China - Southwest Poverty Reduction Program Project

Document of The World Bank Report No: 26132 IMPLEMENTATION COMPLETION REPORT (IDA-27440; TF-26512; CPL-39060; SCL-39066) ON A LOAN IN THE AMOUNT OF US$47.5 MILLION AND A CREDIT IN THE AMOUNT OF SDR128.6 MILLION TO THE PEOPLE'S REPUBLIC OF CHINA FOR A SOUTHWEST POVERTY REDUCTION PROJECT 06/30/2003 Rural Development and Natural Resources Sector Unit East Asia and Pacific Region  CURRENCY EQUIVALENTS (Exchange Rate Effective May 2003) Currency Unit = Yuan Y1.0 = US$ 0.12 US$ 1.00 = Y 8.3 FISCAL YEAR January 1 December 31 ABBREVIATIONS AND ACRONYMS CAE Country Assistance Evaluation CAS Country Assistance Strategy CCCC China Cross-Cultural Consulting Center CWHRDC China Western Human Resources Development Center DECRG Development Research Group DFID Department for International Development FCPMC Foreign Capital Project Management Center GX-MEDs Guangxi Municipal Employment Development subcomponent IBRD International Bank for Reconstruction and Development LGPR Leading Group for Poverty Reduction MCH Maternal and Child Health MIS Information Management System NBS National Bureau of Statistics NHEI National Health Economics Institute NMPRP Northern Mountains Poverty Reduction Project PHRD Policy and Human Resources Development PMO Project Management Office PPA Participatory Poverty Assessment PPAR Participatory Poverty Appraisal Report QAG Quality Assurance Group QBPRP Qinba Mountains Poverty Reduction Project SAR Staff Appraisal Report Strategies China: Strategies for Reducing Poverty in the 1990s SWPRP Southwest Poverty Reduction Project TVE Township and Village Enterprises UNICEF United National International VCMS Village Cooperative Medical Scheme Vice President: Jemal-ud-din Kassum Country Manager/Director: Yukon Huang Sector Manager/Director: Mark D. Wilson Task Team Leader/Task Manager: Alan L. Piazza  CHINA Southwest Poverty Reduction Project CONTENTS Page No. 1. Project Data 1 2. Principal Performance Ratings 1 3. Assessment of Development Objective and Design, and of Quality at Entry 2 4. Achievement of Objective and Outputs 4 5. Major Factors Affecting Implementation and Outcome 13 6. Sustainability 14 7. Bank and Borrower Performance 15 8. Lessons Learned 17 9. Partner Comments 19 10. Additional Information 24 Annex 1. Key Performance Indicators/Log Frame Matrix 25 Annex 2. Project Costs and Financing 26 Annex 3. Economic Costs and Benefits 28 Annex 4. Bank Inputs 31 Annex 5. Ratings for Achievement of Objectives/Outputs of Components 33 Annex 6. Ratings of Bank and Borrower Performance 34 Annex 7. List of Supporting Documents 35 Annex 8. Beneficiary Survey Results 36 Annex 9. Stakeholder Workshop Results 38 Annex 10. Borrower's ICR Summary 39 Annex 11. Village-Based Multisectoral Poverty Reduction Approach 51 Maps IBRD 26524, 26525 and 26629  Project ID: P003639 Project Name: Southwest Poverty Reduction Project Team Leader: Alan L. Piazza TL Unit: EASRD ICR Type: Intensive Learning Model (ILM) of ICR Report Date: June 27, 2003 1. Project Data Name: Southwest Poverty Reduction Project L/C/TF Number: IDA-27440; TF-26512; CPL-39060; SCL-39066 Country/Department: CHINA Region: East Asia and Pacific Region Sector/subsector: General agriculture, fishing and forestry sector (40%); Other social services (32%); General education sector (12%); Health (12%); Central government administration (4%) Theme: Social risk reduction (P); Rural markets (P); Rural services and infrastructure (P); Poverty strategy, analysis and monitoring (S); Gender (S) KEY DATES Original Revised/Actual PCD: 10/28/1993 Effective: 10/16/1995 10/16/1995 Appraisal: 11/17/1994 MTR: 06/30/1997 10/20/1997 Approval: 06/15/1995 Closing: 12/31/2001 04/30/2004 Borrower/Implementing Agency: PRC/YUNNAN; PRC/GUIZHOU; PRC/GUANGXI Other Partners: STAFF Current At Appraisal Vice President: Jemal-ud-din Kassum Gautam Kaji Country Director: Yukon Huang Shahid Javed Burki - Department Director Sector Manager: Mark D. Wilson Joseph Goldberg - Division Chief Team Leader at ICR: Alan L. Piazza Alan L. Piazza ICR Primary Author: Alan L. Piazza 2. Principal Performance Ratings (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HL=Highly Likely, L=Likely, UN=Unlikely, HUN=Highly Unlikely, HU=Highly Unsatisfactory, H=High, SU=Substantial, M=Modest, N=Negligible) Outcome:HS Sustainability:HL Institutional Development Impact:H Bank Performance:S Borrower Performance:HS QAG (if available) ICR Quality at Entry: S S Project at Risk at Any Time: No  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: This complex and highly innovative project sought to influence national poverty reduction policy and to address absolute poverty in some of China's most highly disadvantaged areas. The specific principal objectives of the project were to: (a) demonstrate the effectiveness of a focused multisectoral rural development project approach to poverty reduction, (b) facilitate a "market friendly" increase in labor mobility from the poor areas to better off rural and rapidly growing urban areas, (c) upgrade poverty monitoring at the national and local levels, and (d) significantly reduce absolute poverty in 35 of the very poorest counties in southwestern China. The project also intended to stabilize or even reverse the trend of upland environmental destruction through land and pasture improvement and soil conservation works, and to encourage greater local community participation by enabling households to make key decisions during project design and implementation. These objectives were consistent with the Country Assistance Strategy priorities of poverty reduction and environmental protection. Importantly, these objectives were jointly determined through the World Bank and the Government's Leading Group for Poverty Reduction (LGPR) collaboration on the 1992 study "China: Strategies for Reducing Poverty in the 1990s" (Strategies). Through the preparation of Strategies and its public release at the "International Conference on Poverty Issues in China" (held in Beijing in October, 1992), close collaboration with the Government was established and the Bank has played an increasingly active role in the development of national strategies for reducing absolute poverty. At the conclusion of the conference, State Councilor Chen Junsheng (at that time, the Government's senior policy-maker for agriculture, rural development, and poverty reduction) and the Country Director announced the proposal to operationalize Strategies' recommendations for reducing poverty in the 1990s through a Bank assisted project to reduce poverty in southwestern China. It was decided at that time that the project would take a multisectoral approach to reducing absolute poverty in the worst affected upland areas of Guangxi, Guizhou and Yunnan. Consistent with one of the principal themes of the conference, it was also agreed that the project would encourage community participation during preparation and implementation and that all interested international agencies, both official and nongovernmental, would be invited to participate in project preparation, funding and implementation. 3.2 Revised Objective: The original project objectives remained unchanged.1 3.3 Original Components: The project's objectives were to be realized through the following package of multisectoral components: (1) Basic Education (6.7% of Appraisal total project cost). This component intended to improve poor children's access to elementary school education through (a) primary school renovation and construction, (b) support for tuition assistance and nutritional supplements for poor students, (c) the provision of textbooks, instructional equipment and furniture, and (d) teacher and management training. (2) Basic Health Services (5.4% of Appraisal total project cost). This component was designed to increase the accessibility and enhance the quality of health services provided to the poor at the village and township level. This was to occur primarily through facility construction and renovation, the establishment of a village cooperative health fund, training of health personnel, strengthening of public health programs in disease prevention and control and maternal and child health (MCH), and enhancing the management capabilities of health administrators at the county and provincial level. 1 In June 2003, about SDR 0.9 million of the Credit was allocated to a new SARS and Infectious Diseases Response Program (see Section 10). This program is still under implementation and not part of this ICR. - 2 -  (3) Labor Mobility (29.9% of Appraisal total project cost). This highly innovative component sought to increase the upland poor's access to off-farm jobs through the establishment of a voluntary system of enhanced rural labor. The labor mobility system comprised a program of initial orientation, skill training through an enterprise reimbursement system, a computerized demand-driven job placement system emphasizing local markets, monitoring of worker safety and living conditions, and reporting of abuses and grievances. The labor mobility component was also to be supported by the Guangxi Municipal Employment Development subcomponent (GX-MEDs, or the "hard loan"). GX-MEDs was intended to generate employment opportunities in three municipalities in Guangxi (namely, Nanning, Beihai and Fangcheng) through support for a "poverty-reduction enterprise development zone" in Beihai, low-cost housing schemes, and mariculture, agriculture and tree plantation activities. (4) Rural Infrastructure (10.5% of Appraisal total project cost). This component intended to improve the poor's access to transport, drinking water, electricity and other basic services through the labor-intensive construction of rural roads, safe drinking water supply systems, small scale irrigation, agricultural drainage works, biogas digesters, and rural electrification. (5) Land and Farmer Development (30.0% of Appraisal total project cost). This major component was designed to increase upland agricultural productivity and reverse the trend of environmental degradation. Developed on a participatory basis, poor farm households were to be provided with a menu of crop and livestock activities to select from. In addition, the component supported (i) extension of improved agriculture and smallholder livestock management techniques and inputs, (ii) establishment of tree crop nurseries and provision of saplings and tree crop management extension (for high altitude and red soil varieties), (iii) intensification of staple crop production on the modest amount of available flat land, (iv) terracing for improved water and soil conservation, and (v) the support of provincial and regional applied agricultural technologies. (6) TVE Development (14.4% of Appraisal total project cost). This component was designed to augment the development of local township and village enterprises (TVE) through investment credit for labor-intensive commercially viable agro-processing, mineral, service and handicraft industries which had strong backward linkages to poor households and which met environmental safeguards. (7) Institution Building (1.7% of Appraisal total project cost). Successful implementation of this complex and innovative project would require strengthening all levels of the project management office (PMO) system. The institution building component was designed to improve the LGPR and PMO system's capabilities in project design and implementation, procurement, accounting, and monitoring and evaluation. (8) Poverty Monitoring (1.5% of Appraisal total project cost). This component was intended to improve national level poverty monitoring capability and to more objectively monitor the physical progress and impact of the project. The poverty monitoring component would establish an objective poverty monitoring system for the project areas by using improved National Bureau of Statistics (NBS) survey instruments to provide a credible and comprehensive poverty profile, analyze the accuracy of the targeting of benefits to the absolute poor, and measure and evaluate the impact of the project's individual components. 3.4 Revised Components: Two of the project components performed poorly during the initial years of implementation and were therefore significantly revised during the latter years of project implementation. First, implementation of the TVE development component was adversely affected by a number of difficulties including (a) relatively cumbersome procedures and requirements necessary for the clearances of TVE proposals, (b) uncertain ownership of the enterprises and inadequate counterpart financing, and (c) inadequate market analysis, poor product quality, and/or collapse of the market for certain products during the enterprise proposal - 3 -  development period (see Annex 7 Document 5.6). As a result, the TVE development component got off to an extremely slow start during the initial years of project implementation and was substantially downsized during the Mid-Term Review and again during the final years of project implementation. In addition, attempts were made to improve the viability of project-supported TVE enterprises by establishing linkages (through joint venture arrangements) to existing enterprises with stronger marketing channels, management skills, and technical expertise. Secondly, implementation of the GX-MEDs activities in Beihai were adversely affected by a shortage of counterpart funding, inadequate support and commitment from the Beihai Municipal Government, and limited poverty reduction impact (see Annex 7 Document 11 "GX-MEDs Follow-Up Evaluation Report"). After several failed attempts to revitalize the Beihai subproject, it was finally decided to cancel the Beihai subproject and redirect all the remaining IBRD funds (about US$16.6 million) from the Beihai subproject to water resources infrastructure development (about US$8.6 million) in the Guangxi project counties and to other agriculture activities (another US$8 million) in Nanning and Fangcheng. The use of about US$8.6 million of the Loan for the construction of small water cisterns ("ditou shuigui") in the upland poor counties was considered a major project innovation (see Annex 7 Document 12 "Adjustment Plan of Guangxi Hard Loan Project"). 