Sri Lanka - Poverty Alleviation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No.: 17998 IMPLEMENTATION COMPLETION REPORT SRI LANKA POVERTY ALLEVIATION PROJECT (Credit 2231-CE) Juae 15, 1998 Education Sector South Asia Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY AND EQUIVALENTS UNIT Currency Unit = SLRupees (Rs) Average Yearly US$1.00 Equivalent 1991 1992 1993 1994 1995 1996 1997 41.4 44.3 47.8 49.7 52.0 55.4 57.8 GOVERNMENT OF SRI LANKA - FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AGA Assistant Government Agent ARTEP Asia Regional Team for Employment Promotion CAS = Country Assistance Strategy CBSL Central Bank of Sri Lanka CENWOR Center for Research on Women CF = Credit Fund CFS = Consumer Finance Survey CMED = Credit & Micro-Enterprise Development CRBs = Cooperative Rural Banks DCS = Department of Census and Statistics DSA Divisional Secretariat Area DSD Divisional Secretariat Division EDP Entrepreneurship Development Program EPPU = Employment and Poverty Policy Unit FSP = Food Stamp Program GDP Gross Domestic Product GND = Grama Niladari Division GOSL = Government of Sri Lanka GRO = Grassroots Organization HRDF = Human Resource Development Fund IDA International Development Association IFAD International Fund for Agricultural Development ILO = International Labor Organization IRDP = Integrated Rural Development Program JSP Janasaviya Program JTF = Janasaviya Trust Fund KfW Kreditanstalt fur Wiederafbau MIS = Management Information System MOF = Ministry of Finance MOH = Ministry of Health MPCS = Multi-Purpose Cooperative Societies MPIPA = Ministry of Plan Implementation and Parliamentary Affairs MPPI = Ministry of Policy Planning and Implementation MTRID = Ministry of Textile and Rural Industrial Development MYAS Ministry of Youth Affairs and Sports NDTF = National Development Trust Fund  FOR OFFICIAL USE ONLY NF = Nutrition Fund NGO/PO Non-Governmental Organization/Partner Organization NNCC = National Nutrition Coordination Committee NPD National Planning Department NSB = National Savings Bank NYSCO = National Youth Services Cooperatives PA Public Affairs Division PHRD Policy and Human Resource Development Grant PMU Poverty Monitoring Unit PNN Praja Naya Niymaka ("Loan Agents") PPU = Provincial Planning Unit PS Pradeshiya Sabha QLFS Quarterly Labor Force Survey RDA Recommended Daily Allowance RDD = Regional Development Division RDS = Rural Development Society REDS = Rural Enterprise Development Service RRDB = Regional Rural Development Banks RWF = Rural Works Fund SAL Structural Adjustment Lending SAR Staff Appraisal Report SEEDS = Sarvodaya Economic Enterprise Development Service SK Sahaya Kandayama (JSP "Support Team") SM = Social Mobilization TCCS = Thrift and Credit Cooperative Society Trust = Janasaviya Trust Fund'National Development Trust Fund UNDP = United National Development Program UNICEF = United Nations Children's Fund USAID = United States Agency for International Development WCNF Women and Child Nutrition Fund WCND = Women and Child Nutrition Division WERC = Women's Education and Research Center Vice President Mieko Nishimizu Country Director Roberto Bentjerodt Sector Leader Ralph Harbison Team Leader Jacob Bregman The Implementation Completion Report is based on the supervision and ICR mission that took place from September 1-22, 1997. The mission team was composed of Mmes/Messrs. Jacob Bregman (Team Leader, SASED) and consultants Harsha Aturupane (Economist and Poverty Specialist), David Pyle (Senior Nutrition Specialist), Henk Op Het Velt (Rural Works and Social Mobilization Specialist), Mallika Samaranayake (Participatory Development Specialist, SACSL), Julitta Rasiah (Financial Analyst, SACSL), Nandini Gunewardena (Gender and Social Mobilization Specialist) and Kamal Siblini (Monitoring and Evaluation Specialist). The report is also based on data collected during the joint GOSL-KfW-IDA Mid-Terrn Project Review mission (May 1995). This ICR was prepared by Jacob Bregman (Team Leader, LCSHD) and Harsha Aturupane (Economist and Social Poverty Specialist) with inputs from Nandini Gunewardena (Gender and Social Mobilization Specialist) and David Pyle (Senior Nutrition Specialist). Tables and graphics were compiled and edited by Kamal Siblini. The draft text was proofread and formatted by Julie-Anne Graitge (SASED). Comments from the GOSL, KfW and UNDP (Colombo Office) are included. Peer reviewers were Mmes. Soniya Carvalho (PRMPO) and Judith McGuire (LCSHD). This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  TABLE OF CONTENTS PREFACE EVALUATION SUMMARY Introduction ...................................................i Project Objectives ................................................... ii Implementation Arrangements .................................................. ii Implementation Experience and Results ................................................... ii Project Sustainability .................................................. vii Future Operations .................................................. vii Key Lessons Learned .................................................. viii PART I - PROJECT IMPLEMENTATION ASSESSMENT A. Introduction . B. Project Rationale and Objectives .3 C. Achievement of Objectives and Major Factors Affecting Performance .6 D. Project Sustainability .15 E. Bank Performance .16 F. Borrower Performance .18 G. - Assessment of Outcome .19 H. Future Operations .20 1. Key Lessons Learned .20 PART Ila - GOVERNMENT CONTRIBUTION A. Introduction .23 B. Project Objectives and Components .23 C. Problems Faced by the Project and Contributory Factors .26 D. Strengths and Achievements .27 E. Major Issues Encountered .28 F. Lessons to be Learned .29 G. Conclusion .29 PART Ilb - CO-FINANCIERS CONTRIBUTION A. Credit Component ......................................................... 