ICRR 11153 Report Number : ICRR11153 ICR Review Operations Evaluation Department 1. Project Data: Date Posted : 04/10/2002 PROJ ID : P038885 Appraisal Actual Project Name : Rural Poverty Alleviation - Project Costs 60.0 57.4 Sergipe US$M ) (US$M) Country : Brazil Loan/ Loan US$M ) 36.0 /Credit (US$M) 36.0 Sector (s): Board: RDV - Power (23%), Cofinancing Water supply (23%), Roads US$M ) (US$M) and highways (23%), Irrigation and drainage (22%), Sub-national government administration (9%) L/C Number : L3919 Board Approval 95 FY ) (FY) Partners involved : Closing Date 12/31/2000 06/30/2001 Prepared by : Reviewed by : Group Manager : Group : John R. Heath Ridley Nelson Alain A. Barbu OEDST 2. Project Objectives and Components a. Objectives "The project would assist the State Government of Sergipe to alleviate rural poverty and its consequences by : (a) providing basic social and economic infrastructure and employment and income -generating opportunities for the rural poor; (b) supporting rural community groups in planning and implementing their own subprojects; (c) providing a safety net for the rural poor during a period of strong fiscal adjustment when the Government's ability to finance essential investments and services is constrained; and (d) leveraging revenue mobilization at the community and municipal levels . The project would ensure that funds are targeted and reliably transferred to the poorest communities, and would promote further decentralization of decision -making to state, municipal and local levels " . (Staff Appraisal Report, pp. 9-10) b. Components (a) Community subprojects (US$55.1 million, 93 percent of base cost), which would support small-scale investments selected and subsequently operated and maintained by the beneficiaries themselves; (b) Institutional development (US$2.4 million, 4 percent of base cost), which would provide all implementing entities and communities with technical assistance and training to increase their capacity and improve implementation of the project; (c) Project administration, monitoring and evaluation (US$1.8 million, 3 percent of base cost), which would finance project coordination and activities to provide feedback on project performance and impact . (SAR, p. 10) c. Comments on Project Cost, Financing and Dates The actual cost of the components was : (a) US$52.4 million, (b) US$$1.5 million, and (c) US$3.4 million. The overall project cost was 4 percent less than expected, mainly because of the post -1999 devaluation of the Real against the US dollar. To bring it in line with similar projects in other states, the Bank's share of the project cost was increased from 60 to 75 percent from February 12, 1999. Counterpart financing was sound for all except the last six months . 3. Achievement of Relevant Objectives: All four objectives were highly relevant to the task of combating rural poverty . Based on data contained in Annexes 1 and 2 of the ICR, achievements were as follows : Objective (a). Given that this was a demand-driven project, no appraisal target was set, either for the number of subprojects or the number of beneficiaries . However, it was estimated that 2,130 subprojects would be implemented, benefiting 330,000 people. The actual numbers were: 1,810 subprojects and 300,000 beneficiaries. Thirty percent of beneficiaries were women. Most of the subprojects (84 percent) were for rural infrastructure, particularly rural electrification and water supply; productive and social subprojects accounted for respectively 11 percent and 5 percent of the total. The average subproject cost was US$ 27,000 (the ceiling was US$50,000). Objective (b). The number of community associations that participated was 931, compared to an indicative estimate of 1,065. Subprojects were planned and implemented via three separate channels, distinguished by the level to which decision making was decentralized . The most decentralized channels (FUMAC and FUMAC-P) involved municipal councils, with 80 percent of members drawn from the community . These two channels handled 78 percent of subproject disbursements, slightly more than projected at appraisal (75 percent). However, Annex 1 indicates that, compared to projections at appraisal, there were far fewer than expected training events delivered to councils and beneficiaries. Objective (c). The ICR notes that this is one project in a multi -state program that "became the single most important source of funding for protection of the poor in the rural Northeast " (p. 6). This statement is not quantified : it is not clear how either the number of beneficiaries, or the volume of resources per beneficiary, compared with other programs (e.g. pensions for rural workers). Objective (d). Based on notional estimates of the value of family labor, cash and materials supplied by beneficiaries, it is estimated that they contributed 9 percent of project costs. Municipal governments were not required by the loan agreement to co-finance subprojects but about 60 percent of them contributed an amount equivalent to US$ 3.0 million, or 6 percent of the total cost of subprojects . 4. Significant Outcomes/Impacts: This is one of a series of Northeast poverty reduction projects that, collectively, have delivered substantial resources to the rural poor of that region, and have significantly influenced the Bank's approach to community development . Based on lessons learned during the early phase (1985-93), these projects were restructured to incorporate the following principles: Place money for implementing approved subprojects directly into the hands of beneficiary communities Decentralize decision making and involve local authorities as participants Maintain transparent decision-making at all levels Stress community participation in planning, financing, executing, operating and maintaining the investments Use simple, explicit and monitorable poverty targeting mechanisms . (The following outcomes are not specific to the state of Sergipe but refer to the program as a whole ) Following restructuring, there was a substantial smoothing of disbursement for subprojects (the previous counterpart funding problems were largely resolved and decisions were taken much more rapidly ) and the share of project resources transfered to beneficiaries increased from one -third to 91 percent. Although no rate of return was estimated for each project as a whole, analysis of 12 subproject types shows that rates of return exceed the opportunity cost of capital, ranging from 15 percent (feed production equipment, fish farm, manioc mill ) to over 50 percent (small-scale irrigation, cashew processing ). The approach is highly cost -effective. Subprojects implemented directly by beneficiary groups cost 10-40 percent less than similar subprojects implemented by third parties . They are also of better quality and were more rapidly executed . Analysis of a random sample of subprojects (including ten categories representing about 80 percent of types financed ) found costs to be 30-50 percent lower than subprojects of similar quality implemented by the state . There is some positive early evidence of subproject sustainability . A survey conducted in 2000 found that 89 percent of a sample of 8,123 subprojects financed between 1995 and 1998 were still operational. (In Sergipe specifically) The project had a positive fiscal impact . It is estimated to have saved over R$ 430,000 per year in a year of "normal rainfall" by eliminating, through investments in water supply, the need to send water tankers to communities without drinking water. The incremental revenue from taxes on domestic appliances (purchased after electricity became available) is about R$2.0 million. (In December 2000, US$1.00=R$1.95). 5. Significant Shortcomings (including non-compliance with safeguard policies): The ICR notes that "Targeting was successful under the Sergipe project " (p. 8) but the statement is not quantified . In general , these projects have spread resources thinly : they have put money in the hands of large numbers of poor people but it is not clear how solid a contribution they have made to the long -term growth of incomes and employment in the Northeast (where development is severely constrained by poor soils and lack of water ). Communities have expressed less demand for productive projects and these projects have tended to be less sustainable than infrastructure projects . It is not clear from the ICR what percentage of community associations received funding for more than one subproject; and what percentage were able to finance new subprojects from alternative sources. 6. Ratings : ICR OED Review Reason for Disagreement /Comments Outcome : Satisfactory Satisfactory Institutional Dev .: Substantial Substantial Sustainability : Likely Likely Bank Performance : Satisfactory Satisfactory Borrower Perf .: Satisfactory Satisfactory Quality of ICR : Exemplary 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: From the many listed in the ICR, the following stand out : (a) By assigning greater responsibility to local organizations and municipal goverments, community demand -driven projects can help build institutional capacity in poor rural areas; (b) Community-driven approaches can reduce bureaucracy and administrative obstacles and reinforce accountability for project performance by reducing the distance between decision -makers and beneficiaries. (c) Decentralization of the Bank supervision effort ---in this case, directed from the Recife office ---can greatly enhance implementation, by ensuring rapid response to problems, timely advice giving, and careful monitoring . (d) Screening of projects should be conducted by municipal councils (rather than sending proposals to be vetted by a technical agency in the state capital ) because this helps to build local capacity, with benefits extending beyond the project (potentially citizens acquire more voice in municipal planning ). (e) Productive subprojects should be subjected to rigorous selection, preparation, technical assistance and supervision criteria and linked to a graduation strategy which facilitates beneficiary communities access to formal credit. 8. Assessment Recommended? Yes No 9. Comments on Quality of ICR: This is a model ICR. The report is very well organized, providing a good overview of the program as a whole (including an excellent annex on economic costs and benefits ) with adequate details on achievements specific to the state of Sergipe. There is a good section on lessons learned . The report is very clearly written .