3.5 Quality at Entry: Quality at Entry is rated satisfactory, based on the strength of (a) consistency of the project objectives with the CAS priorities for poverty reduction and environmental protection and with the Government's stated strategy for poverty reduction, (b) lessons learned from previous broad-based rural development projects, (c) project design, which accounted for the institutional challenges for such a complex and innovative operation, and (d) consistency with the Bank's safeguard policies. A Quality Assurance Group (QAG) "Post-Approval Quality at Entry Assessment" completed in November, 1997 rated Quality at Entry as fully satisfactory with several best practice features, citing "China's exceptional commitment to poverty reduction and impressive implementation capacity and record," effective targeting of the poorest households and villages, and participatory approach. The SAR thoroughly documented the project background and highlighted the key risk factors of complexity and innovative approach. These risk factors were addressed by the project design and plans for management and organization. Most importantly, LGPR, which was an interministerial office of the State Council, was to be responsible for coordinating project design and, through the provincial and county PMOs and Leading Groups, monitoring implementation. Project preparation was supported by a large PHRD Grant and through collaboration with Ford Foundation and CARE Australia. With the benefit of hindsight, quality at entry could have been even better had the following been established prior to implementation: (a) a mechanism to more fully continue the participatory process from the project preparation period into implementation, (b) a computerized system for recording household loan receipts and repayments, (c) additional explicit project activities in support of environmental protection, and (d) more user-friendly and appropriate community-based procurement methods. 4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: Achievement of the project objectives and outputs was highly satisfactory. First, the project had a major positive impact on national poverty reduction policy by (a) demonstrating the effectiveness of a new multisectoral poverty reduction model in China's most severely affected areas, and (b) directly involving senior policy makers in the design and implementation of this new model. The new integrated rural development project approach to poverty reduction -- which the Government calls the "Southwest Project" model -- was first extended to the 26 counties of the ongoing Qinba Mountains Poverty Reduction Project (QBPRP), and the Chinese Government continues to extend the lessons of the "Southwest Project" approach to the poverty reduction program throughout China's poor counties. In addition, the "Southwest" - 4 -  model has now been adapted to northern Vietnam, and the Vietnam Northern Mountains Poverty Reduction Project (NMPRP) became effective during late-FY02. (Staff of the SWPRP Guangxi PMO assisted in the preparation of NMPRP, and a team of NMPRP project staff undertook a Study Tour to the SWPRP project areas in 1999.) The Village Development Plan model advanced under SWPRP has now been extended to 146,000 poor villages in China. The 1997 QAG review concluded that SWPRP's "actual and potential impact at the highest policy levels can already be characterized as best practice," and the 2003 Country Assistance Evaluation (CAE, see Annex 7 Document 7) similarly identifies SWPRP as an "outstandingly successful" poverty reduction project which "served as an example of best-practice methods of investment for poverty reduction." Second, the labor mobility component became a large-scale success story, with outstanding benefits for the project's absolute poor and significant policy implications at the national level. The project assisted about 280,000 people to secure seasonal and long-term off-farm jobs. Third, NBS completed annual sample surveys of households in project and nonproject areas, and this has helped upgrade poverty monitoring at the national and local levels. NBS (see NBS, 2003) concluded that "SWPRP plays a very important role in the process of China improving its poverty monitoring capacity." Fourth, available information supports the conclusion that the project has had a strongly favorable impact on the income levels, grain and other agricultural production, and overall well-being of the majority of project area poor households. (The NBS survey data show per capita net income increased from Y 939 to Y 1422 and the headcount incidence of absolute poverty declined from 32% to 13% between 1995 and 2001 in the project area. The survey data also show that the share of the project area population with grossly insufficient grain production declined from 18% to 9% over these same years.) Fifth, the project's environmental impact has been mostly positive overall, and impressive achievements were made in soil conservation and increasing vegetation coverage. Finally, while community participation -- particularly during project implementation -- was not as great as had been desired, SWPRP certainly contributed to popularizing participatory project design and preparation. 4.2 Outputs by components: Basic Education ($32.4 million SAR, $29.7 million actual): This component is rated highly satisfactory. The component's key outcome was a significant increase in school attendance and an improvement in educational quality brought about through primary school construction, the provision of tuition assistance and school equipment, and teacher training. The NBS project survey data indicate that sixth grade completion rates in the project area increased from 46% in 1995 to 78% in 2001. Primary school renovation and construction targets were successfully completed in each of the three project provinces. The improved and new school buildings have provided adequate, safe and hospitable space for village children to attend school, and focus group discussions with project area parents, teachers and principals interviewed have indicated that the new buildings encouraged children to attend and stay in school. Similarly, the provision of desks, chairs, and a variety of other teaching equipment and tools has contributed to the quality of schooling and encouraged children to attend and stay in school. In some of the poorest counties, the provision of tuition assistance was the principal reason for increased enrollment rates. The training activity was also very successful and the number of qualified teachers increased considerably during project implementation. Finally, an unexpected payoff to the basic education component was that its early implementation provided a highly favorable introduction to the project in many of the poorest and most remote project villages. These villages became very enthusiastic about the overall project and were more eager to access other project activities and benefits (see SWPRP Component Evaluations). Although rated highly satisfactory, the basic education component did suffer several shortcomings and design flaws. First, the nutrition supplementation activity faltered during the initial years of implementation and was finally eliminated at the time of the Mid-Term Review. The nutrition - 5 -  supplementation activity was ineffective because of the complicated nature of distributing the food and social constraints to anything but a fully equal distribution (that is, targeting only the very poorest children for nutritional supplementation was not found to be an acceptable practice at the local level). Second, there were also difficulties with implementation of the tuition assistance activity during the early years of the project. Most importantly, supervision missions during the first two years of implementation found that some of the tuition assistance in some project counties was provided to children already attending school instead of to the families of children who were not attending school because they were too poor to cover even the very modest costs of tuition. Third, the primary school building construction program was also found to have some unexpected negative consequences. Although not a common occurrence, a Department for International Development (DFID) funded Participatory Poverty Assessment (PPA) team found that the construction of a primary school in a central location induced the county Education Bureau to eliminate a "teaching point" in a distant ethnic minority village and, as a result, several ethnic minority children could no longer attend primary school (see Annex 4 of China: Overcoming Rural Poverty). Basic Health Services ($26.4 million SAR, $20.7 million actual): On balance, the component is rated satisfactory. Although the basic health services component very successfully achieved its primary objective of increasing the accessibility and enhancing the quality of health services to the poor at the village and township level, it was not successful in establishing a village cooperative medical scheme (VCMS). The project monitoring report indicates that the infant mortality rate declined from 5.0% to 2.0% in the Guangxi project area and from 8.2% to 5.0% in the Guizhou project area between 1995 and 2001. The National Health Economics Institute's (NHEI) reviews of the basic health services component in three project counties (Lingyun and Donglan Counties in Guangxi, and Yanjin County in Yunnan) concluded that: "The health subdivision of the project obviously improved the health care for the rural population in the project areas and the accessibility of the farmers in poor areas to health care services. The major indicators for population health like maternal mortality, infantile mortality and incidence of infectious diseases were greatly improved, indicating that the health subdivision of the project has played a significant role in relieving poverty of the population in the project areas" (see SWPRP Component Evaluations 2A and 2B). First, the NHEI staff found that the program of facilities construction and renovation had been very successful and achieved a remarkable improvement in basic health care infrastructure in the project area at the local level. Second, the training program provided short-term and long-term training for large numbers of township and village health workers leading to a sustainable increase in the quality of health care. Third, the project counties emphasized the strengthening of maternal health care, planned immunization of children, and the prevention and management of key diseases as key project activities, and this led to significant decreases in the maternal mortality rate, infant mortality rate, and incidence of infectious diseases. Although the component made great contributions to basic health services at the local level, efforts to establish the VCMS were not successful. The NHEI reviews laud the good intentions of the project's VCMS experiments, and note that the severe poverty in the project areas almost certainly doomed these experiments from the outset. Similarly, UNICEF staff found (during field visits to the project counties of Luquan and Wuding in Yunnan in 2000) that the VCMS was not viable and that financial incentives for medical workers in the very poorest villages were still not sufficient (even with the project's assistance) to adequately upgrade basic health care services. Overall, it is now increasingly recognized that funding for health services in China's rural areas is inadequate and that the problem is most severe in the poorest rural areas. While the SWPRP basic health services component made a valuable contribution to basic health in the project area, it did not establish a clear replicable model for upgrading basic health services on a sustainable basis for the rural absolute poor. - 6 -  Labor Mobility ($145.2 million SAR, $133.3 million actual): This component is rated highly satisfactory. The key outcomes were the establishment of a voluntary system of enhanced rural labor, and the placement of about 280,000 workers in safe off-farm employment. Placement and monitoring were critical to the success of the component. A powerful, extensive and flexible placement mechanism succeeded in finding jobs for the low skill workers over the life of the project despite the rapidly changing labor market. The extensive monitoring effort