ICRR 11880 Report Number : ICRR11880 ICR Review Operations Evaluation Department 1. Project Data: Date Posted : 09/21/2004 PROJ ID : P003649 Appraisal Actual Project Name : Shanxi Poverty Alleviation Project Costs 182.8 204.3 Project US$M ) (US$M) Country : China Loan/ Loan US$M ) 100 /Credit (US$M) 90.9 Sector (s): Board: RDV - General Cofinancing agriculture fishing and US$M ) (US$M) forestry sector (68%), Flood protection (11%), Water supply (10%), Roads and highways (10%), Agro-industry (1%) L/C Number : C2834 Board Approval 96 FY ) (FY) Partners involved : Closing Date 12/31/2002 12/31/2003 Prepared by : Reviewed by : Group Manager : Group : Kavita Mathur Nalini B. Kumar Alain A. Barbu OEDSG 2. Project Objectives and Components a. Objectives The objective of the project was to help alleviate poverty in the 20 poorest counties of Yuncheng and Luliang Prefectures in Shanxi Province. b. Components The project components included: improvement of the Yuncheng Irrigation System (appraisal US$ 35.1 million, actual US$ 17.1 million); the development of some 13,000 ha river beach-land along the Yellow River (appraisal US$ 15.4 million, actual US$ 16.9 million); rural water supply (appraisal US$ 9.2 million, actual US$ 11.4 million); rural roads (appraisal US$ 17.1 million, actual US$ 27.2 million); soil and water conservation including terracing, afforestation and check dam construction (appraisal US$ 14.7 million, actual US$ 23.3 million); horticulture development (appraisal US$ 26.7 million, actual US$ 53.7 million); livestock raising (appraisal US$ 14.5 million, actual US$ 23.6 million); expansion of agro-processing (appraisal US$ 17.3 million, actual US$ 18.0 million); All China Women's Federation (ACWF) (appraisal US$ 0.6 million, actual US$ 4.5 million); and project management support (appraisal US$ 13.5 million, actual US$ 8.6 million). c. Comments on Project Cost, Financing and Dates The actual project costs were US$ 204.3 million, slightly higher than the appraisal estimate of US$ 182.8 million. The IDA disbursements were US$ 90.8 million compared to the appraisal estimate of US$ 100 million. This was due to significant devaluation of the SDR against the US$. The credit was extended for one year to allow time for implementation of the ACWF component in Yuncheng. 3. Achievement of Relevant Objectives: The project objective appears to have been largely achieved, though the evidence provided in the ICR is weak. While on one hand the ICR reports that an estimated 4.1 million people directly benefited from the project as compared to an appraisal target of 3 million, on the other it notes that this number includes double-counting of people which benefited from several components. Subsequent information provided by the Region shows that actual number of people that benefited were 3.2 million. Further, the ICR reports a substantial increase in per capita income but it is not clear how much of that increase can be attributed to the project. Annex 1 notes that per capita grain production was seen as an indicator at the appraisal stage but as grain prices dropped non-grain based income became much more effective in generating income. However at project closing per capita grain production was actually lower than the base scenario for 1994. There were also no indicators that report on how much animal husbandry and horticulture components contributed to increase in incomes. In addition, though the ICR also reports a high ex-post Economic Rate of Return (ERR of 27% for the whole project) the calculations do not take into account the agro-processing component which was unsatisfactory and accounted for 17 percent of the total project cost. 4. Significant Outcomes/Impacts: The project is reported to have developed about 16,820 ha (compared to the SAR target of 13,000 ha) of highly fertile land along the Yellow River. Access to markets is reported to have increased with the construction of 318 km of rural roads. The number of cars and small trucks owned by residents along the roads has increased. More than 11,000 women were trained in poultry, rabbit and pig enterprises in Luliang. This component was extended to Yuncheng where 2,600 women were trained under the ACWF training model. ACWF trained direct recipients to train other households in enterprise management. Terraces were built on 16,107 ha compared to the target of 13,400 ha. 5. Significant Shortcomings (including non-compliance with safeguard policies): The project was classified an Environment Category B project.The Environment Impact Assessment at the appraisal stage described the main environmental concerns associated with each subcomponent and the appropriate mitigation strategy. The ICR contains no details whatsoever on how the environmental concerns were dealt with and whether compliance with safeguard procedures was adequate (the region subsequently provided evidence of substantive compliance with key safeguards regarding resettlement and dam safety, although some questions remain regarding the extent of compliance with environmental safeguards). It is not even clear whether the project established the Environmental Monitoring Unit envisaged in the SAR. Several agro-processing plants failed because of poor management skills, large unproductive work force and inability to adjust to rapidly changing commercial environment. There is no information in the