Document of The World Bank Report No: ICR1835 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-73100 TF-54747 TF-55699) ON A LOAN IN THE AMOUNT OF US$100.0 MILLION TO THE PEOPLE‟S REPUBLIC OF CHINA FOR A POOR RURAL COMMUNITIES DEVELOPMENT PROEJCT December 20, 2011 China and Mongolia Sustainable Development Unit Sustainable Development Department East Asia and Pacific Region CURRENCY EQUIVALENTS (Exchange Rate Effective as of December 20, 2011) Currency Unit = Renminbi (RMB) or Yuan (Y) Y 1.00 = US$ 0.17 US$ 1.00 = Y 6.4 FISCAL YEAR July 1 – June 30 ABBREVIATIONS AND ACRONYMS ACWF All China Women's Federation CAS Country Assistance Strategy CBR Community Based Rehabilitation CCCC China Cross-Cultural Consulting Center CDPF China Disabled Persons Federation CDD Community Driven Development CDPF China Disabled Peoples Federation CLD Community Learning Center CLG Central Liaison Group CMS Cooperative Medical System CPMO County PMO CPS Country Partnership Strategy DFID Department for International Development DMIS Disability Management Information System DRC Development and Reform Commission EA-EMP Environmental Assessment-Environmental Management Plan EMDP Ethnic Minority Development Plan FCPMC Foreign Capital Project Management Center HI Handicap International IBRD International Bank for Reconstruction and Development IDA International Development Association KPI Key Performance Indicator LGOP State Council‟s Leading Group Office for Poverty Reduction MOF Ministry of Finance M&E Monitoring and Evaluation MTR Mid Term Review NSB National Statistics Bureau O&M Operation and Maintenance PADO Poor Area Development Office PDO Project Development Objective PHRD Japan Policy and Human Resources Development Fund PIM Project Implementation Manual PM Participation Manual PLG Project Leading Group PM&E Participatory M&E PMO Project Management Office PP Participatory Planning PPA Participatory Poverty Assessment PPMO Provincial Project Management Office PRCDP Poor Rural Communities Development Project (Poverty 4) PwD People with Disabilities QAG Quality Assurance Group QBPRP Qinba Mountains Poverty Reduction Project (Poverty 2) SDPRAP Sustainable Development in Poor Rural Areas Project (Poverty 5) SWPRP Southwest Poverty Reduction Project (Poverty 1) TPWS Township Project Work Station VDP Village Development Plan VPEG Village Project Execution Group YDPF Yuexi Disabled Peoples Federation Vice President: James W. Adams Country Director: Klaus Rohland Sector Manager: Paul Kriss (Acting) Project Team Leader: Alan Piazza ICR Team Leader: Alan Piazza   INTRODUCTION The Poor Rural Communities Development Project (PRCDP) was the fourth in a series of five household and village level poverty reduction projects designed and implemented by the State Council‟s Leading Group Office for Poverty Reduction (LGOP) system with support from the World Bank and other development partners. PRCDP was a complex multisectoral project including assistance for agricultural production, rural infrastructure, basic education and health, and community capacity building. It was also a highly innovative project which spearheaded the development of a robust participatory approach, and the project‟s outreach to the poorest and most disadvantaged ethnic minority people in western China was extraordinary. The United Kingdom‟s Department for International Development (DFID) provided very generous grant assistance to the overall project, and separate additional DFID grants provided vital assistance for the development of the project‟s participatory approach, the rigorous and cross-cutting monitoring and evaluation system, the Free Education Pilot, and an extraordinary assistance for people with disabilities activity in Yuexi County in Sichuan. The project would not have happened without DFID‟s grant assistance, and the DFID team fully participated in all of the project preparation, supervision, and evaluation missions. The core World Bank and LGOP system project team was maintained for a full decade from the beginning of project preparation in 2001 through to the conclusion of this ICR in 2011. Most importantly, the provincial, county and lower level project management teams designed, implemented and evaluated the project with tremendous resolve and intensity. Overcoming several substantial obstacles along the way, the provincial, county and lower level project management teams were most directly responsible for PRCDP‟s overall success.   CHINA POOR RURAL COMMUNITIES DEVELOPMENT PROJECT CONTENTS Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph 1. Project Context, Development Objectives and Design ............................................... 1 2. Key Factors Affecting Implementation and Outcomes .............................................. 6 3. Assessment of Outcomes .......................................................................................... 18 4. Assessment of Risk to Development Outcome......................................................... 31 5. Assessment of Bank and Borrower Performance ..................................................... 31 6. Lessons Learned ....................................................................................................... 34 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners .......... 35 Annex 1. Project Costs and Financing .......................................................................... 36 Annex 2. Outputs by Component ................................................................................. 38 Annex 3. Economic and Financial Analysis ................................................................. 52 Annex 4. Bank Lending and Implementation Support/Supervision Processes ............ 55 Annex 5. Beneficiary Survey Results ........................................................................... 57 Annex 6. Stakeholder Workshop Report and Results................................................... 58 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ..................... 59 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ....................... 60 Annex 9. List of Supporting Documents ...................................................................... 65 MAP   A. Basic Information POOR RURAL COMMUNITIES Country: China Project Name: DEVELOPMENT PROJECT IBRD-73100,TF- Project ID: P071094 L/C/TF Number(s): 54949,TF-70354 ICR Date: 12/20/2011 ICR Type: Core ICR PEOPLE'S REPUBLIC Lending Instrument: SIL Borrower: OF CHINA Original Total USD 100.00M Disbursed Amount: USD 99.71M Commitment: Revised Amount: USD 99.71M Environmental Category: B Implementing Agencies: Ministry of Finance, State Council‟s Leading Group Office for Poverty Reduction, and the Guangxi, Sichuan, and Yunnan Foreign Capital Project Management Offices Cofinanciers and Other External Partners: Department for International Development (DFID) of the United Kingdom provided grant support for the project amounting to US$37 million. B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 10/09/2001 Effectiveness: 12/15/2005 12/15/2005 Appraisal: 11/01/2004 Restructuring(s): Approval: 06/21/2005 Mid-term Review: 06/16/2008 06/17/2008 Closing: 12/31/2010 06/30/2011 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Moderate Bank Performance: Satisfactory Borrower Performance: Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Satisfactory Government: Moderately Satisfactory Quality of Supervision: Satisfactory Implementing Highly Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry No None at any time (Yes/No): (QEA): Problem Project at any Quality of Yes Moderately Satisfactory time (Yes/No): Supervision (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General agriculture, fishing and forestry sector 65 44 General education sector 6 7 General water, sanitation and flood protection sector 6 7 Roads and highways 13 34 Sub-national government administration 10 8 Theme Code (as % of total Bank financing) Education for all 11 11 Indigenous peoples 22 22 Other rural development 22 22 Participation and civic engagement 23 23 Rural services and infrastructure 22 22 E. Bank Staff Positions At ICR At Approval Vice President: James W. Adams Jemal-ud-din Kassum Country Director: Klaus Rohland David Dollar Sector Manager: Paul Kriss (Acting) Mark D. Wilson Project Team Leader: Alan Piazza Alan Piazza ICR Team Leader: Alan Piazza ICR Primary Author: Alan Piazza F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) Improve livelihoods security and achieve sustained participation of the poorest rural people in project design, implementation, and M&E in eighteen counties of the three project provinces. Revised Project Development Objectives (as approved by original approving authority) The objectives of the project remained unchanged throughout the implementation period. (a) PDO Indicator(s) The original Project Design Summary (Annex 1 of the PAD) included more than 50 Key Performance Indicators (KPI). Many of the Project Design Summary KPI were difficult or impossible to quantify, and that list of KPI was not implemented or monitored by the Bank team during implementation. Instead, Schedule 7 of the Loan Agreement specified a more practical list of 14 KPI for each of the project provinces. The three project provinces carefully considered and selected the Loan Agreement KPI, and regularly reported these Loan Agreement KPI to the Bank during the entire implementation period and at project completion. These Loan Agreement KPI have been recorded in all of the project archived ISRs, are reported below, and are used in this ICR for assessing project performance and outcomes. The Loan Agreement KPI for Sichuan and Yunnan are identical, but the KPI for Guangxi are different and are reported separately below. The source of data in the table below is the project‟s internal MIS and M&E systems. Additional detailed results from the external monitoring system are reported in the main text of this ICR. As explained in Section 2.3 of the main text, this additional evaluation information comes from Modules 1, 2 and 3 of the external monitoring system funding by the project and with the assistance of DFID grants.1 1 The external independent impact assessment system monitored the project's participatory approach and its impact on poverty and well-being through three “modules� including both quantitative and qualitative approaches. Module 1 comprised annual quantitative surveys of 1800 project and nonproject households (100 households in each of the project‟s 18 counties) undertaken by China‟s National Statistics Bureau (NSB). Module 2 comprised annual qualitative surveys of the project‟s process and outcomes undertaken by the ITAD Company. The Module 2‟s Participatory Poverty Monitoring was based on assessments by local communities which tracked poverty impacts and provided feedback on people‟s demand and project performance. Module 3 provided direct support for a participation specialist to join all of the Bank/DFID project preparation and supervision missions. Sichuan and Yunnan KPIs Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years PDO Indicators 1 and 2 (Participation): Number of villages None (Sichuan) 519 (Sichuan) conducting (Date: 2002) 516 (Sichuan) (Date: 2010) -- participatory None (Yunnan) 187 (Yunnan) 187 (Yunnan) poverty (Date: 2001) (Date: 2010) assessments (PPA) and participatory planning Number of None (Sichuan) 106,190 (Sichuan) households (Date: 2002) 105,846 (Sichuan) (Date: 2010) participating in 420 (Yunnan) 91,995 (Yunnan) 76,450 (Yunnan) productive (Date: 2001) (Date: 2010) activities Highly Successful: All project villages conducted PPA and participatory planning. The project rolled out the participatory approach on an Comments unprecedented scale, involving a very substantial number of remote natural villages in participatory planning and decision making, and the project became an influential platform for innovative practices in China. PDO Indicators 3 and 4 (Gender): 0% (Sichuan) 51.0% (Sichuan) Percentage of (Date: 2002) 50.0% (Sichuan) (Date: 2002) women -- 3.3% (Yunnan) 41.7% (Yunnan) 45.1% (Yunnan) participating in (Date: 2001) (Date: 2011) participatory planning 0% (Sichuan) 33.8% (Sichuan) Percentage of (Date: 2002) 50.0% (Sichuan) (Date: 2002) women receiving -- 4% (Yunnan) 50.0% (Yunnan) 51.8% (Yunnan) training (Date: 2001) (Date: 2011) Successful: Gender equality was emphasized during all stages of the project. The KPIs for gender equity were the percentage of women taking part in participatory planning and the percentage of women receiving training, and the values reported for these indicators showed significant increases (and mostly exceeded the agreed upon project goals in each of the three project provinces). In addition, the M&E Module 2 reports note that (a) women Comments participated in the selection of component activities, discussions about the implementation of those activities to be undertaken by the villages, and formulating subsequent management and maintenance systems for these village project activities, and (b) the project effectively addressed women‟s preferences as expressed through the participatory process (schools, village clinics, and drinking water schemes were among women‟s priorities), and women benefitted from training (which included training on women‟s healthcare), subsidies for pregnant and delivering women, and loans for livestock. PDO Indicator 5 35.4% (Sichuan) (Improved (Date: 2002) 5.0% (Sichuan) 6.3% (Sichuan) livelihoods): % of 62.0% (Yunnan) 22.8% (Yunnan) 32.2% (Yunnan) households below (Date: 2001) poverty line Successful: The two provinces‟ monitoring and information system all documented significant reductions in poverty incidence. Results from M&E Module 1 (undertaken by the National Statistics Bureau) more rigorously demonstrate that the project generated broad and significant positive impacts Comments on the beneficiaries‟ livelihoods. For example, from 2006 to 2010, project farmers‟ per capita incomes increased by 12% per year, which is greater than that of non-project villages (9%), the national average for rural areas (9.2%), and all nationally designated poor counties (10%). PDO Indicator 6 (Sustainable 14.0% (Sichuan) 4.8% (Sichuan) Mountain (Date: 2002) 1.5% (Sichuan) (Date: 2010) Agriculture): % -- 13.4% (Yunnan) 2.6% (Yunnan) 2.5% (Yunnan) of villagers with (Date: 2001) (Date: 2010) less than 150 kg of grain Satisfactory: The incidence of poverty in project villages was significantly reduced in these two provinces, food security in project villages greatly Comments improved, and increased income from cash crops and livestock financed under the project were significant contributors to household income growth. PDO Indicators 7, 8 and 9 (Rural 203,070 (Sichuan) 54,367 (Sichuan) Infrastructure): 94,990(Sichuan) (Date: 2002) (Date: 2010) Number of 37,155 (Yunnan) -- 112,221 (Yunnan) 28,020 (Yunnan) villagers with (Date: 2001) (Date: 2010) drinking water problems 263 (Sichuan) 2,382 (Sichuan) Km of village (Date: 2002) 2,653 (Sichuan) (Date: 2010) -- access roads 3,983 (Yunnan) 5,180 (Yunnan) 5155 (Yunnan) (Date: 2001) (Date: 2010) 54.0% (Sichuan) 100% (Sichuan) 95.0% (Sichuan) % of villages with (Date: 2002) (Date: 2010) 93.2% (Yunnan) -- road access 68.9% (Yunnan) 100% (Yunnan) (Date: 2001) (Date: 2010) Satisfactory: The outcomes were satisfactory. The rural infrastructure contributed to improvement of living conditions, supported development of livelihood activities, and improved access to social services through provision of basic rural infrastructure including village roads, drinking water supply, Comments irrigation works, land improvement, rural energy and improved housing. Key success factors include the component‟s focus on addressing basic household needs, farmer participation in infrastructure planning and management, training for farmers in infrastructure operation and maintenance, satisfactory technical quality of construction, and interaction with other components to generate added value from the infrastructure. 81.5% (Sichuan) PDO Indicators 99.4% (Sichuan) (Date: 2002) 10 and 11 (Basic 98.0% (Sichuan) (Date: 2010) 94.5% (Yunnan) -- Education): 98.6% (Yunnan) 99.6% (Yunnan) (Date: 2001) Enrollment rate (Date: 2010) 81.1% (Sichuan) 96.7% (Sichuan) (Date: 2002) Sixth Grade 95.0% (Sichuan) (Date: 2010) 88.3% (Yunnan) -- Completion Rate 99.5% (Yunnan) 98.7% (Yunnan) (Date: 2001) (Date: 2010) PDO Indicators 4.48% (Sichuan) 0.4% (Sichuan) 2.5% (Sichuan) 12 and 13 (Basic (Date: 2002) (Date: 2010) 0.78% (Yunnan) -- Health): Infant 2.72% (Yunnan) 1.09% (Yunnan) mortality rate (Date: 2001) (Date: 2010) 0% (Sichuan) 97.7% (Sichuan) % of households 98.0% (Sichuan) (Date: 2002) (Date: 2010) receiving basic 65.5% (Yunnan) -- 9.2% (Yunnan) 72.4% (Yunnan) health education (Date: 2001) (Date: 2010) Satisfactory: The outcomes were satisfactory. The Basic Education and Basic Health components were particularly important to the successful Comments outreach to the poorest and most remote project villages, ethnic minority communities, women and other vulnerable groups. PDO Indicator 14 (Project 0% (Sichuan) 99.7% (Sichuan) 100% (Sichuan) Management): % (Date: 2002) (Date: 2010) 84.6% (Yunnan) -- of villages 3.7% (Yunnan) 94.5% (Yunnan) receiving full (Date: 2001) (Date: 2010) project support Guangxi KPIs Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years PDO Indicators 1 and 2 (Participation): 13.6% (Guangxi) 86.9% (Guangxi) Percentage of (Date: 2004) 80% (Guangxi) -- (Date: 2011) households participating in project planning Percentage of remote natural 17% (Guangxi) villages 87.7% (Guangxi) (Date: 2004) 80% (Guangxi) -- participating in (Date: 2011) community project decisions PDO Indicator 3 (Gender): 2.8% (Guangxi) Percentage of 59.1% (Guangxi) (Date: 2004) 30% (Guangxi) -- women (Date: 2011) participating in project activities PDO Indicator 4 (Environment): Relatively No baseline value Significantly Community improved reported improved -- natural (Guangxi) (Guangxi) environment (Date: 2011) improved Guangxi regularly reported on this KPI indicator (number 4) for environment, Comments but this was not an agreed upon project objective and the values reported are difficult to understand. PDO Indicator 5 (Improved 80.8% of households Significantly livelihoods): % below poverty line 31.2% (Guangxi) Reduced of households (determined by PPA in (Date: 2011) (Guangxi) below poverty 2004 in Guangxi) line PDO Indicators 6 and 7 (Sustainable Mountain 0% (Guangxi) Agriculture): 78.6% (Guangxi) (Date: 2004) 100% (Guangxi) -- % of (Date: 2011) households receiving project’s assistance % of 0% (Guangxi) households 70.7% (Guangxi) (Date: 2004) 100% (Guangxi) -- receiving (Date: 2011) training PDO Indicator 8 (Rural Infrastructure): 6.3% (Guangxi) 92.2% (Guangxi) % of villagers (Date: 2004) 80% (Guangxi) -- (Date: 2011) benefitting from rural infrastructure PDO Indicators 9 and 10 (Basic 3 girls (Guangxi) 0 girls (Guangxi) Education): (Date: 2004) 0 girls (Guangxi) -- (Date: 2011) Girl’s Drop Out Rate % of households 4.2% (Guangxi) receiving 76.8% (Guangxi) (Date: 2004) 100% (Guangxi) -- applied (Date: 2011) technology training PDO Indicators 11 and 12 (Basic 0.1% (Guangxi) Health): % of 70.2% (Guangxi) (Date: 2004) 60% (Guangxi) -- households (Date: 2011) receiving medical assistance 11 (Guangxi) Infant 0 (Guangxi) (Date: 2004) 0 (Guangxi) -- mortality rate (Date: 2011) PDO Indicators 13 and 14 (Project 0 (Guangxi) One third More than one Management): (Date: 2004) (Guangxi) -- third (Guangxi) Number of (Date: 2008) (Date: 2008) poorest villages covered by the project % of poor 0% (Guangxi) households 91.7% (Guangxi) (Date: 2004) 100% (Guangxi) -- covered by (Date: 2011) project Guangxi achieved all of the KPI targets for numbers 5, 8, 9, 12 and 13 above. The KPI targets for numbers 6, 7, 10, 11 and 14 above (all 100%) were probably set too high at the time of appraisal, and achievement of Comments these target coverage rates was adversely affected by the decline in the US$ value of the IBRD Loan during implementation. Overall, the achievement levels for KPI numbers 6, 7, 10, 11 and 14 above is believed to have been substantial and satisfactory. G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 12/15/2005 Satisfactory Satisfactory 0.00 2 07/14/2006 Satisfactory Satisfactory 2.24 3 03/07/2007 Satisfactory Satisfactory 12.57 4 06/22/2007 Satisfactory Satisfactory 15.07 5 02/12/2008 Satisfactory Satisfactory 27.79 6 09/15/2008 Satisfactory Satisfactory 42.27 Moderately Moderately 7 01/31/2009 52.22 Unsatisfactory Unsatisfactory 8 06/25/2009 Moderately Satisfactory Moderately Satisfactory 59.31 9 01/13/2010 Satisfactory Satisfactory 72.44 10 04/07/2011 Satisfactory Satisfactory 98.77 H. Restructuring (if any) Not Applicable I. Disbursement Profile   1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1.1.1 At the time of appraisal, China was widely recognized for its achievements in reducing absolute poverty since the adoption of a broad program of rural economic reforms beginning in 1978. The Chinese government had a strong commitment to poverty reduction. The scale and funding of its poverty reduction program, and the sustained dramatic reduction of absolute poverty over the last thirty years of reform, were exemplary by any standards. Based on the government‟s austere rural poverty line (US$0.66/day of income), official estimates indicated that extreme poverty declined from more than 30 percent of the rural population in 1978 to about 3 percent by 2003, or from about 250 million rural poor in 1978 to 29 million in 2003. Estimates based on the Bank‟s international poverty line (using the 2005 International Comparison Program results) strongly confirmed the continuing remarkable decline in poverty incidence, but documented much greater numbers of poor in all years. The Bank‟s estimates at the time of project appraisal indicated that the proportion of China's poor population living below US$1.25 per day of consumption declined from about from 835 million in 1981 to 208 million in 2005. 1.1.2 Even with this great success, China's poor still comprised the second largest national concentration of extremely poor in the world after India. Almost all of China‟s poor resided in rural areas, and these poor were heavily concentrated in natural resource deficient mountainous areas of the central and western provinces where single sector poverty reduction measures have had limited sustainable impact. Certain disadvantaged ethnic minority groups, people with disabilities, and women were known to be heavily overrepresented among the poor. Overcoming the remaining rural poverty required innovative approaches and continued strong support. 1.1.3 Participatory approach and other innovations. Development work around the world has found that giving stakeholders a voice in project design, management, and evaluation improves results. At the time of project appraisal, the Government recognized the importance of strengthening the participatory approach in poverty reduction work. Internationally-assisted and national projects had built a body of experience and knowledge in participation in China. Most of China's achievements with participation had been supported by grant finance and led by national participation experts, conducted at relatively small scale and focused on poverty reduction project identification and design. Consequently, there was a need to consolidate work in participatory implementation and monitoring and evaluation and to ensure the quality of a participatory approach when it is used at scale and facilitated by county and township stakeholders. 1.1.4 Targeting. China's poverty reduction program and funding was targeted to 592 nationally-designated poor counties, and focused on improving access to assets, services and markets. Government estimates indicated that the proportion of the rural poor residing in the nationally-designated poor counties has declined from 72% in 1994 to about 60% in 2000. This meant that about 40% of the rural poor resided outside the poor counties and received very little or no central government assistance. In addition, since central government assistance to the nationally-designated poor counties was not well targeted to the absolute poor within those counties, there was a tremendous leakage of available poverty reduction funding to the nonpoor within those counties. The effectiveness and efficiency of available poverty reduction funding could be significantly improved through better targeting. 1 1.1.5 Addressing the poverty of ethnic minorities and people with disabilities. Ethnic minority groups and people with disabilities were known to represent a highly disproportionate share of the rural poor. Ethnic minority groups made up less than 9 percent of the total population, but were believed to account for about 40 percent of the remaining absolute poor, and often lived in the deepest poverty. Minority autonomous counties also accounted for more than 40 percent of the nationally-designated poor counties. The poorest ethnic minority groups often resided in extremely remote townships and villages, and the outreach of China's poverty reduction program to these isolated communities needed to be further strengthened. Similarly, people with disabilities and households with disabled members represented a large and rising share of the poor. It is believed that people with disabilities comprised about one third of the remaining rural poor at the time of appraisal. In poor areas, there was a mutually reinforcing relationship between poverty and disability: the inadequacy of effective public health measures and medical care in poor areas increased the chances of suffering a disabling injury or illness, and people with disabilities (and their families) were often pushed into extreme poverty by medical costs related to their injury or illness and by limited or no employment opportunities other than farming. 1.1.6 Improving access to basic education and health services. The educational, health and nutritional status of the remaining absolute poor in the 1990s was unsatisfactory. As many as half of the boys in many of China‟s poorest villages and, particularly in the most remote ethnic minority areas, an even greater share of the girls did not attend or complete elementary school and would not achieve literacy. Infant mortality rates and maternal mortality rates in very poor counties exceeded 10 percent and 0.3 percent respectively (or at least 50 percent to 100 percent greater than the national averages at that time), and were greater yet in the poorest townships and villages. The incidence of several infectious and endemic diseases, including tuberculosis and iodine deficiency disorders, remained concentrated in poor and remote areas. Roughly half of children in households at or below the absolute poverty line were at least mildly malnourished (stunted), and iron, vitamin A, and other micronutrient deficiencies remain a severe problem among the poor. 1.1.7 Improving productivity of mountain agriculture. At the time of project appraisal, the majority of China's remaining absolute poor resided in mountainous areas with either low rainfall, limited arable land, or both. In many cases the farm population greatly exceeded the limited carrying capacity of these uplands. In Yunnan, for example, the provincial government selected 506 key townships as the core focus of its poverty reduction programs beginning in 1995. Almost all of these poor townships were distributed in high mountain ranges and uplands, and in minority and border areas. About 86 percent of Yunnan‟s remaining poor in 1996 reportedly resided in these 506 poor townships. Achieving even subsistence levels of agricultural production in many of these mountainous areas was a difficult challenge. Existing single-sector strategies for mountainous area development tended to underestimate the integrated nature of the problems and challenges, and had not led to adequate and sustainable improvements in the well-being of the poor in these areas. Instead, there remained a need to better integrate individual sector strategies into a more comprehensive multisectoral approach to promote economic development and sustainable poverty reduction in mountain areas. Investment priorities within agriculture included strengthening agricultural training and technical support services, and providing greater support for expanding animal husbandry, land improvement, food crop production, tree and cash crops, and watershed management. 1.1.8 Further strengthening basic infrastructure. China's poverty reduction program emphasized the establishment and construction of a wide variety of basic rural infrastructure including water resource works, arable land, transport, power and communications. Between 2 1984 and 2000, for example, nearly 100 million mu (or 6.7 million ha) of basic farmland were improved or established in the poor areas, and drinking water systems were established for more than 77 million people and 84 million head of large animals. However, at the time of project appraisal, about 8% of poor area administrative villages were still not accessible by road, more than 25% of poor area administrative villages were not accessible by telephone, and there were about 21 million people in the nationally-designated key poor villages with difficult or inadequate local supplies of drinking water. Although great strides had been achieved, further improvements in basic infrastructure in China's poorest areas remains an effective and rapid means of improving people's lives and reducing poverty. 1.2 Original Project Development Objectives (PDO) and Key Indicators 1.2.1 The PDO was to improve livelihoods security and achieve sustained participation of the poorest rural people in project design, implementation, and M&E in eighteen counties of the three project provinces. 