helped ensure that the workers were safe, and that any difficulties or concerns could be quickly addressed. Because of the impressive poverty impact of the component, which was evident even from the early years of implementation, great efforts were made to ensure that the component benefited the very poor. In the early years of the component, there was a tendency to target workers somewhat better off than the average. However, as implementation proceeded, the targeting of the component improved, and the component was better able to reach the poorer villagers. Overall, some 1798 project villages participated in the project. The high coverage rate was instrumental in spreading the benefits of the component beyond its immediate beneficiaries. Over the life of the project, workers who had participated in the component, returned to their villages and brought out their neighbors and relatives, who otherwise would not have had the opportunity. The impact of the component was impressive, and most county PMO and project township Work Station staff stated that the labor mobility component was the single most effective poverty reduction activity in SWPRP, produced the quickest returns, and was the component that was most warmly welcomed by the primary beneficiaries. Based on a 2% sample survey of participants in the component, workers averaged 1135 yuan per year in remittances over the life of the project. These funds were invested in farm inputs (24%), housing (20%), children's education (10%) and healthcare (6%). The remittances also were used to reduce the household's debt burden (11%) and to start small businesses in rural areas (4%). In addition to the cash remittances, the workers brought back substantial amounts of consumer goods (averaging 900 yuan per worker) to their families. Major efforts were made to ensure that women and ethnic minorities had adequate opportunities to participate in the component. The success of these initiatives varied among locations, partly due to cultural influences. All in all, the component did a good job in balancing the benefits from participating with the particular risks inherent when these populations take off-farm jobs. Much effort was made to compare the benefits for workers participating under the component with those who found off-farm jobs on their own (spontaneous migration). Overall, it appears that participating in the component brought the workers higher salaries and more stable jobs than would have been possible through spontaneous migration. Despite the overall success of the component, a number of difficulties were encountered. The management structure for the component was perhaps, in hindsight, somewhat detrimental to the success of the component. The emphasis on the independence of the labor agencies meant that the component suffered disproportionately from a shortage of counterpart funds, and from slow disbursements of project funds to the implementing agencies. Also, the emphasis on autonomy for the labor companies, while providing some incentive, at the same time may have distracted management from its key objectives. The overall impact of the SWPRP project would possibly have been better if the component had been better integrated with the main project. For example, the impact of the labor mobility component on the agriculture component was not well-monitored, and in some (although likely the minority) of cases, the component may have inadvertently reduced the effectiveness of the agriculture component (be reducing the amount of available labor). In addition, the monitoring effort, while extensive, did not adequately capture the medium-term impact on the home village. In particular, the social impact on the home village may not have been investigated in sufficient detail. It is also unfortunate that the experiment is not replicable in other parts of China. The high cost of setting up a reliable monitoring system has discouraged local municipalities from using public funds to replicate the Southwest system. - 7 -  The GX-MEDs subcomponent was a large scale experiment with creating employment opportunities for the project area's upland poor in labor intensive enterprises in three municipalities in Guangxi. As noted in Section 3.4 above, GX-MEDs performed poorly -- especially in Beihai Municipality -- and about 20% of the subcomponent's funding was eventually redirected toward the construction of small on-farm water cisterns in the project's upland counties. Overall, GX-MEDs provided fewer job opportunities than anticipated, the investment cost per job created was often excessive, and many of the enterprises and activities supported were not financially viable. By comparison, using a portion of the GX-MEDs funding for the construction of small on-farm water cisterns was highly successful. These water cisterns provide drinking water and supplemental irrigation in the Karst mountain region, and have now been very widely adopted (see SWPRP Component Evaluation No. 4). Rural Infrastructure ($50.9 million SAR, $64.0 million actual): Project beneficiary interviews consistently confirmed that the project's poor households were extremely supportive of the rural infrastructure component and that improved access to roads, drinking and irrigation water, electricity, and other basic infrastructure were powerful and "fast-acting" measures to improve well-being and reduce poverty. Unfortunately, constraints imposed by the counterpart funding mechanism and repayment requirements adversely affected the timeliness of the physical completion of these works and limited the total scale to considerably less than what was actually desired by the primary beneficiaries (see discussion on counterpart funding mechanisms and loan repayment issues in Section 5.4 below). On balance, the component is rated satisfactory, and would have certainly been rated highly satisfactory had funding and repayment mechanisms allowed more timely completion of an expanded program of rural infrastructure. Project rural road construction improved the flow of goods into and out of remote poor areas, and increased the access to outside jobs, health care, and education services. Solving household water supply shortages both improved household hygiene and living standards and reduced the time and hard labor previously expended by (mostly) women and girls on transporting water. Small-scale irrigation works greatly improved food security. All of the SWPRP rural infrastructure activities had important synergistic effects on other project activities (for example, with road access, school and health clinic construction was made much easier and it became possible to bring in chemical fertilizer and other materials for agricultural activities). Finally, a major innovation during the latter years of the project, led by the Guangxi PMO, was support for natural village access roads and tracks. Typically only suitable for farm tractors, light utility vehicles, and animal-drawn carts, these roads were built with labor volunteered by the villagers and supported with materials and technical assistance provided through the project. Land and Farmer Development ($145.1 million SAR, $160.3 million actual): This component is rated satisfactory. At least 80% of project households received assistance for a very wide variety of field crops, tree crops, and livestock activities. Some of these activities were very effective means of improving household food security and increasing incomes. The component supported an environmentally friendly integrated land use system, and proved that a farming population can sustain itself over the long term in most of the Karst region of southwest China. The farmer training programs were effective overall, though they could have been improved had specific training programs been tailored for women and young farmers. Several factors constrained the land and farmer development component from achieving its full potential. First, the outreach of the project to the very poorest and most remote households was delayed and never reached the levels received by less poor households closer in to the road system. It is understandable that local leaders and project staff either did not, or could not, reach the poorest and most remote project villages in the first two years of the project, and a sustained push in the latter years of the project did extend benefits to most all households and to extremely remote areas. However, much of this latter - 8 -  assistance was limited in value, too simplistic, and not done on a participatory basis (for example, each farm household in extremely remote villages would receive a cow in year 4 or 5 of the project). Since ethnic minorities often reside in the most remote villages of the project areas, some ethnic minority families did not receive an adequate or fair share of the benefits of the project. Second, many of the land and farmer development program activities were effectively "hijacked" by the County technical bureaus who were more interested in implementing county plans for agricultural development than providing poor households what they really wanted. For example, a Bank supervision mission which visited an extremely remote village in Luquan County (in Yunnan) found that most village households had been obliged to plant chestnut tree saplings under the project because the County Forestry Bureau had a county-wide chestnut tree planting program. However, these villagers reported to the Bank mission that they already had an adequate cash income from growing tobacco and would have been far more interested in improving primary school education and increasing staple crop production had they been encouraged to express their preferences. Clearly, the technical bureaus' emphasis on county plans sometimes undermined the participatory implementation process (see Section 5.3 below). Third, the benefits from heavy investments in a number of tree crops in the early years of the project were not realized because (a) the quality of the planting material was unsatisfactory, and (b) the markets for most of these tree crops were in a state of severe decline or complete collapse due to heavy saturation of the regional or national markets for these products. Since the land and farmer development component was the single largest project component, the impact of these three constraints on the overall success and effectiveness of the entire project was substantial. TVE Development ($70.1 million SAR, $42.3 million actual): This component is rated unsatisfactory. The inadequate infrastructure base (including poor transport and telecommunications) and paucity of technical and management skills in the project's poorest townships and villages hampers the development of township and village enterprises in much of the SWPRP project area. For this reason, the total scale of the TVE development component was substantially reduced during project design and preparation. With the slow and unsuccessful implementation of the component during the first years of the project, the total scale of the component was further reduced during the latter years of implementation (see Section 3.4 above). While most of the medium and larger TVEs supported by the project were eventually found not to be financially viable or to have had significant poverty reduction impacts, small scale enterprises and small scale farmers' markets were typically far more successful. These enterprises, which were often privately owned, provided goods and services for the local market and relied on simple technologies. Most of these enterprises were found to be financially viable, have had favorable poverty reduction and local fiscal impacts, and have provided marketing channels for farm and other products produced by the poor. Medium and larger scale TVEs (that is, in the context of the SWPRP project area, enterprises with investments of Y 2 million or more) supported by the project faced several important constraints. First, with the takeoff of light industrial production throughout China, the SWPRP TVEs had to compete with much better funded enterprises with superior technology, management, and marketing systems. Second, most of these enterprises found the approval and clearances process and procurement procedures to be very time consuming, confusing and difficult. Third, there were conflicts of interest among the shareholders of larger enterprises. Fourth, the strengthening of the Government's regulations on worker safety and environmental protection restricted the existence of township and village mining and mineral processing enterprises. Fifth, the shortage of funding often left these enterprises without sufficient working capital to undertake full production, and thereby crippled attempts to reach a profitable scale of production. Unfortunately, expecting to be able to augment fiscal revenues, most project County PMOs (which were all closely linked to the County Governments) were inclined to use SWPRP to support only these medium and larger scale enterprises. Experience suggests that most of these enterprises were not financially viable, became a drain on county fiscal resources, and had little or no sustained poverty reduction impact. The unsatisfactory performance of the TVE development component (see Annex 7 Document 5.6) does not - 9 -  mean that enterprise