ICR on the agro-processing plants that were set-up and the monitoring of environmental impacts and the mitigating measures used. Low government commitment towards investment in Yuncheng Irrigation System. There were significant delays in the implementation of the Yuncheng Irrigation System component due to lack of counterpart funding. This led to elimination of some elements of the Langdian Complex. The sediment basin to protect Jimamakou and Xiaofan was not built. The 10 km section of feeder channel from Xiaofun to Zuncun to supply the Zuncun area was not completed. Therefore, much of the area served by Zuncun continues to rely on tube-wells. There is significant shortfall in afforestation under the soil and water conservation component. The area for afforestation was 1,677 ha compared to the target of 4,200 ha. Project monitoring was weak. 6. Ratings : ICR OED Review Reason for Disagreement /Comments Outcome : Satisfactory Moderately Satisfactory OED rates a project as moderately satisfactory [a rating which does not exist under the ICR's 4-point rating scale] when it achieves most of its major relevant objectives but with significant shortcomings. The shortcomings noted under section 5 are significant. Institutional Dev .: Substantial Substantial Sustainability : Highly Likely Highly Likely Bank Performance : Satisfactory Unsatisfactory Monitoring of the project objective was extremely weak. No income based indicators were developed for non-grain based income sources. For several of the output indicators presented in Annex 1 in the ICR it is not clear how they relate to Table 2 (page 198) of the PAD. As a result, during implementation, even though non-grain based sources of income became more important for generating income, it is difficult to tell how much they contributed to increased incomes in the project area. In addition, it appears from the information reviewed by OED that the environmental categorization for the project should have been an 'A' instead of a 'B': inter alia, the project proposed to include 30,000 ha of new irrigation development and safeguard policies which were or should have been triggered included natural habitat, dam safety, pest management and resettlement. Borrower Perf .: Satisfactory Satisfactory Quality of ICR : Unsatisfactory NOTE: NOTE ICR rating values flagged with ' * ' don't comply with OP/BP 13.55, but are listed for completeness. 7. Lessons of Broad Applicability: Based on the ICR the lessons are: At appraisal key indicators to evaluate the achievement of project objectives should be developed for each component and during implementation the performance indicators should be monitored. Otherwise it becomes difficult to assess how far each component contributed to the achievement of the overall objective. Investment in agro-processing can only be successful, if management and ownership structures are equipped to commercially compete in today’s business environment. Commitment to changed ownership and management is critical. An additional lesson drawn by OED is that, where the appraisal notes the need to monitor compliance with Bank safeguard policies these should be addressed explicitly during implementation and the results should be reported in detail during supervision and completion. As well as demonstrating due diligence by the Bank, such oversight would enrich and disseminate experiences that could be replicated elsewhere. 8. Assessment Recommended? Yes No Why? For three reasons: (i) To verify the outcome, institutional development and sustainability ratings, and (ii) to determine impact on poverty alleviation; (iii) to assess compliance with the Bank's safeguard policies. 9. Comments on Quality of ICR: The quality of the ICR is unsatisfactory in light of its significant shortcomings: The ICR does not report on safeguard compliance. Even though the appraisal document noted the need for environmental monitoring and pointed to the adverse environmental impact of some components, there is no information in the ICR on how these were followed up during implementation. The ICR also does not discuss whether any resettlement occurred especially in the case of construction of rural roads and the Langdian pumping complex. Further, while the ICR notes that in some cases roads were built to a higher standards than Class 4, it provides no information on whether road widening was required or whether this meant the acquiring of additional land. The ICR overstates the project's impacts. For instance it notes that an estimated 4.1 million people benefitted directly from the project as compared to the appraisal estimate of 3 million. However this number is misleading as it includes double-counting of people, which benefited from several components (the regions subsequently confirmed that the actual number of beneficiaries was 3.2 million). Further, even though it claims that the ERR calculation is for the whole project, it leaves out the unsuccessful agro-processing component from the calculation. The ICR provides information mostly on physical outputs. Except for a few indicators, it does not report fully on the limited key performance indicators noted in the SAR (Annex 15, Table 2). The ICR quotes a 2003 survey as evidence that 90 percent of the water supply schemes are well managed and maintained, but provides no basic information on the survey or the methodology it used.