1.2.2 The original Project Design Summary (Annex 1 of the PAD) included more than 50 Key Performance Indicators (KPI). Many of the Project Design Summary KPI were difficult or impossible to quantify, and that list of KPI was not implemented or monitored by the Bank team during implementation. Instead, Schedule 7 of the Loan Agreement specified a more practical list of 14 KPI for each of the project provinces. The three project provinces carefully considered and selected the Loan Agreement KPI, and regularly reported these Loan Agreement KPI to the Bank team during the entire implementation period and at project completion. These Loan Agreement KPI have been recorded in all of the project archived ISRs, and are used in this ICR for assessing project performance and outcomes. The Loan Agreement KPI for Sichuan and Yunnan are identical, but the KPI for Guangxi are somewhat different: Objective KPI Sichuan and Yunnan Provinces Participation 1. Number of villages conducting participatory poverty assessments (PPA) and participatory planning; 2. Number of households which have participated in productive activities Gender equity 3. Percentage of women participating in participatory planning; 4. Percentage of women receiving training Improved livelihoods 5. Poverty incidence Mountain agriculture 6. Percentage of villagers with less than 150 kg grain/person Rural infrastructure 7. Number of villagers with drinking water problems; 8. Km of village access road; 9. Percentage of villages with road access Basic Education 10. Enrollment rate; 11. Sixth grade completion rate Basic Health 12. Infant and maternal mortality rates; 13. Percentage of households receiving basic health education Project management 14. Percentage of villages receiving full project support Guangxi Autonomous Region Participation 1. Percentage of households participating in project planning and decisions; 2. Percentage of remote natural villages participating in community project decisions 3 Gender equity 3. Percentage of women who have participated in project activities Environment 4. Community natural environment improved Improved livelihoods 5. Poverty incidence Mountain agriculture 6. Percentage of households receiving project assistance; 7. Percentage of households participating in training activities Rural infrastructure 8. Number of villagers benefitting from improved rural infrastructure services Education 9. Girls‟ drop out rate; 10. Number of households receiving at least one applied technology training Health 11. Infant mortality rate; 12. Number of households receiving medical assistance Coverage 13. Number of poorest villages covered by the project; 14. Percentage of poor households covered by the project 1.3 Revised PDO 1.3.1 There were no revisions made to the PDOs or to the KPIs during the life of the project. 1.4 Main Beneficiaries. 1.4.1 The PAD stated that the project would improve the income levels and well-being of about 1.4 million absolute poor in 18 counties in Sichuan, Yunnan, and Guangxi. The two provinces and one Autonomous Region are in China's lagging western region, and all 18 counties are nationally-designated poor counties. All of the project villages in Yunnan and Guangxi are designated key poor villages, and two-thirds of the project villages in Sichuan are designated key poor villages. Ethnic minority people comprise two thirds of project‟s primary target group, and include people of the Yi, Wa, Dai, Lahu, Zhuang, Miao, Molao, Dong, Maonan and other groups. Eleven of the project‟s 18 counties are ethnic minority autonomous counties. 1.4.2 The causal links between the project investments and the benefits expected to flow to the beneficiaries were clear. The main project benefits were to be (a) a greater sense of community cohesion and empowerment through the participatory project design, implementation and M&E process, (b) increased income levels and food security from improved upland agricultural productivity, (c) greater utilization of basic social and other infrastructure services from increased access to markets, safe drinking water, transportation, electric power, education, health and other social services, (d) stabilization and reversal of environmental destruction through land improvement and natural resource conservation works, and (e) improved project design and implementation capacity at different levels that could be mobilized for the Government's own poor areas development programs. A special pilot activity, implemented by DFID (Department for International Development) through separate parallel institutional and funding channels, was intended to benefit people with disabilities in Yuexi County in Sichuan. 1.5 Original Components 1.5.1 Participatory Approach. The project included a multisectoral package of six components intended to achieve the development objectives. Central to the design and implementation of all six components was the development of the project‟s participatory approach. The project‟s participatory approach built on the Government‟s Village Development Plan initiative and other participation experiences, and aimed to empower local communities and 4 promote inclusion of all disadvantaged social groups throughout the project cycle. The participatory approach was intended to be an empowering process that emphasized the voices and choices of different groups in the community, and built their ownership of, and capacities for, self-reliant community development. A Participation Manual (PM) was developed during the project preparation period in order to provide guidance to management staff and county and township facilitators on how to implement and monitor the participatory approach. The PM describes the basic principles of the participatory approach, details the main steps in the participatory project cycle, elaborates institutional and management issues in using the approach, discusses its costs and benefits, and provides some resource material on methods and tools. 1.5.2 Component 1: Sustainable Mountain Agriculture. The component supported agricultural development and sustainable livelihoods in each of the 1,063 project villages. There was a direct causal link between the component and the objectives of improved livelihoods, food security and participation. The main activity was to provide project households with financial support to improve, diversify and develop their productive activities. These activities were to be determined by the communities and households themselves from a highly flexible menu of options (including cash crops, tree crops, food security and animal husbandry) through the project‟s participatory process. Other component activities were intended to support improved household agricultural productivity and diversification through the provision of information, training and small demonstrations. Public extension services, private enterprises, training centers and farmer-to-farmer services were to provide agricultural technology that was to be appropriately adapted to local environments, to the small scale of household operations, and to indigenous cultures. 1.5.3 Component 2: Basic Rural Infrastructure. The component supported: improvement in village access through construction and improvement of roads, tracks and paths; provision of potable water supplies for humans and livestock; construction of small water conservancy works (including irrigation and drainage) with associated land improvements; extension of rural electrification and communications networks; and household energy and sanitation improvements. There was a direct causal link between the component and the objectives of improved access to basic infrastructure services and participation. The infrastructure was to be limited to small-scale works that directly benefit the project communities which were to be involved in their prioritization, construction and maintenance through a rigorous community participation process. 1.5.4 Component 3: Basic Education. The component supported: rehabilitation of dilapidated schools in poor and remote villages where existing buildings present health and safety risks to pupils and staff; training primary school teachers with innovative teaching methodologies and improving their ability to provide effective and relevant training and education to community members; integrating the development of the schools and communities with the transformation of the village schools to be community learning centers; and providing training services to the adults in the project villages with a focus on functional literacy education and practical technical skills. An additional DFID Free Education Pilot (FEP) grant piloted a nine-year compulsory education program in selected project villages which covered all charges for all children in the selected pilot areas in order to reduce the cost burden for basic education for poor families. There was a direct causal link between the component and the objective of strengthening the livelihoods and well- being of the project area population by improving access to primary education for poor and vulnerable children and adult education and training. 1.5.5 Component 4: Basic Health. The component supported: financial assistance for medical care for the poorest to ensure that they can utilize basic health services; essential public health interventions including maternal and child health care, immunizations for all children, 5 children‟s nutrition and control of parasitosis, and health education for communities on basic health knowledge and HIV/AIDS; and improvement of basic health service capacities at the township and village levels through training for health workers, building village clinics and rehabilitating township health facilities, provision of basic equipment and essential drugs, and adoption of clinical protocols, an essential drug list, and a mutual referral system. There was a direct causal link between the component and the objective of strengthening the livelihoods and well-being of the project area population by improving their health status. 1.5.6 Component 5: Community Capacity Building. The component supported activities to: improve basic production and living skills of the poor farm households; strengthen community leaders‟ basic skills; address the demands of some specific disadvantageous groups, particularly in ethnic minority areas, and including special training for women; and promote and strengthen community participation in the decision-making process of local public development affairs, including training in participatory approach methodologies and the establishment of farmer participatory monitoring groups. There was a direct causal link between the component and the objective of strengthening the livelihoods and well-being of the project area population by promoting project implementation effectiveness and sustainability. 1.5.7 Component 6: Project Management and Monitoring. The component supported capacity building for project management at the provincial, prefectural, county and lower levels, and the design and operation of an “external� independent impact assessment system. A generous DFID grant played an essential role in the design and operation of the independent impact assessment system which included extensive quantitative and qualitative survey work at the village and household levels throughout the life of the project (see paragraph 2.3.1). There was a direct causal link between the component and the objective of strengthening the livelihoods and well-being of the project area population through effective project management and rigorous M&E. 1.6 Revised Components 1.6.1 No significant revisions were made to any of the project‟s components. 1.7 Other significant changes 1.7.1 The scale of all project activities was reduced to some degree during the 2008 Mid Term Review (MTR) in order to adjust for the declining value of the IBRD US dollar loan (the US dollar depreciated by about 20% against the Chinese yuan during the first three years of project implementation) and rising costs for most all project activities. However, while there was a great deal of rebalancing of emphasis on individual subcomponent activities during the MTR (for example, Sichuan decided to provide extra support for grain crops, Guangxi chose to reduce investment in some cash tree and field crops, and Yunnan decreased support for grain crops and increased investments in livestock), no changes were made to the scope or content of project components. Furthermore, the Provincial Project Management Offices (PPMO) endeavored to minimize the overall decrease in the scale of project activities to the extent possible by seeking additional counterpart funding, and the changes in emphasis between individual project activities were identified and finalized through a participatory approach. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 6 2.1.1 Long preparation period. The project preparation period extended some 45 months from October 2001 to June 2005. Most of this extreme delay was the result of (a) a lengthy and contentious process to determine the project location, and (b) a lengthy disagreement within the central government over the role of the central government in the project and which arm of the government would take lead responsibility for the project. First, during the early stages of project preparation, the government initially proposed to locate the project in two northern provinces near Beijing which were subject to significant soil erosion problems. This project area selection was expected to strengthen central government support for the project, but was rejected by the Bank/DFID team because it was believed that the depth of poverty in the two northern provinces was not sufficient for a focused poverty reduction project. DFID subsequently required that the project be located entirely in Yunnan province in southwestern China. This single province proposal was in turn rejected by the central government because of its limited scale. Following a delay of at least twelve months, all parties agreed that the project would include Guangxi, Sichuan and Yunnan provinces in western China. 2.1.2 Second, the central government decided to move responsibility for the project from the State Council‟s Leading Group Office for Poverty Reduction (LGOP) to the three provincial Development and Reform Commissions (DRC). This switch would have moved control of the project from the LGOP system to the DRC system. Fearing a reduction of the project‟s poverty focus if designed and implemented under the DRC system, the Bank team actively resisted this switch. These fears were confirmed at a project workshop held in Chengdu in May 2002 where it became clear that the Sichuan DRC‟s proposed project concept was not consistent with the project‟s original poverty focus.2 Following an extensive delay and the near collapse of project preparation, an imperfect compromise solution was agreed upon in which (a) a Central Liaison Group (CLG) was established at the central level under the LGOP system, and (b) the provincial DRCs and Poor Area Development Offices (PADO, under the LGOP system) would share responsibility for project management. This imperfect solution led to a significant design flaw which adversely affected project implementation (see paragraphs 2.1.6 and 2.2.4 below). 2.1.3 Design. The background analysis supporting the project was strong and benefitted from an ongoing major study of rural poverty in China3 and the experience of two earlier multisectoral poverty reduction projects in China.4 The PAD provided a clear summary of sectoral issues at that time including the need for a strong participatory approach, targeting of the poorest villages and most vulnerable groups (including ethnic minority people, women, and people with disabilities), and a multisectoral strategy with support for agricultural production, improved basic rural infrastructure, greater access to basic education and health services, and assistance for local level capacity building. The project design conformed exactly to these sectoral priorities by including: an extremely strong focus on the participatory approach; an extraordinary outreach to the poorest, most remote, and most disadvantaged villages and ethnic minority people, a strong focus on gender equity, and a special DFID Grant to provide support for people with disabilities; and a full multisectoral approach including components providing support for agricultural production, rural infrastructure, basic education and health, and community capacity building. Overall, the project was well designed with clear development objectives and strong linkages between these objectives and the project components. 2 The Sichuan DRC proposed a project concept centered on constructing two toll highways to a tea and a livestock production base in poor areas in northern and southern Sichuan. 3 The World Bank (2001): “China: Overcoming Rural Poverty� (The World Bank, March 2001). 4 The projects were the Southwest Poverty Reduction Project (SWPRP, P003639) and the Qinba Mountains Poverty Reduction Project (QBPRP, P003590) which began implementation in 1995 and 1997 respectively. 7 2.1.4 Several key lessons were learned from the successful implementation of the earlier SWPRP and QBPRP multisectoral poverty reduction projects in China and, with one major exception, these important lessons were fully reflected in the project design. The single most important lesson learned from the experience of the SWPRP and QBPRP projects was the vital importance of strong support from the central government to the success project preparation, implementation and M&E. Unfortunately, the Government chose to disregard this key lesson and required (despite the Bank‟s objections) that central government support and leadership for project preparation and implementation be kept to a minimum (see paragraph 2.1.2). 2.1.5 The other key lessons from the successful implementation of the earlier SWPRP and QBPRP projects were carefully reflected in project design. First, it had been determined that a highly effective organizational and management structure was necessary to meet the institutional demands of this highly complex and innovative series of projects, and the Poor Rural Communities Development Project (PRCDP) included two components intended to strengthen capacity within the PMO system and within the project communities themselves. Second, community participation was understood to be essential to successful project design, implementation and M&E of the project, and detailed plans were established to sustain a robust participatory process throughout the life of the project. Third, good synergy could be achieved by providing a full range of poverty reduction activities down to the township and village level. In particular, early provision of support for rural infrastructure and for basic education and health would quickly extend the outreach of the project to the poorest villages and greatly encourage those villagers to participate in all aspects of the project. The project design therefore included a full multisectoral package of poverty reduction activities, and the Loan Agreement required that project expenditures within each province for the Rural Infrastructure, Basic Education, Basic Health, and Community Capacity Building components would not be less than half of total cumulative project expenditure over the life of the project. 2.1.6 Design flaws. Although the overall project design was sound, the project preparation team was aware of two key design flaws. First, as noted above (paragraphs 2.1.2 and 2.1.4), the sharply reduced leadership role of the central LGOP in project design and implementation adversely affected the project preparation process and would adversely affect the quality of project implementation. Second, the Government required that the project provinces, counties and primary beneficiaries repay the project loan. Since no repayment arrangements could be put in place for public goods such as village roads, drinking water systems, and other basic rural infrastructure, basic education, and basic health, Government therefore required that these public goods investments be kept to a minimum in the project. Instead, based on the understanding that the project loans could eventually be recovered only from investments in household and village level farm production, Government required that the majority of project funding be directed to household and village agricultural production schemes through the Sustainable Mountain Agriculture component. This top-down planning bias in favor of what the Government considered to be “productive� investments in the Sustainable Mountain Agriculture component directly led to two severely adverse consequences for the quality of project implementation and achievement of the PDO. First, the participatory process was substantially undermined since the demands of the project‟s primary beneficiaries as expressed through the participatory process – which were overwhelmingly in favor of basic infrastructure, basic education and health, and other public goods – could not be adequately addressed by the project. Secondly, the largest share of project funds were instead forced into agricultural production schemes which were not appropriate in many instances in the poorest project areas where the carrying capacity of the agricultural resource base was already fully exhausted or even exceeded. This severe design flaw is recorded in the PAD‟s risk assessment, and the corresponding mitigation measure is also 8 reported. Through extensive discussions, an imperfect solution was agreed upon which limited total cumulative investment in the Sustainable Mountain Agriculture component in each project province to no more than half of each province‟s total project expenditure. 5 This agreement was also formally recorded in the Loan Agreement (see paragraph 9 of the Annex to Schedule 5). 2.1.7 Assessment of risks. The PAD recorded seven significant risks to achieving the project‟s intended outputs and objectives. First, the PAD‟s risk assessment identified the two key design flaws discussed in paragraph 2.1.6 above, and reported the corresponding mitigation measures of establishing the CLG and capping total investment in the Sustainable Mountain Agriculture component. Second, the PAD noted the substantial risk that insufficient start-up funding and the inadequacy of counterpart funding might lead to delays in the start-up and progress of project implementation. This risk was to be mitigated through dialogue with the central government and close monitoring of the availability of counterpart funding. Third, the PAD reported a significant risk that participation would not be mainstreamed at adequate quality levels throughout the life of the project. Seven mitigation measures were put in place to reduce this risk including the establishment of sound implementation guidelines, engaging technical specialists, provision of training and support for local facilitators, and careful monitoring of the participation process. Lastly, the three other three risks identified in the PAD were the possibility of (a) insufficient funding for operation and maintenance (O&M) costs for the project‟s many small scale civil works, (b) the sector bureaus (including the Agriculture, Construction, Education and Health bureaus) taking over control of the project and not adequately following the participatory approach, and (c) insufficient capacity and commitment to follow through on addressing gender inequality. A number of mitigation measures were put in place to reduce these three risks including establishing guidelines (a) requiring the estimation of O&M costs as part of civil works designs, (b) specifying the roles of the sector bureaus, (c) including sector bureaus in the discussion of supervision results, (d) compiling gender disaggregated project data, (e) adopting remedial actions such as gender quotas, and (f) assigning gender focal points at the provincial and county PMOs. 2.1.8 Adequacy of the Government’s commitment. During the early stages of project preparation, the Government demonstrated strong commitment to and ownership of the project. In May 2000, the Government, with the assistance of the World Bank, successfully hosted an International Conference on China's Poverty Reduction Strategy in the 21st Century, and the project concept originated at that Conference. The Vice Premier, Wen Jiabao, subsequently discussed the project with the UK Government during a visit to the UK, and the UK agreed in principle to provide major grant funding in support of project preparation, implementation, and monitoring and evaluation. The provincial governments also had a strong commitment to the project, and the provincial PMOs played the leading role in project preparation. Unfortunately, during the later stages of project preparation, the Government‟s decision to minimize LGOP‟s role in the project greatly undermined the central government‟s support and interest in the project.6 Also, the attempt to turn over management of the project to the three provincial DRCs led to confusion over the control of the project and a weakening of the chain of command. 5 The three provincial DRCs initially proposed that the Sustainable Mountain Agriculture component would comprise at least 70% of total project cost. Following extensive discussion, it was agreed that the component would be capped at no more than 50% of total project cost. 6 LGOP had originally intended to support project implementation with significant central government grant funding. However, plans for LGOP financial support for the project were withdrawn when LGOP‟s role in the project was minimized. 9 2.1.9 DFID Partnership and Grant Support. DFID was a full partner in all of the project activities from the outset of project preparation in 2001 through to the completion of this ICR in 2011. The DFID task team participated in all project preparation and supervision missions. DFID‟s grant support for the project totaled about US$37 million. Most importantly, the Chinese Government would not have undertaken preparation or agreed to implementation of the project in the absence of DFID‟s cofinancing grant (TF54949, administered through a Tripartite Agreement) of US$32.5 million which softened the terms of the IBRD Loan. The DFID cofinancing grant was “made available to China for prepayment of part of the principal amount of the Bank Loan.� An additional DFID grant (TF70354) of about US$2.7 million provided vital assistance for the development of the project‟s participatory approach, the rigorous and cross- cutting monitoring and evaluation system, and the Free Education Pilot. In addition, DFID provided and executed separate grant funds to (a) review China‟s previous experience with participation, and (b) in close coordination with government staff and the Bank team, draft the project‟s Participation Manual (PM). Lastly, in connection with the project but arranged directly between DFID and Handicap International, a separate DFID grant of about US$2 million supported an extraordinary activity to provide assistance for people with disabilities in Yuexi County in Sichuan. DFID‟s role in the project was extremely successful, the project would not have happened without DFID‟s grant support to soften the IBRD Loan, and DFID‟s additional grants played a key role in the successful design, implementation, supervision, and monitoring and evaluation of the project. The Bank team believes that DFID‟s extensive collaboration in the design and supervision of the project, and DFID‟s generous financial support for the project, were exemplary and should be repeated. The central and provincial governments also greatly appreciated DFID‟s strong technical and financial support throughout the project‟s preparation, implementation, and completion review periods. 2.2 Implementation 2.2.1 Summary. Most of the design flaws and risks anticipated and identified in the PAD came into play at the outset of project implementation: initial implementation was slow; the Sustainable Mountain Agriculture component significantly exceeded more than half of total project expenditure in two of the three project provinces; the participatory process was undermined by the project‟s limited flexibility to respond to the priorities expressed through participatory planning; outreach to the poorest natural villages and most disadvantaged ethnic minority people was in many cases delayed and insufficient; there was limited attention to gender equity issues; and there were instances where the sector bureaus took control of project implementation and where O&M support for the project‟s civil works was inadequate. This relatively bleak assessment of the initial pace and quality of implementation and the correspondingly dim outlook for achieving the PDO was confirmed in the 2008 QAG review of the project (which was undertaken very soon after the 2008 MTR). However, over time, the provincial and county PMOs were able to successfully address and overcome each of these significant challenges. With consistent support and, in many cases, strong pressure from the Bank/DFID supervision team, each project issue was identified, carefully examined, and satisfactorily resolved. A key element of the eventual success of the project was the Bank/DFID team‟s sustained focus and support for the participatory process, and the reciprocal determination of the provincial and county PMOs to fully realize a high quality and extensive participatory process. Though substantially constrained by the imposed limitations on their role in project management, the CLG did participate in all supervision missions and thereby played a consistent and vital role in resolving these several key project issues. In the end, all of the design flaws and obstacles to quality implementation were satisfactorily overcome, and the project was highly successful in achieving the PDO. 10 2.2.2 Inadequacy of counterpart funding. The inadequacy of start-up funding greatly slowed the pace of initial project implementation, and the consistent shortage and late arrival of counterpart funding led to a (a) growing lag in loan disbursement, and (b) widening imbalance between component implementation during the first three years of project implementation (2006- 08). The disbursement lag grew from 12 months in 2006 to 18 months in 2007 and reached more than 20 months in 2008. Even more troubling, the shortage of specific types of counterpart funding contributed to the extremely slow pace of implementation in Sichuan and Guangxi of the rural infrastructure, basic education, basic health, and community capacity building components in the first three years of project implementation (see paragraph 2.2.3 below). Fortunately, the counterpart funding problems were finally resolved in the final two years of project implementation (2009-10) and a massive increase in funding for rural infrastructure, basic education, basic health, and community capacity building completely eliminated the imbalance between component implementation, and the project was successfully completed by the revised closing date of June 30, 2011. 2.2.3 Excessive investment in the Sustainable Mountain Agriculture component. In combination with the Government‟s preference for “productive� investments in agricultural production (see paragraph 2.1.6), the shortage and late arrival of specific types of counterpart funding led to excessive levels of investment in the Sustainable Mountain Agriculture component.7 (Furthermore, the project county PMOs were reluctant to use the project loan for the rural infrastructure, basic education and basic health components since alternative sources of government funding for these public goods were generally provided on a grant basis, whereas the counties would be held fully responsible for the repayment of the entire project IBRD loan and the project beneficiaries would be held responsible for the repayment of the sustainable mountain agriculture component expenditures.) This imbalance between the rate of implementation of the Sustainable Mountain Agriculture and all other components was a serious problem in 2006 and 2007, and reached the crisis stage in mid-2008 when expenditure on the Sustainable Mountain Agriculture component comprised about 80% of Sichuan‟s cumulative project expenditure and 57% of Guangxi‟s cumulative project expenditure. With the strong support and exhortations of the Bank/DFID supervision team, the Sichuan and Guangxi provincial PMOs were finally able to resolve their counterpart financing problems in 2009 and 2010, and the imbalance between component implementation was fully resolved by the end of 2010. 2.2.4 Project management. The absence of a strong central management/leadership unit hampered the pace and quality of project implementation in several ways. For example, a strong central management/leadership unit could have played a key role in quickly eliminating the shortage of start-up and counterpart funding. Most importantly, it was feared that the absence of a strong central management/leadership unit would limit the transfer of experience from PRCDP to the national poverty reduction program. (The strong leadership role of the Central PMO and the transfer of experience to the national poverty reduction program was a hallmark of the success of the earlier Southwest and Qinba Mountains Poverty Reduction Projects during 1995-2004. The 1997 QAG review of the Southwest Poverty Reduction Project found that the project was best practice in terms of directly engaging senior policy makers in the hands on practical testing 7 Only certain types of counterpart funding, such as Yigongdaizhen (Food for Work) funding and funding from the sectoral bureaus (including education and health), could be used for the project‟s public sector investments. 11 of innovative poverty reduction measures.) However, although the CLG was required to play a limited role in guiding project implementation, LGOP did remain deeply interested in and supportive of the successful implementation of the project. 2.2.5 Particularly in Guangxi, the experiment with the Autonomous Region‟s DRC and PADO sharing joint responsibility for project management (see paragraph 2.1.1) led to confusion over the control of the project and a weakening of the chain of command. The October 2005 First Supervision Mission found that institution building work in Guangxi had faltered badly and that the county PMOs and township Work Stations had not been established. This uncertainty regarding the day-to-day control of the project hobbled effective project management in Guangxi during the early years of project implementation. Although this problem was never fully resolved, by the final years of project implementation it appeared that the Guangxi DRC and PADO had established a high level of collaboration and that institution building at the county PMOs and township Work Stations had become fully satisfactory. 2.2.6 Challenges to the participatory process. The quality and outreach of the participatory process were considered to be the key measures of the project‟s overall success and of the achievement of the PDO, and it was clear from the outset of project implementation that the PMO system at all levels was committed to getting the poor to participate in all phases of the project. However, the limited flexibility within the government‟s planning system and funding structures seriously constrained the project‟s ability to adequately deliver the public goods which were most highly desired by the primary beneficiaries, and there were concerns that this would undermine community support for the participatory process. Priorities of the poor that were documented in the participatory plans included village access roads, drinking water systems and other basic infrastructure, improving basic education and health, and community capacity building. Implementation of these four public goods components was weak and slow during the first three years of the project (due to the lack of start-up funding and delays in the provision of counterpart funds). The apparent discrepancies between the expressed priorities of the poor and actual investments were a major cause of concern in the 2008 QAG review of the project which questioned the viability of the participatory approach. However, the QALP-1 Panel advised that, even if the project could not fully satisfy the primary beneficiaries‟ top priorities, a satisfactory participatory process was still possible so long as the communities became able to determine their own development and poverty reduction approach and to seek appropriate levels of assistance through the project and from other sources of support. Monitoring of the participatory approach showed that once the initial funding problems were overcome (beginning in the third year of implementation) the project responded better to the priorities of the poor. Careful monitoring also documented a number of cases of project institutions‟ more responsive practices and behavior. A major milestone was the adjustments made during the 2008 MTR, which included a participatory feedback process and, among others, led to provision of more drinking water schemes. In addition, in a number of project counties and villages in Yunnan and Guangxi where the demand for public goods exceeded what the project was able to provide, local project staff were able to acquire additional funding from non-project sources based on the participatory planning. Overall, implementation of the project‟s public goods components accelerated dramatically during 2009-10 and, in recognition of the success of the participatory approach, the project received the Bank‟s 2009 Second Place “People First� award. The success and impact of the participatory process continued to accelerate during 2009-10 and, by the time of project closing in 2011, it was clear that the project had become the cutting edge of large scale participatory development in rural China. 2.2.7 Outreach to the poorest ethnic minority people. The project‟s outreach to the poorest ethnic minority people is also considered to be a key measure of the project‟s overall success and 12 essential to the achievement of the PDO. Certainly the project area covered the greatest concentrations of some of the most severely disadvantaged ethnic minority people in China including people of the Yi, Wa, Lahu, Miao, Yao, Molao, Dong, Maonan, Zhuang and other groups, and the PMO system at all levels recognized that successful outreach to these poorest communities was a top priority of the project. While there was adequate outreach to the poorest ethnic minority people in many cases during the first years of project implementation, there were also several significant instances of (a) inadequate outreach to these groups, and (b) the excessive provision of support for farming activities under the Sustainable Mountain Agriculture component when the participatory process had clearly determined that the poorest ethnic minority communities much more strongly preferred assistance through the Rural Infrastructure, Basic Education and Basic Health components. The Bank/DFID task team understood that the project county PMOs were caught in the difficult position of coping with inadequate counterpart funding and the requirement imposed by higher levels of government that project loan repayment by the counties would depend upon the recovery of project funds provided for farm households‟ investments in agricultural activities. While the county PMOs were doing the best they could under these circumstances, it was clear in 2006 and 2007 that the project‟s outreach to the poorest ethnic minority villages was in some cases both limited and skewed toward activities which were not the top priorities of the primary beneficiaries. The Bank/DFID team responded by pressuring the county and provincial PMOs to improve and extend the project‟s outreach to the poorest ethnic minority communities, and also worked in close partnership with the provincial PMOs to resolve the underlying counterpart financing issues. The provincial and county PMOs worked diligently to overcome these obstacles and were able to steadily improve the quality and extend the outreach of project assistance to the poorest and most remote ethnic minority communities. By the conclusion of the project, the scale and quality of the project‟s assistance to the most disadvantaged ethnic minority villages had become exemplary. 2.2.8 The Bank/DFID supervision team found two notable cases which demonstrate the provincial and county PMOs‟ successful efforts to overcome the initially limited and poor quality outreach to ethnic minority people. First, the October 2006 Third Supervision Mission found a worrisome mismatch between the high priority demands of project farmers and the project‟s actual support in the poorest Yi ethnic minority villages in Meigu County in Sichuan: “infrastructure construction (road, water, and electricity) was clearly identified by farmers as their highest priority demand; however, due to lack of counterpart funding (as well as certain constraints on the uses of available counterpart funding), the actual interventions from the project were walnut, Sichuan pepper, and kidney beans. Also in Meigu County, due to the sourcing of counterpart funds from private sector companies investing in tree crop development, the mission noted a troubling substitution of tree crops for grain in some grain-deficient villages.� Second, the December 2007 Fifth Supervision Mission found that better off Dai ethnic minority villages in Mengma Township in Menglian County in Yunnan had received an excessive share of project benefits while the poorest Lahu and Wa villages (in Mengma Township) had received only very limited benefits. Subsequent Bank/DFID supervision team field visits found that the county PMOs had done an excellent job of correct these problems. First, the October 2009 Tenth Supervision Mission observed that the Meigu County PMO in Sichuan had newly included three extremely poor and remote villages to the project in Ninha Township, and that “there has been a substantial level of investment in these villages, including village roads, drinking water, agricultural productive activities, health clinic facilities and equipment, and village learning centers (averaging about RMB 2,000 per household), which has brought about an impressive change in the level of income and the living conditions in these villages.� Second, the October 2010 Twelfth Supervision Mission found that outreach to the poorest and most disadvantaged ethnic minority people in Menglian County in Yunnan was excellent, and that investments in the 13 poorest Lahu and Wa ethnic minority villages had greatly increased (while all further project investments in the better off Dai villages had been cancelled) in Mengma Township. 2.2.9 Sector bureau engagement. The multisectoral nature of the project required the active engagement of several sectoral bureaus (including the agriculture, transport, water resources, education, and health bureaus) at the county and township levels. Successful project implementation depended on the support of these sectoral bureaus, but the risk that these bureaus would take control of the project and, in particular, not adhere to the participatory process was recognized at the design stage (see paragraph 2.1.7). Although the need to strengthen coordination with the sectoral bureaus was a constant theme of the first six supervision missions during 2005-08, overall sectoral bureau support during project design and implementation was substantial. Also, while some project counties were found to have relied too heavily on the sectoral bureaus to organize and manage some project components, regular supervision did not find that the sectoral bureaus had established widespread control of the project.8 Finally, while the sectoral bureaus‟ overall uptake of the participatory approach was less than desired, the sectoral bureaus did not significantly impede the project‟s participatory process. The greatest risk was in the Sustainable Mountain Agriculture component, and the Bank/DFID team found some cases in which county agriculture bureaus had undermined and partially overridden participatory planning by imposing their own county or township agricultural development plans. By comparison, the Bank/DFID team found that levels of participation were particularly high in small-scale infrastructure schemes (such as community drinking water activities) implemented through a community-led approach. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 2.3.1 M&E design and funding. The project‟s M&E system took advantage of the successful experience of the earlier Southwest and Qinba Mountains Poverty Reduction projects and comprised both an "internal" physical monitoring and information system (MIS), and an "external" independent impact assessment system. The internal MIS (a) monitored disbursement and recovery of project funds and physical progress of the project, and (b) included surveys of a small number of project households with Village Monitoring Groups and other measures to gauge the quality and success of the participatory approach. The external independent impact assessment system monitored the project's participatory approach and its impact on poverty and well-being through three “modules� including both quantitative and qualitative approaches. Module 1 comprised annual quantitative surveys of 1800 project and nonproject households (100 households in each of the project‟s 18 counties) undertaken by China‟s National Statistics Bureau (NSB). Module 2 comprised annual qualitative surveys of the project‟s process and outcomes undertaken by the ITAD Company. The Module 2‟s Participatory Poverty Monitoring was based on assessments by local communities which tracked poverty impacts and provided feedback on people‟s demand and project performance. Module 3 provided direct support for a participation specialist to join all of the Bank/DFID project preparation and supervision missions. A generous DFID Grant provided crucial support for this comprehensive and rigorous external M&E system. 8 Perhaps the worst case of excessive sectoral bureau control of the project was found in Menglian County in Yunnan by the October 2010 Twelfth Supervision Mission which observed that key project progress, financial management, and procurement records were maintained by the sectoral bureaus in some townships and were not readily available to the County PMO. However, the overall quality of Menglian County‟s project implementation was excellent, and it appears that the excessive role of the county sectoral bureaus did not result in any significant adverse effects. 14 In particular, the DFID Grant (a) made it possible for the NSB surveys to be undertaken in all years of the project, (b) covered all the costs of the annual qualitative surveys of project impact and results, and (c) provided funds for the participation specialist who worked closely with the Bank/DFID project preparation and supervision missions. The Bank/DFID team believes that the DFID Grant was essential to the success of M&E system and that, most importantly, the consistent presence of the DFID Grant-supported participation specialist was absolutely vital to the outstanding success of the rollout and quality of the project‟s participatory approach. 2.3.2 M&E implementation and utilization. The well funded M&E system made possible comprehensive updates on the status of project implementation and provides the basis for a rigorous evaluation of the project‟s impact. Although one of the project provinces (Yunnan) was late in establishing and operationalizing its MIS, all three provinces regularly provided the Bank/DFID team with KPI agreed to in the Loan Agreement (see paragraph 1.2.2). DFID‟s Grant funding for the M&E system was not confirmed until soon after project negotiations, and this resulted in a delay of the baseline survey work. However, the National Statistics Bureau (NSB) quantitative surveys and the ITAD qualitative surveys were undertaken in each year of project implementation. This independent M&E mechanism provides a rich and unbiased picture of project impact. Although the project provinces did not access the NSB survey results on a timely basis, all parties agreed that the NSB quantitative surveys represent the most rigorous and credible data for project evaluation. 2.4 Safeguard and Fiduciary Compliance 2.4.1 Environmental impact. The project‟s Rural Infrastructure and Sustainable Mountain Agriculture components had modest environmental impacts. The October 2010 Twelfth Supervision Mission observed that the Rural Infrastructure component was generally well executed and that the Environmental Management Plan (EMP) was followed in all three provinces. The adverse impacts of rural infrastructure activities mostly occurred during their construction period and thus were of a temporary nature with limited scope due to their small and scattered scale. The Sustainable Mountain Agriculture component had the potential for more long-term environmentally adverse impacts through the improper use of fertilizer and pesticides and from certain agronomic practices (such as growing sugar cane on steeply sloped land). It is believed that a greater emphasis on agricultural training could have minimized such adverse impacts and could have improved returns by reducing agricultural production costs. 2.4.2 Environmental monitoring. Except for the monitoring of water quality in the project‟s drinking water supply schemes (which was conducted regularly in the project areas in accordance with domestic regulations and the EMP), project environmental monitoring was not undertaken. Instead, the provincial and county PMOs relied on existing domestic monitoring network for other indicators to be monitored according to the EMP. The poor performance of the project‟s environmental monitoring was related to two factors: (a) given the scattered and small scale nature of the project‟s agricultural activities, changes in some indictors (such as surface water quality and soil erosion) could not be attributed solely to the project‟s activities. Therefore, it was inappropriate to include such indicators in the EMP as a means of measuring the project‟s impacts on the environment; and (b) other indicators included in the EMP (such as soil fertility and contamination, and pesticide residues in agricultural products) can be influenced more directly by project activities when properly sampled. However, even in the suburbs of big cities in China, these items are rarely tested due to the complex testing capacity required (for organic substances in a solid substrate), high cost, and the number of years needed to have any detectable change. It is thus unrealistic to monitor these indicators in remote and poor rural areas. The 15 Bank team‟s environmental review concluded that, for this type of rural poverty reduction project where impacts are predominantly from non-point sources, the design and focus of environmental safeguards should shift from monitoring end-of-pipe impacts to prevention and minimization at the source. This could have been done by more heavily supporting farmers‟ training in the areas of environmentally friendly and resource efficient agronomic practices such as the appropriate use of chemical fertilizer and integrated pest and disease control. In order to have been more specific to the project and to have increased the likelihood of follow-up during project implementation, the design of the EMP and especially its monitoring scheme should have better taken into account both the academic soundness and practicality of its approach. 2.4.3 Land acquisition. Project land acquisition activities in Guangxi were carried out to the satisfaction of the affected people through participatory approaches. Most of the land needed for the village-level infrastructure works was voluntarily provided by the affected villages and villagers. Case study investigations suggest that the project affected persons (PAPs) were satisfied with the land acquisition arrangements and that their income levels increased by at least 50% during 2004-10 on average. Over 8,000 project households in Guangxi were affected by the acquisition of about 3,000 mu of land for the construction of village access roads, health clinics and drinking water facilities. All land acquisition was generally done in accordance with Chinese regulations and Bank policies. 2.4.4 Use of the participatory approach, early advance notice, and adequate community mobilization made land requisition much easier, lower cost, and more efficient. The project successfully mobilized the affected communities, and these communities and villagers voluntarily offered land for the construction of project civil works when these people and communities understood that they would be the direct beneficiaries of these village access roads, drinking water facilities, health centers, and other civil works. However, the Bank‟s safeguards specialist noted that agreement in writing was not always reached in advance with the PAPs when it was necessary to take their land in order to complete project civil works, and that compensation standards should be (a) adequately discussed with and supported by the PAPs, and (b) in line with the local government regulations. 2.4.5 Procurement. Based on a sample review of goods and civil works contracts procured through NCB and shopping procedures in Yunnan, it appears that procurement procedures agreed upon in the legal agreements were followed in the procurement processes. In all cases, the lowest evaluated responsive bidders, suppliers, and contractors were awarded contracts and no major irregularities were identified during the ex-post review. However, the final supervision mission found there were some procurement actions which were not fully in line with World Bank procedures. First, none of the Yunnan county PMOs had a Procurement Plan approved or updated as required under the Loan Agreement (the county PMOs only issued internally approved project implementation plans on an annual basis, and these plans were used as the basis for the implementation of procurement actions). Second, in Menglian County in Yunnan, each of the related line bureaus had taken responsibility for contract administration, and the County PMO did not have control over their work or access to the related documentation. In addition, payment records and related financial documents were not kept together with the procurement documents. Lastly, it was found that the payments schedule under the contracts was not always followed. Some of the most common deficiencies found were (a) advance payment guarantees stipulated under some contracts were not provided by the contractors, (b) retention deposits to be deducted and withheld by the client were not normally withheld, and (c) some payments were made as long as one year later than the date specified in the contract conditions. 16 2.4.6 Financial Management. Although no significant financial management issues were noted during project implementation, there were still some financial management weaknesses that needed to be highlighted. For instance, since all project counties were in remote locations and were in weak fiscal positions, the committed counterpart funds for some of the project counties were either not fully provided for or not delivered on a timely basis. This had a negative impact on project implementation. In addition, contractors and beneficiaries were not always paid on a timely basis due to the shortfall of counterpart funds. Also, the capacity of project financial staff was often quite weak and some training was provided to them during project implementation to address this weakness. Although the quality of the financial staff improved during project implementation, the performance of the financial staff in some of the project counties was generally unsatisfactory. Finally, the slow and inefficient disbursement review procedures delayed the payment process. 2.5 Post-completion Operation/Next Phase 2.5.1 Sustaining reforms and institutional capacity. The project (a) helped to explore, develop and promote a number of approaches to poverty reduction in rural China, and (b) invested in a variety of civil works and other activities to improve livelihoods security and access to basic services. The improved approaches to rural poverty reduction promoted under the project included the participatory approach, strengthened outreach to the poorest villages and most disadvantaged ethnic minority people, gender equality, and assistance for people with disabilities. The exploration, development, and promotion of these improved approaches was highly successful and very advanced in each of the project provinces. Since the PADOs in each province have been heavily involved in all aspects of the project from the outset of design through their final project completion reports, the (a) transition arrangements for the continuation of these improved approaches are fully in place, and (b) expansion of these approaches province- wide in Guangxi, Sichuan and Yunnan is already well underway. Furthermore, the CLG has reported that key approaches, including the complete participatory approach, are being adopted in other development programs in China (see Annex 7 paragraph 7.2). The ultimate test of the promotion of these new approaches will be the extent to which they are included in the Government‟s new poverty reduction strategy for 2011-2020. Although the Government‟s new poverty reduction strategy for 2011-2020 has not yet been released, it is believed that the participatory approach and other key approaches to rural poverty reduction promoted under the project will be highlighted in the Government‟s new poverty reduction strategy. 2.5.2 The project‟s participatory approach has already been strengthened and expanded through the Community Driven Development (CDD) component of the Sustainable Development in Poor Rural Areas Project (SDPRAP, P099751/P101844) which began implementation in 2010. SDPRAP corrects this project‟s two key design flaws (see paragraph 2.1.6),9 and the CDD component fully empowers village communities to determine the final use of available project poverty reduction funding. LGOP has proposed that it collaborate with the Bank on a practical review or study of recent experience with the participatory and CDD approaches in this project and under SDPRAP. Coupled with a review of this project‟s experience with outreach to ethnic minority people and of some emerging concerns (including the well-being of the elderly remaining in remote poor villages and of rural migrants residing in urban areas), this practical review could assist with the operationalization of the Government‟s new poverty reduction strategy through a follow-up Bank-supported poverty reduction project. 9 In SDPRAP, the leadership role of the central LGOP in project design and implementation has been restored, and the project‟s villagers and counties are not required to repay the IBRD loan. 17 2.5.3. Maintenance and operation of civil works and other activities. A second concern is the post-completion maintenance and operation of the project‟s many small-scale and scattered civil works and other activities to improve access to basic services. For the Sustainable Mountain Agriculture component, project farmers will continue to need technical support and ongoing training from line bureaus over the next few years. Many farmers are still learning how best to manage new crops and livestock breeds to achieve optimal production levels. Many farmers participated in irrigation schemes for the first time through the project, and they are still learning the transition from dryland farming practices to irrigated agriculture. As a result, they are not yet achieving the full production potential offered by irrigation investments under the project. These farmers will continue to need ongoing training and technical support to gain full benefit from the irrigation investments. The Rural Infrastructure component‟s investments now form important community assets and good post-construction operation and maintenance is essential to maintaining the functionality and life of the infrastructure. Many villages have organized themselves well to undertake infrastructure management and maintenance, but others still need help and ongoing technical assistance from the relevant line bureaus. For example, the water supply schemes are a new technology in many villages and the villagers are still learning to manage them correctly, and such villages will need to continue receiving support from the Water Resources Bureaus over the next few years, especially for system maintenance. To ensure sound ongoing operation and maintenance of infrastructure, county PMOs and the relevant township line bureaus should ensure that (a) all villages put in place suitable management and maintenance systems, (b) further infrastructure operation and maintenance training is provided to villagers where required, and (c) ongoing line bureau support is provided when required in the post-project period. 2.5.5. For the Basic Education, Basic Health, and Community Capacity Building components, the biggest challenge during post-completion will be the availability of the financial resources to sustain the operation of teaching points and remote schools, health clinics and health services delivery, and the CLCs. Funding support for teachers and health care workers will also be a significant challenge. Although village level governments are now receiving more and more intergovernmental transfers for a variety of purposes, it will be essential that financial support for basic education, basic health, and community capacity building be recognized as a top priority for poverty reduction and community development. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation 3.1.1 The project‟s objectives were and remain highly relevant, were an important element of the County Assistance Strategy (CAS), and are an important element of the current Country Partnership Strategy (CPS). At the time of project appraisal, one of the central goals of the Bank's CAS was to assist "the poorer and disadvantaged people and lagging regions through investment lending in rural development, infrastructure and social sectors, as well as training." The project directly supported this CAS goal by targeting the poorest villages and the most severely disadvantaged ethnic minority groups in eighteen of China's poorest counties with a multisectoral package of assistance including improved agricultural productivity, basic infrastructure, primary education and public health services, and community capacity building. Many of the ethnic minority groups who benefitted from the project have received very little international assistance in the past. The project provinces and counties are all located in the lagging region of western China. Similarly, Pillar 2 of the current CPS is to “reduce poverty, 18 inequality, and social exclusion.� The project directly supported the CPS Pillar 2 by targeting the poorest villages and most severely disadvantaged ethnic minority groups. The project‟s participatory approach also greatly contributed to the goal of shifting from top-down poverty reduction programs to community driven development. 3.2 Achievement of Project Development Objectives 3.2.1 New Approaches. The outcome indicator for the project development objective was “new approaches to poverty reduction explored and promoted in the strategies and plans of the three project provinces.� The project helped to explore and promote a number of approaches to poverty reduction in the three project provinces including the participatory approach, strengthened outreach to the poorest villages and most disadvantaged ethnic minority people, gender equality, and assistance for people with disabilities. The exploration and promotion of these improved approaches was highly successful, and the expansion of these approaches province-wide in Guangxi, Sichuan and Yunnan is already well underway. Although the provincial governments‟ new poverty reduction strategies for 2011-2020 have not yet been released, it is believed that the participatory approach and the focus on gender equity promoted under the project will be highlighted in their new strategies. Furthermore, the Central Liaison Group has reported that key approaches, including the complete participatory approach, are being adopted in other development programs in China. Also, evidence from all three provinces collected by M&E Module 2 shows that the participatory practices developed through the project are spreading within the provincial Poor Area Development Office system. 3.2.2 Participation. The first of the project‟s two development objectives was the “sustained participation of the poorest rural people in project design, implementation, and monitoring and evaluation.� The key performance indicator for this development objective was the number of project villages conducting participatory poverty assessments and participatory planning, and the values reported for this indicator exceed the agreed upon project goals in each of the three project provinces. In addition, based on the evidence collected through the M&E system (Modules 2 and 3), the participatory approach is rated highly satisfactory. The project not only fully achieved its objectives with regard to participation, but has also become an influential platform for innovative participation practices in China:  The project rolled out the participatory approach on an unprecedented scale, involving a very substantial number of remote natural villages in participatory planning and decision making. For example, in Guangxi, 88% of the project‟s remote natural villages had been directly involved in project decision making by the end of the project, and 399 project implementation groups had been set up in the poorest natural villages. Yunnan set up 893 project implementation groups in the project‟s poorest natural villages.  All project villages were required to conduct a participatory planning exercise prior to implementation. Spot checks during supervision confirmed that the participatory process was well documented and that participatory plans were prepared in all project villages visited. M&E Module 2 confirmed that project activities were generally implemented in line with community priorities, and in a number of cases community priorities were also addressed through additional investments from other funding sources.  The project‟s form of participation differed fundamentally from other government-led projects. Detailed analysis undertaken by M&E Module 2 shows that project communities substantially participated in decision making at critical stages as well as in the monitoring and final evaluation of project activities. Communities generally reported that, unlike in other government projects, project information was publicized and 19 decisions were taken in a transparent way under the project. Levels of participation have been particularly high in small-scale infrastructure activities (such as drinking water schemes) implemented through a community-led approach.  M&E Module 1 (NSB) data show that during 2006-2010, the proportion of rural households participating in the selection of poverty reduction activities in project villages rose from 19.5% to 46.2% and the proportion of project households participating in the determination of poverty reduction activities increased from 18.6% to 43.3%. The proportion of rural households believing that the poorest households should have preferential access to project poverty reduction activities increased from 34.7% to 50%. Also, rural households‟ knowledge of the project within the three provinces‟ project villages rose from 51.6% to 87.1%.  The participatory approach was an important vehicle to increase attention to women‟s needs. Women participated in selecting priority project activities, participated in the discussion of implementation proposals for community-implemented project activities (such as drinking water schemes), and participated in formulating subsequent management and maintenance systems. The project effectively addressed their expressed preferences (including schools, village clinics, and drinking water) through the participatory process. Project staff reported that they substantially changed the way in which they address gender as a result of the project, in particular with regard to motivating and targeting women and considering gender-specific needs.  Through its successful implementation, the project has set an example for decision makers that substantial levels of participation can be achieved on a large scale in government-led projects if sufficient attention is paid to ongoing capacity building and learning. The 2011 M&E Module 2 survey shows that PMO staff have achieved a broader and more inclusive understanding of participation as a process to engage with poor people as key stakeholders, not just as beneficiaries.  The project has developed a number of innovative practices that are seen as useful for strengthening community engagement and pro-poor focus in poverty reduction projects (such as participatory poverty analysis, community procurement, and setting up project implementation groups in remote natural villages). As a result, the project has become an influential platform for innovative participation practices in China10. 3.2.3 The project was successful because it addressed the challenges to the participatory approach related to the sheer size of the project area, the cultural and environmental diversity of the locations covered, the complexity of project‟s multisectoral design, the number of stakeholders involved, and the limited prior experience with participation at the local level through a progressive approach that involved the gradual roll-out of participation and ongoing learning and improvement. The project (a) implemented a rigorous system for monitoring progress on participation, (b) set out to achieve high levels of participation beyond the planning phase, thus extending the understanding of participation to the implementation and M&E phases, (c) encouraged diverse methods, which were adapted to the culturally and socially diverse environment, and (d) supported a strong focus on learning from practice (for example, through 10 See for example Proceedings of the Kunming Workshop on Practices for Participatory Poverty Reduction (2009), the documentation of the Final Workshop on Reflection and Final Evaluation of Participation in PRCDP (Kunming 2011), and the forthcoming issue of the PLA Notes No. 62 “Emerging Practices for Participatory Poverty Reduction in China� (2011). In addition, there is a wealth of (Chinese language) documentation on good practices and experiences from the Provinces. 20 documentation of good practice notes and regular workshops where facilitators could share their experiences). 3.2.4 The flexibility integrated in the design of participatory approach supported high levels of local ownership and sustained learning throughout the project. The wealth of Good Practice Notes which have been produced by the provinces are evidence of this. External monitoring (in particular M&E Modules 2 and 3) was instrumental for supporting the reflective learning on participation. At the same time, it played a pivotal role in safeguarding the quality of the participation approach. Without regular supervision and external monitoring, the project would not have been able to strengthen its outreach to poor natural villages and achieve sustained participation throughout implementation. As a result of intensive supervision of provinces and counties lagging behind in the process, the roll-out of the participation approach visibly improved after the first two years of implementation, and issues of participation and social inclusion remained a focus of attention throughout the life of the project. 3.2.5 The project‟s participatory approach also supported high levels of transparency and ownership at the community level. As a result, the project communities felt more responsible for the management and maintenance of project facilities than in other government-led projects. The participatory approach also strengthened community capacities and social capital. For example, M&E Module 2 reported that many villagers and, in particular, women became more confident about their productive and non-productive skills as a result of the project. Project communities learned the skills necessary to manage and supervise projects and, as a result, many PRCDP villages have been able to attract substantial follow-up investments by nonproject agencies because they already had a plan and the demonstrated capacity to implement projects. 3.2.6 The commitment of local project staff was a key factor in the successful implementation of the participation approach, and it also facilitated lateral changes within and beyond the project. Learning on participation has transformed staff attitudes and behaviour beyond the application of participatory tools and methods. As a result of the participatory process, staff have become increasingly aware of social inclusion issues, such as outreach to marginal groups, gender equality, and respect for cultural diversity. Furthermore, project staff have increasingly applied participatory methods outside of the project. Evidence from all three provinces collected by M&E Module 2 shows that participatory practices developed through the project are spreading within the provincial PADO system. There is also evidence from Guangxi that project practices have been adopted in other agencies including, in particular, county water bureaus and the DRC. 3.2.7 Livelihoods. The second of the project‟s two development objectives was to “improve livelihoods security of the poorest rural people.� The key performance indicator for this development objective was the incidence of poverty (that is, the share of the population below the poverty line) in the project areas. The values reported by the project provinces for this indicator improved dramatically during project implementation (though they did not fully achieve the agreed upon, but perhaps overly ambitious, project goals in two of the three project provinces). Results from Module 1 of the project‟s independent and external M&E system, undertaken by NSB, more rigorously demonstrate that the project generated broad and significant positive impacts on the beneficiaries‟ livelihoods. First, from 2006 to 2010, project farmers‟ per capita incomes increased by 12% per year, which is greater than that of non-project villages (9%), the national average for rural areas (9.2%), and all nationally designated poor counties (10%). During the same period, project farmers‟ per capita consumption increased at 8.7% per year, which is greater than that of the national average (7.5%) and of all nationally designated poor counties (8.1%). Second, according to the NSB Module 1 surveys, the incidence of poverty in project villages registered a strong decline. In the case of Yunnan, it declined from about 23% to 21 9%. Sichuan and Guangxi also recorded big declines although the two provinces enjoyed relatively better initial conditions compared to Yunnan. Third, access to basic infrastructure, services and off farm jobs in the project villages improved significantly. For example, in the case of Guangxi, the percentage of households with running water in the project villages increased from about 23% in 2007 to 47% in 2010; in Yunnan and Sichuan, it increased from 42% to 55% and 29% to 39% respectively during the same period. Indicators of electrification, telephone services, labor education level, and migrant labor outflow also registered significant improvements. Overall, based on the evidence collected through the M&E system (both the internal MIS and the external Module 1), the improvement of livelihoods security is rated satisfactory. 3.2.8 Sustainable Mountain Agriculture component. The Sustainable Mountain Agriculture component was the largest component in terms of investment (45% of total investment) and is rated as satisfactory. The component focused on improving food security and cash income generation to reduce poverty and increase livelihood security. These objectives were achieved through investments in grain production, cash crops, tree crops, livestock and small scale value- added agricultural processing. Production activities were supported by farmer training conducted under the Community Capacity Building component. The incidence of poverty in project villages has been significantly reduced in all three provinces (see paragraph 3.2.7). Food security in project villages has also been improved with significant reductions in the percentage of households with less than 150kg grain/capita/year in Sichuan (from 14.0% in 2002 to 4.8% in 2010) and Yunnan (from 13.4% in 2001 to 2.5% in 2010). Increased income from cash crops and livestock financed under the project were significant contributors to this income growth (see Annex 2). 3.2.9 Rural Infrastructure component. The Rural Infrastructure component (33% of total investment) was very well received by project beneficiaries and is rated as satisfactory. The component contributed to improvement of living conditions, supported development of livelihood activities and improved access to social services through provision of basic rural infrastructure, including village roads, drinking water supply, irrigation works, land improvement, rural energy and improved housing. Road access coverage was extended to 202,066 households (802,264 people). Village access roads accessible to tractors and light vehicles have been effective in linking many natural villages to existing rural road networks and other natural villages drawing such villages into the mainstream of local economic activities often for the first time. Project households demonstrated strong support for road construction and the value of their labor contributions significantly exceeded the cash investment by the project. Construction of drinking water supply was also very well received and the project successfully solved drinking water problems for about 120,154 households (480,000 people). These low cost, simple, and easy to maintain facilities have improved drinking water quality, reduced sickness days, and relieved households from the labor burden of carrying drinking water from remote sources. This has freed up labor to be profitably used for food production and income generation activities. Small irrigation systems have improved security of water supply for agriculture enabling farmers to increase yields, grow cash crops, and additional feed for livestock, thereby contributing to increased income generation. Land improvement through terracing or through improvement of low and medium yield fields increased the area of arable land for crop production in project villages. Improved rural energy supply through biogas and fuel efficient stoves has contributed to labor-saving (less fuel wood collection) and cleaner energy, while supply of electricity has provided opportunity for small scale processing activities. These infrastructure investments all directly underpin agriculture production and other types of economic development, thus contributing indirectly to improved livelihoods security and income generation in project villages. 22 3.2.15 Basic Education, Basic Health and Community Capacity Building components. The Basic Education, Basic Health, and Community Capacity Building components are each rated satisfactory. The key performance indicator for the Basic Education component was the sixth grade completion rate, and the values reported for this indicator increased sharply in both Sichuan and Yunnan.11 The key performance indicator for the Basic Health component was the percentage of project households receiving basic health education, and the values reported for this indicator also increased sharply in both Sichuan and Yunnan.12 The components‟ investment and output targets were achieved, although investment in “hardware� items (such as the construction of school buildings, township hospitals and village clinics, and the provision teaching and medical equipment) proceeded more quickly and comprised more than half of total component expenditure. Investment in “software� items -- including the establishment and operation of Community Learning Centers (CLC), the Free Education Pilot (FEP), the provision of public health services, and the training of teachers and medical workers -- took longer to complete, but played a key role in consolidating the overall achievements of these components. These three components certainly added to the project‟s high level of complexity, and the innovative designs of the CCL, FEP and HIV/AIDS activities were particularly challenging for the provincial and county PMOs. However, all three provincial PMOs confirmed that these three components were (a) valuable additions to the project which were certainly worth the additional complexity, and (b) particularly important to the successful outreach to the poorest and most remote project villages, ethnic minority communities, women and other vulnerable groups. 3.2.16 Although the appreciation of the yuan during the period of project implementation resulted in a reduced project budget, all three provinces were able to complete the establishment and operation of the agreed upon number of CLCs. Longsheng County did an excellent job of implementing the community capacity building component, and in 2007 Longsheng was awarded UNESCO‟s International Award for Reducing Illiteracy. The results of the Completion Review of the FEP have been quite encouraging. The enrollment and completion rates in the FEP pilot villages and townships increased quite significantly. The data also show that the most disadvantaged groups, such as girls and ethnic minorities, benefited considerably from the FEP. However, the data also document that some children were not able to finish their primary education, and that there is a significant way to go to fully realize the target of universal free education in China‟s poorest rural areas. 3.2.17 The Basic Health component strengthened the capacity of the basic health care system in the project areas and provided a variety of public health services including planned immunizations, maternal and child health care, allowances for modern delivery, medical assistance, general investigation and monitoring of epidemic diseases, and HIV/AIDS prevention and control. In Sichuan, for example, the project provided maternal and child health care services to 21,150 women and 22,578 children, modern deliveries for 21,085 infants, and 505,485 person/times of medical assistance. In Yunnan, benefitting from the advanced design and comprehensive methods of an earlier Global Fund activity, the Cangyuan County HIV/AIDS subcomponent included HIV/AIDS education and prevention, methadone maintenance treatment, needle exchange, peer education of commercial sex workers, mother to child transmission control, and AIDS anti-retrovirus treatment (ART) activities. Although representing only a very small share of total project cost, the Cangyuan County‟s HIV/AIDS prevention and treatment 11 The Guangxi KPI for the Basic Education component was the female students‟ dropout rate, and the values reported for this indicator also achieved the agreed upon project target. 12 The Guangxi KPI for the Basic Health component was the number of project households receiving medical assistance, and the values reported for this indicator also achieved the agreed upon project target. 23 subcomponent greatly contributed to the prevention and treatment of HIV/AIDS in Cangyuan and established a truly great model of effective HIV/AIDS prevention, control, case management and treatment in a very poor and remote ethnic minority area. 3.2.18 The original design of the Basic Health component included support for health education for communities on HIV/AIDS. However, none of the project provinces initially elected to support a full HIV/AIDS prevention and treatment activity. In response to repeated requests from the Bank team, the Yunnan PPMO began support of an HIV/AIDS prevention and treatment activity in Cangyuan and two other project counties during the third year of project implementation. Benefitting from the advanced design and comprehensive methods of an earlier Global Fund activity, the Cangyuan County PRCDP HIV/AIDS prevention and treatment activity included HIV/AIDS education and prevention, methadone maintenance treatment, needle exchange, peer education of commercial sex workers, mother to child transmission control, and AIDS anti-retrovirus treatment (ART) measures. 3.3 Efficiency 3.3.1 Rate of return analysis. As confirmed by the results of the M&E system Modules 1 and 2 reports (from the NSB‟s annual quantitative monitoring and evaluation as well as the ITAD Company‟s annual qualitative Participatory Poverty Monitoring surveys), the project generated significant economic and social benefits (see paragraph 3.2.7). In addition, a traditional methodology was used to estimate the project‟s economic rate of return (ERR). First, the streams of costs and benefits for different project activities were estimated, and the net benefits derived from these streams were then used to calculate the net present value (NPV) and ERRs for these activities and for the project as a whole. The individual components‟ ERRs are weighted by their corresponding proportion in the total project investment to determine the overall ERR for the project. The benefits of three major components, that is (a) Sustainable Mountain Agriculture, (b) Basic Rural Infrastructure, and (c) Basic Education, were quantified. Investment in these three components accounts for about 85% of total project investment. The benefits of other three project components (that is, Basic Health, Community Capacity Building, and Project Management and Monitoring) are difficult to quantify and hence ERRs have not been calculated for them. As summarized below, the overall ERR for the project, under a few conservative assumptions, is estimated to be around 30%, which is the same as the estimate in the Project Appraisal Document, and the individual ERRs range from 22% for Basic Education to 31% for the Sustainable Mountain Agriculture components: Project as Sustainable Basic Rural Basic whole Mountain Infrastructure Education Agriculture ICR 30% 31% 29% 22% PAD 30% 35% 27% 19% 3.3.2 No further efforts were undertaken to separately estimate the financial rate of return (FRR). After more than three decades of market-oriented economic reforms, the price distortions due to the government‟s intervention in the Chinese economy have been basically eliminated. Rich and strong empirical evidence consistently show that an integrated domestic market has been established in China. In addition, the Chinese economy has been increasingly integrated with the international market. Against this background, there is essentially no additional value to separately calculate FRRs to capture the impact of price distortions. 24 3.4 Justification of Overall Outcome Rating Rating: Satisfactory 3.4.1 The project is highly relevant to China‟s poverty reduction program and the overall outcome is rated satisfactory. Had two key design flaws (see paragraph 2.1.6) not adversely affected implementation quality during the project‟s first three years, the overall outcome would have been rated highly satisfactory. Indeed, by the time of project closing in 2011, achievement of the PDOs is rated highly satisfactory for sustained participation and satisfactory for improved livelihoods security. Furthermore, outreach to the project area‟s poorest villages and to the most disadvantaged ethnic minority people are also both rated highly satisfactory. The project also made important contributions to gender equality and to assisting people with disabilities in an extremely remote ethnic minority area in rural China. Lastly, the rate of return analysis shows that the project was highly efficient, and the thorough M&E system has provided an unusually rich and rigorous set of quantitative and qualitative data to assess project impact. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 3.5.1 Outreach to poorest villages. The PDO was to improve livelihoods security and achieve sustained participation of the poorest rural people. China‟s poorest rural people are geographically concentrated in the poorest villages, townships and counties within the western provinces, and it was possible for the project area selection process to focus on the poorest townships within the project‟s 18 nationally-designated poor counties. The project selection process was very successful and the project‟s targeting of the poorest rural people was exemplary. However, based on the experience of the earlier poverty reduction projects in China, the Bank/DFID team sought to further guarantee outreach to the poorest people by identifying the poorest administrative villages in the project area and to require that those poorest administrative villages received at least an equitable share of project benefits.13 Many of the most disadvantaged ethnic minority communities are found in these poorest administrative villages. Niha Village in Meigu County in Sichuan, for example, is entirely comprised of Yi minority people and was certainly one of poorest and most remote villages in China before receiving assistance through the project. 3.5.2 Although not recognized as an explicit project objective, it was understood from the beginning of project preparation that outreach to the poorest administrative villages was a core project objective and would be a key measure of the project‟s overall success. No specific key performance indicator was established to measure project performance on outreach to the poorest administrative villages, but Sichuan provided very useful information on this aspect of the project in their final implementation completion report. Sichuan‟s analysis showed that the poorest administrative villages in each of the six project counties received greater amounts of project assistance (on a per capita basis) than the average amount of project assistance per administrative 13 The project area selection process of the earlier Southwest and Qinba Mountains Poverty Reduction projects also successfully targeted the poorest townships of nationally-designated poor counties, but it was determined that the poorest administrative villages within the project area sometimes received a less than equitable share of project benefits. 25 village (on a per capita basis). This focus on assistance to the poorest administrative villages was a priority of the design process and always played a key role in Bank/DFID supervision work. As evidenced by Sichuan‟s analysis and many of the Bank/DFID supervision mission‟s aide memoires, the project‟s emphasis on assisting the poorest administrative villages strengthened over time (see paragraphs 2.2.7 and 2.2.8) and was considered highly successful by the time of project completion. 3.5.3 Gender equality. Gender equality was emphasized during all stages of the project. The project‟s Participation Manual includes principles, indicators and practices to focus attention on the equal participation of men and women during participatory poverty analysis, project planning, implementation and monitoring. In addition, relying in part on the Participation Manual, the project provinces developed their own training materials and guidance which generally paid attention to the participation of disadvantaged groups, including women. The project‟s internal MIS and external (Module 2) systems also adopted indicators to monitor the participation of women. The available data demonstrate that women actively participated in the project and that their priorities were broadly addressed. The key performance indicator for gender equity was the percentage of women taking part in participatory planning, and the values reported for this indicator exceed the agreed upon project goals in each of the three project provinces. In addition, the M&E Module 2 reports note that (a) women participated in the selection of component activities, discussions about the implementation of those activities to be undertaken by the villages, and formulating subsequent management and maintenance systems for these village project activities, (b) the project effectively addressed women‟s preferences as expressed through the participatory process (schools, village clinics, and drinking water schemes were among women‟s priorities), and women benefitted from training (which included training on women‟s healthcare), subsidies for pregnant and delivering women, and loans for livestock (which was another priority of women), and (c) women‟s participation improved during implementation. Provincial data show that women benefitted more than men from training in Yunnan and Guangxi, and that the number of project loans taken out by women increased over time and was particularly high in Yunnan. This reflects the increasing attention that staff have paid to gender equality. During the 2011 M&E Module 2 survey, staff reported that they substantially changed the way in which they address gender as a result of the project, in particular with regard to motivating and targeting women and considering gender-specific needs. However, the available evidence also points toward the need for a continuing emphasis on women‟s participation in decision making. Women made roughly half of the labor contributions, but comprised less than one third of management staff at the village level. 3.5.4 Outreach to ethnic minority people. Although not recognized as an explicit project objective, it was understood from the beginning of project preparation that outreach to the poorest ethnic minority people was a core project objective and would be a key measure of the project‟s overall success. No specific key performance indicators were put in place to regularly measure project performance on outreach to ethnic minorities, but the focus on ethnic minority peoples was clearly a priority of the design process and always played a key role in Bank/DFID supervision work. Project design was informed by a thorough Social Assessment that provided information on (a) ethnic minority composition in the project area, (b) key social organizational features of ethnic groups (including decision making mechanisms and gender relations), (c) the related legal framework and existing governmental assistance programs, and (d) perceptions of ethnic minority development by governmental officials. Ethnic minority peoples comprise two thirds of the project‟s primary beneficiaries, and comprise more than 70% of total county 26 population in 10 of the 18 project counties. During supervision, Bank/DFID missions concentrated on the six project counties which had the highest proportions of the poorest and most disadvantaged ethnic minority people.14 3.5.5 Based on the available data, it is clear that the project was highly successful in its outreach to the poorest ethnic minority peoples. First, targeting of ethnic minorities was achieved through the selection of ethnic minority counties for project implementation and through the prioritization of ethnic minority administrative and natural villages for project implementation. As a result, ethnic minority peoples comprised two thirds of the project beneficiaries and outreach to the very poorest and most remote ethnic minority villages improved during the course of project implementation (see paragraphs 2.2.7 and 2.2.8). Second, inclusion of ethnic minorities in project planning, implementation and monitoring was measured through (a) routine monitoring data on the ethnic minority composition of Project Implementation Groups, and (b) Modules 2 and 3 of the project‟s M&E system indicates that a significant numbers of ethnic minority people were engaged in project consultations and decision making.15 Findings of the M&E Module 2 reports show good outreach to ethnic minority communities and their involvement in key stages of project implementation, including its monitoring. The M&E Module 1 Final Report also documented that, relative to the averages for all rural households in China and for all rural households in the nationally-designated poor counties, ethnic minority households in the project villages enjoyed greater increases in per capita annual net income (during 2006-10) and in elementary school enrollment rates (during 2007-10). 3.5.6 Assistance for people with disabilities. A generous US$2 million DFID Grant supported the Assistance for People with Disabilities (PwD) component in Yuexi County in Sichuan. The component‟s objective was to increase the social inclusion of PwD and help prevent disabilities in the future. The Yuexi Disabled People Federation (YDPF) implemented the component in close cooperation with the Yuexi County Government and with technical assistance and financial management from Handicap International (HI). The component‟s primary outputs were to (a) build capacity at the County level to plan and implement disability action; (b) develop a replicable Disability Management Information System (DMIS) for disability action; (c) establish a community based system for the prevention of disability, rehabilitation and social inclusion (including a sustainable Disability Fund); and (d) develop regulations, training and reference materials for mainstreaming PwD into different sectors of society. The component was successful, and was quite extraordinary in its efforts to assist PwD in an extremely remote ethnic minority area in China: (a) more than 3,000 PwD in 137 villages in Yuexi County received some kind of service or support, (b) PwDs gradually gained better access to services and became more involved in socio-economic activities, while learning about laws and regulations and becoming more aware of their rights; and (c) the capacity of local partners has been greatly enhanced and improved in terms of their understanding and acceptance of the Community Based Rehabilitation (CBR) concept. The project benefited from the constant involvement of the China 14 These six counties are Ximeng, Menglian, and Cangyuan in Yunnan, Meigu and Yuexi in Sichuan, and Huanjiang in Guangxi. Of the 26 Bank/DFID supervision mission field visits to project counties during 2005-10, some 15 visits were made to these six counties. 15 The PAD indicator to measure ethnic minority inclusion in the project is “People from different ethnic minorities are at least proportionally represented in decision-making and project planning, implementation and monitoring.� Although information was not collected by Provincial PMOs in this format, available data shows that in: (a) Guangxi, 80% of people involved in project planning and decision making were from ethnic minority communities, and (b) Yunnan, ethnic minority representatives comprised 61% of the village project management teams. 27 Disabled Persons Federation (CDPF), the main organization at national level dealing with disability issues. A set of reference materials was formally presented to the CDPF at a national conference on CBR in September 2011, and the official Chinese translation of the latest version of the CBR guidelines, developed by WHO, ILO and UNESCO, was made available in late-2011. (b) Institutional Change/Strengthening 3.5.7 Institutional development at both the central and provincial levels and the county and lower levels was a major focus of the project. The central LGOP system was established in 1985 and was already extremely capable and accomplished before the outset of project preparation. During the early stages of project preparation, the central LGOP system was deeply involved in project design, but the central government decided that the role of the central LGOP system in project leadership and support should be sharply curtailed. Fortunately, the Foreign Capital Project Management Center (FCPMC) and its CLG – an operational arm of the central LGOP system – was able to stay heavily involved in the final stages of project preparation and throughout implementation and the final project evaluation. Although a higher profile engagement of the central LGOP system would have greatly benefitted project implementation and better facilitated the transfer of project experience to the wider LGOP system, it is clear that (a) senior FCPMC staff remained constantly engaged in project design, implementation, and evaluation, and (b) the key lessons of the project‟s experience with participation, outreach to ethnic minority people, and gender equality have been fully absorbed into China‟s overall poverty reduction program. Most importantly, senior FCPMC staff participated in all the project supervision and final evaluation missions and were deeply engaged in the “Final Workshop on Reflection and Final Evaluation of Participation in PRCDP� held in Kunming, Yunnan in March 2011. 3.5.8 All three provincial PMOs had participated in the earlier Southwest and Qinba Mountains Poverty Reduction Projects and were also already extremely capable and accomplished before the outset of project preparation. The provincial PMOs played the key role in overcoming the obstacles to successful implementation during the project‟s first two years and in promoting the project‟s rollout of the participatory approach, outreach to ethnic minority people, gender equality, and other innovations. These key lessons of the project‟s experience with participation, outreach to ethnic minority people, and gender equality have been fully absorbed within the three provincial PADOs. Perhaps the best example of the absorption of the project experience in the provincial PADOs is the quality of the Sichuan PMO‟s response to the 2008 Wenchuan earthquake in Sichuan. The United Kingdom‟s International Development Committee visited Sichuan soon after the earthquake and recognized the Sichuan PMO‟s use of participatory methods and strong monitoring system (see Annex 8 paragraph 1.3) in Sichuan‟s response to the disaster. 3.5.9 The project county PMOs, Township Work Stations, and lower level project units were all new to the demands of designing and implementing a multisectoral and innovative poverty reduction program, and the project actively promoted institution and capacity building at the county and lower levels. Particularly in Guangxi, where the joint PADO and DRC management of the project initially confused the line of command, there were significant gaps in capacity at the county and lower levels in the early years of project implementation. All Bank/DFID supervision missions evaluated institution and capacity building at the local level, and problems were highlighted in mission Aide Memoires. Most importantly, the April 2007 Fourth Supervision Mission found extreme deficiencies in institutional capacity in Luocheng County in Guangxi. The Guangxi and Luocheng County PMOs took immediate actions to correct these problems, and the November 2007 Fifth Supervision Mission found that “Luocheng County had 28 made excellent progress on institution building and improving implementation quality.� Overall, by the conclusion of the project in 2011, county and lower level institution and capacity building was found to be excellent in all project counties. Based on the experience of the earlier Southwest and Qinba Mountains Poverty Reduction projects, it is expected that the county level institutions and poverty reduction project implementation capacity will be sustained for years to come. (c) Other Unintended Outcomes and Impacts (positive or negative) None. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 3.6.1 Although a separate beneficiary survey was not undertaken as part of an Intensive Learning ICR, the M&E system Module 2‟s Satisfaction Survey reported high satisfaction levels throughout the project‟s many activities.16 Village leaders generally gave higher ratings than the members of the community. Men gave higher ratings, in particular for roads. Women‟s ratings were high for drinking water, but generally lower than men‟s. Women complained that they were less informed, mostly because information was displayed at central locations (village meeting room) but women often did not have time or opportunity to go to these locations. 3.6.2 Rural infrastructure. All infrastructure activities were rated high on utility (64%-86% were rated “very high�). Road activities were rated highest on utility (86%). Irrigation schemes were rated highest on transparency (83%) and quality (67%). Communities generally felt that they were well informed about the activities, that decisions were taken in a transparent way, and that budgets were publicly disclosed. Only a minority of communities thought that the costs for infrastructure works were very low, and communities had to contribute often substantial (labor and cash) costs to the construction of community infrastructure: Table: Percentage of infrastructure activities rated “very high� per activity (Project villages) Indicator Knowledge Timeliness and Project Usefulness Effective Assets Cost and responsiveness Quality use of maintenance understanding assets and of the project management Irrigation Ditches (6 projects) 83% 50% 67% 67% 50% 17% 33% Drinking water (29 projects) 59% 48% 52% 76% 48% 41% 14% Village road & sub-village road (29 66% 48% 31% 86% 59% 17% 7% projects) Village clinic construction (24 50% 63% 42% 79% 50% 42% 25% projects) Primary school construction (11 45% 55% 36% 64% 55% 45% 45% projects) 16 M&E system Module 2 facilitated group discussions on a pre-defined set of criteria (“community score card�) in project villages and control villages. These are the ratings from the 2011 survey which covered 36 project villages and 12 control villages in all three provinces. Communities have only rated the project activities that were implemented in their village. Community score cards were prepared separately for men, women and village leaders. The ratings presented here are consolidated for all three groups. For detailed ratings refer to the M&E Module 2 reports. 29 3.6.3 Comparison with control villages shows that (in nearby non-project villages) infrastructure works were rated significantly lower on all criteria except utility (50%-100% were rated “very high�). Substantially lower ratings on “knowledge and understanding� (0-33%), “timeliness and responsiveness� (25%-33%) and “assets maintenance and management (0-33%) indicate that the project has added substantial value in terms of transparency, responsiveness and sustainability. The only exception is construction of village clinics, where ratings were similar to those of the project. People in project villages were more satisfied with the construction of school buildings. Communities had previously complained about the lack of information in works implemented by sector departments. The most significant differences were for drinking water and irrigation schemes, which were rated low on most indicators in control villages: Table: Percentage of infrastructure works rated “very high� per activity (control villages) Indicator Knowledge and Timeliness & Project Useful Effective use Assets Cost understanding of responsivene Quality ness of assets maintenance & the project ss management Irrigation Ditches (4 25% 25% 25% 50% 25% 0% 0% projects) Drinking water (9 0% 11% 33% 67% 44% 11% 0% projects) Village road & sub- 25% 13% 0% 63% 38% 25% 0% village road (8 projects) Village clinic 75% 50% 50% 100% 50% 25% 50% construction (4 projects) Primary school 33% 33% 33% 100% 33% 33% 33% construction (3 projects) 3.6.4 Drinking water and irrigation schemes were the infrastructure works where the project added the most value. A detailed review of ratings for drinking water schemes shows that they were particularly well received in Guangxi and Sichuan. Communities in Yunnan have given lower ratings on a number of criteria, but most significantly on the timeliness and responsiveness. People were reportedly also not satisfied with the limited coverage of drinking water schemes in Guangxi and Yunnan. In Sichuan quality was rated very high, and dissatisfaction was reported only for two drinking water schemes in Mabian County, where the quality of drinking water was reportedly unsatisfactory: Ratings for drinking water works in project sample villages 25 20 15 10 5 Sichuan 0 Yunnan low: Guangxi Fair: Fair: Fair: Fair: Fair: Fair: Fair: Fair: Fair: Low: Low: Low: Low: Low: Low: Low: Low: high: High: High: High: High: High: High: High: High: Very low: Very low: Very low: Very low: Very low: Very low: Very low: Very low: Very low: Very high: Very high: Very high: Very high: Very high: Very high: Very high: Very high: Very high: Knowledge and Attitude and Timeliness and Project Quality Practical use Effectiveness of Targeted to the Assets Costs understanding of Behavior of Staff responsiveness built assets use poorest maintenance and the project management 3.6.5 Overall ratings were lower for other non-infrastructure works. On the one hand, ratings for Community Learning Centers (CLC) and subsidies for pregnant women were still relatively high, in particular with regard to utility (83%). On the other hand, the Sustainable Mountain Agriculture loans activity was rated lower on transparency and responsiveness than most other project activities, including infrastructure projects. During the first two years there were many complaints about the lack of transparency in the allocation of agricultural loans. However, since then, the loan activity reportedly became more transparent: 30 Table: Percentage of other activities rated “very high� per activity (project villages) Indicator Knowledge and Timeliness and Quality Usefulness Cost understanding responsiveness of the project Loan (30 projects) 43% 33% 27% 53% 10% Training (21 projects rated) 48% 57% 29% 71% 62% Free Education (2 projects) 0% 0% 0% 50% 0% Biogas tank (3 projects) 0% 0% 33% 33% 0% Community Learning Center (6 projects) 50% 50% 33% 83% 67% Subsidies for pregnant women (8 projects) 38% 50% 63% 88% 38% 3.6.6 Among the criteria rated in M&E Module 2 was “attitude and behavior of staff.� Ratings were significantly higher in project villages than in the control villages. Attitude and behavior of staff were rated “very high� in 59% of the project villages (21% of control villages). Some 17% of the communities in control villages were clearly not satisfied with the attitude and behavior of staff compared to 6% in project villages. Feedback from beneficiaries seems to support the overall findings from M&E Module 2 that the mindsets, attitudes and behavior of project staff have changed as a result of the project. Overall, staff were found to be more concerned about the beneficiary needs, communicated better, and became more responsive: Attitude and Behaviour of Staff very high high fair low very low PRCDP villages 59% 36% 4% 1% 1% Control villages 21% 62% 15% 2% 4. Assessment of Risk to Development Outcome Rating: Moderate (see Section 2.5) 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 5.1.1 Project quality at entry was moderately satisfactory and the project preparation process benefitted from a major study of rural poverty in China and the experience of the earlier Southwest and Qinba Mountains Poverty Reduction projects in China (paragraph 2.1.3). The PAD provided a clear summary of sectoral issues at the time of project preparation, and these sectoral issues remained important to China‟s rural poverty reduction program throughout the life of the project. The project design conformed exactly to these sectoral priorities. Although the project design was highly complex, the Bank/DFID project preparation team included all the necessary sectoral, participation, safeguards, fiduciary, and M&E staff sufficient to ensure good quality project preparation. Unfortunately, (a) the project preparation period extended some 45 31 months from October 2001 to June 2005 (paragraph 2.1.1), and (b) two significant design flaws were clearly recognized during the project preparation period (paragraph 2.1.6). The delay in project preparation was certainly excessive, but the causes of the delay were outside the control of the Bank team. The Bank team (including the Country and Sector Managers at that time) played a proactive role in negotiating solutions to the problems which had led to the delays in the preparation process, and it is likely that preparation would have failed entirely in the absence of this proactive stance. 5.1.2 Similarly, the two substantial design flaws (namely, the reduced leadership role of the central LGOP in project design and implementation, and the requirement that project counties and the primary beneficiaries repay the project loan) were outside the control of the Bank team. The Bank team engaged in extensive discussions on these issues, and partial solutions to these design flaws were eventually agreed upon and formally recorded in the Loan Agreement. These imperfect solutions (namely, establishing the CLG, and limiting total cumulative investment in the Sustainable Mountain Agriculture component) allowed the project preparation process to move forward, but it was understood at the time that there would be consequent significant risks to the pace and quality of project implementation. These design flaws and their associated risks were identified in the PAD and, unfortunately, came into play during the first two years of project implementation (2006-07). However, the Bank team did a moderately satisfactory job of reaching the best possible compromises on these contentious issues and establishing mitigation measures to contain the risks associated with these design flaws. With hindsight, it is now possible to conclude that the Bank team made the correct decision to proceed with the project despite these known design flaws rather than to have abandoned project preparation. (b) Quality of Supervision Rating: Satisfactory 5.1.3 The Bank team‟s intensive and thorough supervision of the overall project and of each of the project components was highly satisfactory. However, supervision of financial management and procurement during 2006-08 was less intensive and is rated only moderately satisfactory, and therefore the overall rating for the Bank‟s quality of supervision is rated satisfactory. There were twelve supervision missions (averaging just over two supervision missions per year), and the large Bank/DFID team included individual sectoral specialists (for participation, ethnic minorities, gender, agriculture, infrastructure, education, health, and capacity building) on almost every mission.17 Each of the supervision missions established an updated project action plan, and a detailed Mid Term Review was completed in 2008. This intensive and comprehensive supervision process focused on the achievement of the PDO and of each of the component objectives. Most importantly, the Bank/DFID supervision missions very actively identified problems with project implementation quality and explicitly highlighted these problems in the mission Aide Memoires. The Bank team strongly believes that this intensive and comprehensive supervision process was an essential element of the eventual success of the project. Most importantly, the very candid and explicit identification of project implementation issues played a vital role in resolving these problems and improving the quality of project implementation over 17 As the Bank Country Director informally agreed with the Head of DFID‟s China Office during the project preparation period, the Bank provided a full double allocation of supervision budget during each year of project implementation. In addition, DFID‟s strong budgetary support for the project covered most or all of the costs of fielding the participation, gender, infrastructure, education, and health specialists. 32 time. For example, the December 2007 Fifth Supervision Mission found that better off Dai ethnic minority villages in Mengma Township in Menglian County in Yunnan had received an excessive share of project benefits while the poorest Lahu and Wa villages (in Mengma Township) had received only very limited benefits, and the mission applied very strong pressure on the provincial and county PMOs to improve outreach to the poorest ethnic minority people. Three years later, the October 2010 Twelfth Supervision Mission found that outreach to the poorest and most disadvantaged ethnic minority people in Menglian County in Yunnan was excellent, and that investments in the poorest Lahu and Wa ethnic minority villages had greatly increased (while all further project investments in the better off Dai villages had been cancelled) in Mengma Township. The Menglian County Magistrate advised the Twelfth Supervision Mission that this corrective action would not have been taken in the absence of the strong constructive pressure applied by the Fifth Supervision Mission. 5.1.4 The 2008 QAG review of the project found that the Bank‟s supervision of project financial management and procurement was inadequate. Although there were no complaints and no cases were referred for higher level review, the QAG review concluded that there should have been two financial management and procurement reviews per year. The frequency of financial management and procurement review increased during 2009-11, and financial management and procurement staff joined the regular supervision missions instead of undertaking separate reviews. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory 5.1.5 The overall performance of the Bank is considered to have been satisfactory. In the absence of the lengthy delays in project preparation and the two significant design flaws, and with somewhat greater attention to the supervision of fiduciary policies, the Bank‟s performance would have been highly satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Moderately Satisfactory 5.2.1 China is widely recognized for its tremendous achievements and great determination in reducing absolute poverty and, during the early stages of project preparation, the Government demonstrated strong commitment and support for the project. Unfortunately, the Government chose to sharply constrain the level of LGOP‟s support and leadership for the project. Instead, during the project preparation process, Government required that project leadership be devolved to the provincial DRCs and PADOs, and central government financial support for the project was eliminated. This directly and indirectly led to the two significant design flaws which adversely affected the initial pace and quality of project implementation. Though substantially constrained by the imposed limitations on their role in project management, the CLG was able to play a consistent and vital leadership role and very much helped to overcome the initial deficiencies in project implementation. Overall, Government performance is rated moderately satisfactory. (b) Implementing Agency or Agencies Performance Rating: Highly Satisfactory 5.2.2 The implementing agencies‟ performance is rated highly satisfactory. Successful implementation of this highly complex and innovative project would have been an enormous challenge in the best of circumstances. The challenge was made much greater by the low level of 33 support and guidance from the central government and the initial lack of start-up counterpart financing. In close partnership with the CLG and the Bank/DFID team, the provincial and county PMOs weathered these significant obstacles and difficulties during the first two years of project implementation and step-by-step resolved these problems. By the latter years of the project, the strong resolve and determination of the provincial, county and lower level project management teams had achieved a highly satisfactory pace and quality of project implementation. With only one exception, compliance with the legal covenants was excellent. Although compliance with the legal covenant requiring that expenditure on the Sustainable Mountain Agriculture not exceed half of total cumulative project expenditure in each province was not complied with in Sichuan and Guangxi for several years, corrective measures were undertaken and this problem was completely resolved by 2010. The project‟s M&E arrangements were extraordinarily strong, and the results generated were effectively used in improving the quality of implementation and in project evaluation. Many of the lessons learned and the experience gained from project design and implementation have been extended throughout the provincial PADO system and beyond (see paragraphs 3.5.7 – 3.5.9 and Annex 2 paragraph A13). (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory 5.2.3 The overall performance of the Borrower is considered to have been satisfactory. Fortunately, the highly satisfactory performance of the implementing agencies was sufficient to correct for the shortcomings of the central government‟s only moderately satisfactory performance. 6. Lessons Learned 6.1.1 The following important lessons can be drawn from the project:  Participation. Substantial levels of participation can be achieved on a large scale in government-led projects if sufficient attention is paid to ongoing capacity building and learning. Participation requires a process-oriented approach with a long-term view on learning, and this was achieved in the project through a flexible approach, focusing on a deep understanding of basic participation principles and ongoing learning from practice. Documentation of practice, process and outcomes is critical for scale-up, and the project‟s participation approach was a success and became influential because it has paid attention to the documentation of practice (Good Practice Notes), process (M&E Module 3) and outcomes (M&E Module 2) from the beginning.  Government support. For a complex, multisectoral, and highly innovative operation, strong support from the central government is essential. Unfortunately, the low level of support and guidance from the central government adversely affected the initial pace and quality of implementation of the project. Although the adverse impact of this low level of support and guidance from the central government was eventually overcome through the strong resolve and determination of the provincial, county and lower level project management teams, the project‟s overall rating would likely have been highly satisfactory had LGOP been allowed to play a greater role in the project.  Intensive supervision. Intensive and thorough supervision can play a vital role in the success of a complex, multisectoral, and highly innovative operation such as this project. The Bank/DFID supervision missions very actively identified problems with project implementation quality and explicitly highlighted these problems in the mission 34 Aide Memoires, and this intensive and comprehensive supervision process was an essential element of the eventual success of the project.  DFID partnership. DFID provided very generous grant cofinancing assistance of about US$32 million to the overall project, and the project would not have happened without DFID‟s grant assistance. In addition to the main cofinancing grant, two separate DFID grants amounting to about US$5 million provided vital assistance for the development of the project‟s participatory approach, the rigorous and comprehensive M&E system, the Free Education Pilot, and the extraordinary assistance for people with disabilities activity. These three separate DFID grants massively contributed to the depth and innovative nature of the project, and the project would not have been successful without this support.  Company plus household model. The Sustainable Mountain Agriculture component had mixed experience with the use of the company plus household model as a poverty reduction mechanism. Successful use of the company plus household model requires (a) stable and profitable companies in poor areas with secure markets for products, (b) product raw materials and inputs that can be produced profitably by poor farmers, (c) fair and equitable working relationships and contracts between companies and households, and (d) provision of quality inputs and technical support to farmers.  Sequencing of activities. In a multisectoral project, early completion of infrastructure works is very important to obtaining the full benefit from related investment activities that rely on such infrastructure. Earlier completion of project village roads would have facilitated project work on agriculture, education, health and community capacity building activities and would have improved the effectiveness of those activities.  Mainstreaming assistance for people with disabilities. The Assistance for People with Disabilities (PwD) component in Yuexi County in Sichuan was successful, but there was consensus between the Sichuan provincial PMO, the Yuexi County PMO and Disabled People Federation, and Handicap International that the component should have been more fully mainstreamed into the overall project. Disability is a cross-cutting issue and activities to assist PwD should be fully mainstreamed as an integral part of poverty reduction and development strategies.  Importance of social services. Implementation of the Basic Education, Basic Health, and Community Capacity Building components was satisfactory and the components made important contributions to the project‟s overall favorable impact on villagers‟ well- being. However, inclusion of these three components added considerable complexity to the project design and the challenge of successful implementation. The three provincial PMOs and the county PMOs all confirmed that these three social services components were (a) certainly worth the additional complexity, and (b) particularly important to the successful outreach to the poorest and most remote project villages, ethnic minority communities, women and other vulnerable groups. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies The Implementing Agencies did not provide comments on this ICR. Their overall assessment of the project is reported in Annex 7. (b) Cofinanciers DFID‟s overall assessment of the project is reported in Annex 8. (c) Other partners and stakeholders See Annex 8. 35 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Sustainable Mountain Agriculture 62.88 71.98 114% Basic Rural Infrastructure 50.38 56.12 111% Basic Education 8.58 11.81 138% Basic Health 6.68 8.01 120% Community Capacity Building 7.00 8.19 117% Project Mngt & Monitoring 6.58 7.74 118% Total Baseline Cost 142.10 163.86 115% Physical Contingencies 0.00 0.00 Price Contingencies 0.00 0.00 Total Project Costs 142.10 163.86 115% Front-end fee IBRD 0.50 0.50 100% Total Financing Required 142.60 164.36 115% Sichuan Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Sustainable Mountain Agriculture 25.1 29.3 117% Basic Rural Infrastructure 15.8 21.3 135% Basic Education 2.2 3.0 136% Basic Health 2.3 2.8 122% Community Capacity Building 2.8 3.5 125% Project Mngt & Monitoring 1.6 1.7 106% Total Project Costs 49.8 61.6 124% Front-end fee IBRD 0.2 0.2 100% Total Financing Required 50.0 61.8 124% 36 Guangxi Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Sustainable Mountain Agriculture 21.6 22.3 103% Basic Rural Infrastructure 12.9 16.1 125% Basic Education 1.8 2.6 144% Basic Health 2.0 2.4 120% Community Capacity Building 2.5 1.8 72% Project Mngt & Monitoring 1.6 2.3 144% Total Project Costs 42.5 47.4 112% Front-end fee IBRD 0.1 0.1 100% Total Financing Required 42.6 47.5 112% Yunnan Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Sustainable Mountain Agriculture 16.2 20.4 126% Basic Rural Infrastructure 21.7 18.7 86% Basic Education 4.5 6.3 140% Basic Health 2.3 2.8 122% Community Capacity Building 1.7 2.9 171% Project Mngt & Monitoring 3.5 3.8 109% Total Project Costs 49.8 54.9 110% Front-end fee IBRD 0.2 0.2 100% Total Financing Required 50.0 55.1 110% (b) Financing Appraisal Actual/Latest Type of Estimate Estimate Percentage of Source of Funds Cofinancing (USD (USD Appraisal millions) millions) Counterpart Borrower 42.60 64.64 152% Funds International Bank for Reconstruction Loan 100.00 99.72 100% and Development DFID Trust Funds Grants 32.45 37.00 114% 37 Annex 2. Outputs by Component A. Participatory Approach A.1 Overall assessment. The project‟s PDO includes “sustained participation of the poorest rural people in project design, implementation, and monitoring and evaluation in eighteen counties of the three project provinces.� Based on the evidence collected through supervision (M&E Module 3) and participatory surveys (M&E Module 2) during implementation, the participatory approach is rated highly satisfactory (see paragraph 3.2.2). A.2 Background. Implementation of the participatory approach was a major challenge for several reasons, including the sheer size of the project area, the (cultural and environmental) diversity of locations covered, the complexity of project design (involving several components), the number of stakeholders involved, and the limited experiences with participation at local levels prior to the start of the project. The project has addressed these challenges through a progressive approach that involved the gradual roll-out of participation and ongoing learning and improvement. M&E was an integral part of this learning approach. The project has implemented a rigorous system for monitoring progress on participation and outreach to the poorest. A.3 Characteristics of the participation approach. The project‟s participatory process was characterised by the following elements: 1. It has built on previous experiences with participation in the three provinces, which have informed the project‟s Participation Manual; 2. Within a “conventional� approach to poverty reduction, it has bridged the gap between community priorities and top-down planning; 3. The project set out to achieve high levels of participation beyond the planning phase, thus extending the understanding of participation to the implementation and M&E phases; 4. It adopted participatory poverty analysis as a method to improve targeting of the poorest beyond the level of administrative villages; 5. The project encouraged diverse methods, which were adapted to the culturally and socially diverse environment; and, 6. The project supported a strong focus on learning from practice including, for example, through documentation of good practice notes and regular workshops where facilitators could share their experiences. A.4 Participation during project planning. The project‟s rigid planning framework limited scope for participation during the planning phase (see paragraphs 2.1.6 and 2.2.6). The project was budgeted with clearly defined sector allocations, which did not permit an open planning process. Project management was required to cascade down these sector allocations to the county and village levels in order to ensure that funding for the agricultural component (which comprised roughly half of all project funding) was fully utilized. Community demand for infrastructure was generally higher than for agricultural loans, but communities had to adjust their priorities to the sectoral availability of funding. The scope of participatory planning was further constrained by the timing of the provincial technical feasibility studies which had to be finalized before project approval. However, at that time, participatory planning was still underway in Guangxi and Yunnan, so that the set of technical options available within the Sustainable Mountain Agriculture component were somewhat limited. During implementation, the provinces moved to a more flexible annual planning process and undertook further adjustments during the Mid Term Review, and this helped lead to a more responsive project plan. 38 A.5 Through the participatory planning process, the project activities were generally aligned with community priorities. M&E Module 2 findings show that the project activities were implemented largely according to what the original Participatory Plan (PP) proposed. Where the survey found divergence from the original plan, this was often because certain activities (such as drinking water schemes) were in the end funded from other sources or added later. Demand for infrastructure, and in particular for drinking water, was generally higher than what the project could provide. However, communities largely felt that the project has addressed their needs and that all activities implemented were “much needed.� A.6 The Participatory Poverty Analysis (PPA) supported effective targeting of the poorest natural villages and households. Based on the poverty criteria and groupings developed through a participatory process, PPA helped identify the poorest natural villages and households. The PPA process led to greater transparency and facilitated consensus within the community on who should benefit on a priority basis. During the final workshop in Kunming, participants agreed that PPA was among those practices that should be disseminated for more effective targeting of the poorest. A.7 Levels of participation. The project achieved substantial levels of participation during all phases of the project cycle including planning, implementation and monitoring. M&E Module 2 systematically assessed levels of participation in each project phase. The findings show that participation was generally very high during planning, and this is consistent with the fact that participatory plans were a requirement set by the project. Participation levels were also high during implementation and follow-up maintenance. However, participation during the M&E phase was low, particularly in Sichuan and Yunnan. In Guangxi, the allocation of clear roles for M&E in the community enabled higher participation levels. A.8 Levels of participation were generally lower in project activities implemented by line agencies. Being multisectoral, the project had to cope with the challenge of aligning multiple stakeholders behind the participation approach. Capacity building and learning on participation at the provincial level was mostly limited to the PADO and Development Reform Commission staff, and awareness among other departments‟ staff was insufficient. This has affected levels of participation during the implementation stage, which were lower in the Basic Education and Basic Health components and in irrigation works. M&E Module 2 reported that participation in education and health activities was low, and communities also complained about the lack of transparency in larger infrastructure works implemented by external contractors. A.9 Integration into project components. The participation approach has been systematically integrated into all project components. Levels of participation have been particularly high in small infrastructure, which was implemented mainly using community procurement methods. For example, Guangxi used community procurement for 70% of all procurement and achieved high levels of participation in the Basic Rural Infrastructure component as a result. Even where other procurement methods were used, the communities were systematically involved in the decisions on design and procurement, supervision of construction, and final acceptance (yanshou). Participation was less integrated into the Sustainable Mountain Agricultural component. M&E Module 2 reported a lack of transparency in the allocation of loans and insufficient targeting of poor households during the first two years. Transparency improved with the better documentation of loans at community level during the project‟s final three years of implementation. It is also believed that the project missed opportunities to build in options for a community-driven approach in the Sustainable Mountain Agriculture component. In particular, farm production loans were operated in a rather conventional way (through RCCs or through the PMOs). The M&E Module 2 reports have highlighted the fact that communities, and 39 in particular women, were least satisfied with the Sustainable Mountain Agriculture component and complained about the lack of transparency in the allocation of loan funding during the first two years. In a few places (such as Yuexi County), the project overcame these issues and developed (“home-grown�) innovative practices, like the operation of rotating funds and group loans which became successful and reportedly experienced high repayment rates. A.10 Rollout of the participation approach. The slow start-up of the project undermined the motivation of communities and project staff alike and undermined participation during the first two years of implementation. However, with the acceleration of project implementation beginning in 2008, participation quickly gained momentum. The project rolled out the participatory approach over the entire project area, reaching a high number of remote natural villages. With settlements scattered over a larger area, Guangxi and Yunnan were very successful in their outreach to remote natural villages. For example, in Guangxi, 88% of the remote natural villages have been directly involved in project decision making by the end of the project, and 399 project implementation groups were set up in the poorest natural villages. Yunnan set up 893 project implementation groups in the poorest natural villages. Project implementation groups, appointed by the villagers affected, were in charge of managing and/or supervising specific project activities. Bringing projects closer to the people was a major step towards greater motivation and ownership. A.11 Community capacities. Evidence from M&E Module 2 suggests that communities were satisfied with the levels of transparency achieved and feel more empowered to take charge of their project activities. The project was better than other government-led activities in engaging communities in the selection of project activities, management and supervision of infrastructure projects, and disclosure of financial information. As a result, communities feel more responsible for the management and maintenance of project facilities. Feedback from county project staff (collected through the M&E Module 2 surveys) indicates that follow-up maintenance in the project activities is generally better than in other village level development activities. However, M&E Module 2 also found that, although all villages surveyed have discussed and agreed on rules for maintenance, they often fail to collect fees to cover unexpected expenses (in Guangxi in particular). A.12 The participation approach has built community capacities and social capital for sustainable poverty reduction. M&E Module 2 findings show that, through the participation approach, communities have learned skills to manage and supervise projects. Women seem to have benefitted from capacity building even more than men; a majority of them reported that they have improved their skills in agricultural production and marketing as a result of agricultural training. They also feel that they improved their financial management skills as a result of the farm production loans. The men reported that they have become more confident in communicating with village leaders and organizations outside the village. Respondents also reported that they communicate more and better within the village, and that their ability to deal with conflicts has improved. The M&E Module 2 studies found that famers‟ organizations have been set up in most villages in Sichuan and Yunnan, where cash crops were supported by the project. The Guangxi and Sichuan studies also reported that PRCDP villages have been able to attract substantial follow-up investments by other agencies because they already had a plan and the capacity to implement investment activities. A.13 Institutional learning and change. The commitment of local project staff was a key factor for the successful implementation of the participation approach, and it also facilitated lateral changes within and beyond the project. Learning on participation has transformed staff attitudes and behaviour beyond the application of participatory tools and methods. The 2011 40 M&E Module 2 survey shows that PMO staff have achieved a broader and more inclusive understanding of participation as a process to engage with poor people as key stakeholders, not just as beneficiaries. All staff interviewed agreed that the poor must participate in decision making about project activities. PMO staff found the use of tools to support inclusive discussion and decision-making processes particularly useful. Staff also reported significant changes in their views on gender. They also reported that they substantially changed the way in which they address gender as a result of the project, in particular with regard to motivating and targeting women and considering gender-specific needs. M&E Module 2 found wide-ranging evidence that staff are increasingly applying participatory methods outside of the project, and that the project‟s experience with participation is starting to influence other government departments. Evidence from all the three provinces collected by M&E Module 2 also shows that participatory practices from PRCDP are spreading within the provincial PADO systems. There is also evidence from Guangxi that project practices have been adopted in other agencies, in particular county water bureaus and reform commissions. B. Sustainable Mountain Agriculture Component B.1 Component outputs. The Sustainable Mountain Agriculture (SMA) component aimed to support agricultural development and sustainable livelihoods in project villages through (a) improving household productivity and diversification, (b) developing farmer-to-farmer services, and (c) strengthening agricultural training and technical support services. The provincial Progress Reports show that a total of 303,454 households participated in activities under the component resulting in an average coverage rate of 87%. Of these, 186,667 households participated in Sichuan, 55,354 in Yunnan and 61,433 in Guangxi. Household coverage targets for component activities were achieved in Sichuan, but actual coverage fell short of target in Yunnan (83% of target number of households achieved) and Guangxi (79%) (Provincial Progress Reports, March 2011). This shortfall is partly due to increases in unit costs for production inputs during the project period resulting in fewer households being covered by the investment activities. Project financial support for agricultural investments was made to 189,100 households (62% of total households participating in the component). These households comprised 105,884 households in Sichuan (56% of households participating in the component), 42,406 households (77%) in Yunnan and 40,810 households (66%) in Guangxi. Of these, 41,908 households (40%) in Sichuan received two investments while some households in Yunnan and Guangxi also received two investments. B.2 Agricultural investments included grain production, cash crops, tree crops, livestock, and small scale value added agricultural processing. The number of households investing in livestock production totaled 130,434 of which 95,242 households were in Sichuan, 11,166 households in Yunnan, and 24,026 households in Guangxi. The average agriculture investment per household ranged from RMB1,565 in Sichuan, RMB2,400 in Yunnan to RMB2,933 in Guangxi. Farmer training in agricultural and livestock practices comprised 245,008 person times in Sichuan, 94,248 person times in Yunnan, 44,724 household times in Guangxi. Each participating project household has received technical training at least once. Women comprised 34-52% of trainees across the three provinces. B.3 Contributions to Outcomes. Activities under the Sustainable Mountain Agriculture component have contributed to achievement of the project‟s outcomes through improved livelihoods security. The NSB survey reports (M&E Module 1) and the provincial Progress Reports show that the component contributed to reductions in the incidence of poverty, improved food security, and increased incomes. The incidence of poverty in project villages has been significantly reduced in all three provinces. The average incidence of poverty in project villages 41 decreased from 15.9% in 2007 to 9.3% in 2010. This 6.6% decrease in poverty incidence is faster than the decrease in poverty incidence over the same period in national key poverty counties where it has decreased from 13.0% in 2007 to 8.3% 2010, a decrease of 4.7% (NSB 2011). Poverty incidence in project villages in Sichuan has decreased from 14.5% in 2007 to 10.7% in 2010; in Guangxi from 10.6% in 2007 to 7.6% in 2010; and in Yunnan from 23.1% in 2007 to 9.2% in 2010. Food security in project villages has been improved with significant reductions in the percentage of households with less than 150kg grain/capita/year in Sichuan (from 14.0% in 2002 to 4.8% in 2010) and Yunnan (from 13.4% in 2001 to 2.5% in 2010). (There is no data available for Guangxi.) B.4 All three provinces show significant increases in net income per capita in project villages during the project period (Table 1). Average net income per capita in project villages increased from RMB1,752 in 2006 to RMB3,215 in 2010, an increase of RMB1,463 or 83.5%. The actual net income growth after correcting for pricing factors and inflation is 57.4%. The annual rate at which net income per capita (deducting pricing factors) has increased in project villages from 2006 to 2010 (12% per annum) is faster than the annual rate of increase in key poverty counties (10%) and the national average (9.2%). In 2010, the net income per capita in project villages was 54.3% of the national average compared to 48.8% in 2006, an increase of 5.5%. Table 1. Annual Net Income (RMB) Per Capita. (Source: NSB M&E Module 1) Year National Key PRCDP Project Villages Average Poverty Average Guangxi Sichuan Yunnan Counties Average 2006 3,587 1,928 1,752.0 1,983.2 1,846.9 1,414.4 2007 4,140 2,278 2,153.2 2,331,2 2,353.7 1,757.8 2008 4,761 2,611 2,556.2 2,813.1 2,642.6 2,196.3 2009 2010 5,943 3,215.0 3,217.0 3,408.0 3.022.0 B.5 In the Guangxi project villages, net income per capita increased 38.7% from 2006 to 2010, an average annual growth of 8.6%. In Sichuan, net income per capita has increased 58.5% over the project period at an average annual rate of 12.2%. These annual rates of net income/capita increase are similar to the provincial averages in Guangxi and Sichuan. In the Yunnan project villages, net income per capita increased by 83.3% from 2006 to 2010 with an average annual increase of 16.4%. This is 11.2% higher than the provincial average for Yunnan. B.6 Sources of Income Changes in Poor Households. At the beginning of the project, households in project villages were classified into low, medium and high relative income categories. Project monitoring of project households in 2006 and 2010 by NSB included income and its sources for project households in each category within each province. Monitoring results for Guangxi, Sichuan and Yunnan are given in Tables 2 to 4. Growth of total income in low income households from 2006 to 2010 is higher than that of medium and high income groups in Guangxi and Yunnan, but slightly slower in Sichuan. Income of the poorest households (that is, the low income group) grew 71% in Guangxi, 87% in Sichuan, and 129% in Yunnan from 2006 to 2010. Table 2. Income Sources in Guangxi Project Households. (Source: NSB M&E Module 1) 42 Item 2006 (RMB) 2010 (RMB) % Change 2006 - 2010 Low Medium High Low Medium High Low Medium High Total income 1,360.4 2,212.7 4,444.4 2,325.3 3,944.3 6,780.3 70.9 78.3 52.6 1. Wages 201.9 499.6 927.1 611.3 1,104.7 1,915.6 202.7 103.1 106.6 2. Operational 1,090.3 1,630.2 3,343.6 1,618.8 2,697.4 4,171.8 48.5 65.5 24.8 Cropping 569.9 779.2 1,666.0 872.9 1,284.7 1,593.2 53.2 64.9 -4.4 Livestock 366.1 518.2 976.2 456.8 860.1 1,399.0 24.8 66.0 43.3 Net Income 934.5 1,765.8 3,482.2 1,498.0 3,056.5 5,649.7 60.3 73.1 62.2 % income from 68.8% 58.6% 59.5% 57.2% 54.4% 44.1% cropping/livestock Table 3. Income Sources in Sichuan Project Households. (Source: NSB M&E Module 1) Item 2006 (RMB) 2010 (RMB) % Change 2006 - 2010 Low Medium High Low Medium High Low Medium High Total income 1,439.2 2,116.2 4,066.9 2,687.3 4,426.9 7,781.6 86.7 109.2 91.3 1. Wages 148.6 265.1 879.9 357.3 764.4 1,571.7 140.4 188.3 78.6 2. Operational 1,191.5 1,690.6 2,719.7 1,990.7 3,289.7 5,504.5 67.1 94.6 102.4 Cropping 652.5 836.6 1,313.1 1,071.5 1,701.5 2,467.1 64.2 103.4 87.9 Livestock 506.0 693.8 1,137.6 810.7 1,212.8 1,908.2 60.2 74.8 67.7 Net Income 921.4 1,624.6 3,307.9 1,717.4 3,205.6 6,171.1 86.4 97.3 89.6 % income from 80.5% 72.3% 60.3% 70.0% 65.8% 56.2% cropping/livestock Table 4. Income Sources in Yunnan Project Households. (Source: NSB M&E Module 1) Item 2006 (RMB) 2010 (RMB) % Change 2006 - 2010 Low Medium High Low Medium High Low Medium High Total income 1,174.4 1,854.4 3,605.7 2,691.5 3,884.9 7,000.6 129.2 109.5 94.2 1. Wages 34.9 137.1 278.0 263.4 572.2 1,118.8 655.0 317.3 302.4 2. Operational 1,046.6 1,583.4 3,134.4 2,097.0 2,893.1 5,291.0 100.4 82.7 68.8 Cropping 730.3 1,094.2 2,118.8 1,531.7 1,842.2 3,291.7 109.7 68.4 55.4 Livestock 246.5 334.8 569.8 365.7 723.6 1,046.4 48.4 116.1 83.7 Net Income 559.1 1,180.9 2,583.4 1,676.2 2,642.6 4,902.2 199.8 123.8 89.8 % income from 83.2% 77.1% 74.6% 70.5% 66.0% 62.0% cropping/livestock B.7 Wages (off-farm work) has generally been the most rapidly growing contributor to income growth for all categories of households in each province (except for the high or “better off� category of households in Sichuan). The growth rate for wages (off-farm work) is generally fastest for the low and medium household categories, but especially for the low income (poorest) households in Guangxi and Yunnan. Growth rates in wages (off-farm work) are much higher in Yunnan compared to Guangxi and Sichuan, but they started from a much lower base for all categories of households in Yunnan in 2006. B.8 Operational activities in Tables 2 to 4 are largely comprised of cropping and livestock production. All provinces and all household categories show significant growth rates in income 43 from operational activities, as well as cropping and livestock activities, except for the high income group in Guangxi (Table 2). Income growth from both cropping and livestock in Guangxi and Sichuan was fastest in the medium income group. In Yunnan, income growth from cropping was fastest in the low income group while that from livestock was fastest in the medium income group. It is difficult to disaggregate income growth in cropping and livestock that resulted from the project as compared to any other (government, NGO) programs that might have been carried out in the project villages sampled by NSB. However, discussions with provincial PMO staff indicate that few project villages had any other agriculture programs other than those of the project. As a result, it can be reasonably concluded that PRCDP was a significant contributor to the income growth from cropping and livestock as shown in Tables 2 to 4. B.9 In 2006, income from cropping and livestock was by far the major source of household total income in all categories of project households in all three provinces, exceeding 80% in Sichuan and Yunnan. The relative contribution of cropping and livestock to household income decreased in all household groups in each province in 2010. However, it continued to contribute more than 50% in all households in 2010 except for the high income group in Guangxi. In 2010, cropping and livestock in Sichuan and Yunnan contributed more than 70% of total income in the low income households and more than 65% in the medium group in 2010. This emphasizes the continued importance of agriculture to the household economy in the poorer household groups despite the rapid growth in wage income in all groups and provinces. Success Factors B.10 Access to finance. Benefits from the component have arisen from poor farmers gaining access to loan funds for investment in agricultural activities for food security or income generation. This included new agricultural activities or expanding areas of existing ones; improved access to new varieties, better quality seeds, seedlings and livestock breeds; and improved training and technical extension. Many poor farmers did not have access to finance for investment in agriculture development prior to the project. Through the project many farmers were able to gain experience in borrowing and repaying loans for agricultural investments, building a credit track record through repayment of due loans on time, thus enabling them to take a second loan from the project. In the longer term, this track record will also make them potentially more attractive to lending organizations following the project. B.11 Access to technical training. Strong technical support through farmer training was provided by line bureaus in support of agriculture investments. This enabled farmers to improve their crop and livestock management and increase yields and product quality, which in turn contributed to improved food security and increased income generation from cash crops and livestock production. B.12 Flexibility in farmer selection of investment activities. An important success factor was that farmers were generally able to select their own agricultural investment activities from a long list of project-approved activities. Additional flexibility was provided through an unearmarked funds subcomponent in Guangxi that enabled farmers to respond quickly to market and other changes in selecting investment activities. Such flexibility is particularly important where there is a long project design period or where farmers may not receive investment funds until later in the project. The unearmarked funds subcomponent comprised 32% of the component investment in Guangxi and was very well received by farmers. Future poverty reduction projects involving agricultural investment activities should include a high proportion of 44 unearmarked funds to allow flexibility and farmer response to the market and other changes. B.13 Integration with rural infrastructure. Agriculture development in project areas benefitted greatly from improved road access, irrigation infrastructure, and other rural infrastructure investments under the project. The advent of roads to natural villages has resulted in a change in the agriculture structure in many project areas from largely subsistence agriculture (food crops) to cash crop and livestock production for income generation. An example is in Pannei Village, Longsheng County, Guangxi, where farmers have changed from largely food crop production to lohan fruit production because village road investment has improved access to buyers, improved production inputs (seedlings, varieties, fertilizer), and market information. Investment in irrigation and water storage infrastructure enabled farmers in many villages to change from dryland grain crops to paddy rice, and/or to higher value cash crop production because of increased security of irrigation water supply. This in turn resulted in improved grain production for human or animal consumption and income generation from production and sale of cash crops. Similarly, labor savings and improved health from improved drinking water supply enabled farmers, especially women, to invest more time in productive activities. Shortcomings B.14 Lack of start-up funds. A major weakness of the component was the lack of startup counterpart funds which resulted in minimal implementation progress during the first 12 months of the project. All three provincial PMOs developed innovative ways of obtaining and channeling funds to implement the component, including use of Rural Credit Cooperatives (RCC), township Finance Bureaus, line bureau (Agriculture, Livestock, Forestry) funds, bridge funds through company plus household contracts, as well as significant farmer funds contribution. The various funding approaches eventually enabled the component to be implemented satisfactorily. However, this required a large input of extra time and effort from PMOs at all levels and made implementation of the component much more complex than it needed to be (often with more than one funding channel being used in a county). Use of some funding channels also constrained farmers‟ choice of investment activities in some cases. For example, if county agriculture bureau funds were used, then farmers were often constrained to investing in activities that complied with the county agriculture development plan. Similarly, where company plus household models were used, farmers were often only able to invest in activities and locations supported by the company. As a result, use of different channels impacted on the range of investment choice of many farmers and in many cases was not always consistent with the project‟s participatory approach to activity selection. B.15 Company-Household contracts. The “company + household� approach was applied in many counties, particularly for crops such as tea, sugarcane, and tree crops, with many different types of company + household models used. The three provinces reported different experiences with the company + household model as a poverty reduction mechanism in poor rural areas. Sichuan reported a generally positive experience, while Guangxi reported a mixed experience and Yunnan a negative experience. Benefits of the company + household model in Sichuan included access to bridge funding to initiate the component in the absence of startup funding; availability of quality inputs through the company; farmer training and technical support by specialist company technicians; a guaranteed buyer for the agricultural output produced; and improved farmer access to market information and the market chain. The company + household approach in Guangxi depended on finding profitable and stable companies with good products that were able to develop equitable working relationships with poor households. However, few such 45 companies were present in poor areas in the Guangxi project counties and this lending channel was not a major part of the project in Guangxi. In Yunnan, the company + household approach was not successful because of (a) deficiencies in contracts between PMOs and companies, and in company-household contracts, (b) variations in input and product prices compared to those in farmer contracts, and (c) collapse of companies. B.16 In some counties the companies acted as intermediaries of the project funds rather than project funds going direct to the village group/household. In such cases, the companies supplied farmers with inputs and technical assistance (rather than cash) and usually required farmers to sell their products back to the company assuring them a fair price at harvest. While this gave farmers a potential buyer for their product, it also restricted farmer options in what they could produce and in sale of product. B.17. Limited range of agricultural investments in some counties. The actual range of agricultural investments was quite limited in some counties despite the project approach of promoting diversification and farmer choice in activity selection. For example, in Menglian County, Yunnan the only productive agricultural investments supported by the project have been tea and pigs. In some villages all beneficiaries only received one of these two types of investment. The reasons for this limited range of investments may have been related to funding channel, limited farmer knowledge of alternatives, proven local buyers for these commodity products, and limited local market for alternatives. This limited range of investment options increases farmer vulnerability to external shocks that increases risks to price and income fluctuations, natural disasters, pests and disease, food supply and poverty levels. In the longer term, a more diversified range of investment options is necessary to reduce vulnerability of farm households to these risks. Lessons Learned B.18 Start-up counterpart funds. A key lesson learned is that start-up government counterpart funds are essential to enable a smooth start to implementation of agriculture (and rural infrastructure) activities in poor remote ethnic minority villages. Absence of such funds resulted in slow initial progress of agriculture activities, complexity of funding sources to get activities started, additional complexity and time required for project management at all levels. It also inhibited targeting of the poorest villages and households, especially where commercial lending or company funds are used because of perceived risks involved with lending to the very poor. B.19 Loan repayment. The project provinces reported highly variable recovery rates of due loans. Some counties achieved 100% repayment of due loans while repayment in others has been very low. Reasons for the level of repayment varied, but an important lesson learned from the household loan activities is that to achieve good loan repayment there needs to be (a) good publicity and farmer awareness about loans and the need for repayment, (b) a loan channel (for example, an RCC) that emphasizes and follows up on repayment, (c) successful investments that achieve good financial return, (d) clear statement of loan obligations in farmer contracts, and (e) low incidence of natural disasters during the production/investment period. Repayment of loans provides the opportunity to revolve funds for further lending to extend coverage and increase impact from loan funds. B.20 Company + Household, farmer associations and cooperatives. PRCDP had mixed success with company + household financing channels as a poverty reduction mechanism. Lessons learned for success in applying this approach include the need for (a) stable and 46 profitable companies in poor areas with secure markets for products, (b) product raw materials and inputs that can be produced profitably by poor farmers, (c) fair and equitable working relationships and contracts between companies and households, and (d) provision of quality inputs and technical support to farmers. A promising alternative to the company + household approach is the development of village- and product-based farmer associations and shareholder cooperatives that have developed successfully in many project areas. These were not directly financed under the project but many project farmers receiving project agriculture investments have become members of such associations and cooperatives. Such associations and cooperatives are providing technical support and cheaper bulk purchased inputs, and assisting with product marketing and sales. It is likely that such associations and cooperatives will become increasingly important in China given the agricultural policy trend to „scale‟ and „industrialized‟ agriculture development in the future. Post-Completion Operations B.21 Ongoing farmer training. Project farmers will continue to need technical support and ongoing training from line bureaus following the closure of the project. Many farmers are still learning how best to manage new crops and livestock breeds to achieve optimal production levels. For example, many farmers that have received irrigation water for the first time under the project are still learning the transition from dryland farming practices to irrigated agriculture. As a result, they are not yet achieving the full production potential offered by irrigation investments under the project. They will continue to need ongoing training and technical support to gain full benefit from the irrigation investments. B.22 Revolving of repaid loan funds. The repayment of household loans provides an opportunity to revolve funds for re-lending to new households or making follow-up loans to project households that have repaid. Such revolving of loan funds is being implemented in some project provinces and counties, but not others. Revolving of repaid loan funds should be implemented in all provinces and counties following project closure to enable wider coverage and impact to be achieved from the Bank loan funds. C. Rural Infrastructure Component C.1 Component outputs. The rural infrastructure component aimed to improve living conditions, support development of livelihood activities, and improve access to social services in project areas through provision of basic rural infrastructure, including village roads and paths, water supply, water conservancy works, rural electrification, rural energy and sanitation. The provincial Progress Reports indicate that approximately 501,690 households benefitted from activities under the component. In Sichuan 121,081 households have directly benefitted from rural infrastructure investments, 241,714 households in Yunnan, and 138,895 households in Guangxi. C.2 Rural roads. The road infrastructure investments benefitted a total of 202,066 households. The project constructed 295 km of new off-grade standard roads between administrative villages and natural villages, which benefitted about 65,693 households. In Sichuan it also rehabilitated 18 km of village roads benefitting 879 households. The project also constructed 1,826 km of roads between natural villages benefiting approximately 134,694 households. This work included 56 linear meters of bridges in Yunnan serving 800 households. C.3 Drinking water supply. Overall, approximately 120,154 households benefitted from improved drinking water supply through project investments in water supply systems. Such 47 systems included stream intake structures, filters, wells, water storage tanks, pipe systems, and reticulation to households. C.4 Irrigation. The project constructed 121 water storage ponds for irrigation, 299 km of new irrigation canals, rehabilitated 374 km of existing irrigation canals, and installed 5,538 water storage tanks. These improved irrigation facilities benefitted approximately 95,412 households in the three provinces. C.5 Land improvement. Terracing and improvement of low and medium yield land was undertaken in Yunnan and Sichuan. In Yunnan, 6,100 mu (407 ha) of new terraces were built, and 13,450 mu (897 ha) of low and medium yield land was improved. Smaller amounts of land was improved in Sichuan with 140 mu (9.3 ha) of terracing and 235 mu (15.7 ha) low and medium yield land improvement undertaken. In total, 64,738 households benefitted from this activity. C.6 Rural energy. A total of 6,880 biogas systems were constructed in Yunnan and Sichuan providing energy for 8,880 households, while 2,202 households obtained fuel-efficient stoves in Sichuan. Electricity was supplied to some villages in Yunnan and Sichuan with the project installing 20 transformers and constructing 34 km of transmission lines benefitting about 10,40 households. C.7 Other infrastructure. Other infrastructure activities undertaken include construction of 122 new houses and rehabilitation of 399 houses in Guangxi; development of 4,420 mu (295 ha) of improved grassland in Yunnan that benefitted 489 households; and construction of 10,793 m2 of village markets in Guangxi that benefitted 39,996 households. C.8 Training. Developing the capacity of project communities to operate and maintain the infrastructure correctly was an important activity under the Community Capacity Building component. In Guangxi, 44,724 person-times of training in infrastructure operation and management was provided, comprising about one-third of project infrastructure beneficiaries in the province. C.9 Contributions to Outcomes. Activities under the component have contributed to the achievement of the project‟s outcomes by providing basic rural infrastructure necessary to underpin improved livelihoods security and meeting basic household needs in project villages. Road access to many natural villages has stimulated economic development in project villages, improved access to many different types of services (such as health, education, markets), reduced travel times, and resulted in a change in the agriculture structure from largely subsistence agriculture in many villages to cash crop and livestock production. Similarly, labor-saving and reduced sickness days due to improved water quality from water supply systems have resulted in households having more time to invest in productive and other activities. Guangxi reported that 91.4% of households now spend less than 30 minutes/day fetching drinking water. C.10 Water conservancy works (irrigation works, water storage) have provided security of irrigation water supply in many villages. This enabled farmers to increase their grain yields, change to paddy rice, or grow cash crops thus contributing to income generation. Land improvement through terracing or low and medium yield field improvement has increased the area of arable land for crop production in project villages. Improved rural energy supply through biogas and fuel efficient stoves has contributed to labor-saving (less fuel wood collection) and cleaner energy, while supply of electricity has provided opportunity for small scale processing activities. These infrastructure investments all directly underpin agriculture production and other 48 types of economic development, thus contributing indirectly to improved livelihoods security and income generation in project villages. Success Factors C.11 Focus on basic household needs. The activities under the component were selected by the communities themselves to address their most important basic infrastructure needs. These generally dealt with important constraints to household well-being and village development, such as poor water supply or lack of road access. Most of the infrastructure activities had an immediate impact on household well being (such as labor savings and improved health through water supply and rural energy) and livelihood development (such as market access due to village roads, improved productivity due to irrigation works). As a result, the infrastructure investments were well received and supported by most communities. C.12 Farmer participation in infrastructure planning and management. Field visits and farmer interviews during supervision missions and participatory monitoring surveys (M&E Module 2) indicate that households were actively involved in the selection and planning of infrastructure in most villages. This ensured that community-agreed needs were addressed by the infrastructure investments. The provincial PMOs reported that infrastructure management committees and management rules have been set up in many villages to manage ongoing operation and maintenance of public infrastructure (such as water supply, roads, and irrigation works). As a result of farmer participation in infrastructure selection, planning and management, there is strong villager ownership of the infrastructure works in most project villages. C.13 Farmer training in operation and maintenance. Most villagers recognize that the public infrastructure investments now form important community assets and good post- construction operation and maintenance is essential to maintaining the functionality and life of the infrastructure. In most villages, designated farmers have received training in basic operation and maintenance of the infrastructure and are responsible for its ongoing management. This enables basic maintenance to be done quickly and efficiently within the village, and gives villagers ownership of the ongoing maintenance function. Technical support is provided by the relevant line bureau for larger or complex maintenance works. C.14 Technical Quality. The technical quality of infrastructure works was generally satisfactory with design drawings, statements of quantities, and costs prepared in advance for most civil works. Checking and acceptance procedures were implemented to ensure that construction of civil works was carried out according to design. Village representatives were members of the checking and acceptance process to ensure that villager views were represented in the technical quality control process. This investment in quality control has been important to ensure that civil works have been constructed to the appropriate technical standards so that such works can achieve their design life with appropriate operation and maintenance. C.15 Interaction with other components. The Village Development Plan has been an important tool to integrate the planning of rural infrastructure with activities under the project‟s other components. This has enabled benefits from other components to be increased through integration with the rural infrastructure component. Examples include increased crop production and cash crop production resulting from irrigation and access to market/buyers through road improvements; improved household health resulting from improved water supply; and improved access for children to go to school resulting from improved village roads. 49 Shortcomings C.16 Delays in counterpart fund provision. Slow delivery of counterpart funds delayed the start and early progress of the component with many of the infrastructure activities only being completed in the last three years of the project. Ideally, rural infrastructure activities such as roads, water supply, irrigation and water storage works should be conducted as early as possible in rural development projects so that investments in other related components, such as agriculture, can leverage off the benefits brought by the infrastructure investments. Improved access to markets, input supplies, buyers, improved production conditions and additional labor time to invest in productive activities are all benefits resulting from the infrastructure activities that can then be leveraged into income generation through appropriate agriculture investments. However, under the project, the lack of start-up counterpart funding for the Sustainable Mountain Agriculture component and slow delivery of counterpart funds for the Rural Infrastructure component meant that counties had to access a range of funding sources to start the Sustainable Mountain Agriculture component, obtain reimbursement of these funds quickly, and then use reimbursed funds to start the rural infrastructure activities. As a result, many rural infrastructure activities followed agriculture activities when ideally the sequence should have been the other way around. For this reason, the full potential benefit from the infrastructure-agriculture combination of activities was probably not achieved in PRCDP because of this lack of startup counterpart funding for the two components. Lessons Learned C.17 Timely provision of counterpart funds. An important lesson learned is that counterpart funds for rural infrastructure investments should be made available at the very start of the project. Early completion of infrastructure works is essential to obtain full benefit from related investment activities that rely on the infrastructure, notably agriculture activities. The Village Development Plans provided for integrated village development, but the long time frames required for infrastructure design and slow delivery of counterpart funds meant that much of the infrastructure was delivered late in the project when other complementary investments (especially agriculture) were largely completed. C.18 Technical quality control. Technical design and construction of civil works to the correct technical standards supported by a rigorous checking and acceptance process is essential for good quality construction of works. Involvement of villager representatives as part of the design, and checking and acceptance teams is important to ensure that villager views are fully represented at all stages. Such involvement helps to ensure that the community‟s needs are fully met and that they have ownership of the built infrastructure. Post-Completion Operations C.19 Post-project operation and maintenance. The rural infrastructure investments now form important community assets and good post-construction operation and maintenance is essential to maintaining the functionality and life of the infrastructure. Many villages have organized themselves well to undertake infrastructure management and maintenance, but others still need help and ongoing technical assistance from the relevant line bureaus. For example, the water supply schemes are new technology in many villages and the villagers are still learning to manage them correctly. Such villages will need to continue receiving support from the Water Resources Bureaus over the next several years, especially for system maintenance. To ensure sound ongoing operation and maintenance of infrastructure, county project management offices and the relevant township line bureaus should ensure that (a) all villages put in place suitable 50 management and maintenance systems, (b) further infrastructure operation and maintenance training is provided to villagers where required, and (c) ongoing line bureau support is provided when required in the post-project period. 51 Annex 3. Economic and Financial Analysis 3.1 Overall evaluation. As confirmed by the results of the M&E system Modules 1 and 2 reports (from the NSB‟s annual quantitative monitoring and evaluation as well as the ITAD Company‟s annual qualitative Participatory Poverty Monitoring surveys), the project generated remarkable economic and social benefits. First, the project has successfully reduced the incidence of poverty in the project area. From 2006 to 2010, the incidence of poverty declined from 15.9% to 9.3% in the project area, registering a total reduction of 6.6 percentage points. This reduction outperformed the national average which was reduced by just 1.8 percentage points during the same period. Second, net per capita income in the project area increased 57.4% in real terms from 2006 to 2010 (reaching RMB 3215). This is equivalent to an annual growth rate of 12%, which again outperformed the national average increase of 10%. Consistent with this increase in income, real per capita consumption increased 39.8% during 2006 to 2010. Third, infrastructure conditions in the project area have been considerably improved. By the end of 2010, 98.8% natural villages had access to electricity, 83.9% to road, and 98.2% to telephone services. Fourth, farmers‟ participation in local governance and public affairs has improved. For example, in 2006, only about 19% of farmers reported that they had played an important role in selecting and determining village development activities; however, the percentage had increased to 46% in 2010. Project targeting has also been improved due to farmers‟ active participation. The NSB data showed that 50% of community members believed that the poorest households in their communities received favorable and timely assistance from the project. In 2010, close to 90% of farmers reported that they were aware of the project to a certain extent and, of these, almost 53% said they had very good knowledge about the project. Compared to the Southwest and Qinba Mountains Poverty Reduction projects, the considerably improved farmers‟ participation in this project ensured a higher degree consistency between the project contents and the farmers‟ preferences. Table 3.1 Selected Indicators of the Project‟s Economic and Social Benefits 1 2 3 4 5 6 7 8 Year Poverty Per capita HH with Villages Labor Per capita Villages Participa income running w/ migration with clinic tion rate Incidence (yuan) access consumption water (%) (%) (%) (%) (%) roads (yuan) (%) 2006 15.9 1752 32 79 17 1524 88 20 2008 15.8 2556 38 85 19 2171 92 34 2010 9.3 3215 46 84 24 2484 97 46 3.2 Rate of return analysis. The overall economic rate of return (ERR) to the project, under a few conservative assumptions, is estimated to be around 30%, which is the same as that estimated in the Project Appraisal Document. A traditional methodology was used to estimate the ERR for the project. For different project activities, the estimated future stream of costs and benefits were used to derive net benefits and to calculate the net present values (NPV) and ERRs. The individual components‟ ERRs are weighted by their corresponding investment proportion of the total project cost. Data for calculating the ERRs were collected by the provincial PMOs specifically for the purpose of the ICR (see Annex 9, Document No. 8). Information from the NSB‟s annual M&E Module 1 surveys as well as some secondary research on China‟s rural economy were also taken advantage of. The analysis quantified the benefits of three major project components: (a) Sustainable Mountain Agriculture, (b) Basic Rural Infrastructure, and (c) Basic Education. Investment in these three components accounts for about 85% of total project investment. The benefits of the other three project components (that is, Basic Health, Community Capacity Building, and Project Management and Monitoring) are difficult to quantify and hence 52 ERRs were not calculated for them. Given that the analysis‟ estimates covers 85% of total project investment, the analysis provide a reasonable estimate for the overall project ERR. Table 3.2 presents estimated ERRs for the whole project as well as for the three specific components: Table 3.2 Rates of return to the project and different components Project as Sustainable Basic Rural Basic a whole Mountain Infrastructure Education Agriculture ICR 30% 31% 29% 22% PAD 30% 35% 27% 19% 3.3 Sustainable Mountain Agriculture component. Following the conventional methodology, the ERRs for the three sub-components of the Sustainable Mountain Agriculture component were calculated: livestock sub-component (including goat, pig, silkworm, ox, and chicken), cash-crop and tree crop sub-component (including vegetables, sugarcane, and tea), and the grain sub-component (including maize, rice, and potato). The overall ERR for the Sustainable Mountain Agriculture component is about 31%, ranging from 23% (rice) to 43% (chicken). 3.4 Rural Infrastructure component. The Rural infrastructure component was most welcomed by the project beneficiaries. In terms of the degree of warmth of reception by the project villagers, the NSB‟s annual M&E Module 1 surveys showed that the Basic Rural Infrastructure component was ranked second only to the agriculture component. There are many different rural infrastructure activities, and the analysis has focused on estimating the benefits from small irrigation schemes. The NSB‟s Poverty Monitoring Report in 2001 shows that small irrigation schemes can generate an increase of about 1.35% in annual per capita income. Based on this important parameter and a few moderate assumptions, the rate of return to the small irrigation activity is estimated around 29%. This ERR can be treated as the lower bound of the real ERR for the whole Basic Rural Infrastructure component since, when asked to rank the different activities of infrastructure component, farmers always give the first two priorities to road access and drinking water systems. Farmers generally rank small irrigation as the third priority. Therefore, we can reasonably believe that ERRs for roads and drinking water schemes should be greater than that of small irrigation schemes in order to be consistent with farmers‟ preference order. Moreover, other researchers (such as Rozelle and Huang, 2003) have also found similar results, and this further supports the analysis‟ findings. 3.5 Basic Education component. Ravallion and Chen (1997) showed that the per capita income of those who have completed elementary education is about 20% higher than those who are illiterate in rural China. By using (a) this key parameter and assuming it is applicable to 9- year compulsory education (that is, not only applicable to 6-year elementary education), and (b) the NSB‟s per capita income data of the three project provinces from 2006 to 2010, the rate of return to the education is estimated to be around 22%. This ERR can be assumed to be the lower bound of the real ERR for the Basic Education component. This is mainly because of the implementation of the government‟s free education policy, particularly in poor areas, since 2005. As a result, poor farmers‟ expenditure on education has been significant reduced. In addition, the analysis did not quantify the benefits of teacher‟s training as well as classroom construction, while these investments were included as costs. Hence, the real ERR for the Basic Education component could be much higher. For example, research conducted by the International Food Policy Research Institute (IFPRI) in 2001 showed that investment in education is the most profitable investment in terms of its return to rural GDP in western rural China. IFPRI estimated that a RMB 1 investment in education would generate RMB 5 in return. 53 - 3.6 Other calculations. No further efforts were undertaken to separately estimate the financial rate of return (FRR). After more than three decades of market-oriented economic reforms, the price distortions due to the government‟s intervention in the Chinese economy have been basically eliminated. Rich and strong empirical evidence consistently show that an integrated domestic market has been established in China. In addition, the Chinese economy has been increasingly integrated with the international market. Against this background, there is essentially no additional value to separately calculate FRRs to capture the impact of price distortions. Detailed calculations, including methodologies, specific assumptions, crop budgets, and activity models are provided in the project files (Annex 9, Document 9). 54 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending/Supervision/ICR Alan Piazza Senior Economist SDV TTL Senior Financial Management Yi Dong EAPFM Financial Management Specialist Yiren Feng Environmental Specialist EASCS Environment Guzman P. Garcia-Rivero Consultant, Operations Adviser EASER Operations Yuan Wang Procurement Analyst EAPPR Procurement Ede Jorge Ijjasz-Vasquez Sector Manager EASCS Manager Guo Li Senior Agriculture Economist AFTAR Co-TTL Navin K. Rai Adviser SDV Ethnic Minorities Xin Ren Environmental Spec. EASCS Environment Ulrich K. H. M. Schmitt Senior Natural Resources Econ. EASCS Co-TTL Jinan Shi Senior Procurement Specialist EAPPR Procurement Chaogang Wang Senior Social Development Spec SDV Participation Meixiang Zhou Social Development Specialist EASCS Safeguards Patricia Fernandes Social Development Specialist EASER Participation Bruce Trangmar Consultant Agriculture Johanna Pennarz Consultant Participation Arjan deHaan DFID TTL DFID Co-TTL Holger Grundel DFID TTL DFID Co-TTL Catherine Martin DFID TTL DFID Co-TTL Alistar Wray DFID TTL DFID Co-TTL Aimin Hao Social Development Adviser DFID Social Development Xiaochun Ma Social Development Adviser DFID Social Development Hu Wenbin Consultant DFID Education Kunlun Ding Consultant DFID Rural Infrastructure Yao Lan Consultant DFID Health Yunguo Liu Consultant DFID Health (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY01 55.60 FY02 89.33 FY03 77.25 55 FY04 307.92 FY05 325.80 FY06 0.58 FY07 0.00 FY08 10.85 Total: 867.33 Supervision/ICR FY01 0.00 FY02 0.00 FY03 0.00 FY04 0.00 FY05 0.00 FY06 160.99 FY07 181.67 FY08 219.21 FY09 172.00 FY10 210.00 FY11 171.00 FY12 23.26 Total: 1138.13 56 Annex 5. Beneficiary Survey Results See Section 3.6 and M&E Module 2 Reports (Annex 9 ITAD Reports and Documents). 57 Annex 6. Stakeholder Workshop Report and Results See Section 3.6 and M&E Module 2 Reports (Annex 9 ITAD Reports and Documents). 58 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR 7.1 General project assessment by the Central Liaison Group. The Poor Rural Community Development Project (PRCDP) in China is the fourth extensive collaboration between Chinese Government and the World Bank, which adopted integrated measures to directly target the poor households across regions, covering 1.4 million poor populations in 18 counties of Sichuan, Yunnan and Guangxi. The project had started preparation since 2002 and has been carried out since 2005 with its completion at the end of June, 2011. The practice of PRCDP not only contributed to the total investment of poverty reduction and the decrease of the poor population, but also offered experience and lessons for the mechanism innovations of the rural poverty reduction and development for China. Meanwhile, it expedited the realization of the goal of the Chinese Poverty Reduction and Development Outlines (2001-2010). 7.2 Many innovative designs and interventions could be found as for PRCDP. It is the first time that the participatory approach has been carried out through the whole process including project design, implementation and supervision. The project also explored concrete proceedings for those poor people to participate in poverty alleviation project throughout and provided experience as well as lessons for domestic participatory poverty alleviation. By now, PRCDP‟s comprehensive process of participatory approach has been adopted by some other domestic projects. Besides, the project especially focused on ethnic minority groups as a poverty reduction project. At the stage of design, the Ethnic Minority Development Plan has been developed; at the stage of planning, the development demands of the ethnic minorities have been fully respected; at the stage of the implementation, the privilege has been given to the ethnic villages to firstly commence the implementation, and the ethnic people have been prioritized as loan recipients and their loan amount has been under special considerations. The implementation of the project has addressed the poverty issues of the ethnic minority groups effectively. 7.3 The implementation of PRCDP is successful. The foregone practice has proved that the achievement of the project depends on the Grant for purpose of softening and the advanced managerial ideology introduced by the World Bank. Equally important, that the decision of involving the co-participations of DFID and World Bank in project implementation and management has been approved to be correct, which provided a solid basis for the success of the project. 7.4 When we are concluding the successful experience, we should also notice the improvements that could be made. Firstly, some issues of significance, such as the arrangement of the counterpart funds, received negative impact because of the absence of a strong Central Project Management Office. Secondly, poverty reduction aid to the disabled group is a brand new issue under Chinese domestic context, which made it an innovative highlight of PRCDP. However, the component was designed and implemented by an international NGO independently, and separated itself away from the project and management office, which also made it difficult to be replicated and extended to other domestic projects. 59 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders DFID 1.1 DFID comments. DFID has continued to be closely involved in the project, till the counterpart funding came to an end. Collaboration between World Bank and DFID has been excellent, among the best that the DFID programme manager Arjan de Haan has experienced in ten years of working for DFID. Also, we believe that the modality of co-funding was productive, as the blending of funding combined with donor collaboration in monitoring contributed to promotion of key aspects of the new approach developed in the project (notably, the participatory approach). 1.2 While programming priorities for DFID changed over the period, and it became increasingly difficult to justify grants to China (given its fabulous economic growth, reserves, and display of technology to the rest of the world), PRCDP continued to be one DFID‟s key programmes able to show that international collaboration with China continued to contribute to achieving the MDGs. At the same time, the collaboration with LGOP provided a basis for the UK in its efforts to promote collaboration with China on China‟s international development programme (which was not part of the original objectives, but became an important spin-off). 1.3 The strength of the project was clearly demonstrated during the visit of the UK International Development Committee, which visited Chengdu, in the months after the Wenchuan earthquake. The committee was visibly impressed by the quality of the work and presentation of the Sichuan PMO, including the strength of its monitoring system (which has been acknowledged earlier during a visit of the UK National Audit Office). There was also much recognition for the way in which participatory methods developed under PRCDP was used to work in the extremely challenging task of reconstruction of thousands of villages destroyed or damaged by the earthquake. 1.4 DFID has filed extensive reporting that documents the progress under PRCDP. The following points will be repeating much of the assessment in the World Bank ICR text, but are still worth listing as independent observations from the DFID monitoring and supervision:  First, project implementation has reached 100% three months before end of the project. This followed significant challenges regarding counter-part funding, which had been overcome after year two/three (and despite the distraction to program implementation that the earthquake could have caused at least in Sichuan).  Second, the integrated approach has been achieved, ensuring for example enhanced access to health care by poorest, and important experience gained through free education and disability pilots. At least for parts of the project, this implied progress in coordination across departments, or at least the realization that development progress requires multi- sectoral thinking and approach.  Third, targeting of project funding has been very good, overcoming initial difficulties, reaching the poorest among over 1000 villages. 1.5 Project partners all agree that the participatory approach has been the defining feature of PRCDP. Substantial awareness of its benefits has been created, and participation has been mainstreamed throughout all project stages including – as became clear in the last year of the project – post-project management. The participatory approach has directly led to significant direct benefit for poorest communities and marginalized groups. There is convincing evidence 60 that the PRCDP experiences have influenced national and provincial policies, including NDRC approaches to earthquake reconstruction. Handicap International 2.1 Handicap International -- Yuexi disability project. The Yuexi Disability Project is a component of the Poor Rural Community Development Program (PRCDP) initiated in 2004 by the World Bank, the UK‟s Department for International Development (DFID) and the Leading Group Office for Poverty Alleviation and Development of the State Council of China. The program covers activities in the fields of poverty alleviation, education, agriculture and infrastructure. The majority of persons with disabilities are poor, so it was decided to add a disability component to this program. Yuexi County, located in Liangshan Yi Autonomous Prefecture in Southwest Sichuan was selected to implement a pilot project aimed at improving the social-economic inclusion of people with disabilities (PwD). Handicap International (HI) provided the technical assistance and financial management to implement this project, in close cooperation with the Yuexi County People‟s Government. The Yuexi Disabled People Federation (YDPF) implemented the project. The project was implemented over 5 years (from August 2005 to June 2010) and covered 137 villages in 18 townships in Yuexi County. 2.2 Objectives. The overall objective of the project was the social inclusion of persons with disabilities in rural China and the prevention of disability, leading to overall poverty reduction. The specific objective was to develop and implement in Yuexi County duplicable methods and tools, which show positive results in the social inclusion of persons with disabilities and the prevention of disabilities. The outputs were: to build the capacity at local (Yuexi) county level to plan and implement disability action; to develop a replicable Disability Management Information System (DMIS) for disability action. To establish a community based system for the prevention of disability, rehabilitation and social inclusion; a sustainable Disability Fund was established in Yuexi County, managed by the YDPF. Regulation, training and reference materials were developed so they could be used in other counties. Guidelines for mainstreaming PwDs into different sectors of society were put in place and tools and methods for disability action were disseminated to policy makers for disability action and poverty alleviation in China. 2.3 Yuexi was a very ambitious project as it focused both on improving the social inclusion of persons with disabilities in Yuexi County itself as well as making sure that the deliverables of the project could be replicated in Yuexi and other rural areas in China. It therefore had to ensure the “buy-in� of both local and national authorities. 2.4 Impacts. The impacts of the project were:  On the final beneficiaries - more than 3,000 PwD in the 137 villages of the 18 townships received some kinds of service or support. PwDs have gradually gained services and become more involved in socio-economic activities, while learning about laws and regulations and becoming more aware of their rights.  On local partners: the capacity of local partners has been greatly enhanced and improved in terms of their understanding and acceptance of the CBR (Community based Rehabilitation） concept. The understanding and use of Logframe and Participatory approaches, especially in terms of income generation and Social inclusion activities has been enhanced. The partners‟ capacity and skills in project implementation and management was greatly improved. A team of well-trained staff capable of carrying out 61 sustainable development actions and a sustainable development mechanism has been implemented in Yuexi.  At national level: The project benefited from the constant involvement of the China Disabled Persons Federation (CDPF), the main organization at national level dealing with disability issues. A set of reference materials will be formally presented to the CDPF at a national conference on CBR to be organized in September 2011 and with the financial support from DFID, the official Chinese translation of the latest version of the CBR guidelines, developed by WHO, ILO and UNESCO will be available in late 2011. 2.5 Lessons learned. The project was too ambitious compared to the local context and capacity of partners and faced with such an ambitious project, the local Yuexi team felt that many actions undertaken by HI were very difficult to promote due to the poor economic context, remoteness of the location, the limited resources and the local partners‟ weak capacity. In addition:  Quality - the Yuexi DPF, the main implementer of the project, is in charge of administrative issues related to disability. Although their staff have been trained on all the technical components of the project, the DPF still lack qualified technical staff to follow- up quality issues.  Partnerships were difficult due to the differences of expectations between HI and the YPDF. While HI focused more on the social inclusion activities, YDPF was more interested in hardware and concrete project outcomes, such as new houses for persons with disabilities and the new comprehensive centre. 2.6 Recommendations:  More leverage on Chinese policies and regulations: While China has ratified the UN Convention on the Rights of Persons with Disabilities in 2008 and has adopted a series of national regulations regarding disability which comply with the convention. These regulations are not well known in rural China and more dissemination could help in motivating the local DPF to implement them. HI should have leveraged more on those national regulations instead of using the CRPD to motivate the local stakeholders.  Better customized CBR concepts to local context: In particular, the social mobilization and empowerment components were difficult to implement in a context like Yuexi, as many of the basic health, education and livelihoods services for persons with disabilities were not yet available, at the start of the project. It would have been better to focus on those basic services first rather than trying to implement all concepts at the same time. It helps both partners and beneficiaries to understand the direct benefits they could get from such an approach.  Mainstreaming: There were only few opportunities for mainstreaming the concepts of the project offered by the PRCDP. Disability is a cross-cutting issue and should be dealt as part of the main poverty reduction and development strategies. It is recommended in the future to include disability as a component of Chinese poverty reduction strategies. Johns Hopkins University 3.1 Johns Hopkins University – Comments on the visit to Guangke Village. A team of 18 graduate students and professors from Johns Hopkins University visited Beijing and Yunnan and undertook field visits in the Meigu County project area in January 2011. 62 3.2 Background. From January 2 to 17, 2011, 18 graduate students and professors from the Johns Hopkins University School of Advanced International Studies (SAIS) and the Johns Hopkins University-Nanjing University Center for Chinese and American Studies visited Beijing and Yunnan to research poverty reduction and sustainable development efforts in China. Participants had the opportunity to meet with Country Director Klaus Rohland and Senior Economist Ulrich Schmitt at the World Bank‟s Beijing Office before visiting project sites in Yunnan province that were part of the Bank‟s Poverty IV project. Because few outside groups had been to these sites, Senior Economist and Task Team Leader Alan Piazza invited the Hopkins trip participants to provide feedback on their observations and conversations with residents of Guangke Village in Menglian County, Yunnan province. 3.3 The World Bank‟s program in Guangke is administered locally by the Menglian County Poor Rural Communities Development Project (PRCDP) Office. PRCDP also has a Guangke Village office staffed with five or six Guangke residents. Guangke is a remote village located in the mountains 50 kilometers from the Burmese border that has, according to Menglian County PRDCP Director He Xiangxin, about 3,000 people and an annual per capita income of US$700 to $900. Yunnan is one of China‟s most ethnically diverse provinces and Guangke residents primarily belong to the Dai, Lahu, and Wa ethnic minority groups. 3.4 Observations.  Impact on poverty and well-being. In general, residents spoke positively about aspects of the Bank‟s program, such as the microloans that allowed them to raise pigs and the improved transport infrastructure that has boosted crop sales, including those of rubber, bananas, and corn. Families that trip members spoke with seemed to live comfortably and felt little need to seek employment in the cities. Many villagers also said they felt richer and more satisfied than if they had gone to the city to work as migrant laborers.  Infrastructure development. Guangke lies in a mountainous area three hours from the county seat of Menglian, so villagers have limited access to centers of commerce and trade. Although many residents expressed confidence in the state-sponsored pension system and the New Rural Cooperative Medical System, gaining access to medical services was sometimes a challenge since the nearest clinic was several kilometers away on unpaved roads. Participants noted recent improvements, however, including the paving of an access road to Guangke. Motorbikes were common in the village and communications infrastructure seemed well-developed. Each family that trip participants spoke with had a cell phone and at least one household had a television signal. Significant effort had also been expended to bring clean water and electricity across the mountains to Guangke.  Participatory approach. Arriving in Guangke, Hopkins trip participants were interested in seeing how PRCDP leaders implemented the participatory approach to community development projects, given its key role in recent World Bank poverty reduction programs in China. Representatives from the PRCDP facilitate meetings with Guangke residents, many of whom seemed quite interested in discussing and voting on development projects for their community. Local officials and village leaders expressed support for the success of the Bank‟s projects and were impressed by the community's active response to the participatory approach to development.  While there are obvious benefits to this approach, discussions with participants offered insights into some of the challenges in implementation. For example, although there is a quota for the percentage of meeting attendees that must be female, some women explained that they were not able to attend community meetings because of work or household commitments. After a welcoming ceremony for the Hopkins group, 63 participants saw a man write his sister‟s name on the attendance sheet even though she was not present. Villagers said they typically receive US$3 to $4 for attending World Bank project-related events, including PRCDP meetings, to compensate them for their time. Many villagers also expressed uncertainty about what projects would benefit them the most. Several said that they tended to follow the suggestions of village government leaders while others maintained that the perceived ability to successfully implement a project was a key factor in Guangke‟s project selection.  Educational needs. Village children attended elementary school up to grade four in Guangke but went elsewhere for middle and high school. This already signifies an important step forward, as in the past even the youngest students had to travel long distances to attend school. Grade five and middle school students attend school in Gongxin town, which draws students from the 12 surrounding villages, while high school-age children attend school several hours away in the county seat of Menglian. Guangke is composed of several ethnic communities and the school is located closest to part of the village where the Dai community is concentrated, while most Wa and Lahu families live further away. Teachers at Guangke elementary school said learning was most difficult for Wa children because of language and cultural barriers. There is high turnover among teachers, many of whom come from less remote areas and find it difficult to live in Guangke. 3.5 Lessons Learned.  Infrastructure, especially road construction, provides a vital link between rural residents and essential public services. Many villagers repeatedly stressed their desire for a completed circuit of roads. Future projects should continue to prioritize road infrastructure so that village residents can have maximum access to the commercial, social service, and educational benefits of the larger community.  The participatory approach enjoys strong support from villagers and government officials, but there is a need for better monitoring and more inclusive participation. Although women comprised roughly half of the population, it seemed that men were the more active participants in discussions and decisions. The PRCDP meeting facilitators should encourage and enable women to have a voice in project selection and community priorities, and concrete monitoring and evaluation systems should be implemented to ensure the integrity of the participatory approach.  In project selection, village leaders should play more of a facilitating role and less of a decision-making role. Many participants noted that village leaders‟ opinions dominate development planning. While these leaders may be well-meaning, further efforts should be made to empower individuals and households to participate actively in the project decision-making process. 64 Annex 9. List of Supporting Documents All of the documents (including ITAD‟s Reports and Documents) are available in WBDocs. Documents 1. World Bank. 2005. Poor Rural Communities Development Project Project Appraisal Document. 2. Quality Assurance Group. 2009. QALP Final Report. 3. Supervision Mission Aide Memoires. 4. Sichuan FCPMC. 2011. Sichuan Project Yanshou Evaluation Report. 5. Guangxi FCPMC. 2011. Report of PRCDP Implementation in Guangxi. 6. Yunnan FCPMC. 2011. Report of PRCDP Implementation in Yunnan. 7. World Bank. 2001. China: Overcoming Rural Poverty. 8. World Bank. 2011. Rate of Return Calculations (spreadsheet) and Underlying Data (in Chinese). 9. NSB. 2011. Final Report of Poverty Monitoring Survey (and each year 2006-10). M&E Module 1 Reports. 10. Li, Guo. 2011. Economic Analysis Calculations. 11. Mott MacDonald (Beijing) Limited. 2011. PRCDP FEP Final Survey Data Analysis Report. 12. Handicap International. 2010. Yuexi Disability Project Component of PRCDP. 13. World Bank. 2011. Table of Project Performance Indicators. ITAD Reports and Documents ITAD. Final Report on “Development of Participation Techniques for Use in PRCDP� (Preparation Phase). 2005 ITAD. PRCDP Participation Manual, January 2005 (http://www.itad.com/PRCDP/) ITAD. Report on Beijing Workshop on “Participatory Approaches for equitable and inclusive poverty reduction in Guangxi, China�. 2005 ITAD. Report on Kunming Workshop on “Good practices and perspectives on participatory poverty reduction in China�. 2009 ITAD with Yunnan University. Proceedings of Kunming Workshop on “Good practices and perspectives on participatory poverty reduction in China�. 2009 Johanna Pennarz, Song Haokun, Deng Weijie and Wang Jianping (Eds.). “Emerging practices in participatory poverty reduction in China�. PLA Notes No 62. 2011. First Supervision Mission: Progress on Participation (2005) Second Supervision Mission: Progress on Participation (2006) Third Supervision Mission: Progress on Participation (2006) 4th Supervision Mission: Progress on Participation (May 2007) 5th Supervision Mission: Progress on Participation (January 2008) 6th Supervision Mission: Progress on Participation (April 2008) 7th Supervision Mission: Progress on Participation (July 2008) 8th Supervision Mission: Progress on Participation (December 2008) 9th Supervision Mission: Progress on Participation (July 2009) 10th Supervision Mission: Progress on Participation (December 2009) 65 11th Supervision Mission: Progress on Participation (May 2010) PRCDP Final Report on Participation (July 2011) Module 2 Participatory Poverty Monitoring (PPM Report) - Year 1 (2007) Module 2 Participatory Poverty Monitoring (PPM) Report - Year 2 (2008) Module 2 Participatory Poverty Monitoring (PPM) Report - Year 3 (2009 Module 2 Participatory Poverty Monitoring (PPM) Report - Year 4 (2010) Module 2 Participatory Poverty Monitoring (PPM) Report - Year 4 (2010) Module 2 Participatory Poverty Monitoring (PPM) Report - Year 5 (2011) Workshops Major workshops on participation in PRCDP Workshop Name18 Date Purpose Participants First “Three April 2004, First workshop to Participants from the Provinces Sharing Kunming share approaches on three PRCDP Workshop� PPA/PP between provinces; provinces during representatives from preparation phase. DFID and World Purpose was to agree Bank on common principles to integrate into PRCDP Participation Manual. Workshop on December 2005, Workshop to Representatives Participatory Beijing disseminate learning from FCPMC, Approaches for from Guangxi‟s Guangxi PADO, equitable and approach to PPA Guangxi FCMPC inclusive poverty and PP, developed and the Guangxi reduction in during the Commissions for Guangxi, China preparation phase. Economic Development. Representatives from World Bank and other international donors and projects. Workshop on March 2009, Workshop to Practitioners from “Good practices and Kunming support learning on eight provinces 18 In addition PRCDP has conducted a series of workshops with individual provinces on PPA and PP methodology during preparation (see Final Report on Preparation Phase), on sharing of good practices during supervisions (see Reports on Supervision) and as part of the feedback of findings from the annual Module 2 surveys. 66 perspectives on participation among (Sichuan, Yunnan, participatory a wider group of Guangxi, Guizhou, poverty reduction in practitioners Shanxi, Beijing, China� working in DFID Gansu, and Jiangsu) and World Bank and representative of funded projects in the Leading Group different parts of on Poverty China Alleviation and the Ministry of Water Resources Interim sharing April 13. 2010, Workshop Participants from the workshop on Nanning organized as part of three PRCDP “PM&E and case the 11th Supervision provinces; studies in PRCDP� to strengthen representatives from learning on PM&E DFID and World and case studies Bank Final sharing March 2011, To critically review Participants from the workshop Kunming experiences and three PRCDP “Reflection on draw conclusions provinces; participation in for further scale-up representatives from PRCDP� of participatory DFID and World practices from Bank PRCDP 67 IBRD 33882 100° 104° 108° 112° CHINA POOR RURAL COMMUNITIES SHAANXI hui Han S DEVELOPMENT PROJECT PROJECT COUNTIES 32°N Yal PROJECT PROVINCES ong 32°N Han Shui MAIN RIVERS M in PROVINCE CAPITALS Pingchang PROVINCE BOUNDARIES HUBEI INTERNATIONAL BOUNDARIES SICHUAN Chang Jian g (Yangtze) Tuo Lan CHENGDU can Jinsha g Dadu He Nu CHONGQING CHONGQING n Yua Mabian Pingshan Zi Shui Yuexi Meigu 28°N 28°N Xuyong GUIZHOU HUNAN Jins Wu ha GUIYANG u gto 0 100 200 300 Kilometers Lo n 0 100 200 Miles g X i an Longsheng This map was produced by the Map Design Unit of The World Bank. Sanjiang The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any Rongshui endorsement or acceptance of such boundaries. KUNMING Rong'an G ui Huanjiang Luocheng Changning RUSSIAN FEDERATION Longling Nu YUNNAN G U A N G X I KAZAKHSTAN HEILONGJIANG 24°N N 24° Yongde L ui MONGOLIA JILIN Sea Pearl KYRGYZ (Zh of You REP. LIAONING D.P.R. OF u ji a n XINJIANG NGO L KOREA Japan g) I MO BEIJING NE BEIJING Cangyuan TIANJIN REP. OF JAPAN Lan HEBEI KOREA NANNING can Yellow NINGXIA SHANXI SHANDONG MYANMAR g QINGHAI Sea Ximeng JIA AREA OF GANSU HENAN NG MAP SU SHAANXI Z uo SHANGHAI XIZANG ANHUI East HUBEI G IN ZHEJIANG SICHUAN China Menglian Q NG O CH JIANGXI Sea HUNAN FUJIAN GUIZHOU National Capital YUNNAN GUANGDONG TAIWAN VIETNAM Province Boundaries International Boundaries GUANGXI HONG KONG MACAO LAO VIETNAM Bay of B e n g a l LAO P.D.R. HAINAN 100° P.D.R. 104° 108° PHILIPPINES APRIL 2005