development cannot play an important role in poverty reduction, but does strongly indicate that a different approach should be adopted. Institution Building ($8.2 million SAR, $6.1 million actual): This component is rated highly satisfactory. It was recognized at the outset of preparation that successful implementation of this innovative multisectoral project would be dependent upon greater-than-average institutional capacity, and that existing capabilities and organizational strengths were limited and already stretched. It was therefore determined that a stand alone institution building component was essential. The project's legal documents included (a) the standard requirement that the central and provincial PMOs "would be staffed at levels, and maintained with functions and responsibilities, acceptable to the Bank Group," and (b) a special requirement to establish a monitoring and reporting system for the labor mobility component. Adequate project support for capacity and institution building, coupled with extraordinarily strong support and leadership from the highest levels of the Central Government, led to a rapid buildup of preparation and implementation capacity at all levels (from the Central PMO all the way down to project township and village Work Stations). The outcome was a highly capable and complete PMO system with intense ownership and commitment to the project concept and successful implementation. With the establishment of quality work space and equipment early on in the project cycle, combined with training opportunities and a clear sense that the project was being closely watched by senior Central Government staff, the PMOs at all levels were able to attract and retain very capable and highly motivated staff. The highly effective PMO system became a model for the remainder of the poverty reduction program in the three project provinces and, eventually, throughout the country. Staff of the Central and provincial PMOs have stated that the strong institution building component was one of the best "ideas" of the project conceptualization process. Poverty Monitoring ($7.1 million SAR, $7.2 million actual): This component is rated highly satisfactory. In 1992, Strategies highlighted the need to upgrade China poverty monitoring system: "It is essential that an independent and objective poverty monitoring system be established and the true location of the poor be more 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." Achieving this key recommendation was the objective of the SWPRP poverty monitoring component. With ongoing technical assistance of DECRG staff, the NBS undertook annual sample surveys of project and non-project households throughout the implementation period. The outcome was a very significant increase in NBS' poverty monitoring and analysis capability and an unusually good data base for detailed project monitoring and evaluation. The SWPRP Component Evaluation Paper No. 7 (see Annex 7 Document 5.7) notes: "Significantly, the SWPRP poverty monitoring exercise provided the impetus for China to establish a nationwide poverty monitoring system. Since 1997 China has conducted a national poverty survey that covers all nationally designated poor counties. This national poverty monitoring system was modeled on and drew heavily from the experience of the SWPRP. Staffing of the national poverty monitoring offices at the central and provincial levels also has drawn from the SWPRP personnel. Questionnaire design, sampling and data collection build on methods developed under the SWPRP. Following the model of the SWPRP monitoring reports, the NBS has published an annual Poverty Monitoring Report of Rural China since 2000. In this regard the SWPRP Monitoring Component has had an impact that goes far beyond the project itself." The poverty monitoring component could have been improved through greater and earlier emphasis on the establishment of an information management system (MIS) for monitoring all aspects of the physical progress of the project. The project MIS system was intended to be separate and distinct from the NBS poverty monitoring work, was included in the project design and sufficient funding for this activity was provided for under the institution building component, and the Guangxi PMO did eventually develop an excellent "computerized household loan management system." However, the Guangxi MIS system was - 10 -  fully established only in the latter years of project implementation and was not effectively extended to the other project provinces. Project implementation, monitoring and evaluation would have been significantly enhanced by the establishment of a good MIS system (similar to that developed by the Guangxi PMO) prior to implementation in all three project provinces. Status of Women and Female Participation. The project helped project area women and girls overcome poverty and improve their well-being, but should have focused more intensively on gender issues. The April 2003 SWPRP Participatory Poverty Assessment Report (PPAR) found that all of the project activities benefited women and girls and, in particular, that the project directly assisted with improving girls' education, providing training to women teachers and rural women medical workers, and strengthening maternal and child health care services. However, the PPAR also found that women's work load in rural areas was excessive and increasing, there were only limited efforts to plan special assistance activities for project female beneficiaries, and that project area women did not have equal access to agricultural technology training activities and project loans. While participation in project activities was always on a completely voluntary basis, the PPAR's women's forums found that participation in some project activities pushed women's work load beyond sustainable levels. Ethnic Minorities. Ethnic minorities comprise less than 10 percent of China's population, but are heavily overrepresented among China's absolute poor. The project directly targeted poor ethnic minority townships and villages, and ethnic minority people comprise more than half of the total beneficiary population. The PPAR found that ethnic minority project beneficiaries believed that the project had greatly helped them overcome poverty, and that ethnic minorities had participated in all project activities. However, the PPAR found four areas where the project could have improved its outreach to the most disadvantaged ethnic minority people. First, some minority-specific cultural values restricted their participation, and more creative means should have been devised to better include these groups in the project. Second, some of the most disadvantaged ethnic minorities were deemed to be too poor to participate in project activities. Third, in some cases ethnic minority people did not participate because they were afraid of taking on the responsibility of loan repayment. Lastly, education attainment levels among some ethnic minority groups are relatively low and they could not fully benefit from agricultural technology training. Overall, the PPAR found that better methods of reaching the ethnic minority groups could have further improved the project's success in improving the well-being of these project beneficiaries. In addition, the PPAR suggests that the project should strive to better understand and adopt ethnic minority villages' own view of what well-being and development are: "Any poor community has its own criteria of value for the judgment of poverty and happiness, and the project vision drawn up by the project planners is probably not the entire happy life desired by the poor communities." 4.3 Net Present Value/Economic rate of return: As evidenced by a variety of available data, the project achieved substantial economic and social benefits. These data consistently confirm the improvements of a series of key monitoring indicators. For example, the incidence of poverty in the project area declined from 31.5% in 1995 to 12.9% in 2001, and per capita income in the project area increased from 939.5 yuan to 1422 yuan during the same period. Project villages outperformed the non-project villages in many aspects, while the latter generally enjoyed much better initial conditions than that of project villages. The beneficiaries obtained more development opportunities as evidenced by enhanced market participation, the strengthened ability to manage risks, increased girls' enrollment rate, and the reduced gender disparity: - 11 -  poverty p.c. food p.c. % village with access gender labor incidence income security* saving gap** migration (%) (yuan) (%) (yuan) electricity road telephone (%) (%) 1995 31.5 939.5 17.7 93.2 75.2 82.3 37.2 86 7.9 2001 12.9 1,422.1 4.0 351.9 99.0 99.0 74.0 97 12.5 * food security is measured by the percentage of population with per capita grain production less than 150 kg/year; ** gender gap is measured by the ratio of girls' enrollment rate over boys' enrollment rate for the age group of 7-15 years old. Project costs and benefits have been quantified for the five major components of (a) education, (b) labor mobility, (c) rural infrastructure (for the irrigation subcomponent), (d) land and farmer development, and (e) TVE development (for the rural market construction subcomponent). Total investment for these five major components accounts for about 75% of total project investment. Based on these five components, the overall economic rate of return (ERR) to the project, under several quite conservative assumptions, is estimated to be around 34%. All calculations are based on actual numbers and under several conservative assumptions. Therefore, the estimated ERRs for most components can reasonably be considered to be the lower bound of the real ERRs. As shown below, the results indicate that the labor mobility, rural infrastructure, and land and farmer development components have higher ERRs than other components. This is consistent with the field observations and results from various economic and social assessments. Detailed explanations and calculations, including methodologies, specific assumptions, crop budgets, and activity models are provided in Annex 3 of this ICR and the project files (see Annex 7 Document 14). Project as Education Labor Rural TVE Land and farmer ERR whole mobility infrastructure development ICR 34% 18% More than 50% 20% 16% 42% SAR 36% n.a. More than 50% n.a. 32-45% 40% 4.4 Financial rate of return: No effort was made to separately estimate the financial rate of return (FRR). Following more than two decades of market-oriented economic reform, price distortions in China have been largely eliminated. In addition, an integrated domestic market has been established and the domestic market is increasingly integrated with the international market. The SAR found no significant difference between the economic and financial rates of return for the three project components examined. On this basis, it is believed that the estimated ERRs are representative of the FRRs. 4.5 Institutional development impact: The project's institutional development impact was high. Project preparation was preceded by Strategies, which was a collaboration between the Bank and LGPR's senior policy makers on China's national poverty reduction program. These senior policy makers were then directly involved in the work of project design and implementation. The 1997 QAG Review (see Annex 7 Document 2) found this policy and institutional development impact to be best practice (see Section 4.1 above), and many of the recommendations and policy initiatives of Strategies and SWPRP are now explicitly included in the Government's principal poverty reduction program pronouncements. For example, the "National Seven Year Plan for Poverty Reduction" calls for strengthening the institutions responsible for implementing poverty reduction projects and programs, the concentration of available funding in the poorest counties, improved access for the poor to outside employment opportunities, greater investment in basic education and health services, and continued investment in poor-area basic infrastructure and agriculture. The SWPRP poverty reduction - 12 -  project approach has now become part of the foundation of national poverty reduction policy. In addition, the SWPRP poverty monitoring component greatly strengthened the NBS' poverty monitoring and analysis capability (see Section 4.2 above). 