32 PART III - STATISTICAL AND PROJECT DATA TABLES Table I Summary of Assessments ......................................................... 34 Table 2 Related IDA Loans/Credit ......................................................... 35 Table 3 Project Timetable ......................................................... 35 Table 4 Credit Disbursements: Cumulative Estimated and Actual .............................................. 36 Table 5 Key Performance Indicators by Component ......................................................... 41 Table 6 Studies Included in Project ......................................................... 43 Table 7.1 Project Costs ......................................................... 46 Table 7.2 Project Financing ......................................................... 47 Table 8 Status of Legal Covenants ......................................................... 48 Table 9 Bank Resources: Staff Inputs ......................................................... 50 Table O Bank Resources: Missions ......................................................... 51  ANNEXES Annex I Aide Memoire ..................................................... 52 Annex 2 Project Objectives ...................................................... 78 Annex 3 Monitoring and Evaluation ..................................................... 79 Annex 4 Targeting, Participation and Social Mobilization ..................................................... 82 Annex 5 Gender Outcomes ..................................................... 87 Annex 6 Institutional Development and the NGO Partners ..................................................... 91 Annex 7 Nutrition Interventions ..................................................... 93 Figure I NTDF Organizational Structure ..................................................... 99 Figure 2 Project Funds and SOE Flow: IDA, GOSL and NDTF ............................................. 100 Map No. 27509  SRI LANKA POVERTY ALLEVIATION PROJECT (Credit 2231-CE) IMPLEMENTATION COMPLETION REPORT PREFACE This Implementation Completion Report (ICR) for the Sri Lanka Poverty Alleviation project, for which Credit No. 2231-CE in the amount of US$57.5 million equivalent was approved on April 3, 1991, became effective on September 26, 1991 and was closed on December 31, 1997, after a one year extension in December 1996. The SAR IDA contribution was US$57.5 million. However, in December 1996, IDA canceled approximately US$10 million (SDR6.5 million). Cofinancing for the project's credit and micro-enterprise development component was provided by KfW from Germany (US$10 million). The SAR estimate of the Government of Sri Lanka contribution was US$17.5 million. Under the technical assistance component of the project, the UNDP provided US$2.5 million for technical assistance to the Trust. Although this was not an integrated part of the project (SAR, paras 5.45 and 6.4), the UJNDP grant disbursed about US$1 million during the project life. The project also benefited from a Japan PHRD Grant for studies and technical assistance on Gender and Women Migration Issues. The final disbursement date was April 21, 1998. There are still some outstanding disbursement problems related to the advances made to the Special Account. Extemal audit reports for 1991-1996 were all unqualified. The 1997 Audit Reports are still outstanding. Government of Sri Lanka US$17.5 KfW cofinancing US$10.0 IDA US$57.5 Total US$85.0 This ICR was prepared by Messrs. Jacob Bregman (LCSHD) and Harsha Aturupane (Economist Consultant). Substantial contributions were made by Ms. Nandini Gunewardena (Gender and Social Mobilization Consultant) and Mr. David Pyle (Senior Nutrition Consultant). Peer reviewers were Mmes. Soniya Carvalho (PRMPO) and Judith McGuire (LCSHD). The GOSL contribution to the ICR is included as part IIA and the KFW contribution as Part IIB and by commenting on the draft Part I. Comments from UNDP (Colombo Office) and NGOs are included in the main ICR text. ICR preparation began during the Bank's final supervision/completion mission, conducted from September 1- 22, 1997 and was completed in April 1998. The ICR is based on data in the project files, records maintained by the project implementation unit in Sri Lanka, information collected in the field by supervision teams, and beneficiary surveys conducted between December 1995 to September 1997. A comprehensive NGO-Beneficiary Assessment, jointly supervised by GOSL and IDA was completed from September 1997 to January 1998. This survey was executed by external consultants from the University of Colombo.  SRI LANKA POVERTY ALLEVIATION PROJECT (Credit 2231-CE) IMPLEMENTATION COMPLETION REPORT EVALUATION SUMMARY Introduction 1. The Poverty Alleviation Project (Credit 2231-CE) evaluated in this report was prepared by the Government of Sri Lanka (GOSL) and the World Bank between October 1988 and April 1990. The project became effective on September 26, 1991 with IDA credit of US$57.5 million, and was closed on December 31, 1997. In December 1996, IDA canceled approximately US$ 10 million (SDR6.5 million) of credit. As of June 10, 1998, the total disbursed amount of the IDA credit was US$38,068,536 (SDR26,390,073) or 83% of the original credit amount. There are still some outstanding disbursement problems related to the adivances made to the Special Account. IDA has requested a refund of US$100,461 representing unused funds remaining in the Special Account. The audit reports for 1997 are still outstanding. All audit reports for 1991-1996 were unqualified. The project was the first of its kind in Sri Lanka, based on a production-oriented, participatory poverty strategy in place of previous top-down, welfare-oriented poverty reduction approaches. The project was consistent with the objectives of the Sri Lanka Country Assistance Strategy (CAS) and the Government of Sri Lanka's own policy objectives for poverty reduction. 2. The project was prepared and implemented during a period of economic and political difficulties in Sri Lanka, which affected its performance and outcome. A civil war in the North Eastem Province increased in intensity in 1987. The President of the country was assassinated in 1993. In August 1994, the government in office changed after 17 years causing major changes in policies and administration that affected project implementation. The escalating civil war led to increased violence, including bomb attacks in Colombo, one of which damaged the offices of the project implementation agency in 1997. Adverse political factors hampered economic performance. From 1992-97 inflation was high, averaging 12% per year; the exchange rate depreciated by 29% and the government budget deficit averaged 9% of GDP per year. Such economic and political trends were likely to have had an adverse impact on the incidence, depth and severity of poverty.' However, child malnutrition fell from 38% in 1987 to 33% in 1995, and unemployment declined from 14% in 1985-90 to 10% in 1997. Project activities are likely to have contributed to these improvements in malnutrition and unemployment. Poverty estimates for overall Sri Lanka after 1990/91 are not yet available. However, there is evidence of worsening poverty in some areas of the country in recent years; See Aturupane, Rodrigo and Perera, Poverty Among Female Headed Households, (1996); Ministry of Plan Implementation, Poverty Monitoring Unit.  - II - Project Objectives 3. The major project objectives were to: i) increase income earning opportunities among the poor; and ii) improve the malnutrition status of pregnant and nursing mothers, and children under 3 years of age (raised at project mid-term to 5 years of age). To achieve these goals, the project included five sub-objectives: a) re-orient and expand existing institutional capacity to serve the poor, and create additional capacity; b) develop credit and other services for promoting self- employment and micro-enterprise development through group-based lending and entrepreneurial development; c) expand productive wage employment for the poor through technically, economically and socially viable rural works projects; d) develop programs for nutrition interventions for malnourished children and pregnant and nursing mothers; and e) create policy research and program formulation capacity within GOSL (Ministry of Policy Planning and Implementation - MPPI) to take greater account of poverty and unemployment issues in overall growth policies and public investment projects. In addition, the project aimed to emphasize gender concerns by promoting female participation in project activities and benefits. Implementation Arrangements 4. Administration of the Funds (CF, HRDF, RWF, NF) was the responsibility of the Janasaviya Trust Fund (JTF), renamed in 1995 the National Development Trust Fund (NDTF). The JTF (NDTF) was established as an apex agency as a result of collaboration with GOSL and IDA and with the technical assistance from UNDP. The Trust received and disbursed designated project funds to Partner Organizations and NGOs (NGOs/POs) according to criteria agreed with IDA. Thie establishment of the Employment and Poverty Policy Unit was the responsibility of the MPPI. The Trust was chartered in January 1991. Its charter specified its non-profit character and its authority to enter into relationships with foreign sources of funding that have received GOSL approval. Its objectives were employment generation and poverty alleviation. The Trust comprised three elements: (i) a Board of Trustees; (ii) an NGO advisory Board and (iii) an executive arm. It linked with the GOSL through the Ministry of Policy Planning and Implementation (in 1995 this Ministry was incorporated into the Ministry of Finances). A technical assistance team assisted the Board and a Managing Director of the Trust. The Managing Director and the Board were all appointed by the Government, although the NGO sector was represented as stipulated in the original agreements with IDA. However, in practice the Trust operated very much as a government body and did not attain the necessary flexibility and autonomy to become a true partner of the NGO partner organizations. The lack of autonomy and flexibility of the Trust proved to be one of the major weaknesses throughout the project life in developing strategies for poverty alleviation with the support and through NGOs. Implementation Experience and Results 5. The achievement of project objectives was partial and unsatisfactory. The overall objectives, expansion of employment