annually during the 1990s. Even if urban enterprise reform induces efficiency gains sufficient to reduce employment growth below these relatively modest expectations, increases in la'bor demand would still exceed the slow expansion of labor suppl'. Contrary to government projections of a very sharp increase, rural underemployment should therefore subside scomewhat during the 1990s. The early succesrb of family planning measures in urban areas in the 1970s, in particular, 'aa already resulted in a sharp reduction in the number of urban- born entrants to the urban work force. A steady stream of migrants from rural areas will therefore be needed during the 1990s to offset the declining numbers of urban-born work force entranta and replace large numbers of retiring urban workers. (xiii) The greatest potential for the absorption of surplus rural workers is believed to be in the multitude of small and medium sized cities and urban towns. Those urban areas have experienced rapid development in recent years, often fostered by a relaxation of controls on in-migration by rural people. While many urban jobs require a middle school education, construction and service sector jobs often do not. Those jobs, and peri-urban agriculture, provide employment opportunities to the poor, who are typically less well educated. Realizing the potential for improved employment and income prospects for the poor will require maNimizing the range of economic activities available to them, both on and off the farm, within and, in particular, outside the poor areas. Efforts to facilitate the out-migration o: excess labor from the poor areas could play a key rnle in translating this potential into reality. (xiv) Recommended Strateoies for the 1990s, Meeting the challenge of poverty reduction in the 1990s will, in addition to overall economic reform, require: (i) investment in the development of human capital, including greater central government funding for education, health and relief services in the poorest areas; (ii) institution building and policy reform, including a strengthening of the institutions responsible for implementing explicit poverty alleviation progr&ms, the establishment of an independent and objective poverty monitoring system, and an improvement in the access of the poor to employment opportunities outside the poor areas; and, (iii) infrastructure development, with poor area development assistance programs revised to include only those agricul'ural, rural enterprise, road and other rural infrastructure development projects which generate reasonable market-determined returns to investment. More specifically, the government should undertake the follow'.ng initiatives during the 1990s:  xi Increase Fundino for Social Services: Now that the incidence of absolute poverty has been reduced to lese than 10% of total population, the government is in the position to address the most critical education, health, and relief needs of the poor at acceptable costs. Recent experimental programs to improve access of the poor, including poor minorities and females, to education and health services provide strong evidence that effective interventions exist at a modest per capita cost. Annualized incremental funding requirements for programs to improve education and health services and status in the poor areas are estimated to be about Y2 billion. Lith this, it shoull be possible to universalize six-year primary education for 15 million poor children and to reduce currently excessive levels of infant, child and maternal morbidity and mortality in the poor areas by at least one-third. Failure to adopt such programs will preclude realization of the government's goals of universal nine-year primary education and "Health- For-All in 2000." Incremental funding for the programs, which must be provided by the central and provincial governments, could be partially defrayed through efficiency gains in planned expenditure on education during the 1990s. Even at full cost, however, incremental funding requirements would represent only about 0.5% of current total government expenditure and only 4% of the government's fiscal losses on consumer food subsidies it now provides almost exclusively to the well-off registered urban population; R Strenathen Poverty Alleviation Agencies: The government should recognize that sustaining the effectiveness of poverty reduction efforts during the 1990s and beyond vill require a permanent and much stronger institutional structure. Accordingly, the permanent status of the LGEDPA system should be formalized and its central, provincial and local offices expanded to in-lude staff trained in applied poverty research and assessment, project appraisal, supervision and monitoring, and, consistent with its mandate to coordinate all poverty alleviation programs, education and health. The LGEDPA should also take a more active role in facilitating the movement of surplus labor out of the poor areas, and new staff should be appointed to research and implement means to this end; 1/ Although education, health and welfare standards are nationally mandated, social services rely primarily on local finance. This situation leaves program availability and quality much lower in poor than well-off rural areas. Relief programs in the poorest villages are capable of sustaining life, but only at the most minimal levels -- principally or solely through the provision of relief grain. Additional resources must be made available if adequate education, health, and other social consumption by the poorest and most vulnerable groups is to occur. Some of those resources can be found through efficiencies to be gained in, or reorienting of, existing programs, but the bulk must come from increased central government transfers to poor areas.  xii uParade Poverty Monitorina: Currently, the LGEDPA has principal responsibility for both monitoring absolute poverty at the national and provincial levels and coordinating the implementation of programs to reduce poverty. This conflict of interest is associated with an optimistic bias in the LOEDPA's estimates of the incidence of poverty. More importantly, the LGEDPA's assessment of the location of poverty indicates that nearly two-third. of the absolute poor reside in southwestern China -- a finding which is inconsistent with more reliable information for the la'.e-1980s from the State Statistical Bureau's independent large-scale rural household income and expenditure surveys. The LGEDPA's assessment, which is based on Ministry of Agriculture county-average rural income data, plays a key role in determining the focus of the nation's poverty alleviation effort and targeting of government assistance. X It i8 essential that an independent and objective poverty monitoring system be established and the true location of the poor be accurately determined as soon as possible. Fortunately, the State Statistical Bureau has both the necessary data and skills to undertake objective monitoring of the incidence of poverty; Imorove Labor Mobility: Obstacles to rural labor mobility, including resistance by county and municipal governments, inadequate availability cf labor market information, and social and cultural prohibitions, have heretofore limited intercounty and interprovincial out-migration of surplus labor -- particularly from the poor areas. County governments have used admninistrative and punitive economic measures to impede the inflow of labor from lower-income areas in order to maximize local employment and incomes. Similarly, the household registration system and the urban qcain ration system have been used to limit rural-to-urban migration in order to keep urban unemployment below planned targets. These obstacles notwithstanding, organized and unorganized labor migration did increase somewhat during the 1980s, and demographic trends and government policy favor a continuation of that increase during the 1990s. Favorable experience with provincial labor export programs during the 1980s has documented the feasibility of improving labor mobility, particularly from the poor areas. The State Council also recently approved 22 experimental labor export and mobility projects, lending high level support to efforts to expand and improve rural labor mobility. Consistent with overall plAns to deconitrol the labor market and remove remaining barriers to labor mobility, the LGEDPA should take an active role In facilitating a market friendly increase in out- migration of surplus labor from the poor areas to better off rural and urban areaos and, Foster Rural EnterDrLse and Agricultural DeveloDment in the Poor Areas: Small-scale private enterprise (PE) appears to be best suited to the underdeveloped markets of the poor areas. Government initiatives to augment the development and poverty alleviation impact of rural 2.1 The State Statistical Bureau sample survey data, by comparison, indicate that at least half of the poor reside in northwestern and northern China.  xiii enterprise in the poor areas have included the introduction of appropriate technologies, the subcontracting of simple manufacturing or production activities to poor households and villages, and the provision of subsidized credit to enterprises which agree to employ a specified minimum number of poor. The recently enacted Regulations on Rural Collective Enterprises, which seek to limit local government's excessive extractions from collective enterprises and I. otect the rights and interests of such enterprise, are also of particular importance to creating a favorable environment for poor area rural enterprise development. v The sheer numbers of the rural poor make it inevitable that most will continue to reside in remote mountainous lands during the 1990s and be forced to rely primarily upon agricultural production on those lands for their subsistence. Therefore, every effort must be made to take advantage of opportunities to augment productivity in poor area agriculture. Though they are generally not easy to find nor necessarily suitable for large- scale government support, real opportunities for improving agricultural productivity and incomes in the upland areas do exist. While increasing the incomes and productivity of poor families and poor areas is a laudable objective, the government must recognize the limits to the effectiveness and efficiency of assisting the poor through rural enterprise and agricultural development projects within the poor areas. (xv) Agenda for International Assistance. International assistance can play an important role in China's poverty alleviation program through support for operational research and rural development programs. Operational research priorities, to both evaluate and improve the effectiveness of China's poverty alleviation program, include (i) the establishment of an independent poverty monitoring system, (ii) the development of effective programs of education and health interventions for the poor areas, (iii) an examination of measures to augment out-migration from the poor areas, (iv) a careful assessment of explicit poverty alleviation programs, and (v) an ac_urate appraisal of the size, composition and well-being of the growing unregistered urban and rural migrant population (para ii). First, the poverty monitoring system (para xiv), once established, should be extended to include education, health and other indicators of welfare. Monitoring should also be extended to cover the unregistered urban population and migrant laborers more generally. It is essential that national and provincial poverty assessments be validated through detailed field checks in a namber of villages around the country on a regular basis. Second, the most efficient and effective programs of education, health and nutrition interventions for the poor areas must be identified through careful experimentation. Extension to progzessively wider segments of the poor population would be undertaken as these intervention packages are refined on the basis of demonstrated cost-effectiveness and popular acceptance. Third, the remaining obstacles to poor area labor mobility must be assessed and the effectiveness of the 22 experimental labor export and mobility projects evaluated (para xiv). It is also important that A/ Similar regulations covering private enterprises have now been separately issued.  xiv the 1986 Chinese Academy of Social Sciences survey (of 222 villages and 11 provinces), which provides the best documentation of rural labor transformation and mobility, be updated and expanded. Fourth, there is a need to undertake a detailed examination of the poverty alleviation impact and economic viability of China's various poverty alleviation programs. Initially, attention could be directed to the State Planning Commission's Food-for-Work Program and an evaluation of relative benefits to investment in road and riverine transport, drinking water and sanitation systems, small- scale irrigation and soil conservation works, primary schools and village health stations, and other capital construction (para viii). Last, there is an urgent need for an improved understanding of the well-being of the migrant population. Though most migrants may enjoy improved living standards, there are signs that a significant number are not faring well. (xvi) In balance with the pricrity attached to combating poverty, international assistance for rural development programs should be focussed on the inland regions where most of China's poor reside. Until now, however, relatively little international assistance has reached the inland poor. Greater efforts could therefore be made to target international assistance to poor provinces and, while recognizing the inherent difficulties and costs to reaching small communities situated in distant areas, to poor counties and even poor townships and villages. Assistance for education, health and other social sector development programs should be concentrated primarily in the poorer rural areas where achieving and maintaining minimal welfare standards is most at risk. A substantial share of agricultural development programs should be directed to poor upland areas to resolve the interrelated problems of low productivity and environmental degradation. -Reader's Guide (xvii) This report provides an analysis of the incidence of absolute poverty during the 1978-90 reform period (Chapter 2 and Annex 1) and proposes macroeconomic and sectoral strategies (consistent with the overall reform agenda) necessary to further reductions of poverty in the 1990s. The macroeconomic developments which helped determine the reduction of poverty during 1978-84, and its stagnation thereafter, are reviewed in Chapter 1. Trends in labc, supply and demand during 1978-90 and expected during the 1990s are examined in Chapter 3. Strategies for augmenting labor mobility, rural enterprise development, and agricultural productivity in the uplands are also explored. Education, health and relief services in the poor areas are discussed in Chapter 4. Financ3al support for these services during the 1980s is analyzed, and a program to bolster education and health in the poor areas during the 1990s is proposed. The government's poverty alleviation programs during the 1980s and under the Eighth Five Year Plan are described and evaluated in Chapter 5. The report is supported by the following five working papers (available upon request): "Demographic Aspects of Poverty," nA Program to improve Health Services in the Poor Areas," "Economic Analysis of the Loess Plateau Land Rehabilitation Program," "Poverty in Southwest China" and "Rural Poverty: Characteristics, Causes and Policies in Baise Prefecture." -- -II I I I 1  1. POVERTY ALLEVIATrON AND MACROECONOMIC TRENDS IN TEE 19803 A. Absolute Poverty 1.01 Overview. Both as an accompaniment of overall economic growth and also through a strong commitment to improving standards of education, health and nutrition for all, China has compiled an impressive record in reducing absolute poverty over the last four decades. Perhaps the most telling indicator of the improvement in the overall well-being of the Chinese population is the increase in expectation of life at birth, which nearly doubled from 34 years in the early-1950s to 67 years at present. Other indicators of social development, reported in Table 1.1, also document tremendous improvement. The crude death rate and infant mortality both declined by about three-quarters since the early-1950s, and illiteracy is estimated to have dropped from 80% of the adult population in the early-1950s to about 30% at present. Average per capita availability of food energy increased by nearly 40%, from about 90% of food energy requirements in the Table 1.1: Selected Indicators of Social Develomment: China and Other Countrios Infant Expectation Crude Mortality Per Capita of Life Death Rate Adult Availability GNP at Birth Rate \a (under age 1) Illiteracy of Food Energy Per Capita ........... ........ . ...... ------ ----- .......... . .. -------------- ----------................................ ............... ..... Country/Region years per 1000 per 1000 X calories/day 1988 USS ..................................................................................................... China: 1988 67 8 58 31 2637 330 1950-55 34 31 236 80 1894 na India: 1988 56 12 91 57 2104 340 Indonesia: 1988 59 11 79 26 2631 440 Brazil: 1988 64 9 71 22 2703 2,160 1988 Average: Low Income Countries 60 10 72 44 2442 \b 320 Middle Income Countries 66 8 52 26 1,930 ..................................................................................................... Sources: World Bank (1990a) for adult illiteracy, GNP, and averages for low and middle income countries, FAO for food energy, Banister (1992) country-specific data for 1988, and World Bank (1984 and 1991a) and Piazza for figures for China for 1950-55. \g Figures for China, India, Indonesia and Brazil for 1988 are age standardized. \k Average for all developing countries (including China).  2 early-1950o to 115% by the late-1980s. These indicators of social development in China compare favorably with the estimates shown in Table 1.1 for India, Indonesia, Brazil, and the averages for other low income and middle income countries. 1.02 Urban Poverty. " The Chinese government has limited urban poverty to extremely low levels through full employment linked to narrow ranges of earned income, price controls on basic foodstuffs, and near universal access to education and health care. Full employment has been maintained by severely restricting migration to urban areas, a restriction strengthened until the mid-1980s by the lack of a rental market for housing and the requirement that grain coupons be used when purchasing either grain or restaurant meals. The urban food ration system has insured that Chinafs urban population enjoys preferential access to staple foods, especially grain and vegetable oil, at below market prices. Urban enrollment ratios at the primary and secondary school levels have risen faster and are now higher than those of most other developing countries. Widespread availability of primary health services, provided at little or no cost, brought the most serious contagious and parasitic diseases under control in urban areas by the early-1960s. At present, the urban population exhibits few signs of absolute poverty and enjoys more than twice the per capita income, significantly better access to education and health services, and greater levels of food intake than do rural inhabitants. 1.03 Rural Poverty. In rural areas, the redistribution of productive ausets, adoption of a basic human needs program, and greatly inereased food production have been of key importance in reducing absolute poverty. In the early-1950s, the National Land Reform redistributed large amounts of arable land, livestock, and other farm assets, and established the basis for significantly increased income among (previously) landless laborers and other elements of the rural poor. The basic human needs program adopted in the 19S0s sought to guarantee floor levels of food consumption, income and health services adequate to maintain minimum levels of physical well-being. The objective of the rural relief component of the basic human needs program was to prevent outbreaks of starvation or severe malnutrition, and to guarantee the welfare of the elderly. Rural relief was for the most part funded locally, and regio.ial or state funding was made available only in cases where local poverty was so severe and widespread that it precluded the possibility of local self-reliance. The rural primary health care component of the basic human needs program, which has included support for childhood immunization, accessible primary health care, infectious disease control, sanitation and safe water supply, became increasingly effective beginning in the late-1950s 1/ In this report, China's urban population is defined to include only (i) the registered population of city and town neighborhood committees and (ii) the unregistered long term (that is, one year or longer) residents of those neighborhood committees (para 1.19). The "floating" population of unregistered migrants is considered part of the rural population for the purpose of estimating the incidence of absolute poverty.  3 and contributed to significant reductions in morbidity and mortality during the 1960s and 70s. 1.04 The positive impact of these rural programs has been complemented by large increases in national average food availability and, especially in the grain deficient provinces, by increased grain production. Increased agricultural production helped push up average per capita daily food energy and protein availability by 35% and 25% respectively, from less than 1900 kcal of food energy and 53 grams of protein in 1950-52 to 2570 kcal and 66 grams in 1980-82. An analysis of provincial per capita grain production shows that seven provinces (Hebei, Henan, Jiangsu, Liaoning, Shandong, Shanxi, and Qinghai), with a total population of 200 million, were grain deficient (less than 275 kg of unprocessed grain production per capita) in 1953/57. By comparison, only five province. (Gansu, Guizhou, Qinghai, Tibet, and Yunnan), with a population of 85 million, were grain deficient in 1979/82. Overall, the number of people residing in grain deficient provinces declined from 35% to less than 10% of total population (Piazza). The improvement in the physical well-being of the rural population, however, was punctuated by severe dislocations. The most striking failure occurred during the agricultural crisis of 1960-61, which brought about an estimated lose of life of between 16 to 30 million excess deaths (Coale and Ashton. et al). These serious difficulties notwithstanding, it is certain that China's rural population in the early-1980s enjoyed much higher standards of health and nutrition than they did in the early-1950s. Overall, China's efforts to reduce the number of poor and to ameliorate the most extreme manifestations of residual poverty must be judged to have been highly successful. 1.05 Poverty Durina 1928.90. This report's estimates of the incidence of absolute poverty in Ch.-a, which are presented in Table 1.2 and detailed in Annex 1, show that the number of poor declined sharply from about 265 million in 1978 to 97 million in 1985, or from about 28% to less than 10% of total population. However, the reduction in poverty was not sustained during the second half of the 1980s -- the number of poor stagnated between 86 to 103 million during 1985-90 and the proportion of total population living in absolute poverty remained roughly constant. These temporal trends in the incidence of absolute poverty have been matched by trends in per capita availability of grain and average levels of nutrition. Availability of grain increased steadily from 305 kg per capita in 1978 to 401 kg in 1984, then dipped to between 359 kg to 385 kg during 1985-89 before recovering to 400 kg in 1990. E The overall reduction An poverty ainco 1978 was also matched by appreclable improvements in ezpectation of life and reductions in mortality (since China had achieved fairly low levels of mortality by the late-1970s, continued sharp declines in the 1980. could not be expected). Data from the 1990 census indicate that survival chances have either improved or stayed the same since 1981 for males and females at almost all ages in rural as well as urban areaa. The 1988 nationwide fertility survey suggests that infant mortality for both sexes was approximately constant from 1977 to 1987, though VI Figures are calculated as domestic production net of international grain imports and exports.  Table 1.2: Macroeconomic Indicators and Estimated Incidence of Absolute Povery Year 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 ~~~~~~~~~~. .. . .. .............. .. . ............ .... ...... .... ...... ... .. ... _................................. 197G 100 100 108 116 121 132 145 167 188 203 226 251 261 274 Anmual Growth Rate (X) 7.6X 7.9X 4.4K 8.8X 10.4X 14.7K 12.8K 8.1X 10.9X 11.0X 4.0K 5.2K Per Capita Grain AvailabiLity (kg): 326 354 338 340 367 391 401 359 368 385 368 378 400 Retail Prices (K annual change): National Average 0.7K 2.0K 6.0K 2.4K 1.9X 1.5K 2.8K 8.8X 6.0X 7.3K 18.5K 17.8X 2.1K Urban Call food) 2.5X 1.9X 8.1K 2.7K 2.1K 3.7K 4.0X 16.5K 7.2K 12.0X 25.2K 14.4% -1.2X (grain) 2.0X -0.5K 0.2X 0.7K -0.8X 2.1K 0.0X 3.3K 4.1X 6.2X 14.1K 16.9K -6.5K Rural (all food) 0.3K 3.8X 7.1K 3.0K 4.4K 0.6K 1.4K 12.5K 7.5K 8.4K 20.9K 18.0X 1.7X (grain) 0.2K 5.7X 7.1K 1.1K 3.5K 0.2K -0.4K 17.6K 13.0K 6.1X 14.0K 25.0X -3.3K Per CaWita Income (Yuan): Urban (current Yuan) -- -- 500 535 573 660 749 910 1012 1192 1388 1523 Rural (current Yuan) 134 160 191 223 270 310 355 398 424 463 545 602 630 Urban (978 Yuan) )a -- -- 446 467 490 550 557 633 647 632 632 685 Rural (1978 Yuan) \b 134 157 180 202 241 274 311 324 325 335 336 313 319 writ iAgetPopultin 485 -- - -- 567 583 602 621 641 656 670 684 697 Total E owayment 402 410 424 437 453 464 482 499 513 528 543 553 567 Urban Economy d 95 100 105 111 114 117 122 128 133 138 143 144 147 Rural Economy \e 306 310 318 327 339 347 360 371 380 390 401 409 420 -Agriculture \ 275 278 283 290 301 304 301 304 305 309 315 324 333 -Nonagriculture \ 31 32 35 37 38 44 59 67 75 81 86 85 87 C-Rural Enterprise \_ 28 29 30 30 31 32 52 70 79 88 95 94 93) Estimated Incidence of Absolute Poverw: Registered Urban (milttion peopte) -- -- -- 4 2 1 1 1 1 1 1 1 1 (K of urban population) -- -- -- 1.9K 0.9K 0.6K 0.3K 0.4K 0.2X 0.2X 0.2 0.3K 0.4K Rural (million people) 260 -- 218 194 140 123 89 96 97 91 86 103 97 (K of rural population) 33.0X -- 27.6K 24.3K 17.4K 15.2X 11.0K 11.9X 11.9K 11.1K 10.4K 12.3K 11.5K Total (million people) -- -- -- 198 142 125 90 97 97 91 86 103 98 (K of total population) -- - -- 19.8K 13.9X 12.0K 8.6K 9.2X 9.0X 8.3K 7.8X 9.2X 8.6K Source: SSB (1991a) for GNP, prices, income and employment, Uorld Bank (1991b) for grain, and Annex 1 for poverty estimates. \a De7Ttted by the urban cost of living index. Deftated by the rural retail consumer price index. SE Officially defined as wcmen ages 16 to 54 and men ages 16 to 59 (exctuding military, prisoners and the disaibled). Includes urban staff and workers and urban individual Laborers in urban and (whiLe they retain their urban residency registration) rural areas. 'W Excludes urban staff and workers employed in rural agriculture and rural nonagriculture. Rural enterprise enployment, including some part time workers. A definitional change, to inctude all kinds, of rural enterprise, explains part of the sharp increase in empLoyment in 1984.  5 the rising sex ratios observed for cohorts born in the 1980e strongly suggest -- to the contrary -- that female infant mortality ha. been rising (Banister, 1992). 1.06 There are important differences in the incidence of poverty between sectors and regions. Absolute poverty amongst the urban population, which was already at extremely low levels by the late-1970s, is estimated to have declined from about 1.9% of the registered urban population in 1981 to only 0.3% by 1984, or from about 4 million to only 1 million urban residents, and to have remained at about that level throughout 1984-90. The number of rural poor is estimated to have declined from 260 million in 1978 to about 100 million by 1985, or from 33% to 12% of total rural population. The incidence of rural poverty then remained roughly constant at between 10.4% to 12.3% of the rural population during 1985-90, ranging between 86 to 103 million poor. Most of China's remaining absolute poverty is now concentrated in a number of resource-poor rural areas, primarily in the northern, northwestern and southwestern provinces. With the exception of a growing residual of households, disadvantaged by physical or mental disabilities, age structure, ill health or other factors, who have not fared well in the transition to a more market oriented economy, absolute poverty has been largely eliminated from better endowed rural areas. B. Impact of Macroeconomic Trends -Economic Reform and Growth 1.07 Economic Reforms. Leveraging investments in agriculture and industry made during the pre-reform period, economic reforms initiated in the late-1970s led to rapid economic growth and structural change during the 1980s. E Rural economic reforms, including the adoption of the production responsibility system (PRS), the dismantling of the commune system, agricultural product price increases and market liberalization, were key elements of the first stage of economic reform during 1978-84. Industrial reforms initiated in 1983, including the adoption of profitability and retained earnings as principal management criteria and the introduction of an enterprise profit tax, allowed enterprises more latitude in the disposition of resources and permitted greater retention of profits. Enterprise reform was furthered in late-1984 with the introduction of management contracting and increased autonomy for rural enterprise. Increased investment demand was accommodated through expanded credit as newly established specialized state banks began making long-term enterprise investment loans. 1.08 Growth. These reforms helped China sustain high rates of real economic growth throughout most of the 1978-90 reform period. Official figures (Table 1.2) indicate that real GNP increased by 174%, at an average 3/ Economic performance and reforms during 1978-88 are discussed in detail in World Bank (1990b).  6 annual rate of nearly 9%, during 1978-90. As summarized below, significant mid-decade shifts in sectoral growth trends underlie this aggregate growths Averae Annual Growth Rate.: OMP and Agriculture \a Gross Value of Agricultural GNP Output Per Capita Production ---------------~~~------ -------------- (physical output) Agri- Non- -- Total culture Industry Crops Crops Grain Oilseeds Meat ---------------------------------------------------------------__------------ 1978-89 9.2% 5.4% 10.7% 4.5% 9.4% 1.3% 7.9% 7.6% 1978-84 9.1% 6.6% 9.7% 6.2% 9.6% 2.8% 14.5% 8.2% 1985-89 8.9% 3.5% 11.2% 2.4% 8.2% -0.4% -1.6% 6.1% Source: SSB (1990a). \a Average annual real growth based on three year averages and constant 1980 prices. Spurred by the rural economic reforms, the agriculture sector enjoyed real average annual growth of 6.6% during 1978-84. Average annual growth in the sector declined to 3.5% during 1985-89, or 1.9% annual growth in per capita terms. Within agriculture, the growth rate of the gross value of crop production slowed from 6.2% during 1978-84 to 2.4% during 1985-89. The growth rate of the gross value of non-crop output (including animal husbandry, forestry, fisheries and sideline production), however, declined only modestly from 9.6% to 8.2% during these same periods. The impact of industrial reform, by comparison, was felt several years after that of the rural reforms. Average annual industrial growth accelerated from 7.8% during 1978-83 to about 14% during 1983-88. offsetting the slowdown in agriculture sector growth, this strong industrial growth helped drive a surge in GNP growth from 7.7% during 1978-83 to more than 11% annually during 1983-88. 1.09 Broad participation in zeform-driven agriculture sector growth played the key role in the tremndous two-thJrds reduction in absolute poverty during 1978-8E. Rural per capita income grew at an average annual rate of 15% in real terms during this period, and increased by a total of more than 130%. In contrast, modest, but still positive, agricultural growth in combination with strong industrial growth was not associated with any significant further reductions in poverty during the second half of the 1980g. The inability to reduce poverty during 1985-90 is in part explained by mid-decade changes in subsectoral growth in agriculture. The average annual growth rate of per capita production of grain and oilseede -- which are the subsistence crops of most immediate importance to the poor -- declined from 2.8% and 14.5% respectively during 1978-84 to -0.4% and -1.6% during 1985-89 (see figures in para. 1.08). Increased output of other products which few of the poor either produce or consume in significant quantities, including nongrain crops and  7 animal and aquatic products, accounted for more than all of net agricultural output growth during 1985-89. A number of other macroeconomic developments, discussed in the remaining eactions of this chapter, also help explain the failure to reduce poverty despite strong overall economic growth during 1985- 90: (i) sharply increased prices for grain and other subsiatence goods adversely affected the real incomes of the rural poor; (ii) rapid growth of the working age population exceeded the expansion of employment opportunities, contributing to a worsening of rural underemployment; and (iii) economic growth was greater ln the higher income coastal provinces than in the lower income inland northwestern and southwestern provinces. In addition, fiscal decentralization passed an increasing share of the costs of rural social and relief services to local governments. While successful in better off areas, fiscal reform pressured the limited revenues of poor area governments. In the absence of meaningful levels of funding from higher levels of government, poor area local governments have been unable to support either adequate social services or economic growth. 1.10 An equally important determinant of the stagnation of poverty during 1985-90, discussed in detail in Chapter 3, is that the quick and saxy reductions of poverty through agricultural growth were exhausted by end-1984. Many of the rural poor in 1978 resided in less remote and less hilly areas, where increased application of fertilizer, better seed and other modern inputs could bring about rapid productivity gains, and so were able to participate in the rapid agricultural growth of 1978-84. By 1985, however, most of the remaining poor were trapped in more remote upland areas where agricultural productivity gaino are much more problematic. -Prices and Income 1.11 Though achieving unprecedented rates of national income growth throughout the 1980s, policy makers failed to adequately guard against inflation. Excessive credit expansion, and the failure to sufficiently augment supply responsiveness through either national market integration or greater competition, contributed to price spirals first in 1985 and then again beginning in late-1987. As shown in Table 1.2, the growth rate of national average retail prices increased from an annual average of less than 3% during 1979-84 to nearly 9% in 1985, declined moderately in 1986-87, and then surged to about 18% in 1988 and 1989. The acceleration of inflation during 1985-89 adverr dy affected the rural poor, both because it included sharp increases in the prices of grain and other subsistence foods and by necessitating adoptioun of anti-inflationary austerity programs which brought about a retrenchment of off-farm jobs in rural enterprises and the urban economy during 1988-89. 1.12 Prices. Price changes during the 1980s significantly affected the poor, first by improving their terms of trade through 1984 and then by worsening them in critical dimensions through the rest of the decade. As shown below, the growth rate of national average retail prices, which averaged about 7% per annum during 1978-89, increased from an annual average of les than 3% during 1979-84 to more than 12% during 1985-89:  Price !ndice:l Average Annual Chanoa (*) Rural Urban Retail \^ Rural Retail \& Free Market All -------------- ----------- - Farm Retail Total Food Grain Total Food Grain Total Grain Procurement 1978-89 6.7% 7.4% 8.7% 4.1% 6.2% 7.8% 8.2% 7.8% 5.6% 9.9% 1978-84 2.8% 3.1% 3.7% 0.3% 2.3% 3.4% 2.8% 1.2% -4.9% 7.4% 1985-89 11.6% 13.1% 14.5% 10.2% 11.9% 13.5% 14.3% 16.6% 24.8% 13.9% ---------------------------------- Sources sSB (1990a). \a Consumer retail prices. Overall, the increase in urban consumid; prices modestly exceeded that in rural areas. Food prices also increased mot.a r:'idly in urban areas. However, urban consumers have been insulatee by grain oubsidies (para 1.02), and retail grain prices in urban areas have increased at a significantly lesser rate than in rural areas. 1.13 Prior to 1985, the rural population benefitted from average annual increases in farm procurement price. of more than 7%, or more than twice the growth rate of rural consumer prices during 1978-84. Prices in the rural free markets, which are "prices of last resort" for the rural poor, increased at the very modest annual rate of only 1% during 1978-84 -- and grain pricen in the rural free warket8 decliAned by about 5% annually during the period. The bumper grain harvest of 1984 depressed free market prices for grain and other staple foods to their lowest level of the decade. During the second half of the 1980s, however, rural consumer prices increased at a rate more closely approximating the growth of farm procurement prices. Increased demand for feed grain, in the context of reduced total availability of grain (para 1.05), tightened supplies, and rural grain prices increased at a rate greater than overall rural inflation during 1985-89 -- and jumped 18% and 25% in 1985 and 1989 respectively. The growth rate of the rural free market price of grain, in particular, increased at an average annual rate of 25% during 1985-89 -- or more than twice the rate of overall rural inflation. Most of the rural poor are net purchasers of food grain, and were adversely affected by increased prices of staple foods. 9 The rural poor suffered an additional price shock in 1985, furthermore, when rural "resales" of grain were increased from the old quota price to the new contract price (World Bank, 1990d). Those poor not entitled to receive sufficient quantities of free relief grain paid an 4/ However, a significant proportion of the poor are surplus producers of grain. Zhanai (1992), for example, has shown that the poor of Baise Prefecture in Guangxi are net producers of grain and hence benefitted from the increase in free market grain prices.  9 additional 35% for resale grain beginning in 1985. & Fnrt:her increases in contract procurement prices for grain in 1989 and 1992 were also associated with substantial increases in the planned sales prices paid by some of the poor for resale grain. E 1.14 Income. While there is no doubt that rapid expansion of GNP raised living standards throughout the country, it is also evident that the increase in retail prices during the second half of the 19809 slowed, and even brought to a halt, increases in average per capita income. As shown in Table 1.2, nominal average per capita income data, from the State Statistical Bureau's (SSB) annual urban and rural household income and expenditure surveys, indicate steady increases in income levels throughout the -3form period. However, deflating by the urban cost of living and the rural retail consumer price index, the official data show that real: -urban income increased a total of 45% during 1981-87, stagnated during 1988-89, and then increased in 1990 to about 6% above the 1987 level; and, -rural income increased a total of more than 130% during 1978-84, at an average annual rate of 15%, and then increased by a total of less than 3% during 1985-90. & -Employment Levels and Demographic Change 1.15 EmDlovment. Underemployment has been a persistent problem in rural and, to a lesser extent, urban China. Previously camouflaged by the commune administrative structure and the work-point system, the full extent of the rural labor surplus became apparent with the introduction of the PRS in the early-1980s. A number of sources have reported that roughly one-third of the rural work force was underemployed during the 1980s, with absolute estimates ranging between 70 million to 156 million underemployed rural workers (Taylor, 5/ The majority of rural resales are provided as disaster relief and to farmers specializing in the production of cash crops and other nongrain commodities. About 7 million tons of grain are provided annually in disaster relief and to the chronic absolute poor at no cost or at the contract price (para 4.55 and 4.64). 6/ The 1992 increase brought contract grain procurement prices, and hence rural resale prices, closely in line with both domestic free market and world market grain prices. Consequently, any fu her revision or decontrol of grain prices should at most have a minimal adverse effect on the rural poor since they now pay the equivalent of market prices for any grain purchased in excess of the relief grain they receive from the government at no cost. 7/ Using the implicit GNP deflator (which increases less than the rural retail consumer price index during 1978-90), real rural income increases each year 1978-89 before declining modestly in 1990. Nevertheless, the mid-decade shift in real growth of rural income remains dramatic -- using the implicit GNP deflator, average annual real rural per capita income growth declined from about 15% during 1978-84 to 2% during 1984-90.  10 1988). Similarly, though the government's full employment policies have kept open unemployment in urban areas t. very low levels, disguised underemployment is widespread, perhaps amounting to as much as 20% of the work force of urban state enterprises (World Bank, 1992a). 1.16 Official employment and working age population figures, summarized in Table 1.2, do not evidence growing underemployment during the 1980s. Total employment reportedly increased at an average annual rate of 2.9%, from 402 million in 1978 to 567 million in 1990, and remained nearly constant as a share of the total working age population. Employment in the urban economy increased at an average annual rate of 3.7% from 95 million to 147 million during 1978-90. Employment in the rural economy increased aS the more moderate average annual rate of 2.7% during 1978-90, from 306 million to 420 million. Within the rural economy, employment growth in agricclture, increasing at an average annual rate of 1.6%, was significantly below that in nonagriculture, which increased at an average annual rate of 9.0%. The official figures show that, of the 114 million "rew jobs" created in the rural economy during 1978-90, 58 million were in agriculture and 56 million were in the nonagricultural pursuits of industry and agro-processing, construction, transport, commerce and other activities. A careful review of farm labor norms and the scale of farm oper5tions (para's 3.39 and 3.40), however, indicates that agricultural labor requirements increased by at most 10% during 1978-90, or less than half the increase suggested by the official figures. If true, even the dynamic growth of urban and rural off-farm employment would not have been sufficient to prevent growing underemployment of the rural working age population during the 19809. Assuming agricultural employment increased by only 10% during 1978-90, total employment declined from 83% of the total working age population in 1978 to 77% in 1990. 1.17 The sharp increase in inulation in late-1987 (para 1.11) prompted the government to introduce strong deflationary measures beginning in late- 1988 including (i) a decrease in state investment in services, office conetruction, processing industries and other (perceived) low priority sectors, (ii) contractionary monetary policy, in part through selective tightening of administratively allocated credit, and (iii) a stiffening of direct controls on prices and marketing. v These measures were instrumental in reducing the annualized rate of inflation from 26% in December 1988 to under 1% in the first quarter of 1990. However, they also resulted in a decline in industrial growth rates beginning in the final quarter of 1988 and negative growth during the final quarter of 1989. Though industrial growth recovered to moderately positive levels by the third quarter of 1990, significant declines in rural nonagricultural employment have been reported.1 8/ Economic developments in late-1988 and 1989 are reviewed in World Bank (1990c). 2/ An alternative measure, the rural enterprise employment series reported in parentheses in Table 1.2, indicates that employment in rural off-farm activities increased at the more rapid pace of 13% annually during 1978-88 before declining by about 2% during 1989-90.  11 The government specifically targeted rural enterprioes (RE), which were believed to be a major source of excess demand, for selective credit tightening. Though the actual reduction in credit to REa was not as great as originally intended, one report has suggested that as many as 8 million people lost their nonagricultural jobs and were "sent back to the land" in 1989. M The official data reported in Table 1.2, on the other hand, indicate (i) a less severe loss of about 1 million rural nonagricultural jobs in 1989, and (ii) that this loss was more than offset by 9 million new jobs in agriculture and 1 million new jobs in the urban economy during that year. 1.18 Demographic Chanae. Demographic changes during the reform period include a significant increase in the urban population's share of total population, substantial rural-to-rural and rural-to-urban migration, and a 30% decline in the growth rate of the total working age population during the late-1980s. The SSB recently released a new disaggregation of total population by residence which shows that China's urban population increased from 17.9% of total population in 1978 to 26.4% in 1990 (SSB, 1991a). These new figures correspond to the population of city and town neighborhood committees, and provide a consistent estimate of the urban population which is reasonably congruous with internationally accepted definitions. "' The disaggregation also accords almost exactly with the definitions of urban and rural used in the SSB household income and expenditure surveys, and so provides the precise figures for total urban and rural population used in this report's estimates of the incidence of absolute poverty. Since the incidence of poverty amongst the urban population is much less than that of the rural population, the sharp increase in the urban population's share of total population accounts for about 10% of the decline in the number of absolute poor during 1981-89. a 1.19 Rural-to-urban migration, which amounted to at least 20 million people during the 1980s, has accounted for a significant share of the increase 10/ China Daily, January 6, 1990. Development of China's REe during 1986-90 is reviewed in detail in Odv. ill Prior to the release of these newly compiled figures, official SSB statistics had indicated that the urban population had increaeed from 18% of total population in 1978 to 52% in 1989 (see 1990 edition of fin, 1991a). The previous statistics were widely considered to overetate the true increase in urban'e share, and led to considerable speculation and confusion over the true size of the urban population. 12/ Holding urban's share of total population constant at the 1981 level, the number of absolute poor in 1990 would have been 106 million instead of the 98 million reported in Table 1.2. The increase in urban's shara also explains part of the difference between the real growth rates of per capita GNP (98%) and urban (54%) and rural (58%) per capita income during 1981-90. Accounting for the increase in urban's share of total population and the fact that average per capita urban income is more than twice that of rural areas, the weighted average of urban and rural real per capita income increased nearly 70% during 1981-90.  12 in total urban population. People living outside of the place where their residence records are maintained are known as China's "floating population," and remain outside their place of residence recordation on either a "long- term" or seasonal basis. The 1990 census indicates that in July, 1990 the long-term floating population numbered about 30 million, of which dt least 10 million are presumed to have been living in urban areas. Available evidence suggests that t:he size of the seasonal floating population at least equals that of the long-term floating population (paras 2.18 and 2.19). Unfortunately, very little is known about the income levels and well-being of the members of the floating pop.21.ation. It is certainly possible that some significant share of this growing segment of the population is absolutely poor, and it is known, for example, that rates of neonatal tetanus are significantly greater among the infants of the floating population than the rates for other segments of the population (UNICEF). 1.20 In contrast to the increase in the annual population growth rate from less than 1.3% during 1978-80 to more than 1.5% during 1985-89, the growth rate of the total working age population declined from 3.6% during 1978-85 to 2.3% during 1985-90 (para 3.05). A delayed response to fertility trends of the 1970s, the absolute annual increase in the total working age population declined from about 19 million during the first half of the 1980s to about 14 million during the second half. Growing underemployment of the rural working age population would have been considerably worse during the late-1980s without this decline in the growth rate of the total working age population. -Reaional Growth Trends 1.21 Coastal Development. Adoption of the "open coastal strategy" at the outset of the reform period provided special privileges, incentives, and investment allocations to coastal provinces and counties. 21 The relationship between the coastal and interior provinces was formalized in the Seventh Five Year Plan (1986-90), which established three "economic belts." The Plan document stated that most foreign investment and technical transfer would benefit the "coastal belt" comprising all the coastal provinces. The "central belt" of provinces, stretching from Heilongjiang through Henan south to Anhui and Jiangxi, would receive a "trickle through" of foreign investment benefits. The "western belt" was, for the time being, to depend largely on 13/ In 1979, Guangdong and Fujian were given special authority to accept large foreign investments without central government approval, and 4 special economic zones (SEZs) were established in coastal counties facing Hong Kong, Macao, and Taiwan (China). The strategy was extended in 1984 to 14 additional coastal cities, and in 1985 to the Yangtze and Pearl River Deltas as well as to a "southern coastal triang'e" in Fujian. In 1988 an additional 140 coastal cities and counties were given "open" status, bringing the total to 288. Hainan was provided SEZ status in 1988, and Guangdong and Fujian were given special authority to establish foreign exchange markets and development banks, privatize housing, and undertake price, wage, and labor reforms. Most of the open coastal initiatives were at least partially curtailed in 1989-91.  13 itB own resources. The coastal strategy took advantage of China's many harbors on one of the world's busiest great-circle shipping routea and encouraged foreign technology acquisition. But it also left more isolated provinces and counties in the interior with fewer resources than would have been the case under a development strategy giving less emphasis to the coast. Furthermore, greater local policy autonomy on the coast permitted greater retention of local tax and foreign exchange revenues and greater freedom to use expansionary bank loans for local investments. In interior provinces, state pricing and procurement procedures remained relatively authoritarian, maintaining the flow of raw materials and other real resources to the coast -- often at below market prices. 1.22 Not surprisingly, the open coastal strategy has encouraged a much greater role for international trade in the economies of the coastal provinces relative to their inland neighbors. The coastal provinces accounted for 77% of China's international exports of US$41 billion in 1989. Export value from these provinces amounted to US$76 per capita in that year. By comparison, the northwestern and southwestern provinces accounted for only 4% and 6% respectively of 1989 exports, and per capita exports averaged about US$12 -- less than one-sixth that of the coastal provinces. 1.23 Impact on Regional Growth. China's different regions shared the national experience of rapid real economic growth in the 1980s, but more isolated and poorer regions grew markedly more slowly than coastal areas. As reported in Table 1.3, real national income increased at an average annual rate of 9.1% during 1978-89, having slowed from an average of 9.5% annual growth during 1978-85 to 8.3% during 1985-89. National average economic growth per capita during 1978-89 was 129%. Real per capita national income grew 142% in the coastal provinces during 1978-89, or about 30% more than in the central and western belts. Overall, average annual growth in the coastal region was more than 1% greater than in the central and western belt provinces. Total per capita growth was greatest in the higher income provinces of south China (178%) and along the middle and lower reaches of the Yangtze River (138%), slightly less than the national average in north China (126%), and least in northeastern (102%) China and the lower income provinces of northwestern (107%) and southwestern (110%) China. L/ 1.24 The regional growth patterns of the 1980. have led to provincial income variation akin to that experienced at oth"r times in the previous three decades. Using newly available data series on net material product and JA/ These growth rates document increased interprovincial income disparity during the reform period -- growth in the low income interior provinces was less than in the higher income coastal provinces. Denny (1991) report. a convergence of per capita provincial output from 1977 to 1988, but his analysis gives equal weight to all provinces. The regional growth rates presented in this report are weighted by provincial population, and correct for the excessive weight attached in Denny's analysis to the slow growth in Beijing, Shanghai and Tianjin.  14 Table 1.3t Provincial Average Annual Real Income Growth t \) Geographic Sub-Regions \c Economic Belts \b ---------------------------------------- ------------------ North- North- Yangtze South- Total Coast Center West North east west River South west --------------------_.--------------------------------------------__---------_ Average Annual Real Growth (4) 1978-89 9.1% 9.7% 8.5% 8.3% 9.1% 7.4% 8.3% 9.5% 11.5% 8.3% 1978-85 9.5% 10.0% 9.2% 8.6% 9.4% 7.4% 8.7% 10.5% 11.0% 8.9% 1985-89 8.3% 9.1% 7.2% 7.9% 8.6% 7.3% 7.8% 7.9% 12.3% 7.3% Total Per Capita Growth (4) 1978-89 129% 142% 112% 109% 126% 102% 107% 138% 178% 110% National Ineome Per Capita (1980 yuan) 1989 761 1077 633 498 771 1029 614 908 872 457 Average Annual Real Agriculture Growth (4) 1978-89 5.5% 5.9% 5.2% 5.4% 6.5% 4.0% 5.9% 5.1% 6.9% 5.0% 1978-85 7.4% 8.1% 7.1% 6.7% 9.0% 6.3% 7.4% 7.0% 7.8% 6.6% 1985-89 2.3% 2.1% 2.0% 3.0% 2.2% 0.0% 3.5% 1.7% 5.4% 2.2% Total Per Capita Agricultural growth (4) 1978-89 57% 64% 52% 54% 72% 41% 62% 50% 76% 49% Population (million) 1989 1108 414 395 296 242 98 128 325 96 217