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: Factors outside the control of government and the implementing agency had a mostly favorable impact on project implementation. Most importantly, China's continuing strong economic growth in all years since the beginning of project implementation provided a highly positive environment for all of the poverty reduction activities supported by the project. For example, robust economic growth in the coastal provinces and in the project provinces provided ample employment opportunities for the labor mobility component to be successful. Regional and provincial large scale transport projects, including road, rail, air and water transport, were also undertaken during the project implementation period. This greatly improved transport capacity and provided additional job opportunities in the project area. The completion of the Nanning (Guangxi) to Kunming (Yunnan) railway in 1998 was particularly important. Project implementation was adversely affected by some natural disasters, including earthquakes in Yunnan and unusually cold weather in Guangxi. However, the impact of these natural disasters was generally modest, temporary, and localized. Declining market prices also adversely affected the profitability of a number of the project's agricultural products. The profitability of the production of silk, tea, and some livestock products was sharply reduced by adverse market price swings, but the price swings and loss of profitability for these crops was generally only temporary. However, severe overproduction at the local and national levels led to protracted and even catastrophic price declines for some tree crop and medicinal products -- including some fruits (such as litchi), some nuts (such as chestnut), and some other niche products (such as eucommia and gallnut). 5.2 Factors generally subject to government control: The influence of factors subject to government and implementing agency control was mostly favorable. Most importantly, the government and implementing agency maintained its strong commitment to effective poverty reduction throughout the period of project preparation and implementation. This strong commitment and the high priority attached to successful implementation was essential to overcoming the many obstacles which confronted the project during the initial years of project implementation. For example, a key implementation obstacle during the first year of implementation was the difficulty of coordinating and integrating the wide variety of project activities to be undertaken by the many technical bureaus involved. The strong leadership of the central and provincial governments and implementing agency, and the high profile they gave to the project, was essential to overcoming this problem of "bureau coordination." Similarly, the national program in support of achieving universal basic education provided a strong foundation for the project's basic education component. Some factors subject to government control adversely affected project implementation and outcome. Most importantly, the government's emphasis on cost recovery led to a bias during the project design stage in favor of those activities which were believed to be most likely to generate cash returns and against those activities which were perceived to have little or no likelihood of providing recoverable cash returns. This bias probably explains the government's preference, for example, for investments in larger scale TVEs and tree crops (which, in fact, did not generally result in the desired cash and fiscal revenues), and aversion to investments in basic education, basic health, and rural infrastructure. - 13 -  5.3 Factors generally subject to implementing agency control: The entire project management system was very strong and contributed greatly to successful project outcomes. In particular, the Central PMO and the Guangxi PMOs' performances were excellent. Performance at the county and lower levels was also exemplary, and steadily improved with assistance from the institution building component. In most cases, supervision missions found that county, township, and village level project staff were extremely dedicated to quality implementation. These staff's dedication to the well-being of the local population (and SWPRP's primary beneficiaries) was admirable. The PMO system developed a practice of "cross-supervision," under which provincial and county project staff regularly supervised the implementation performance of other project provinces and counties, which proved to be a very effective learning tool. In some limited number of cases, however, county PMOs were inadequately staffed and delegated too much control and responsibility to county technical bureaus. This facilitated the "hijacking" of project components and contributed to the erosion of the participatory process during implementation. 5.4 Costs and financing: Actual project costs were US$463.6 million, or about 5 percent less than the SAR estimate. This small difference is mostly explained by fluctuations in the Dollar/SDR and Dollar/Yuan exchange rates. As noted in Section 3.4 above, budget reallocations during implementation moved funding from the poorly performing TVE development component and GX-MEDs activities to increased scale and new activities under the land and farmer development and rural infrastructure components. 6. Sustainability 6.1 Rationale for sustainability rating: The overall project sustainability is rated highly likely, but varies considerably between the project's policy objectives and by component. The sustainability of the core policy objective (that is, of influencing national poverty reduction policy, as discussed in Section 3.1 above) is highly likely since the key reforms are now clearly part of government policy. The recent "Outline of China's Rural Poverty Alleviation" ("Outline") reconfirms the government's support for a comprehensive development approach to poverty reduction and calls for the "actively expand labor export from poverty-stricken areas." The Outline also calls for "enhancing the statistical and monitoring work in poverty reduction and development" and it is clear that the NBS has taken a lead role in the annual assessments of poverty in China. It is also very likely that the achievements of several project components will be sustained. First, the government is now firmly committed to universal basic education through sixth grade, and it seems very unlikely that the improvements in educational attainment in the project area would be allowed to diminish. Instead, the question now appears to be whether universal ninth grade education can be realized in the project area. Second, the labor mobility patterns established through the project are expected to be maintained and even expand over time. Although the fate of the project organizations which supported the labor mobility component during implementation is uncertain, the villagers who took advantage of the labor mobility component have established links to outside enterprises where they and other villagers will continue to seek off-farm employment. Third, many of the land and farmer development component activities, such as the adoption of improved varieties of staple foods and the extension of "fodder trees," will certainly continue. Similarly, most rural roads and village tracks built under the project will be maintained by the villagers who so greatly benefit from the improved access to transport. - 14 -  On the other hand, it is anticipated that some project activities may not be sustained beyond the implementation period. Most troubling, it seems likely that much of the improvement in basic health care services will deteriorate without continued government support. Although the medical workers trained through the project will not lose their skills, the lack of adequate financial incentives may cause them to move to better off areas. Village health clinics will likely be maintained, but funding for medicines will be curtailed and medicines and health services may become too expensive for the poorest villagers. Second, some of the rural infrastructure established under the project may not be well maintained. While China has a demonstrated ability to maintain roads, it is less certain that community drinking water facilities will be properly maintained. Efforts to establish proper maintenance practices and funds during the period of project implementation were not always successful. Lastly, the markets for some of the tree crops and other niche agricultural products supported under the project have already experienced sustained declines or even collapsed (see Section 5.1 above), and these activities are certainly not sustainable. 6.2 Transition arrangement to regular operations: Transition arrangements vary widely by province and even by project county. In Guangxi, most project counties have worked out arrangements by which a portion of the funds recovered from project activities can be used by the existing county PMOs for continuing poverty reduction work. In Yunnan, it was not possible to formalize these arrangements and cost recovery has been very limited. It appears likely, therefore, that the Yunnan county PMOs will be reabsorbed into the county governments and that little or no recovered funds would be available to extend the project activities beyond the formal completion of project implementation. It appears likely that many of the provincial PMO staff will be reabsorbed into the provincial governments and the provincial PMOs downsized. However, one of the key lessons of SWPRP has been the cost-effectiveness of enhanced supervision of poverty reduction expenditures, and provincial and local level supervision is being strengthened within and outside the project area. Therefore, many of the poverty reduction activities supported by SWPRP will continue in the project provinces with government financial support under the leadership and management of strengthened poverty reduction institutions through their regular operations. The Central PMO -- FCPMC -- has now achieved a self-sustaining level of operation, and has increasing contacts with a wide variety of international and multilateral donors. FCPMC will therefore continue beyond the SWPRP implementation period. 7. Bank and Borrower Performance Bank 7.1 Lending: The Bank's performance in the identification, preparation and appraisal of the project is rated highly satisfactory. Following on the recommendations of Strategies and supported by a PHRD Grant (of about US$1.2 million, and which was Bank executed), the Bank's preparation team was able to meet the challenge of assisting with the preparation of this complex and highly innovative project. Ford Foundation, CARE Australia, and the New Zealand Government provided extremely valuable assistance during the earliest stages and throughout project preparation and appraisal. Most of the Bank's preparation team could undertake their work in Chinese, and spent much of their project preparation time at the village and household level gaining a more in depth understanding of the problems and possible solutions. The single important shortcoming of the preparation effort was the failure to complete the Environmental Assessment work prior to appraisal. This delayed the overall project preparation process (an additional small mission was required), and the problem should have been anticipated and avoided from the outset. 7.2 Supervision: The Bank's performance during the implementation of the project is rated satisfactory. There has been an unusual continuity of the Bank team, with most of the core team having been maintained from the start of preparation through the final supervision mission, and there were two full supervision missions during each - 15 -  year of implementation. During the first years of implementation, the Bank supervision missions were comprehensive, and adequately covered all project components and safeguard requirements. The Bank team's relationship with staff of all levels of the PMO system was excellent, and a close collaboration on the common objectives of the project was maintained and expanded throughout the implementation period. Unfortunately, with limited funding for supervision of rural projects in China, and growing demands on the task team's time to prepare other projects, the comprehensiveness of the supervision effort faltered during the latter years of implementation. Most importantly, because of supervision funding constraints, the Bank supervision missions during the middle and latter years of implementation did not adequately cover the education, health, participatory process, impact and environmental monitoring, and financial management aspects of the project. Although this problem was successfully addressed in final three supervision missions when the team did include education, health and social specialists, overall project implementation would have been even better had adequate supervision budgets been available for this complex project and the Bank supervision missions been comprehensive in all years. 7.3 Overall Bank performance: Overall, the Bank performance during project preparation, appraisal and implementation is rated satisfactory. With the exception of a few of the latter years of implementation, the Bank's team was comprehensive and included the required expertise to adequately supervise the project. Most of the Bank's core team was maintained from the beginning of preparation through the completion of implementation, and the Bank team's collaboration with the staff of the PMO system was consistently excellent. This allowed for the rapid identification of project implementation difficulties and issues. Working closely with the PMO system, these problems and issues were successfully overcome one-by-one during the first years of implementation. Borrower 7.4 Preparation: The Borrower's performance in the preparation of this project is assessed to be highly satisfactory. During the early stages of preparation, staff of the Central PMO greatly facilitated the process by boiling down the project's core objectives and "principles" (that is, an early prototype of the project logframe) into a summary eight page document. With the assistance of a Technical Advisory Group (funded from the PHRD Grant), the project design was completed on a timely basis with adequate coverage of the technical, financial, economic, institutional, environmental and social factors. Although participatory project design was not yet popular in China at the time of project preparation, the Central and provincial PMOs energetically supported this concept. Senior government leaders strongly supported the project preparation process at all times, and this leadership proved essential to the successful preparation of this challenging project. The only significant flaw in Borrower's project preparation effort was the failure to bring capable procurement staff -- with a detailed hands-on understanding of Bank procurement guidelines -- into the Central and provincial PMOs during the preparation stage. 