opportunities and incomes among the poor, and the reduction of malnutrition among mothers and young children, were achieved to a limited extent. There were several reasons for the unsatisfactory project performance:  - iii - a) Political Factors: Political factors worked to the detriment of the project. The project's name in the early years, Janasaviya Trust Fund, caused it to be associated with the government's Janasaviya Program. This introduced a degree of political identification of JTF that hampered project implementation. Opponents of the government actively sought to hinder project operations, while some supporters of the government sought to use project funds to extend political patronage. The new government, which assumed office in August 1994, took time to review the project, and was ambiguous in its commitment to project activities as these were perceived as an arm of the previous government's agenda. b) Troubled GOSL-NGO Relationship: The relationship between GOSL and NGOs was weak throughout the project's life time. This weakened implementation given the large role envisaged for NGOs as partner organizations (POs) of NDTF in the delivery of project services. * In 1991, at project inception, the government in power had appointed a commission to investigate NGO activities. This alienated the NGO sector and made them unwilling to establish a partnership with NDTF which was perceived as a government institution. During project preparation the largest and most experienced NGO, Sarvodaya, had been expected to play a prominent role in project implementation. However, when the NGO commission was in progress, Sarvodaya was banned from participating in NDTF activities. These factors caused project disbursements to be initially delayed by more than one year. * After 17 years of opposition the new government had come into power in mid-1994 pledging to alleviate poverty through its flagship Samurdhi Program [see Box 2 in the main text]. Samurdhi was to be implemented by a cadre of workers recruited directly by the government. It only accorded a more limited role for NGOs/POs than was intended during project preparation. Thus there was no significant move towards adopting a more participatory approach as envisaged in the project. * After the new government assumed office, there were again allegations of political bias and corruption against NGOs/POs, especially those involved in the HRDF. It appeared that some NGOs might have used project funds to promote political election campaigns for the pre-1994 government. These allegations culminated in GOSL appointing another committee of inquiry in 1995. NDTF agreed with IDA that external auditors should be contracted (Coopers & Lybrand) and that all NGOs (about 140) having received funds through the Credit Fund would be audited. After a lengthy preparation process the external auditors produced at the end of 1995 a detailed audit report reviewing the 12 largest Credit NGOs (who together represented 87% of outstanding funds from the Credit Fund). This external auditor report, which was jointly accepted by IDA and the GOSL, concluded that (i) most NGOs needed to strengthen and improve their financial accounting procedures, (ii) for most NGOs accounting procedures were in accordance with Sri Lanka's required guidelines, but not necessarily with internationally agreed guidelines, (iii) many smaller NGOs needed substantial training and technical assistance to bring their accounting procedures in line with accepted standards, and (iv) procedures and eligibility criteria for the Credit Fund needed to be revised. * After the change of government in August 1994, doubts about NDTF's continued existence, which originated in the GOSL creation and strong support for the Samurdhi Program, created some uncertainty in the minds of NDTF staff and among NGOs/POs. This led to increased staff  - Iv - turnover, reduced personal incentives for investment in human capital development and caused work flow interruptions and cutbacks in the field. c) Lack of a strategy for institutional development and capacity building: The SAR recommendations for developing NGO/PO capacity and institutional strengthening were inadequately implemented. Although 25% of HRDF funds (about US$2.9 million) was to be allocated for strengthening the conceptual and management skills of NGOs/POs, only a negligible portion of this was actually spent for such activities. d) Limited results from international technical assistance: The IBRD was requested by the GOSL and UNDP to administrate the UNDP grant for tectnical assistance to the JTF (NDTF). Four long term international consultants were contracted under IBRD conditions by JTF (NDTF). However, because their contracts were with the Bank, the NGOs were given the impression that the consultants were working for IBRD. In fact they were selected by and working under responsibility of the JTF (NDTF). In monitoring and evaluation, the technical assistance was partly successful. However, NDTF was not satisfied with results in the area of HRD, management, micro-credit and internal staff training. In addition, one of the long term consultants, who was brought in the first year of project implementation (1991) to assist NDTF with Human Resource Development and Management improvements, joined some NGOs in opposing NDTF and failed to deliver good practices on partnerships. International consultants were recruited between 1993-95 to: (i) help improve