7.5 Government implementation performance: Government implementation performance is rated highly satisfactory. Senior members of the central and provincial governments provided strong support and leadership to the project throughout the implementation period. In a reportedly unprecedented step, the October 1995 SWPRP Implementation Work Conference (a) received a letter from President Jiang Zemin, (b) was chaired by a State Councilor (Chen Junsheng), and (c) required the attendance of the Governor (not the Vice Governor) of the three project provinces and the Directors (not Vice Directors) of the bureaus and agencies involved in project implementation. Known in the SWPRP project circle as the "Three Firsts," this exceptional Project Launch clearly established that successful implementation of the project would be carefully monitored and supported by the highest levels of government. One of the key lessons learned (see Section 8 below) from - 16 -  SWPRP is that this strong central and provincial government support and leadership was the key to successful preparation and implementation. A significant shortcoming of Government implementation performance was that counterpart funding was not always made available on a timely basis. Inadequate counterpart funding was a significant constraint in Yunnan and Guizhou. 7.6 Implementing Agency: The performance of the PMO system at all levels (that is, the "implementing agency") is rated highly satisfactory. Overall, the PMO system did an excellent job of responding to the many challenges inherent to preparing and implementing such a complex and novel project. As discussed in Section 5.3 above, the dedication of the PMO staff at all levels to successful implementation and a real improvement in the well-being of the project's primary beneficiaries is admirable. Nevertheless, there were some actions which could have been taken which could have further enhanced the performance of the PMO system. For example, the Yunnan PMO's effectiveness was hampered at times by the repeated changes in its Director in the early years of implementation, and by the failure to make a firm "yes or no" decision on the Fengguo Irrigation Canal scheme in the project's first years. As discussed in Section 5.3 above, the effectiveness of some county PMOs was diminished when too much responsibility and record keeping was turned over to county technical bureaus. 7.7 Overall Borrower performance: The overall performance of the Borrower is rated highly satisfactory. SWPRP was a complex multisectoral project which pushed the envelope in several directions. Most importantly, the labor mobility component, the focus on participatory design, and the emphasis on improving poverty monitoring must be considered to have been highly innovative in China in the early-1990s. Given this broad and innovative agenda, it was inevitable that initial implementation would be problematic. The Borrower accepted the challenge of preparing and implementing this project, overcame all of the initial implementation problems, and deserves recognition for a job well done. The State Council's LGPR played the single most important role in successfully meeting this challenge. The project would not have been successful without LGPR's very strong and consistent support throughout the entire project cycle. 8. Lessons Learned Although the project was highly satisfactory overall, at least as much can be learned from the project's shortcomings as from its achievements. Positive Lessons. Government Support. Perhaps the most important lesson is that, with strong support from the central and local governments, this multisectoral poverty reduction model can be made to work even in the poorest and most disadvantaged areas. Fortunately, the Central Government at the highest levels and the State Council's LGPR provided strong support and leadership for the project. This support was matched by the institution building component which strengthened the project management system at all levels (from the Central PMO down to the township and village Work Stations). This strong sustained support from the highest levels coupled with the hard work of the strengthened PMO system at all levels has been essential to the successful preparation and implementation of SWPRP. As stated in Section 7.7 above, the project would not have been successful without LGPR's very strong and consistent support throughout the entire project cycle. Synergy. There was good synergy from bringing in a great variety of poverty reduction activities down to the township and village level. In particular, early provision of support for basic education and health (in - 17 -  the form of new classrooms and health clinics, training for teachers and health workers, and tuition assistance and increased incomes for health workers, etc.) extended the outreach of the project to the poorest villages and greatly encouraged those villagers to participate in all other aspects of the project. There was a real organizational and management cost to providing such a wide range of benefits to the poor, but the PMO system rose to this challenge and found that the benefits and synergy from a multisectoral approach outweighed these costs. Effectiveness of Greater Supervision and Management. The project quickly illustrated that the effectiveness and efficiency of rural poverty reduction activities can be greatly enhanced through more support, supervision and management by the central and provincial levels of government. The active involvement of higher levels of government in the allocation and utilization of poverty reduction funding can greatly enhance its effectiveness, and a certain portion of poverty reduction program funding should be allocated to independent and rigorous supervision and monitoring of the use of those funds. In particular, the Directors of the provincial PMOs directly experienced improved results from strengthened supervision and monitoring under SWPRP, and they moved to extend this strengthened supervision system to poverty reduction work province-wide. "Negative" Lessons. False Expectations and Funding Mechanisms Lead to Design and Implementation Distortions. During SWPRP design and implementation, a number of less effective poverty reduction activities were pushed forward too forcefully and did not achieve the desired poverty reduction impact. Most notably, these ineffective activities included some (or even most) tree crops and rural enterprises. It seems that the inappropriately strong support for these activities stemmed from the ready availability of counterpart funding for these activities, the false expectation that these activities would produce cash returns at the farm and local level, and/or the old approach of top-down planning. At the same time, there has been inadequate support for some of the project activities which have been particularly well received by project beneficiaries and had the greatest poverty reduction effect. Such activities include all sorts of rural infrastructure and the applied agricultural research necessary to determine and prepare agricultural activities most suitable to the mountainous areas where most of China's absolute poor reside. It seems that the weak support for such activities has been a result of the inadequate availability of counterpart funding for these activities, uncertainty about how these funds would be repaid, and inadequate support for the participatory approach during implementation. The lesson is that funding and repayment arrangements should be designed so that (a) the project beneficiaries receive what they themselves say they want, and (b) the scale of the most cost-effective poverty reduction measures is not reduced or eliminated solely because it is difficult to determine who or what level of government should repay. Participatory Approach. Participatory project design and preparation was excellent and path breaking for its time. Unfortunately, the focus on community empowerment and the participatory approach was allowed to subside during implementation, and in some cases was overruled by county technical bureaus used to top-down planning. It is essential that mechanisms be developed to ensure that villages and poor households have more direct control over decision making during program implementation and that the participatory process continue into implementation, monitoring and evaluation. Ethnic Minorities. Ethnic minorities comprise an ever greater share of China's remaining absolute poverty, and their needs, aspirations, and background are often quite distinct from majority culture. Although SWPRP gave great emphasis to assisting ethnic minorities, careful analysis at the village level has shown that the most disadvantaged and poorest ethnic minorities did not always receive a fair share of project benefits. It is essential that better mechanisms be developed to ensure that ethnic minorities have - 18 -  full access to the entire range of project benefits and, in particular, that ethnic minority villages be in charge of their own project design, preparation and implementation. Basic Health. China has recognized the serious problem of the inadequacy of the public health care system in rural areas. This problem and its consequences are particularly severe in the poorest rural areas. China is now organizing a response to the inadequacy of the rural health care system, but a special program must be established to improve public health in the poorest rural areas. Given the extreme challenges of improving access to basic health services in the poorest areas, the health component of a multisectoral poverty reduction project such as SWPRP should receive special attention and support from the Central Ministry of Health. 9. Partner Comments (a) Borrower/implementing agency: As the first integrated large-scale and trans-regional poverty reduction project that directly targeting the poor farmer cooperated by the Chinese Government and the World Bank, the China Southwest Poverty Reduction Project (SWPRP) is considered a very successful one. Its design demonstrates the mainstream of anti-poverty theme nowadays. Under the strong leadership of Chinese Government and the sincere collaboration of relevant departments and agencies, the project has achieved the planned objectives through the great effort of project management agencies and the active participation of the broad poor farmers as well. The success of the project has provided instructive experiences in poverty reduction models, institutional innovation and other aspects, which have been used for China's poverty alleviation and development strive since nineties of last century and the formulation of poverty reduction policies for the new century. In addition, the project has greatly improved the development environment and the production and living conditions of the poor farmers in the project area, as well as contributed the income increasing of the poor. The major achievements of the project can be illustrated as follows: - Both the enrollment and completion of primary education have enjoyed a remarkable raise from 89% of year 1995 to 93% of year 2001 and from 46% of year 1997 to 76% of 2001 separately; - Because of the project implementation, the basic preventive health care service in the project areas is upgraded. Altogether there are 1727 poor villages realized the "Four Possessions", e.g. every village has a health post of more than 50 square meters, which are equipped with essential clinical equipment and basic medicine and the servicing staffs master basic medical techniques. 3504 health staffs from townships and villages and 3499 birth attendants received training; - The project constructed 2349 km of rural road and 1398 km of rural power transmission lines. The water drinking problem for over 830,000 people and more than 300,000 animals has been solved through construction of various water facilities such as family water tanks, public water ponds and water storage cellars. The construction of water conservancy works including irrigation, flood control, soil and water conservation facilities as well as the improvement of existing reservoirs guarantees the newly enlarged irrigated areas up to 13136 ha; - Households cash income and grain occupation has been also dramatically increased due to development of labor migration and agricultural production activities. The farmers' net income per capita has been raised from RMB 939 in 1995 to RMB 1422 in 2001, the percentage of farm households with grain occupation per capita less than 150 kg has also decreased from 17.7% to only 8.7%, the problem of "lack of cash and food" nagged the local poorest households in a long period in the project area has been basically solved. - Through 7 years project implementation, the poor population in the project area has been reduced greatly and poor situation is also dramatically changed. From 1995 to 2001, the poverty headcount rate is decreased from 31.5% to 12.9%, and the rate of weighted poverty gap fell down from 1.23 to 1.04, while the rate of poverty gap dropped from 5.8 to 3. - 19 -  The preparation and implementation of SWPRP has produced a wide impact upon several aspects below: - The innovation and exploration conducted in the aspect of planning methodology, project management and monitoring providing some good experiences and lessons for formulation of China Poverty Reduction and Development Strategy in the new century; - Various project activities bring direct benefits to the ethnic minorities households due to the most of the project areas located in the ethnic minorities concentrated area where their population counts for more than 50% of the total local population; - Some of project activities are planned for directly benefiting women, such as providing health care for pregnant and delivery women in the health component, stipulating certain proportion of women's participation in various training programs and giving priority to girls in regarding to the tuition assistance. All these efforts through the project implementation have brought lots of positive impacts on women; - The project has produced long-term effect on capacity building at three levels: firstly, the institutional capacities and working skills of poverty reduction and project management agencies at the central, provincial and county level have been enhanced and improved; secondly, the capacity of self-management, self-organization and self-development of the poor communities has been upgraded; lastly, the overall quality of the farmers and their productive skills has been also increased; - At the same time, some project activities were designed specifically aiming at the ecological environment recovery and to relieve the environmental pressure, some exciting achievements is resulted in the aspects of soil protection and vegetation cover increasing, which has produced a positive impact on local environment. The lessons learned from the design and implementation of SWPRP: - Project design could be improved in a more scientific and operational-oriented way. Such as: (1) when education posts were programmed, the impact on building central schools brought by conduction of family planning policy and construction of village road should be taken into full consideration; (2) as for the issues of in-school supplementary nutrition, exemption and reduction of tuition and book fees as well as establishment of village cooperative health care foundation, detailed studies should be given on policy background and the effectiveness of implementation conditions; (3) low unit prices for part of project works, high proportion of self-collected funds brought difficulties on project implementation as well as post management; - During the design stage, the project counties were requested to provide 25% of the counterpart funds. In fact, all these poor counties are financial deficit and not possible to fulfill the task, which affected the implementation progress and project quality to some extent; - Flexibility should be adopted during the project implementation so that to fit in with the change of the market; - The role of TVEs development in poverty reduction should be reconsidered and discussed. As for the only unsuccessful sub-component in SWPRP, there are many lessons and experiences to be summed up. Except the impact of the national policies, we should also be concerned about following issues: (1) small-scaled rural market should be established in the project area for promoting commodities circulation; (2) project scope should not be over large; (3) the enterprises established should meet the requirement of environmental protection and the state industry development policies; (4) the enterprises ownership should be clear and definite; (5) the overall arrangements on the issues of operating funds and credit guarantees for the enterprises should be planed as a whole; (6) in consideration of ensuring raw materials supply, development of the processing enterprises with close relevance of local agricultural production should be encouraged; (7) the purchasing procedures should be further simplified. - 20 -  (b) Cofinanciers: none (c) Other partners (NGOs/private sector): none (d) External Reviewer: - 21 -  - 22 -  - 23 -  10. Additional Information SARS and Infectious Diseases Response Program The amount of IDA financing used for the Southwest Poverty Reduction Project was about US$1.2 million less than the amount estimated at appraisal. This modest reduction in costs stems from exchange rate changes during implementation and per unit cost savings achieved by the beneficiaries. In 2003, the Borrower and IDA agreed to use these cost savings to support a SARS and Infectious Diseases Response Program in China. In combination with savings reallocated from several other projects and in collaboration with other donors, the SARS and Infectious Diseases Response Program in China will (a) address emergency needs for SARS-related diagnosis, clinical management, and personal protection that are necessary to bring the current epidemic fully under control, and (b) support the Government's efforts to strengthen the capacity of the public health system more generally so that China's health system is prepared to combat the possible re-emergence of SARS and, equally important, similar infectious disease threats that might occur in the future. The reallocation for the SARS Program was approved by the Board on June 19, 2003. This ICR for SWPRP will be supplemented by an assessment of the implementation of the SARS Program that will be carried out within six months of the revised closing date of the Credit, in April 2004. - 24 -  Annex 1. Key Performance Indicators/Log Frame Matrix Outcome / Impact Indicators: Indicator/Matrix Baseline End of Project A.) Poverty Reduction Model: (1) Establish Approach did not exist. Approach adopted as a national policy under a focused multisectoral rural development the National Poverty Reduction Plan project approach to poverty reduction in (2001-2010). China. A.) (2) extend this model to China's other Model did not exist. Model extended nationwide. poor counties. B.) Reduce Poverty: Bring as many as Poverty incidence: 31% Poverty incidence: 12.9% possible of the project's primary beneficiaries Per capita income: 939.5 Per capita income: 1422.1 out of absolute poverty as indicated by (3) per capita income levels above the poverty line B.) (4) greater than subsistence levels of Percentage of population with per capita Percentage of population with per capita grain production. grain production less than 150kg/per year: grain production less than 150kg/per year: 17.7% 4.0% Output Indicators: Indicator/Matrix Baseline End of Project C.) Labor Mobility: Establish a new model of Rate of of labor mobility: 7.9% Rate of of labor mobility: 12.5% organized access to off-farm employment for the absolute poor as indicated by (5) the Total number of poor finding off-farm job Total number of poor finding off-farm job numbers of upland poor finding off-farm through the project: 0 through the project: 280,123 employment through the project. C) (6) remittance by these successful migrant laborers to their "home-village" Total remittances: 0 Total remittances: RMB 1590 million yuan families. Per capita: 0 Per capita: RMB 231.5 yuan D. Social services: improve the poor's access to basic social services, resulting in Enrollment rate (7-12 yrs. old group): 89% Enrollment rate (7-12 yrs. old group): 92.5% (7) increased primary school attendance rates. D. (8) Completion rate Completion rate of children aged 15: 77.5% Completion rate of children aged 15: 45.9% Guangxi: 19.5 per thousand D. (9) reduced infant mortality rates Guangxi: 49.8 per thousand Guizhou: 50.0 per thousand Guizhou: 81.5 per thousand - 25 -  Annex 2. Project Costs and Financing Project Cost by Component (in US$ million equivalent) Appraisal Actual/Latest Percentage of Estimate Estimate Appraisal Component US$ million US$ million Basic Education 24.93 29.71 119.19 Basic Health Services 20.49 20.71 101.04 Labor Mobility 112.71 133.33 118.3 Rural Infrastructure 39.04 63.99 163.91 Land and Farmer Development 111.94 160.30 143.19 TVE Development 53.95 42.31 78.42 Institution Building 6.30 6.05 96.01 Poverty Monitoring 5.52 7.15 129.65 Total Baseline Cost 374.88 463.55 Physical Contingencies 20.00 Price Contingencies 91.50 Total Project Costs 486.38 463.55 Total Financing Required 486.38 463.55 Project Costs by Procurement Arrangements (Appraisal Estimate) (US$ million equivalent) 1 Procurement Method Expenditure Category ICB NCB 2 N.B.F. Total Cost Other 1. Works 0.00 70.00 141.20 0.00 211.20 (0.00) (24.30) (28.40) (0.00) (52.70) 2. Goods 42.20 73.00 28.50 0.00 143.70 (42.20) (54.70) (21.40) (0.00) (118.30) 3. Services 0.00 0.00 29.90 0.00 29.90 (0.00) (0.00) (21.40) (0.00) (21.40) 4. Land Acquisition 0.00 0.00 0.00 12.10 12.10 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Other 0.00 0.00 72.80 16.70 89.50 (0.00) (0.00) (55.10) (0.00) (55.10) Total 42.20 143.00 272.40 28.80 486.40 (42.20) (79.00) (126.30) (0.00) (247.50) - 26 -  Other includes Tuition and Medical Assistance, Labor Training and Placement, and Others. Project Costs by Procurement Arrangements (Actual/Latest Estimate) (US$ million equivalent) 1 Procurement Method Expenditure Category ICB NCB 2 N.B.F. Total Cost Other 1. Works 0.00 14.12 173.55 12.82 200.49 (0.00) (3.94) (45.45) (0.00) (49.39) 2. Goods 39.97 46.47 46.62 12.62 145.68 (39.97) (46.47) (34.02) (0.00) (120.46) 3. Services 0.00 0.00 74.28 0.00 74.28 (0.00) (0.00) (53.62) (0.00) (53.62) 4. Land Acquisition 0.00 0.00 0.00 19.43 19.43 (0.00) (0.00) (0.00) (0.00) (0.00) 5. Other 0.00 0.00 4.24 1.34 5.58 (0.00) (0.00) (1.26) (0.00) (1.26) Total 39.97 60.59 298.69 46.21 445.46 (39.97) (50.41) (134.35) (0.00) (224.73) Other includes Tuition and Medical Assistance, Labor Training and Placement, and Others. 1/Figures in parenthesis are the amounts to be financed by the Bank Loan. All costs include contingencies. 2/Includes civil works and goods to be procured through national shopping, consulting services, services of contracted staff of the project management office, training, technical assistance services, and incremental operating costs related to (i) managing the project, and (ii) re-lending project funds to local government units. Project Financing by Component (in US$ million equivalent) Percentage of Appraisal Component Appraisal Estimate Actual/Latest Estimate Bank Govt. CoF. Bank Govt. CoF. Bank Govt. CoF. Basic Education 17.96 14.44 13.94 15.78 77.6 109.3 Basic Health Services 18.39 8.02 12.80 7.91 69.6 98.6 Labor Mobility 80.90 64.31 71.62 61.71 88.5 96.0 Rural Infrastructure 22.18 28.72 13.31 50.68 60.0 176.5 Land and Farmer 59.85 86.20 68.63 91.67 114.7 106.3 Development TVE Development 38.37 31.74 29.26 13.05 76.3 41.1 Institution Building 4.87 3.31 2.98 3.07 61.2 92.7 Poverty Monitoring 4.97 2.17 2.43 4.72 48.9 217.5 - 27 -  Annex 3. Economic Costs and Benefits Overall Evaluation As evidenced by a variety of available data, the project achieved remarkable economic and social benefits. These data (see Annex 7, Document No. 13) consistently confirm the improvements of a series of key monitoring indicators. The incidence of poverty, for example, declined from 31.5% in 1995 to 12.9% in 2001, and per capita income increased from 939.5 yuan to 1422 yuan during the same period (see Annex 3 Table 1 below). The beneficiaries obtained more development opportunities as shown by the enhanced market participation, strengthened ability of managing risks, increased girl's enrollment rate, and the reduced gender disparity. Project villages outperformed the non-project villages in many aspects, while the latter generally enjoyed much better initial conditions than that of project villages. For example, in 2001, the incidence of poverty is almost the same for both the project and non-project villages. However, in 1995, the incidence of poverty was nearly 11 percentage point higher in the project villages. The overall economic rate of return (ERR) to the project as a whole, under several quite conservative assumptions, is estimated to be around 34%. Annex 3 Table 1: Changes of Series Key Monitoring Indicators poverty p.c. food p.c. % village with access gender labor incidence income security* saving gap** migration (%) (yuan) (%) (yuan) electricity road telephone (%) (%) 1995 31.5 939.5 17.7 93.2 75.2 82.3 37.2 86 7.9 2001 12.9 1,422.1 4.0 351.9 99.0 99.0 74.0 97 12.5 * food security is measured by the percentage of population with per capita grain production less than 150 kg/year; ** gender gap is measured by the ratio of girls' enrollment rate over boys' enrollment rate for the age group of 7-15 years old. Rate of Return Estimation Data for calculating the ERR comes