strategies for credit and micro-enterprise development; (ii) design and develop NGO training programs; and (iii) design and implement NDTF's monitoring and evaluation activities. In early 1995 the GOSL decided not to renew the technical assistance contracts and requested the Bank to act accordingly. In the meantime, the Bank, UNDP and GOSL had reviewed the technical assistance financed under the UNDP grant in a joint supervision mission executed in late 1994. It was jointly decided that the administration of the Technical Assistance program by the -Bank and its results left much to be desired. This further weakened the already weak management within NDTF. The administration of the UNDP grant would be returned to the UNDP office in Colombo with the aim to transfer management of the remaining funds to the GOSL (NDTF) at a later stage. The UNDP grant had disbursed at that point about US$1 million. The transfer of management to the GOSL did not take place and UNDP canceled the remaining amount. e) Absence of synergy between trustfunds: The project was expected to develop synergy between the four trust funds in field implementation activities. It was an important assumption in the project design that communities and NGOs would progress to use all four Funds to escape poverty. However, it was not always possible for NGOs to develop all four TF activities, because NGOs were in many cases specialized. This was insufficiently recognized during the early implementation years. NGOs wanted to be selective and build capacity more slowly in new areas. The RWF was intended to be the first initiating activity and it was thought that this would help to activate communities to start savings schemes, mobilize the poorest in the communities, obtain training and use the Credit Fund for micro credit and employment generation activities. This proved to be a difficult task to achieve for several reasons: * Project interventions in one fund failed to lead smoothly to interventions in other funds. According to the SAR, social mobilization activities under the HRDF were expected to prepare beneficiaries for participation in the CMED, RWF and nutrition activities. Alternately, nutrition interventions which proved to be a good entrance into a community, were to lead to credit and rural works activities. However, such progression of activities from one fund to another was limited due to weak understanding within NDTF of the  importance of promoting synergy between activities of the four trust funds and poor coordination of efforts between NDTF trust funds and NGOs/POs; and * Uncoordinated sector specialization by NGOs/POs also weakened synergy effects between project components. For example, at peak project implementation in 1996, 55% of NGOs/POs were only working with one of NDTF's funds, while only 27% worked with all four funds. In addition, the lack of training to develop NGO/PO expertise prevented them from expanding and diversifying their activities. Weak monitoring and evaluation mechanisms during the first three projectyears: Monitoring and evaluation was weak throughout the life of the project. According to the SAR, a management information system (MIS) was to be developed in NDTF to facilitate decision- making and management. In practice this proved to be very difficult. Only in mid-1995, as a result of technical assistance (from IDA missions and international consultants) did an adequate monitoring system at three levels slowly emerge: (1) within NDTF organization, (2) between NGOs and NDTF, and (3) by NGOs at beneficiary levels. This was a complex task and required costly and dedicated work by NGOs. The results from these efforts did not show up but in the last year of project implementation and thus have not contributed significantly to the project. An exception was the monitoring mechanisms established early by the Nutrition Fund. This was very effective, and broke new ground. However, with the change of the Nutrition Director at the end of 1995 the results deteriorated again. Nevertheless, toward the end of project implementation an effective monitoring mechanism had emerged and it was possible (with technical assistance) to analyze the project's impact over the last three implementation years. Due to the initial delays in implementation at the start of the project life, disbursement during the last three project years represented the major part of disbursed project funds. 6. Credit and Micro-Enterprise Development (CMED) Fund. The objectives of the CMED Fund were only partially satisfied. The SAR expected that about 175,000 poor households would be able to increase their earnings from credit and entrepreneurial activities. However, only 117,000 micro-enterprises were established during the project life. Among these, only about 37% survived beyond three years of operation. Net employment creation was 40,000 jobs (approximately 23% of SAR targets). Beneficiary targeting was generally sound, with project selection criteria being satisfied in over 90% of cases. The majority of micro-enterprises did not develop capacity to expand production, improve product quality or diversify into new lines of production. This was partly the result of a lack of training in enterprise development. However, it was also partly the result of difficulties inherent in promoting micro-enterprises in a troubled economic and political climate. The SAR may have also over-estimated the demand for credit. The repayment rate of loans from NGOs/POs to NDTF was about 97%, while the repayment rate of loans from beneficiaries to NGOs was only about 83%, suggesting that NGOs/POs were utilizing its own funds, including the earnings and assistance from other sources, to repay NDTF credits. 7. Human Resource Development Fund (HRDF). The objectives of the HRDF were only partially met. Over 1.1 million low-income households were reached and mobilized, with beneficiaries organized into about 177,000 small groups. Beneficiary targeting was sound, with the majority of beneficiaries coming from households below the poverty line. The social mobilization efforts contributed to enhanced social capital in low income communities, one of the implicit objectives of all social fund projects. The participatory approaches used by NGOs/POs (as compared to previous top-down and paternalistic approaches toward community development)  - vi - resulted in improved bargaining power and self-reliance among the poor. But the HRDF was only partially able to fulfill its main aim of providing skills and competencies to enable individuals to establish successful micro-enterprises. Insufficient preparatory training in the selection and development of micro-enterprises appropriate to rural markets, consumer demand and village technologies led to limited success in micro-enterprise development. 8. Rural Works Fund (RWF). The objectives of the RWF were partially satisfied. Approximately 700,000 households benefited directly from about 7,765 RW projects. The roads, bridges and culverts constructed or rehabilitated lowered travel time and costs to schools, health facilities, post offices, banks and shops, improved access to raw materials and inputs for production, and enhanced access to markets to sell outputs. Irrigation facilities constructed under the project improved the productivity of agricultural activities and increased the availability of irritable land. Hence, the rural works projects constructed are likely to have had a positive effect on social welfare and poverty. However, project targets for employment generation and wage incomes were not met. Only about 3.5 million work days of employment, worth about SLR210 million (US$5 million) in real wages were generated through the rural works program, as compared to the SAR target of 10 million work days, worth SLR600 million (US$15 million) in real wages. The project's financial allocation to RW projects was completely utilized. Hence, the shortfall in the quantity and value of wage employment created may be partly attributed to cost-ineffective methods of project construction. It is also likely that the SAR targets for work days and wage income were over- optimistic. Also, by project closure, sound maintenance arrangements had not been established for rural works facilities. 9. Nutrition Fund (NF). The activities of the nutrition fund were the most successful. The project covered over one-fifth of the population within the conflict-free zones of Sri Lanka, and was operational in 18 of the 25 districts. At peak project operation (1995-96) about 68 NGOs/POs were involved in nutrition interventions in over 1,600 Grama Niladari Divisions (GNDs), with a beneficiary participation rate of 89% of the target population. Beneficiary targeting was sound, with benefits concentrated among the poorest members of village comnmunities. The outreach of the two approaches, the In-Depth Program and the Basic Program, was approximately 700,000 mothers and 2.7 million children, well in excess of the SAR target of 300,000 mothers and I million children. Program success can largely be attributed to innovative interventions focusing on behavioral factors rather than food insecurity. The reliance on participatory approaches succeeded in enhancing nutrition awareness and improving feeding practices, as compared to previous prescriptive approaches which had yielded insignificant results. However, government production and distribution of the nutrition supplementary food, Triposha, was unsatisfactory and had to be discontinued in 1996. Also, the long-term sustainability of improved nutrition behavior is not yet known. 10. Gender Objectives. The gender objectives of the project were partially met. The social mobilization efforts contributed to improving women's self-reliance and bargaining power. However, the entrepreneurship training provided to women was inadequate to enable them to increase incomes through micro-enterprise development. The 1997 NGO-Beneficiary Assessment revealed that only about one third of women participants who received credit were able to operate successful micro-enterprises. Successful credit beneficiaries tended to have family business connections. 11. The Employment and Poverty Policy Unit (EPPU) and the Poverty Monitoring Unit (PMU). The EPPU and PMU were partially successful in meeting their objective of strengthening  - vii - government capacity to develop national policies for poverty alleviation and employment generation. Several studies on poverty monitoring and policy research provided valuable information on critical factors related to labor market functioning and the interrelationships between poverty and unemployment. These were used in policy formulation by a number of government ministries and agencies. However, GOSL was not able to develop a mechanism through which policy research findings could feed consistently into policy formulation and resource allocation for poverty reduction. 12. Institutional Strengthening and Capacity Building. The main objective of expanding institutional capacity to serve the poor and creating additional capacity, was met only to a limited extent. Capacity building in NDTF was weak, partly due to its failure to transfer knowledge gained from international technical assistance recruited for this purpose. Project implementation was hampered by the inability of NDTF to provide adequate training to NGOs/POs. Better training would have strengthened the ability of NGOs/POs to implement activities especially in the HRD component. Consequently, NDTF was also unable to significantly assist NGOs/POs develop their program implementation capacities and skills. However, over the course of project implementation, communication and cooperation among NGOs/POs and between NGOs/POs and the government had improved considerably at the end of the project. Also, many NGOs/POs improved their management capabilities, expertise and field effectiveness, although largely through their own efforts. Project Sustainability 13. The prospects for project sustainability are limited. The experience gained through this project has apparently failed to convince the Government that the principle of publicly financed poverty alleviation activities managed by a flexible, semi-autonomous apex body which contracts NGOs to implement poverty programs is superior to that of direct interventions by specialized government agencies. GOSL has shifted its focus to more direct government implementation of poverty alleviation activities through its Samurdhi poverty alleviation program. 14. While NGOs continue to play a role under several GOSL programs, such as the integrated rural development program, the prospects for continued NGO involvement in poverty alleviation activities are however sound. The NGO-Beneficiary Assessment showed that NGOs have generally expanded their access to sources of funding from external sources (bilateral and private donors), and utilize these resources to sustain their involvement in micro-credit, social mobilization and nutrition programs for the poor. Future Operations 15. GOSL has expressed interest in a follow-up poverty project which would be supportive of and build on its Samurdhi program. While the Bank is committed to funding programs for poverty alleviation in Sri Lanka, the nature and scope of such programs are yet to be determined. GOSL has organized a consultative process, consisting of key government officials, donors and local consultants, to undertake analytical work on poverty related issues. The Bank is supporting this process through its economic sector work on poverty and support for an integrated survey and analysis. This work is expected to provide the foundation for a follow-up poverty project in the next fiscal year.  -viii - Key Lessons Learned 16. There are several important lessons to be drawn from this project. a) Implementing poverty alleviation activities through NGOs operating under an apex semi- government organization is an extremely complex and demanding task The managerial, administrative, research and technical skills and competencies required are formidable. Substantial lead time and resource investment would be required to develop institutional capacity to ensure that project implementing agencies and NGOs/POs function efficiently. b) The apex organization through which assistance is channeled to NGOs/POs needs to have a sound governance structure, and a high degree of autonomy from government. Project activities suffered because the NDTF was perceived as politically linked to the Janasaviya poverty alleviation program of the earlier government. Also, the NDTF was located under the Ministry of Finance and Planning and the Board of Management consisted mainly of public servants, so that administrative and financial procedures tended to be strict. It should recognized, however, that in spending public funds, certain rules have to be followed in order to achieve accountability and prevent misappropriation and waste. This was perceived as bureaucratic by some NGOs/POs. It is important that an apex organization be highly autonomous, so that it can play an effective facilitating role among NGOs/POs. It is also important that NGOs/POs be given considerable decision-making power in program implementation. However, this has to be achieved within the context of adequate financial accountability and transparency. c) The staff of an apex organization like NDTF needs to possess experience with and respect for NGOs. This was lacking in the project. NGOs encountered considerable problems with NDTF staff members because the latter lacked experience of, and respect for, NGO activity. For instance, during one period of project implementation NDTF refused to allow NGOs to utilize funds obtained from NDTF to pay NGO staff salaries, although these NGO staff members were engaged in implementing poverty alleviation activities under the NDTF trust funds. In contrast, there were periods when project components were facilitated by the presence of individual NDTF staff members who respected and sympathized with NGOs/POs activities. For example, the success of activities under the nutrition fund were partly due to the existence of a committed, well-informed Nutrition Fund management. Subsequently, after the management of the nutrition fund changed in late 1995, the quality of activities deteriorated. d) The economic rationale and technicalfeasibility ofproject components need to be analyzed and developed with care. The HRDF failed to provide the poor with adequate technical skills to undertake micro-enterprise activities successfully, as envisaged in the SAR. However, it is highly questionable whether successful entrepreneurship skills, including taking calculated risks, can be developed through training. Also, the demand for employment among the poor was mainly for wage employment, not micro-enterprise opportunities. Action to remove constraints to the growth of wage employment opportunities, especially in off-farm activities, may be a more effective strategy to assist the poor than the supply of credit to promote micro- enterprises.  - ix - e) Pilot projects, as a basisfor testing strategies and approaches, can play a central role in facilitating overallproject success. The project could have benefited greatly from a pilot phase at project inception which would have helped identify best strategies as well as field constraints. The project SAR recommended that implementation models be developed and that networking among NGOs/POs be supported. However, in practice there was little support for such activities. Piloting would have provided valuable experience and models, as well as the opportunity to identify possible shortcomings in design and field constraints early. J) A high quality and effective MIS, including regular beneficiary surveys, is essential to ensure efficient project implementation and achievement of objectives. The quality of project implementation would have been substantially strengthened if a high quality management information system (MIS) had been established early in the project life. A MIS should contain sound baseline survey data, realistic trends and targets to be achieved, and clearly specified input, process, output and impact indicators. g) Importance of exploiting synergy effects. Synergy between the four trust fund components, rural works, nutrition, human resource development and micro-credit, needed to be exploited to derive optimal social benefits from the project. This aim was partially achieved by some large NGOs/POs, with adequate capacity, which implemented all four project components: rural works, nutrition, human resource development and micro-credit. However, the aim was not achieved by other smaller NGOs/POs specializing in one or two activities. Such NGOs/POs should have been related by NDTF to other NGOs/POs operating in complementary areas so that synergy effects could have been exploited through their combined activities. However, although integration of complex poverty alleviation interventions is desirable, it is not always possible or realistic to expect a multi-dimensional approach from small rural NGOs. Unfortunately, the NDTF component trust funds operated relatively independently of each other, so that synergy between project components could not be effectively exploited. h) Importance ofproviding training to NGOs/POs in poverty alleviation activities Project implementation was hampered by the inability of NDTF to provide adequate training to NGOs/POs. Better training would have strengthened the ability of NGOs/POs to implement activities, especially in the HRD component. i) Most NGOs/POs' administrative andfinancial systems did not comply with Bank-required accountability and audit regulations. The complexity of this problem was insufficiently recognized during project preparation and appraisal. NDTF's original four operational divisions adopted compartmentalized approaches because of weaknesses in project design, political climate and the complexity of building institutional capacity. j) The Government's broad strategyfor poverty alleviation in Sri Lanka underwent a major shift in emphasis mid-way through the projects lifespan. Originally, under this project, the strategy of delivering assistance to the poor through NGOs/POs was given a prominent role by the Government. In this context, the NDTF was planned as an apex organization to provide assistance to NGOs/POs to implement poverty alleviation programs. After the new government assumed office in August 1994 the emphasis of the overall poverty alleviation strategy changed. The Samurdhi poverty program gave prominence to the strategy of delivering assistance 'to the poor directly through government channels, (the Samurdhi Niyamakas), rather than through NGOs/POs. NDTF ceased to play an important role in the new poverty alleviation strategy.  SRI LANKA POVERTY ALLEVIATION PROJECT (Credit 2231-CE) IMPLEMENTATION COMPLETION REPORT PART I: PROJECT IMPLEMENTATION ASSESSMENT A. INTRODUCTION 1. The Poverty Alleviation Project (Cr. 2231-CE) evaluated in this report was prepared by the Government of Sri Lanka (GOSL) and the World Bank between October 1988 and April 1990. The project was appraised in October 1990, negotiated in March 1991 and a credit of US$57.5 million approved by the Board in April 1991. The credit was signed in May 1991 and became effective on September 26, 1991. The project was the first of its kind in Sri Lanka, departing from previous welfare oriented approaches to poverty reduction and relying on a production-oriented strategy. The project was consistent with the objectives of the Sri Lanka CAS and the GOSL's own policy objectives for poverty reduction through broad-base economic growth supported by poverty alleviation programs. Figure 1. Major Poverty Indicators Used During Project Preparation 50% l 2 0 % _;a: J|llglli