from a series of Poverty Monitoring Reports (compiled by the National Bureau of Statistics, 1996-2001) and representative village surveys (conducted by provincial PMOs). Project benefits have been quantified for five major components. These are (a) education, (b) labor mobility, (c) rural infrastructure (for the irrigation subcomponent), (d) land and farmer development, and (e) TVE development (for the rural farmers' market construction subcomponent). Total investment for these five major components accounts for 75% of total project investment. The benefits of other components are difficult to quantify and hence ERRs have not been calculated for them. No effort was made to separately estimate the financial rate of return (FRR). After more than two decades of market-oriented economic reform, price distortions due to the government's planning have been significantly reduced and in many cases completely eliminated. An integrated domestic market has now been established and it is increasingly integrated with the international market. Against this background, there is essentially no additional value to separately calculate ERRs and FRRs to capture the impact of price distortion. Annex 3 Table 2 presents estimated ERRs for the whole project as well as for specific components. - 28 -  Annex 3 Table 2: Rates of Return For The Whole Project And Individual Components Project as Education Labor Rural TVE Land and farmer whole mobility infrastructure development ICR 34% 18% More than 50% 20% 16% 42% SAR 36% n.a. More than 50% n.a. 32-45% 40% Basic Education: Ravallion and Chen (1997), by using the data from the project provinces, showed that the per capita income of those who received elementary education is about 20% higher than those who are illiterate. Based on this key parameter, the per capita income data from 1996 to 2001, and a few conventional and conservative assumptions, the rate of return to elementary education is around 18%. We can reasonably argue that the 18% is the lower bound of the real ERR to the education component, given that many other benefits are not included in this calculation. For example, we did not quantify the benefits of the adult vocational training program, local libraries construction, teachers' training, and classroom constructions. Research conducted by the International Food Policy Research Institute (IFPRI, 2001) showed that investment in education is the most profitable investment in terms of its return to rural GDP in the western China. IFPRI estimated that one yuan investment in education would have five yuan return. Labor Mobility: Labor mobility has overwhelmingly been rated by the project beneficiaries as one of their most favorite components. In many cases, farm households can repay what they borrowed from the project within the same year, which makes ERR undefined. NBS's Poverty Monitoring Report also consistently show that improving off-farm job opportunity has considerably contributed to the increase of per capita income. Under a set of very conservative assumptions, by using the data from the representative village survey, the ERR for labor mobility is estimated to be more than 50%. Rural Infrastructure: Rural infrastructure component is also one of components most welcomed by the beneficiaries. The Poverty Monitoring Report (see Annex 7 Document 13) shows that the small irrigation subcomponent results in 1.35% increase of per capita income. Based on this parameter and other three moderate assumptions, the rate of return to the small irrigation sub-component is estimated around 20%. This ERR can also be treated as the lower bound of the real ERR for the whole rural infrastructure component. For example, when asked to rank the sub-components of infrastructure component, farmers always give the first two priorities to the road construction and electrification. Farmers generally rank the small irrigation as the third priority. Therefore, we can reasonably believe that ERRs for road and electrification should be higher than that of small irrigation in order to match farmers' preference order. Moreover, other researches (for example, Rozelle and Huang, 2003) also found the similar conclusion, which gives us additional confidence over our findings. TVE Development: The focus of the TVE development component has considerably shifted after the mid-term review and the component size has been reduced (see Sections 3.4 and 4.2). The component comprises the two subcomponents of small farmers' markets construction (which accounts for 30% of total component cost) and TVE enterprise development (which accounts for the remaining 70%). For the small market place construction sub-component, based on the data from Guangxi (including information of all 52 newly constructed markets), and extrapolating the benefits over 20 years, the ERR for constructing local market is about 42%. For the enterprise development sub-component, we do not have enough data to calculate the ERR; however, we assume it is around 5% given its generally unsatisfactory performance. - 29 -  Weighted by their share in the total project investment, the ERR for the whole TVE component is estimated 16%. Land and Farmer Development: Following the conventional methodology, the ERRs for the three sub-component have been calculated: livestock sub-component (including goat, pig, and silkworm), cash-crop/tree sub-component (including tea, sugar cane, and longan), and the grain sub-component (including maize and rice). The overall ERR for the whole component is about 42%, ranging from 16% (longan) to 66% (silkworm). Detailed calculations, including methodologies, specific assumptions, crop budgets, and activity models are provided in the project files (see Annex 7 Document 14). - 30 -  Annex 4. Bank Inputs (a) Missions: Stage of Project Cycle No. of Persons and Specialty Performance Rating (e.g. 2 Economists, 1 FMS, etc.) Implementation Development Month/Year Count Specialty Progress Objective Lending 07/93 10 E/ML, (2)FI, (2)E, PH, AQ, M/FI/LB, AG, T 01/94 10 E/ML, E, M/FI/LB, NRE, PH, E&BE, (2)AG, E (CA), T 5/94 8 E/ML, E, M/FI/LB, E&BE, PH, (2)AG, NRE (FF) 9/94 11 E/ML, E, M/FI/LB, E&BE, PH, (2)AG, AN, E&IB, (2)UP 11/94 2 AQ, E 11/94 11 E/ML, DC/A, E&BE, 2UP, M/FI/LB, PS E&IB, AG, AQ, PH 3/95 6 E/ML, M/FI/LB, E&BE, PH, AG, PH Supervision 07/95 8 E/ML, M/FI/LB, E&BE, ENV, HS HS E&IB, ENG, PS E 10/95 15 E/ML, M/FI/LB, AQ, ENV, S S E&IB, E&BE, (3)AG, RED, PH, RS, PS AN, E 04/96 12 E/ML, ENV, PS E&IB, E&BE, S HS RED, PH, (3)AG, SDS, WDS 10/96 9 E/ML, PH, ENV, AQ, E&IB, S HS E&BE, RED, (2)AG 01/97 7 E/ML, E, E&IB, LM, E&BE, S HS AN, AG 04/97 13 E, E&BE, EXT, 10IJ 10/97 9 E/ML, EDU, LM, E&IB, (2)AG, S HS PS AQ, T 04/98 7 E/ML, E&LM, E&IB, (2)AG, S HS (2)E&PM 07/98 3 (2)E, VPDEC 11/98 10 E/ML, RED, E&IB, (2)E, (2)AG, S HS E&PM, PH, CP 05/99 8 E/ML, RED, E&IB, EDU, (2)E, S HS AG, RS 10/99 3 E/ML, E&IB, AG S HS 05/00 13 E/ML, E&PM, RED, E&IB, S HS EDU, AG, FM, PH, RS, (3)AN, VPEXT 03/01 4 E/ML, E&LM, E&IB, AG S HS 10/01 5 ML, E&IB, RED, E/EDU, AG S HS 03/02 9 E/ML, E/EDU, DO, E&IB, S HS - 31 -  (2)AG, PS PH, RED (DFID) ICR 10/02 9 E/ML, E&LM, LM, E, AG, S HS E/EDU, ENV, RED, PH AG = Agriculturalist; AN = Anthropologist; AQ = Aquaculturist; CP = Community Participation; DC/A = Division Chief/Agriculture; DO = Disbursement Officer; E = Economist; E (CA) = Economist (Care Australia); E&BE = Economist & Basic Education; E&IB = Engineer & Institution Building; E&LM = Economist & Labor Mobility; E/ML = Economist/Mission Leader; E&PM = Economist & Poverty Monitoring; EDU = Education; ENG = Engineer; ENV = Environmentalist; EXT = External Affairs; FI = Financial Institutions; FM = Financial Management; IJ = International Journalist; LM = Labor Mobility; M/FI/LB = Microenterprise, Financial Institutions & Labor Mobility; ML = Mission Leader; NRE (FF) = Natural Resources Economist (Ford Foundation); PS = Procurement Specialist; PH = Public Health; RED = Rural Enterprise Development; RED (DFID) = Rural Enterprise Development (Department for International Development); RS = Resettlement Specialist; SDS = Social Development Specialist; T = Training; UP = Urban Planner; VPDEC = Vice President Development Economics & Chief Economist; VPEXT = Vice President External Affairs; WDS = Women in Development Specialist. (b) Staff: Stage of Project Cycle Actual/Latest Estimate No. Staff weeks US$ ('000) Lending 349 Supervision 538 ICR Total 887 ICR costs incorporated in Supervision. - 32 -  Annex 5. Ratings for Achievement of Objectives/Outputs of Components (H=High, SU=Substantial, M=Modest, N=Negligible, NA=Not Applicable) Rating Macro policies H SU M N NA Sector Policies H SU M N NA Physical H SU M N NA Financial H SU M N NA Institutional Development H SU M N NA Environmental H SU M N NA Social Poverty Reduction H SU M N NA Gender H SU M N NA Other (Please specify) H SU M N NA Ethnic Minorities - The project made a substantial contribution to the well-being of ethnic minority people. Although more could have been done, the project was path breaking for its time in this aspect. Private sector development H SU M N NA Public sector management H SU M N NA Other (Please specify) H SU M N NA - 33 -  Annex 6. Ratings of Bank and Borrower Performance (HS=Highly Satisfactory, S=Satisfactory, U=Unsatisfactory, HU=Highly Unsatisfactory) 6.1 Bank performance Rating Lending HS S U HU Supervision HS S U HU Overall HS S U HU 6.2 Borrower performance Rating Preparation HS S U HU Government implementation performance HS S U HU Implementation agency performance HS S U HU Overall HS S U HU - 34 -  Annex 7. List of Supporting Documents 1. Southwest Poverty Reduction Project Staff Appraisal Report 2. Quality Assurance Group's Post-Approval Quality at Entry Assessment 3. Aide Memoires 4. Southwest Poverty Reduction Project "Yanshou" Evaluation Report (by FCPMC in Chinese and English) 5. Component Evaluations 5.1 Education Component 5.2A Health Component (Oct 2002) 5.2B Health Component (Mar 2002) 5.3 Labor Mobility Component 5.4 Rural Infrastructure Component 5.5 Farmer Development Component 5.6 Town and Village Enterprises Development Component 5.7 Poverty Monitoring Component by Terry Sicular 5.8 Poverty Monitoring Component by National Bureau of Statistics 6. "Are the Income Gains from a Development Project Consumed or Saved" (a paper by Shaohua Chen and Martin Ravallion) 7. Country Assistance Evaluation by A. R. Khan 8. China: Strategies for Reducing Poverty in the 1990s 9. China: Overcoming Rural Poverty 10. Participatory Poverty Appraisal Report (April 2003) 11. Guangxi MEDS Follow-Up Evaluation Report (September, 1999) 12. Adjustment Plan of Guangxi Hard Loan Project (Guangxi PMO, May 2000) 13. SWPRP Poverty Monitoring Report (NBS, each year 1996-2000) 14. Rate of Return Calculations (spreadsheet) and Underlying Data (in Chinese) 15. Village-Based Multisectoral Poverty Reduction Approach (in PowerPoint) 16. SWPRP Labor Mobility "Yanshou" Reports (by CWHRDC in English and Chinese) - 35 -  Annex 8. Beneficiary Survey Results Participatory Approach and Stakeholder Workshops. Following project completion, the Bank financed an independent participatory study of the project's overall impact and the participatory process. The Participatory Poverty Appraisal Report (PPAR), undertaken by China Cross-Cultural Consulting Center (CCCC) in April 2003, (a) concludes that the project achieved favorable results in each of the five areas examined, (b) identifies a large number of "existing problems," (c) proposes a number of solutions to these problems, and (d) finds that the participatory process could have been strengthened. In addition, the PPAR's findings on female and ethnic minority people's participation in the project are summarized in Section 4.2 above. Overall, the PPAR is both a powerful qualitative evaluation of the effectiveness and impact of SWPRP, and an extremely useful guideline for improving poverty reduction work in future projects. The PPAR reports that the project "actually relieved poverty in the affected areas" through increases and improvements in human resources and environmental, infrastructure, financial and social assets. The project improved the "quality of human resources" through increased consumption of basic goods (including food and other basic necessities), improved labor capacity and skills, higher educational attainment and health status, and the introduction of advanced technologies. Environmental deterioration was reduced or reversed and land productivity enhanced through terracing, water resource development, soil improvement and other methods. Basic infrastructure was strengthened by the construction of roads, drinking water systems, and electricity systems, and farm productivity was boosted through the provision of basic tools, improved seed, and other inputs. Financial assets increased and new financial channels were introduced. Lastly, social assets were strengthened through the build up of the project management system. Overall, the PPAR found that the project helped reduce poverty and that the multisectoral model was successful. The PPAR also identifies a large number of significant problems and difficulties which were not adequately overcome during project implementation. A total of 105 difficulties and problems are listed in the PPAR, including the following key issues: