Uganda - Program for Alleviation of Poverty and Social Cost of Adjustment (PAPSCA) Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 17055 PERFORMANCE AUDIT REPORT UGANDA UGANDA: Program for Alleviation of Poverty and Social Cost of Adjustment (PAPSCA) project (Credit 2088-UG) September 24, 1997 Operations Evaluation Department 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 Equivalents (annual averages) Currency Unit = Ugandan Shilling (U Sh) Year US$ equals USh 1988 106.1 1989 223.1 1990 428.9 1991 734.0 1992 1,133.8 1993 1,195.0 1994 979.4 1995 968.9 1996 1065.9 Abbreviations and Acronyms AIDS Acquired Immunodeficiency Syndrome DANIDA Danish International Development Agency ERP Economic Recovery Program GTZ Deutsche Gesellschaft ffir Technische Zusammenarbeit HIV Human immunodeficiency virus IA Implementing agency ICB International competitive bidding ICR Implementation completion report IDA International Development Association LIDDA Lira District Development Agency M&E Monitoring and evaluation NGO Nongovernmental organization NRM National Resistance Movement ODA Overseas Development Administration OED Operations Evaluation Department PAPSCA Program for Alleviation of Poverty and Social Costs of Adjustment PAR Performance audit report PCMU PAPSCA Coordination and Monitoring Unit PlU Project implementation unit SAR Staff appraisal report SDA Social dimensions of adjustment (a component of PAPSCA) SIDA Swedish International Development Agency SSA Sub-Saharan Africa UNDP United Nations Development Programme UNICEF United Nations International Emergency Children's Fund USAID United States Agency for International Development UVAB Uganda Veterans Assistance Program VAP Veterans Assistance Program VIP Ventilated improved pit latrine Fiscal Year Government of Uganda: July I-June 30 Director-General, Operations Evaluation Mr. Robert Picciotto Director, Operations Evaluation Department : Ms. Elizabeth McAllister Acting Manager, Sector and Thematic Evaluations Mr- Roger Slade Task Manager Mr. Robert van der Lugt FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S.A. Office of the Director-General Operations Evaluation September 24, 1997 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Uganda-Program for Alleviation of Poverty and Social Cost ofAdjustment (PAPSCA) project (Credit 2088-UG) Attached is the Performance Audit Report for the Program for Alleviation of Poverty and Social Cost of Adjustment (PAPSCA) project, prepared by the Operations Evaluation Department (OED). Credit 2088-UG for US$28 million was approved in FY90. Except for a small balance of US$83,000 (which was canceled), the project was fully disbursed and closed in September 1995 after a one-year extension. Cofinancing was not envisaged, but in 1991 Sweden provided US$3.1 million equivalent, managed in trust by the Bank, to fund additional project activities. The PAPSCA project aimed to address the urgent social concerns of Uganda's most vulnerable groups. Its approach was to be participatory, involving communities, nongovernmental organizations (NGOs), and the government. In the medium term, the project would also strengthen the government's capacity to identify and implement its own interventions for assisting these vulnerable groups. Project components covered rehabilitation of primary schools, support for orphans and widows, provision of health services, income-generating activities, construction of small-scale infrastructure and water and sanitation facilities, strengthening of the government's social planning capacities, and support for project administration. The project faced serious problems from the beginning. Project design did not include beneficiary participation, and the substantial subsequent efforts that were needed to motivate communities to participate in the project delayed implementation. No project launch workshop was ever held, a shortcoming regarded as serious by all stakeholders. Shortfalls in government funding left some components entirely dependent on IDA funding. As these prefinancing arrears had not been refunded, IDA was forced to suspend disbursements in mid-1993. No adequate financial system was ever designed or implemented, and serious accounting deficiencies were noted throughout project implementation. Procurement and disbursement procedures were very cumbersome and proved ill adapted to a community development project. The project management unit performed poorly; it failed to become an efficient coordination and monitoring unit and was dissolved when the project ended. Despite these very serious limitations, implementation at the grassroots level in the NGO-executed components was by and large satisfactory, especially orphan support activities. The project produced a few successes, but in general failed to produce substantial results. The physical achievements of the major education rehabilitation component were dismal. The orphan support programs and small-scale infrastructure programs were generally implemented satisfactorily, but without 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. -2- further funding they will be difficult to sustain. Income-generating activities suffered because loans were not repaid, so that the envisaged revolving funds did not materialize. The social planning component was ineffectively managed. Methodological differences hampered the usefulness of the results, and there is little evidence that the studies influenced government policy. The Operations Evaluation Department (OED) and the region both recognize the difficulties in project performance but disagree slightly on the ratings. The implementation completion report (ICR) rated project performance as marginally satisfactory. Given the abysmal implementation record, the fact that four project components accounting for 58 percent of project resources were unsatisfactory, and the limited impact on institution building and poverty alleviation, OED rates outcome as marginally unsatisfactory. OED agrees with the earlier ICR ratings for sustainability (uncertain) and institutional development (modest) but rates both borrower and Bank performance as unsatisfactory. Strong government support weakened considerably and was nonexistent at the project's end. Counterfunding remained a problem throughout, and cumbersome bureaucratic procedures hampered project implementation. Bank project design was rigid and overly complex. It did not include beneficiary participation or establish adequate financial management and control systems. Supervision was inadequate; task managers changed frequently, and the emphasis was on project minutiae rather than development impact. In 1992, a Veterans Assistance Program (VAP) was approved to help demobilize a substantial part of Uganda's army. The Bank agreed to be the leader of the multidonor consortium that funded this effort (ultimately US$45 million). The Bank and government also agreed that funds from Credit 2088- UG could be diverted to finance this new program (ultimately US$2.5 million). The Bank administered the VAP under the PAPSCA project umbrella. The VAP was well managed (much better than PAPSCA) and successfully implemented, although reintegration carried substantial social costs, and some remobilization has occurred. The audit recommends that the Bank evaluate this activity in a separate ICR. The audit confirms the benefits of stakeholder participation in community-based development interventions for the poor and demonstrates the value of NGO involvement in implementation. However, it also highlights the importance of appropriate mechanisms for selecting NGOs, decentralization of procedural authority, and adequate beneficiary participation from the beginning. The audit also underscores the need for more flexible procurement and disbursement procedures adapted to community development projects. Attachment Robert Picciotto by Elizabeth McAllister Contents Ratings and Responsibilities..........................3...... .............3 Preface .............................................................. 5 1. Background .........................................................7 The Setting ...................................................... 7 Poverty in Uganda ........................................ ......... 8 AIDS ........... .................................................. 8 2. Project Design and Implementation...................................... 10 Project Origin and Design ............................................... 10 Project Objectives and Components ...................2.... ............12 Implementation ......................................... ......... 14 3. Project Component Details..............................................18 Primary School Rehabilitation....................... .................18 Orphan and Widow Support ......................... ................19 Health Services..................................................21 Income-Generating Activities........................ ................21 Small-Scale Infrastructure Rehabilitation ................. ...............22 Strengthening Social Planning....................... .................23 Project Administration.............................................24 4. Ratings...........................................................25 Outcome.......................................................25 Sustainability...................................................25 Institutional Development..........................................26 Borrower Performance.......................... ...................26 Bank Performance................................................27 5. Veterans Assistance Program...........................................28 VAP's Objectives and Implementation..................................28 Results of the Demobilization Effort....................................29 Present and Prospective VAP Evaluation.................................30 6. PAPSCA's Development Approach.......................................31 Using NGOs as Implementing Agencies.................................31 Alternatives to the PAPSCA Model.................................... 34 Annex A, Basic Data Sheet............................................... 37 Annex B, List of Main Documents......................................... 41 Map This report was prepared Robert van der Lugt who field audited the project in November/December 1996. Peer reviewers were Christopher Gibbs, Soniya Carvalho, and Ines Girsback. Emily Chalmers edited the audit. Scoff Dineen provided administrative assistance.  3 Ratings and Responsibilities Performance Ratings Audit OED E VM' ICR Outcome marginally unsatisfactory marginally unsatisfactory marginally satisfactory Sustainability uncertain uncertain uncertain Institutional Impact modest moderate partial Borrower Performance unsatisfactory satisfactory satisfactory Bank Performance unsatisfactory unsatisfactory satisfactory 1. Ratings given by OED during the review of the ICR and reported in the Evaluative Memorandum dated December 31, 1996. Key Project Responsibilities Task Manager Division Chief Director Appraisal Gerard Byam K.Y. Amoake Callisto Madavo Midterm Review Pat Walker Jacob van Lutsenburg Maas Francis Colago Completion Gita Gopal Jacob van Lutsenburg Maas James Adams 4 5 Preface This is the Performance Audit Report (PAR) of the Uganda Program for Alleviation of Poverty and Social Cost of Adjustment (PAPSCA) project. Credit 2088-UG in the amount of US$28 million equivalent supported the project. The Credit was approved on February 1, 1990, and closed September 30, 1995, after a one-year extension. When final disbursements took place on September 11, 1996, a balance of US$83,000 was canceled. Cofinancing was not envisaged as part of the project, but the Swedish International Development Agency (SIDA) subsequently made additional donor funding available. The PAR is based on document and file searches, discussions with knowledgeable officials from the government, Bank, other donor agencies, and nongovernmental organizations (NGOs), and a mission in November/December 1996 that included two focus group discussions with stakeholders (NGOs and donors) plus a limited number of field visits. Two other sources provided valuable insights: the government's in-depth final evaluation report prepared as part of the completion process, and a participatory evaluation commissioned with funds from the Africa Region's systematic client consultation fund. The cooperation and assistance of government officials, donor representatives, NGOs, the Bank's resident mission staff, and the other stakeholders and beneficiaries are gratefully acknowledged. The implementation completion report (ICR) provides a satisfactory account of project implementation, despite some weaknesses in detail. The borrower's executive summary of its final evaluation is attached to the ICR. In general, this PAR concurs with the ICR findings, but not with all its ratings. The audit explores in greater detail the involvement of NGOs and alternative development approaches. The audit also presents details on the Veterans Assistance Program (VAP), a separate program established in 1992, financed by many donors but also for a minor part from Credit 2088-UG funds. The draft PAR was sent to the government for comments on June 10, 1997. Despite an extension of the due date, no comments were received.  7 1. Background 1.1 Uganda launched the Program for Alleviation of Poverty and Social Cost of Adjustment (PAPSCA) in 1989 with high expectations of having a serious impact on poverty. But implementation soon turned into a nightmare. The many components and complex implementation arrangements, combined with micromanagement on a large scale, resulted in bureaucratic chaos. Ultimately, the project failed to appreciably improve living conditions for the country's poorest citizens, and both the government and donors abandoned PAPSCA. What went wrong with this well-intentioned program? This audit examines that question and suggests alternative approaches for similar development operations. The Setting 1.2 At independence, Uganda was known as "the pearl of Africa."] With one of the strongest, most promising economies in Sub-Saharan Africa, the country was the envy of its neighbors. Favored with a good climate and fertile soil, the agricultural sector was self-sufficient and generated considerable foreign exchange. The industrial sector, though small, supplied the economy with basic inputs and consumer goods, and its exports of textiles and copper also brought in needed foreign exchange. Uganda's transport system and health services were regarded as among the best in Sub-Saharan Africa. School enrollments were low, but the quality of education was high. 1.3 The 15 years of political unrest that started in 1970 radically reversed all these achievements. Skilled workers fled the country, the parastatal sector grew bloated, and professional standards within the public service eroded. The breakup of the East African Community in 1977 and the destruction and looting that occurred during the 1978-79 war accelerated Uganda's economic decline. 1.4 When the National Resistance Movement (NRM) assumed power in Kampala in January 1986, it inherited a nation tom by ethnic and religious conflicts and an economy shattered by years of civil war, political instability, and physical insecurity. Many of the nation's major trunk roads and its vehicle fleet had been destroyed, and industrial enterprises lay abandoned. Even the remarkably resilient agricultural sector had been disrupted as farmers fled in search of refuge. The war and lack of maintenance also devastated Uganda's once impressive social infrastructure, and the spread of acquired immunodeficiency syndrome (AIDS) aggravated the already poor health situation. Educational facilities deteriorated, and child illiteracy rates climbed. 1.5 The years of political instability and economic decline also severely disrupted the civil service. While the number of civil servants had increased substantially, their real per capita wages had declined precipitously. The result was a civil service in which personal survival took precedent over public service and the capacity to perform was severely constrained. 1.6 The NRM's first priorities were to restore law and order and civil liberties, overcome the remaining insurgencies in the north and east, and begin the difficult task of rebuilding the nation. 1. For more on the history of Uganda, see Thomas P. Ofcansky, Uganda: Tarnished Pearl ofAfrica (Boulder: Westview Press, 1996). 8 In 1987, with donor assistance, the new government developed a comprehensive program of reforms under the Economic Recovery Plan (ERP). 1.7 The economic reforms were accompanied by major education and health projects funded with the assistance of the International Development Association (IDA) and other donors. But these projects did not address the immediate needs of many victims of the civil unrest and economic decline. PAPSCA was developed to meet those needs. Poverty in Uganda 1.8 Uganda's economy has prospered since the present government came to power in 1986. Economic growth has averaged about 5.4 percent, largely a result of bringing land and capital assets back into production. But the poor have yet to feel the beneficial effects. With a per capita income of less than US$240, Uganda remains one of the world's poorest countries. Social indicators have shown only minimal improvements since 1986 and will remain below those of the early 1970s until the next century.2 A Makerere University research report confirms that despite impressive economic growth, the number of poor people has not declined.3 According to the Bank's 1995 economic report, the poor are predominantly rural; the east and the north are the poorest regions; and the urban-rural gap has widened slightly.4 AIDS 1.9 Visitors to some of Uganda's districts (such as Masaka and Rakai) cannot but help notice the many graves, the empty houses, and the orphans (who account for 50 percent of students in some primary schools). These are the emotional reminders of the way that AIDS has affected the social fabric of Ugandan society-and a warning about the consequences for Uganda's future economic, social, and human development. 1.10 Uganda recognized AIDS as an epidemic in the country only in 1984, although the disease had been spreading since the mid-1970s. In 1993, an estimated 1.3 million adult Ugandans, or about 15 percent of the adult population, were infected with HIV.5 More than 100,00 people now die from AIDS each year. The AIDS epidemic has caused the standard quality of life indexes to deteriorate and has affected the availability of labor and labor productivity. It has also hindered capital formation, as scarce resources are being diverted to the health sector to combat the epidemic. So many people have died that rural communities no 2. The Bank's 1995 country assistance strategy placed life expectancy at 47 for men and 50 for women, among the lowest in the world. Uganda's crude death rate (20 per 1,000) is considerably above the average for Sub-Saharan Africa (SSA). School enrollment ratios are low, especially for girls. The primary enrollment ratio is 71 percent (about average for SSA), but the secondary enrollment ratio (about 10 percent) is far below the SSA average. It is especially low for girls (7 percent). Primary cohort survival is low at only 32 percent. 3. Fred Opio, "The Impact of Structural Adjustment Programme on Poverty and Income Distribution in Uganda," Draft Report (Kampala: Economic Policy Research Centre, 1996). 4. World Bank, "Uganda: The Challenge of Growth and Poverty Reduction," World Bank Country Study (Washington, D.C., 1996). 5. Jill Armstrong, "Uganda's AIDS Crisis: Its Implications for Development," World Bank Discussion Paper 298 (Washington, D.C., 1995). 9 longer have the traditional extended family systems that supported orphans and widows, leaving large numbers of these vulnerable groups destitute. 10 2. Project Design and Implementation Project Origin and Design 2.1 In response to concerns expressed at the 1988 Uganda consultative group meeting, PAPSCA was designed with 19 components and a total cost of US$106 million. It would be implemented by nongovernmental organizations (NGOs) and project implementation units 6 (PIUs) within government agencies. Preparation of the Bank project took place in early 1989, and the Bank appraised it in June 1989. The time from start of preparation to Board approval was less than 15 months, but the rapid pace of preparation may have affected quality at entry. For a chronology of project design and implementation see figure 2.1 2.2 Except for the strategy of using NGOs as implementing agencies, project design was based on the engineering approach the Bank traditionally used for infrastructure projects. This "top-down" approach proved inappropriate for a community development project. The financial arrangements were overly complex (there were more than 30 disbursement categories), and as the ICR points out, appropriate financial management and control systems were lacking. Planners designed a new project coordination and management unit without fully analyzing its scope and role and did not involve the required expertise. Project design also did not include an appropriate monitoring and evaluation (M&E) system.7 2.3 There was no pilot project. Although the Bank approved a project preparation Facility in late 1989, the facility assisted only in project start-up. Project planners assumed that communities would participate in project implementation (mostly through in-kind contributions) but did not test this assumption. Communities were not involved in preparation. As the participatory evaluation notes, "The community of beneficiaries did not adequately contribute to the identification of their most pressing concerns."8 Implementing agencies had to spend considerable time and resources motivating communities to participate, and implementation plans had to be modified considerably to conform to reality on the ground. 6. These are jointly referred to as implementing agencies.(IAs). 7. The Staff Appraisal Report (SAR) refers only to semi-annual and completion reporting requirements and does not include key performance indicators. Project M&E became a concern during the midterm review, but little was done about it. Completion reports on individual components, which varied widely in scope, content, and substance, contained little about impact. 8. Midland Consulting Group, "A Draft Report on the Participatory Evaluation for PAPSCA" (Kampala, 1996). 11 Figure 2. 1: PAPSCA CHRONOLOGY Consultative group October 88 meeting 1989.. February 89 Preparation June 89 Appraisal 990 02/02/90 Approval 02/08/90 Signing (Cr. 2088-UG) 06/29/90 Effectiveness i 1991 SIDA provides October 91 US$3.1 million Veterans Assistance March 92 Program included September 92 Midterm review 19941 09/30/94 Original closing date 09/30/95 Closing date ICR mission October 95 Participatory April 96 Evaluation ICR distributed June 96 09/11/96 Final disbursement OED audit mission November 96 PAR issued June 97 _ 12 Project Objectives and Components 2.4 Although it provides detailed descriptions of the components, the SAR presents PAPSCA's overall objectives in rather general and ambitious terminology. PAPSCA was to address the urgent social concerns of Uganda's most vulnerable groups including widows, orphans, and war veterans. In the medium term, the project also aimed to strengthen the government's capacity to identify, formulate, and maintain interventions for assisting these groups (SAR para. 3.03). 2.5 As designed, the Bank-financed project assisted in the implementation of only six of the 19 original PAPSCA program components.9 These six components were the high-priority items on the PAPSCA agenda, providing maximum coverage of vulnerable groups. The components aimed to rehabilitate primary schools, support orphans and widows, provide health services, foster income-generating activities, provide small-scale infrastructure and water and sanitation facilities, strengthen the government's social planning capacities, and support project administration (table 2.1). 9. The SAR (para. 3.03) lists six components, one of which was subdivided into three, making eight components. 13 Table 2.1: PAPSCA Project Components Implementation Type of Share Donor Project Component Agency Agency (%/)a) Agency Evaluated 1. Primary School Rehabilitation in 12 MOE/PIUc) GOVd) 38% IDA Ee) Districts 2. Orphan Support in Gulu, Masaka/Rakai World Vision NGO 15% IDA E 3. Widow Support in Luwero War Widows NGO 4% IDA Foundation 4. Health and Income in Masindi World Vision NGO 4% IDA E 5. Social Dimension of Adjustment MOP Statistics GOV 15% IDA Department' 6. Small-Scale Infrastructure in Kamuli Action Aid NGO 13% IDA E 7. Water and Sanitation in Rubaga, Kampala City GOV 9% IDA E Kampala Council /PIU 8. PAPSCA Coordination and Monitoring PCMU GOV 2% IDA Unit (PCMU) 9. Keep Kampala City Clean CISRg) NGO (0.7) SIDA 10. Orphan Support in Rukungiri Church of NGO (0.4) SIDA Uganda, North Kigezi Diocese 11. Widow and Orphan Support in Lira LIDDAh) LOC (0.4) SIDA E 12. Women in Agriculture in Apac Apac DDFI) LOC (0.4) SIDA 13. Small-Scale Enterprises in 4 Districts World NGO (1.3) SIDA Learing a. For IDA expressed as percent of project costs; for SIDA expressed in original allocation in US$ equivalent. b. These components were evaluated in the 1996 participatory evaluation by Midland Consultants. c. Ministry of Education, Project Implementation Unit. d. GOV stands for government agency; NGO is a nongovernmental organization e. The evaluation was done only in 50 percent of the concerned districts. f. Ministry of Planning. g. Committee for International Self-Reliance. h. Lira District Development Agency. i. District Development Foundation. j. In cooperation with four small local financial intermediaries. 14 2.6 At a cost of US$37 million (with an IDA contribution of US$28 million), these components made up about 30 percent of the total PAPSCA program.10 The parallel funding other donors had been expected to provide for the remaining components did not materialize, with adverse effects.II However, in October 1991 the Swedish International Development Agency (SIDA) provided about US$3.1 equivalent million to finance five more minor components. The SIDA-financed components would assist widows and orphans, improve small- scale enterprises-including those run by women-create jobs, and improve water and sanitary facilities in peri-urban areas of Kampala (table 2.1). 2.7 In 1992, the Veterans Assistance Program (VAP) was launched. As PAPSCA project disbursements had not progressed as expected, the Bank and government decided to divert some Credit 2088 funds to this new program.12 Subsequently, other donors13 provided funding for the VAP under IDA-administered trust fund agreements. The VAP was not part of PAPSCA but was placed under the PAPSCA umbrella because the Bank procedurally was prevented from administering the program otherwise (chapter 5). Implementation 2.8 The Project Coordination and Monitoring Unit (PCMU), an autonomous unit set up in the Ministry of Finance and Economic Planning, provided implementation coordination. The PCMU's main functions were to assist the implementing agencies with budgeting, project planning and disbursement of funds and accounting for their use, procurement of goods and services, and provide M&E. An interministerial policy steering committee chaired by a senior official of the Ministry of Finance and Economic Planning was to oversee and guide the PCMU's activities. 2.9 From the beginning, the project encountered substantial problems and delays (box 2.1). While the project was signed within a week of Board approval, declaring effectiveness required another five months (until the end of June 1990) and was made possible only through temporary appointments in the PCMU. The PCMU did not become fully operational until six months after effectiveness. Disbursements were also subject to specific conditions but, by December 1990, conditionality for six disbursement categories was still not fulfilled. Little start-up support was provided and, in a major oversight, no project launch workshop was ever organized.14 10. The government and NGOs were expected to contribute US$6.8 million and US$2.2 million, respectively, to project financing. 11. For example, the educational materials needed to improve the quality of the rehabilitated schools were never funded (para. 3.3). 12. Initially, about US$1.5 million, and later another US$1 million. 13. DANIDA, SIDA, the United States Agency for International Development (USAID), the British Overseas Development Administration (ODA), and the Netherlands. 14. Some disbursement and procurement workshops did take place, but much later in the project cycle. The ICR, the government's completion report, NGOs, and the participatory evaluation all mention this shortcoming. 15 Box 2.1: Issues Affecting Start-Up and Subsequent Implementation The principal issues that affected start-up included the following: * unfamiliarity with Bank procedures in the PCMU and implementing agencies * slow, time-consuming disbursement mechanisms for government funds - * delays in finalizing agreements between implementing agencies and the government * NGOs' need for training and start-up funds * civil unrest in some districts * slow processing of bidding documents * delays in the processing of documentation for SIDA financing * delays in establishing PIUs * delays in appointing the social policy adviser * frequent turnover in task managers during the first 18 months of operation Although the Bank reports that many of these issues had been resolved by 1992, subsequent issues that affected project implementation emerged, including: * a lack of counterpart funds * delays in processing exemptions from duties and tax * slow procedures in the Uganda Central Tender Board * staff shortages in the PCMU * improperly kept accounts and cash records * an internal government audit that reported irregularities in the accounting procedures of the PCMU * delays in auditing accounts * the government's slow processing of requests to change the disbursement schedules * the lack of a policy adviser for the SDA component * lack of transport for some NGOs that hampered field activities 2.10 The SIDA grant became effective in July 1991. However, disbursements did not commence until February 1992 because of difficulties identifying suitable NGOs to implement the components and delays in finalizing legal agreements between NGOs and the government. Once they started, the disbursements progressed well until mid-1992, when SIDA diverted 5 million krona of the PAPSCA grant to the VAP. This funding change brought the five SIDA components to a halt.15 2.11 Financial matters were problematic throughout. No adequate financial system was ever designed or implemented. Serious accounting deficiencies were noted throughout project 15. The resident mission attempted to find other donor funds to fill the gap, but without any success. 16 implementation.16 The PCMU was continuously reminded of these weaknesses, and even close to the project's end the Bank funded technical assistance to strengthen the accounting system. In contrast, the implementing agencies provided accurate accounting with only minor shortcomings. Audits were always late, partly due to the cumbersome clearance procedures.1 At the time of the OED evaluation, the final PAPSCA audits had still not been completed.18 Counterpart funding also remained problematic throughout the project. The Bank provided prefinancing from the special account because of serious shortfalls in counterpart funding early in the project. While helpful at the time, this gesture later caused a funding crisis when the IDA was forced to suspend disbursements in mid-1993 until these arrears had been refunded. 2.12 Procurement presented major headaches as well. Several key imports, such as vehicles, cement, and iron roofing, were subject to prolonged international competitive bidding (ICB) procedures. Project planners paid little attention to and made no allowances for the logistics of distribution to districts and subsequently to many dispersed sites. When this became apparent during implementation, procurement of additional means of transport was authorized to alleviate these logistical constraints. But cumbersome procurement procedures and delays in obtaining tax exemptions held up the arrival of the needed equipment and limited its usefulness.19 In an extreme case, four new motorcycles were delivered to an NGO only when the OED audit mission was in Uganda. They were destined for the supervision of a component that had run out of money two years earlier. 2.13 Some procurement resulted in serious protests and challenges, such as the award of a contract for iron roofing sheets in May 1991. At least three bidders protested, but the Bank dismissed the protests on the grounds that the contract had been approved, and the decision could not be reversed. The Bank promised to improve procedures in the future, a promise that one unsuccessful bidder, in a letter to the Bank, called a "cop-out." Local procedures were cumbersome as well. Subsequent procurement courses and the establishment of a PAPSCA procurement committee lightened these problems, but not enough. It is clear that procurement procedures were not well adapted to this project. As one NGO wrote, "There is a clear contradiction between an emergency quick-relief assistance program and ICB." Experience with this project shows that procurement procedures for community development projects need 16. A Bank letter (August 5, 1993) noted that the deficiencies included a lack ofpayment vouchers; no procurement records for some goods; no records for expenses at some training workshops; incorrect booking of consultant fees; payment of invoices without checking against received goods; inadequate controls on fuel consumption; improper use of motor vehicles for private activities; cash advances for unaccounted expenses; inaccurate statements in payroll (no allowances for mandatory salary deductions); poor reconciliation of the PAPSCA Task Force Account at the Bank of Uganda; VAP payments that cannot be accounted for; and an absence of audited statements of expenses on funds spent on United Nations Development Program (UNDP) managed purchases. 17. The PIUs/NGOs were required to sign off on the component audit accounts. But this procedure became unworkable with the final PCMU accounts, which legally had to be signed by two PCMU officers. However the PCMU no longer existed and appropriate officials had retired or died. As a result, this bureaucratic impasse still prevailed during the evaluation. 18. But the auditors indicated to the OED mission that they were unlikely to issue a qualified opinion on the PAPSCA accounts, except for one component. 19. Goods for this project were exempted from duties and taxes. Yet obtaining the necessary certificates could take so long that storage and demurrage charges often exceeded the tax benefit. 17 serious reconsideration. Procedures and regulations should as far as possible facilitate and not impede community involvement.20 2.14 The many problems involved the task managers in project minutiae, and most supervision attention focused on implementation problems at the expense of development impact.21 The constant problems undermined Bank and government support for the project concept. As noted in the minutes of a meeting of the PCMU and implementing agencies, "PAPSCA was not well considered by government as a program and was not well designed." 2.15 Despite the serious limitations in overall project management, implementation at the grassroots level-especially in most of the NGO components-was by and large satisfactory. In addition, the project significantly expanded the institutional capacity of NGOs and helped community groups plan, implement, and operate development interventions. 20. For discussion of this issue, see, for example, Gita Gopal and Marc Alexandre, World Bank-Financed Projects with Community Participation-Procurement and Disbursement Issues, Discussion Paper 265 (Washington, D.C.: World Bank, 1994) and Gita Gopal, "Procurement and Disbursement Manual for Projects with Community Participation," World Bank Discussion Paper 312 (Washington, D.C., 1995). 21. For example, in October 1990 a direct payment request for US$4,500 for consultancy services had been agreed on and approved. But paperwork, other bureaucratic requirements, and the fact that the check was sent to the wrong bank added about six months to the transaction. In another case, the task manager was asked to approve payment of one month's salary in lieu of foregone leave for a PCMU staff member. 18 3. Project Component Details 3.1 The project's 13 components, which were executed separately (table 2.1), fall into six categories (para. 2.5): * primary school rehabilitation * orphan and widow support * health services * income-generating activities * strengthening social planning * project administration Primary School Rehabilitation 3.2 This component was planned as a model of school rehabilitation and maintenance that could be replicated countrywide. It aimed to raise the quality of primary education in 12 of the poorer districts by rehabilitating an estimated 4,200 primary school classrooms. It would involve community self-help through contributions of labor and construction materials. 3.3 The component had major design flaws. First, there were few classrooms to rehabilitate. Classrooms needed to be constructed, not improved, and educational materials had to be provided. (The component that would have provided these materials was never funded; see para. 2.6). Second, communities played no role in selecting sites for schools and were not consulted on the question of self-help. Because the government had traditionally provided schools, mobilizing communities to build their own was difficult. "Self-help" required communities to manufacture brick, collect stones and sand, and provide labor. The project provided sheets of iron, roofing timber, and cement. The project also paid the artisans, but the payments were often late. Procurement problems caused long delays. By the time the goods arrived, local materials had deteriorated and communities had become demoralized. Logistics posed another problem. Planners failed to take into account the distance to some sites, making distribution difficult (para. 2.12). The lack of transport, poor accessibility, and long distances also hampered supervision. 3.4 The component achieved very little. Although statistics vary, at the time of project closing no more than 15 percent of the classrooms had been completed,22 30 percent-according to the PCMU final evaluation report23 -had not even been begun, and the rest were under 22. The ICR estimate was 10 percent. However, some reports argued that once a classroom was roofed, it should be regarded as finished, raising the completion ratio to 40 percent. But these roofed classrooms lack concrete floors and walls and can hardly be considered finished. 23. As noted, exact figures vary, and the ICR has used different information. However, whatever the numbers, there is no doubt that the primary school rehabilitation component was substantially incomplete at credit closing. 19 construction.24 The completed classrooms often lack adequate furniture as well as teaching materials. Enrollments in the new schools reportedly are higher than before (partly owing to transfers from the other older schools), but the participatory evaluation found no proof of progress in educational results.25 Two other projects-the Northern Uganda Reconstruction Project and the Primary Education and Teacher Development Project-are attempting to complete the planned schools.26 3.5 The outcome of this component was unsatisfactory. As one education official put it, "[it] was a disaster." The few completed classrooms are likely to be maintained, but completion of the remainder is likely only with funding from other sources. 3.6 Physical achievements were dismal and should be regarded as highly unsatisfactory. There were only a few positive achievements. As regional staff have pointed out, some capacity building has taken place at the community level and among local education officials. Committees for the project schools have been formed and, according to the participatory evaluation, communities are determined to sustain the schools constructed under the project. And it is true that even incomplete facilities are better than sitting under a tree. Orphan and Widow Support 3.7 Civil war and AIDS have created so many orphans and widows in Uganda that many communities can no longer continue the tradition of supporting them through the traditional extended family system. The orphan and widow support programs were developed in this context. The orphan programs aimed to provide an alternative to orphanages by creating an enabling environment for families and communities to care for the growing numbers of children without parents. The programs include basic education and health support for young orphans, income-generating activities for families, outreach to sensitize communities to and support them in caring for orphans, and vocational training for older orphans. Support for war widows included primarily health services and income-generating activities. The project funded several programs for these groups. The orphan support programs were satisfactory but demonstrated that such support often requires large overhead and strong organizational structures. When funding ceased, these programs tended to collapse, with drastic results: orphans dropped out of school, vocational training stopped, and income-generating activities suffered. The war widow programs were far less satisfactory, owing to the limited results of the programs in relation to need. 3.8 The World Vision orphan program. This program, implemented in the Gulu, Masaka, and Rakai districts, was the largest project orphan support initiative. World Vision created a 24. This figure raises the question of what happened to the materials already procured under this component. Reportedly most have been delivered to the school districts, where they may still be stocked. They may also have been diverted to alternative uses, since PAPSCA no longer exists. Some of the local officials interviewed are convinced that these materials are no longer available for their schools. 25. It notes that there was no corresponding improvement in primary school learning exams, but this seems logical: educational results do not depend on buildings, but on the quality of teaching and teaching materials. 26. However, in a letter to the government dated February 6, 1996, the Bank detailed the information it would need to approve financing: the number of incomplete classrooms, their exact locations, the quantity and type of construction materials already available in the district (or with the PIU), and a detailed estimate (by type and costs) of the additional materials, works, and services required, on a school-by-school basis. While this requirement seems reasonable in terms of accountability, it also seems somewhat stringent for community-based primary school construction. 20 substantial organizational structure, financed partly from its own funds, with national and district headquarters, country-level coordinators, and community development workers committees. It also maintained a sizable staff to manage the program (the question is, could similar results have been achieved with lower overheads?). As table 3.1 shows, the results of the program have been impressive. The component met many of its targets, and World Vision has drawn up a five-year follow-up plan. However, with its own funding it can continue its activities only at a much reduced level. Table 3.1: Key Achievements of the World Vision Orphan Program (approximate as of completion) Masaka Rakai Gulu Total Project Staff 65 94 101 260 Foster families assisted 6,000 6,800 2,500 15,300 Tuition support Orphans in primary school 10,000 13,600 7,700 31,300 Orphans in secondary school 200 180 - 380 Primary schools built/renovated 5 5 4 14 Vocational training Centers established 7 6 1 14 Orphans trained 200 460 110 770 Heath workers trained 700 495 170 1,365 Clinics renovated/built 2 5 4 11 Community workers trained 45 39 50 134 3.9 Rukungiri orphan program. The Church of Uganda (North Kigezi Diocese) supported a community-based care approach for orphans in Rukungiri. In 1992, this district in the southwest of Uganda had an estimated 20,000 orphans, most of whose parents had died of AIDS or malaria.27 Over the three years of the program, about 15,000 orphans received educational support, and about 170 chili-producing groups were formed. But funding from SIDA ceased when money was reallocated to the VAP, and the program came to a halt. Its activities are not likely to be sustained in the long term. 3.10 The widow and orphan support program in Lira was executed by a local NGO, the Lira District Development Agency (LIDDA). A US$400,000 grant made available to SIDA was reduced by half when funds were transferred to the VAP. Benefits to individual widows and orphans, while high, have been limited in relation to the district's needs.28 3.11 The war widow subcomponent aimed to establish a community-based health care program that would provide training in preventive health care and income-generating activities for 25,000 widows and their households. The component was executed by a local NGO, which established one project office in Kampala and one in the Luwero district, where it was combined with a health center. Community health workers were trained, sensitization workshops held, and supplies for income-generating activities distributed, although at lower levels than anticipated. More than half the project expenditures went for overhead and related expenditures (see also 27. The percentage of orphans in many primary schools ranged from 50-70 percent of enrollment. 28. PAPSCA assisted only 2 percent of the orphans and 25 percent of the widows in the district. A survey in 1992 conducted by LIDDA found that some 90 percent of the population aged 0-19 were orphaned children and around 18 percent of the total female population widows. 21 para. 6.9). Only around 3,000 widows have participated in this project-again far too few relative to need. Health Services 3.12 World Vision implemented this subcomponent in two remote counties of the Masindi district of western Uganda. Its major goal was to help communities improve their health status by developing a sustainable health care system that emphasized community participation, strengthened government health services, and fostered income-generating activities. Despite poor communications, late delivery of transport and procurement of project inputs, and lack of construction materials in some places, implementation was successful. The component renovated, expanded, and equipped five health units, trained a substantial number of health workers, and improved health services at the community level. It also implemented a substantial health education program, mobilized communities to improve water supply systems, and introduced limited income-generating projects such as pig and fish production. Sustainability seems likely, but there may not be enough financing to replace broken equipment. Income-Generating Activities 3.13 SIDA funded two specific income-generating programs. One established a revolving credit scheme for income-generating agricultural activities by women in the Apac district.29 At completion, about 70 percent of the original US$400,000 grant had been disbursed. A local NGO, the Apac Development Foundation, executed the program. Bad weather, political interference in the selection of beneficiaries, conflicts between the NGO managers and board, 30 underfunding, and late procurement all seriously affected implementation. The sustainability of this component depends on repayments, which are uncertain at best, and its future is dependent on other donor funding.31 3.14 An international NGO, World Learning Inc. (formerly known as the Experiment in International Living), implemented a small-scale enterprise development component. The program was executed in four districts around Kampala, using local financial intermediaries.32 The project suffered from serious management and targeting problems and substantial procurement delays. One example of the delays involved the procurement of motorcycles for supervisory staff. The motorcycles were delivered to the offices of World Learning in Kampala the same week OED's audit took place, by which time the project component had closed. Despite the problems, however, the program's impact on individuals was positive (box 3. 1).The component funds were disbursed quickly, but the sustainability of these credit activities is unlikely because of loan repayment difficulties. 29. These activities included crop production, agro-industries, fish farming, animal husbandry, and cooperative marketing. 30. During the 1993 election campaign, for example, politicians misinformed chiefs about the animal husbandry loans, which the politicians maintained were government compensation for cows lost to cattle rustlers during the insurgency. 31. For the time being, the funding seems to have been assured under the African Development Bank Poverty Alleviation Program. 32. The four districts were Mpigi, Mukono, Jinja, and Iganga. 22 3.15 In addition to repayment problems, there were also targeting problems with more loans going to men and people outside the poverty groups. Box 3.1: The Impact of the Small Enterprise Component Mugomba D., using a loan of U Sh 250,000, planted cassava and repaid the loan. He was able to purchase sheets of iron, timber, and bricks. He received another loan worth U Sh 900,000 to buy oxen and ploughs and plant 7 acres of maize. He has been able to repay the loan and has used the income to build himself a permanent house. Sarah Nabirye received a loan of U Sh 530,000, which she used to grow rice. She was able to pay off the loan and used the balance to purchase a plot of land, pay school fees for her children, and buy household items. The project has extended another U Sh 1 million loan to her. Robert Nsibirano planted pineapples with his loan of U Sh 1.015 million. He has been able to pay off the loan, open a store where he sells the pineapples, and renovate his old house. Small-Scale Infrastructure Rehabilitation 3.16 This component included rehabilitation of a small-scale infrastructure program (implemented by Action Aid in the Kamuli district in eastern Uganda), plus two water and sanitation rehabilitation components in Kampala. The Kamuli program, which was basically a capacity-building operation, was based on initial beneficiary surveys, strong sensitization training and community mobilization, close cooperation with local and community leaders, and effective technical support and supervision. The program provided materials and artisans' wages. Communities were expected to contribute about 30 percent in kind. Action Aid established a strong presence in Kamuli, paying for the overhead of its large office out of its own funds (US$600,000). The program was both labor intensive and expensive (around 40 percent of project expenditures were for overhead and support activities). 3.17 The communities constructed mainly primary schools (about 670 classrooms). They also built eight secondary schools, two houses for teachers, 48 offices and stores, 10 health units, and one feeder road. Enrollments in most of the schools constructed under the project were higher than before. Again, however, some of this increase may have been migration from other older schools. Despite these efforts a substantial number of the district's school-age children still are not in school. More positively, the training in construction and brick making has helped many youths. Some of the affected communities have undertaken subsequent projects, attesting to the viability of the community training program. The constructed facilities are likely to be sustained, but replicating the approach in poor communities will be problematic, as the local councils have few resources. The program has ended, and the Action Aid office in Kamuli has been closed. 23 3.18 The water and sanitation component aimed to improve water supply and sanitation 33 facilities in the Rubaga district of Kampala. The component has protected 58 springs and installed 100 public standpipes. Because users generally pay to access standpipes, sustainability is likely if the pipes are maintained. But spring water is free, and thus springs are less likely to be sustained. About 3,000 ventilated improved-pit latrines (VIPs) have been constructed for individual households and institutions such as schools and markets. The component also purchased a cesspool emptier to be operated for a fee. The clearing and stone facing of major drains (13 km and 3 kin, respectively) that had been used as refuse dumps have improved water drainage, but maintaining the drains is a problem, as people continue to deposit refuse in them. The component implemented a health education program to address this issue. Implementation was affected somewhat by funding delays and cement shortages. Communities participated in most activities and made some in-kind contributions. 34 3.19 Under the motto Keep Kampala City Clean, a local NGO-the Committee for International Self-Reliance-executed another water and sanitation component for about US$700,000.35 The component, which aimed to create temporary employment and improve health and sanitation in parts of Kampala, met a number of targets. It constructed 48 VIPs, protected 38 springs, and installed seven stand pipes. In addition, it desilted a number of major drains, launched a health awareness campaign, and supported a solid waste recycling pilot project (actually a joint venture between PAPSCA and the International Labour Organization). Community contributions have been minimal, but some local committees have been formed to operate and maintain the springs and latrines. However, the sustainability of most activities is problematic in the absence of further external financial and NGO support, especially the desilting of drains, which generates no funds. Drains will become clogged if they are not maintained, and stone lining, rather than just cleaning, would have provided a better long-term solution. Strengthening Social Planning 3.20 The objective of the Social Dimensions of Adjustment (SDA) component was to create the basis for refining the targeting of vulnerable groups and developing a comprehensive social policy consistent with efficient, growth-oriented strategies. The Ministry of Planning executed the component, which consisted of statistical surveys, policy studies, and institutional capacity building. 33. The immediate objectives were (a) to alleviate the unsatisfactory water supply and sanitation status of a representative low-income peri-urban group by implementing low-cost water supply systems and sanitation supported by health education programs; (b) to develop and demonstrate a viable community-based approach toward integrating water supply, sanitation, and health education that would be replicable in other similar peri-urban areas; and (c) to demonstrate the increased efficiency of sector investments through active user contribution, either in kind or as funds. 34. The participatory evaluation reported that renters resisted making contributions for the latrines, which they felt "belonged" to the landlords. Landlords were suspicious that the project wanted to obtain land rights by constructing facilities. 35. When part of the SIDA grant was reallocated to the VAP, the government had to substitute around US$190,000 in funding (about 25 percent). 24 3.21 The SDA was ineffectively managed.36 The first social policy adviser was appointed only a year after credit effectiveness and remained in office only for about four months. Appointing the second social policy adviser took another year. The steering committee that had been envisaged proved ineffective. In terms of output, the SDA undertook a number of surveys, completed 19 policy studies,37 and provided some institutional support in the form of workshops, short-term overseas courses, and equipment. 3.22 Methodological differences between the various surveys reduced the surveys' usefulness in assessing developments over time. No evidence exists to show that the studies influenced government policy. The fact that this program was located outside the main line ministries and agencies responsible for policymaking may explain why it had so little impact. Project Administration 3.23 This component supported the PCMU. Originally envisaged as a small unit, the PCMU started with 14 staff members. Additional staff had to be added in early 1992, and specific technical assistance to strengthen financial management was provided a year later. 3.24 A steering committee was expected to oversee the PCMU to ensure coordination among the ministries and resolve technical and administrative issues. Although the committee was to meet regularly, it met only twice (in June of both 1991 and 1992). Without the steering committee, the PCMU lacked the support it needed to manage the project effectively. The problem was exacerbated by the PCMU's lack of experience with project management in general and with Bank and government procedures in particular. The PCMU proved excessively bureaucratic and suffered from internal conflicts and irregularities. Financial accounting and M&E were poor. In short, the PCMU failed to become an efficient coordination and monitoring unit, and it was dissolved when the project ended. 36. The government evaluation report notes that the procurement of goods and services through the UTNDP/OPS was efficient, although it entailed paying a handling fee. 37. The ICR, table 7, lists the studies. 25 4. Ratings 4.1 The ICR rates project outcome as marginally satisfactory, sustainability as uncertain, institutional development as partial, and borrower and IDA performance as satisfactory. In its initial assessment, OED lowered the rating for project outcome to marginally unsatisfactory and downgraded the Bank's performance rating to unsatisfactory. This audit confirms these ratings, but also downgraded borrower performance to unsatisfactory. Outcome 4.2 The individual PAPSCA project components had differing rates of success (table 4.1). Of the 13 components, seven (accounting for 42 percent of project resources) are rated as satisfactory. The other six are rated as unsatisfactory. Thus, some 58 percent of project resources were absorbed by activities that produced unsatisfactory outcomes. Table 4.1: Component/Subcomponent Ratings Component /Subcomponent Weight (%) Outcome Sustainability Primary Education 34 unsatisfactory uncertain Orphan Support 14 satisfactory uncertain Widow Support 4 unsatisfactory unlikely Masindi Primary Health 4 satisfactory likely Strengthening Social Policy Planning 14 unsatisfactory uncertain Small-Scale Infrastructure 12 satisfactory likely Low-Cost Water and Sanitation 8 satisfactory uncertain Project Administration 2 unsatisfactory unlikely Keep Kampala Clean 2 satisfactory uncertain Orphan Support in Rukungiri 1 satisfactory uncertain Widow/Orphan Lira I satisfactory unlikely Women's Agriculture 1 unsatisfactory uncertain Small-Scale Enterprises 3 unsatisfactory unlikely 4.3 The project's overall objective was to alleviate poverty, provide aid and training for Uganda's most vulnerable groups, and foster institution building. The project failed to substantially alleviate poverty (para. 1.8), accomplished little in the way of institution building and provided an amount of aid to orphans and widows that fell far short of meeting actual needs. The project earns an unsatisfactory rating for failing to meet any of its objectives, but the positive results of some individual components (especially those providing orphan support) support the rating of marginally unsatisfactory. Sustainability 4.4 Sustainability was not discussed in the SAR. the issue may have received little attention at the time because the project was seen as a quick-disbursing mechanism to alleviate, at least temporarily, the traumatic conditions under which the poor and vulnerable groups were living. But, sustainability became a more pronounced issue during the midterm review. It also became a 26 concern for a few of the implementing agencies as they realized that follow-up arrangements needed more systematic attention (at that time the future of PAPSCA already seemed uncertain). In retrospect, sustainability warranted more attention during design and implementation. 4.5 At project completion, the sustainability of all but two of the components is unlikely or uncertain (table 4.1). Many of the activities, such as tuition support for orphans, were humanitarian in nature and in principle cannot be expected to continue beyond the project period without further aid. The income-generating activities did not produce the levels of funding necessary to become self-sustaining. Some NGOs will continue to provide support (for example the orphan programs supported by World Vision), but at levels that will require cutting back on these activities. In other cases (such as the widow program and income-generating activities), programs have simply come to a halt. In addition, most of the infrastructure constructed under the project will be maintained only if communities are motivated to do the work themselves. Where maintenance depends on public funding (such as supporting vehicles to clean latrines) or communities have only an indirect interest (for example, in protecting. springs or clearing drains), sustainability is less certain. The schools can be completed and instructional materials provided only with outside aid. In these cases, sustainability is unlikely unless follow-up is integrated into other donor-supported projects. All these considerations underpin the uncertain rating for project sustainability. Institutional Development 4.6 The project accomplished little in terms of institutional development. The PCMU has been disbanded, there was no appreciable impact on policy, and the education PIU showed little evidence of strengthened capacity. However, the project did have some institutional impact at the NGO and community levels. Some implementing agencies have been able to mobilize resources to continue programs. Some implementing agencies have reached agreements with the government about the future of programs with follow-up activities. To facilitate the lAs' work, the government has permitted these agencies to retain vehicles and equipment purchased under the project.38 4.7 The NGOs themselves have gained experienced implementing projects in cooperation with governments and international organizations. As a result, Uganda now has more in-country capacity to design and implement community-based development interventions. The pool of skilled community development staff has been expanded. Although there was little or no community involvement in the beginning, the project did succeed in motivating communities to participate in and contribute to development. Communities contributed substantial resources, and some organized themselves to support and sustain development interventions. Because the project helped Uganda make more effective use of its human, organizational, and financial resources, institutional development is rated as modest. Borrower Performance 4.8 The strong support the government had provided at the beginning of the project weakened considerably during implementation and was virtually absent at the end. While the government was instrumental in the design and launching of the PAPSCA program, it offered little political or bureaucratic support for the PCMU. In the end, the government abandoned PAPSCA. 38. Government regulations normally stipulate that project-funded equipment and vehicles become government property at project completion. 27 4.9 Counterpart funding was not provided in the first year and was problematic throughout the project period (para. 2.11). The government did nothing to make administrative procedures more flexible, so that most were unnecessarily lengthy and cumbersome. The interministerial committee for the PCMIU and the steering committee for the SDA component failed to meet regularly or to provide guidance and supervision. The PCMU performed poorly, and the PIlus did not do much better. Based on this mainly weak record, the audit rates borrower performance as unsatisfactory.39 Bank Performance 4.10 When the need for a program to alleviate poverty among Uganda's most vulnerable groups became apparent, the Bank acted quickly and decisively, and hence identification was highly satisfactory. But the project that resulted had the verbal but not the financial support from other donors (although they had participated in the 1988 consultative group meeting). Appraisal was deficient for a community-based development project because the communities themselves were not involved, a blueprint approach was used, and there was no piloting. Quality at entry was poor (paras. 2.1-2.3). 4 4.11 Project supervision was also deficient. In the initial years, task managers changed frequently. Implementation procedures and systems were not established up front, and the absence of a project launch workshop was a major shortcoming. After a serious procurement problem (para. 2.13), the Bank appointed a senior implementation specialist as task manager. However, the focus was almost exclusively on procedures and the minutiae of project execution. Supervision missions were often too short; they involved just one person and were generally only a part of much broader mission itineraries. While the supervision provided may have prevented the project from derailing altogether, it did not give sufficient attention to development impact. Based on this performance, and despite good efforts by individual Bank staff, Bank performance is rated as unsatisfactory.41 39. B.P. 13.55 defines borrower performance to be an assessment of borrower performance including that of implementing agencies. Weighting the generally good performance of NGOs, the ICR rated borrower performance as satisfactory. In the end, these performance judgments require balancing many factors. When evaluating borrower performance, the audit considered the policy environment created for the project, the level of commitment of the government and key institutions associated with the project, the provision and reliability of domestic funds, administrative procedures, quality of decision making, quality of management, adequacy of monitoring and evaluation, and extent and quality of participation of beneficiaries. As performance on many of these items was very low, the audit concluded that borrower performance overall must be rated as unsatisfactory. 40. It has been argued that PAPSCA was an early attempt to design a poverty alleviation project. Yet, poverty alleviation has been a Bank concern for much longer. Also, for example, the Bolivia Social Fund had already begun, and the Honduras Social Fund and Senegal-AGETIP projects were prepared at the same time as the PAPSCA project. Bank staff have also argued that while community participation in design would have improved implementation, it would also have slowed it down, and this was undesirable, given the emergency situation and need for swift disbursements. However, other projects in similar situations used piloting effectively and made judicious use of the project preparation facility to start up these projects. 41. Bank staff have argued that the audit should take a more holistic view on Bank performance. However, while identification was highly satisfactory, Bank performance during appraisal and supervision was not, and this is the main reason rating Bank performance overall as unsatisfactory. 28 5. Veterans Assistance Program 5.1 Public expenditure reviews in the early 1990s revealed that Uganda's defense expenditures had reached levels that were not only unsustainable but damaging to the economy.42 The Bank, donors, and government agreed that the time had come to cash in on the "peace dividend" by demobilizing a substantial part of the army, thus releasing funds for priority economic and social infrastructure and services programs.43 5.2 At the government's request, in late 1991 the Bank agreed that a number of preparation studies could be funded under the SDA component of PAPSCA. The studies were conducted between March and July 1992. The Bank subsequently decided that it would be the fund manager for the demobilization effort. The Bank also decided to provide some financial support. It chose not to create a separate project but to transfer funds from the PAPSCA credit. PAPSCA would also serve as the vehicle for the Bank to administer VAP. 5.3 Ultimately, the Bank diverted US$2.5 million from PAPSCA to VAP. 44 SIDA, which was part of the consortium that funded the VAP, diverted another one million krona. The impact on some PAPSCA components was severe, especially on those financed by SIDA. Why did the Bank decide to take this step? Obviously, the VAP was a high priority in 1991, but PAIPSCA had been a high priority just two years before (1988-89). One hypothesis is that the Bank had already become disillusioned with PAPSCA's poor performance and decided that some of the funds could be more usefully employed, although this notion remains a point of internal debate. VAP's Objectives and Implementation 5.4 The VAP aimed to reduce Uganda's armed forces by about one-half (40,000-50,000 soldiers). The program's objectives dealt with more than just releasing the soldiers: it would demobilize, provide "reinsertion assistance" (cash to meet basic needs for an initial period after demobilization), and reintegrate soldiers and their families into the community. The Bank appraised the VAP in September 1992. Rather than having the army carry out the program, in October 1992 government created a parastatal agency, the Uganda Veterans Assistance Board (UVAB), that answered directly to the prime minister's office. 5.5 The Bank provided leadership to this effort and to the donor consortium.45 The German donor agency GTZ and the Danish International Development Agency (DANIDA) provided technical assistance for M&E and project management. The VAP was executed in three phases (December 1992-July 1993; April-July 1994; and October-November 1995). In total, about 36,400 soldiers with an estimated 135,000 dependents were demobilized at a total cost of around 42. Defense spending had risen from 28 percent of government expenditures in 1986 to 43 percent in 1991. Over the same period, capital expenditures on defense had risen from 18 to 38 percent. 43. For key documents on demobilization and the transition from war to peace, see annex B, list of main documents. 44. As the government's completion report points out, the dollar/SDR exchange rate gains compensated somewhat for the diversion of funds. 45. The donor consortium consisted of IDA, the European Union, USAID, Germany, Denmark, Sweden, the Netherlands, ODA, and UNICEF. 29 US$45 million (table 5.1). Thus, the program successfully demobilized close to the target number of soldiers and reintroduced them to civilian life. The "reinsertion" efforts have not been without problems, however. Reintegration measures, such as vocational training, counseling, and emergency health care have started but are constrained by a lack of resources. The prevailing opinion is that these activities should no longer be performed by a specialized agency (UVAB) but should be integrated into other development projects. A number of bilateral programs and projects are being implemented (mainly by NGOs) at the district level to provide the necessary support. Table 5.1: Summary of the Uganda VAP Total Phase I Phase II Phase III I-III 3 Vets demobilized (thousand) 22.9 9.3 4.1 36.4 Program funding (pledges-US$ million equivalent) Multaterl cotiutions IDA/SAC II 2.1 1.0 0.25 3.35 UNDP 0.3 0.18 0.40 0.89 UNICEF 0.25 - - 0.25 Total Multilateral 2.65 1.18 0.65 4.5 SIDA 1.5 2.5 1.8 5.8 DANIDA 2.9 2.5 1.5 6.9 EU 2.9 - - 2.9 GTZ * * Netherlands 2.8 1.4 1.5 5.7 ODA 1.5 3.0 2.62 7.1 USAID 5.5 1.5 - 7.0 Total bilateral 17.1 11.0 7.42 35.45 Government of Uganda 2.40 1.35 1.11 4.87 Grand total 22.1 13.5 9.19 44.80 *GTZ's contribution for Phase II was US$1,333. Source: Uganda Veterans Assistance Board (UVAB), Information Digest 1996. Results of the Demobilization Effort 5.6 In terms of implementation, the VAP was a success. It was very well managed (certainly far better than PAPSCA), mostly because of the dynamic efforts of the UVAB executive secretary, a retired army major-general. A majority of the veterans have now been back in civilian society for a few years. But as a DANIDA evaluation points out, the social reintegration of veterans has had considerable social costs.46 Divorce rates are high, indicating that veterans have been establishing new social ties in their communities at the expense of their original partners. Health indicators suggest that 40-60 percent of the veterans have AIDS or are HIV positive; thus many of their partners are likely to be infected as well. These problems are far beyond the scope of the VAP, and the UVAB is not the most suitable organization to handle the burgeoning social and economic issues. 46. See DANIDA, 1996, "Demobilization of the Ugandan Army: Lessons Learned and Strategy Options for Reintegration of Veteran Families," Denmark. 30 5.7 Some of the positive results of the demobilization have been undone as the country's internal and external security environments deteriorate. At the time of the audit, an estimated 7,000 veterans had been remobilized, albeit with one-year contracts. The donors are somewhat embarrassed by this development, having just spent some US$45 million on the demobilization program. However, donors need to balance the need of sovereign nations to respond to threats to their security against the equally compelling need to reduce defense budgets-a dilemma that will become more pressing as donors are increasingly involved in war-to-peace transitions. Present and Prospective VAP Evaluation 5.8 Bank performance with regard to the VAP was excellent. The program has received much attention as an important example of a war-to-peace transition Yet internally, this important activity-the amount involved in the VAP was greater than what was designated for the entire PAPSCA project-was not recognized as a product separate from PAPSCA. There are no separate files for the VAP, and it is difficult to determine how much time and resources the Bank spent on this activity. There is less official Bank documentation than for a normal project.47 There is no ICR.48 5.9 The audit recommends that the Bank, as task manager, properly and transparently account for this activity through a formal evaluation. A final VAP donor conference is planned for 1997. At that time, the UVAB will present a completion of activity report. Makerere University and DANIDA are also compiling an evaluation that should be completed before the conference. The conference and evaluation should provide a solid basis for a VAP ICR in the second half of 1997. 49 47 . There are no separate files, but the resident mission commented that, by its nature, this operation was managed from the field and that resident mission files contain more correspondence on VAP than on any other project. 48. The PAPSCA ICR did not evaluate the VAP, as the program was still operating. However, the VAP was clearly a distinct program and therefore should not have been covered in the PAPSCA ICR in any case. 49. Since there is probably little "official" documentation, the ICR might have to be a little more extensive than prescribed in OP 13.55. A participatory beneficiary assessment, if not already done, should be part of the process, and donor input should be sought. 31 6. PAPSCA's Development Approach 6.1 Two lessons can be drawn from the failed PAPSCA experience. The first involves the strategy of using NGOs as implementing agencies. The second involves project design. Would a different design have increased the project's chances of success? This chapter explores these two issues in some detail. Using NGOs as Implementing Agencies 6.2 Historically, NGOs and community-based groups have played an important role in development in Uganda. While the government once felt that the public service could effectively perform NGO activities, it now believes that the civil service and NGOs should work together. Without NGOs, PAPSCA could not have been implemented. 6.3 The NGOs that were responsible for implementing most of the PAPSCA components were not familiar with the Bank. Because they were used to dealing with bilaterals directly rather than through the government, the NGOs also did not understand that working with the Bank meant working with the government. The learning process was painful, but most NGOs agreed that it was useful not only in helping them understand the Bank but in forging stronger ties with the government. For its part, the government learned that it needed to develop adequate capacity to deal with NGOs. Finally, the Bank also learned from its experience with the NGOs. The Bank had had little experience working with NGOs in Uganda, and this lack of familiarity helps to explain some of the start-up problems. The Bank has gained experience that extends beyond the PAPSCA framework. The Bank added an NGO liaison officer to the mission staff in Kampala in 1996 and completed a study on NGOs in Uganda in 1994.50 The NGOs' Experience 6.4 The audit mission organized a focus group with NGO representatives to distill their perceptions of PAPSCA's strong and weak points (box 6.1). The NGOs saw the principles of community-based development that are embedded in PAPSCA as the program's strongest point. The major failing involved Bank and government procedures, which the NGOs felt needed to be improved. 50. Eastern Africa Department, Population and Human Resources Division, The Role ofNongovernmental Organizations and Community-Based Groups in Poverty Alleviation, World Bank Report 12262-UG, (Washington, D.C., 1994.) 32 Box 6.1: NGO Perceptions of PAPSCA-Strong and Weak Points Strong Points * Emphasis on community participation and strengthening * Focus on poverty, vulnerable groups, and employment * Partnerships among donors, government, NGOs, and World Bank * Multisector approach Weak Points * Overly complicated procedures for procurement, disbursement, and other activities * No project launch workshop * PCMU that suffered from understaffmg, and inadequate monitoring capabilities hindered project implementation * Inadequate feedback from Bank task managers to partners 6.5 Most of the procedural problems were related to the disconnect between the management of implementation for individual components, which was decentralized, and Bank and government procedures, which were centralized. For example, the provisions for imprest accounting did little to meet the requirements of the implementing agencies. The Bank established a special account (US$3.0 million) with the PCMU, an arrangement that, from the Bank's point of view, facilitated disbursements at the project level. But the NGOs had to prefinance project expenditures and wait a considerable time for reimbursement. Procurement was also centralized, even for small purchases such as individual vehicles, motorcycles, and office equipment. Many of these administrative requirements could have been more efficiently handled at the component level. Advantages and Disadvantages 6.6 An audit is limited in scope and therefore can make fewer specific recommendations than a case study.51 However, using other Bank studies as a reference, the audit has been able to shed some light on the advantages and disadvantages of working with NGOs. 6.7 An earlier study elaborated on the strengths and weaknesses of involving NGOs closely 52 in project implementation. Table 6.1 compares the PAPSCA experience to these findings. 51. The Uganda experience in general and the PAPSCA experience in particular would make an excellent case study of NGO/Bank interaction. 52. Cernea, Michael, M., "Nongovernmental Organizations and Local Development," World Bank Discussion Paper 40 (Washington, D.C., 1988). 33 Table 6.1: Strengths and Weaknesses of Working with NGOs Findings Audit Observations R:fecognizk strength,s .. * Capacity to reach the rural poor and outreach to remote This strength was evident amount the NGOs areas executing PAPSCA components. * Capacity to promote locate participation The PAPSCA NGOs had the capacity to promote participation, but they should not have had to work as hard as they did to promote it. The project should have included communities from the beginning. * Capacity to operate on low cost The PAPSCA experience points to very high NGO overhead and operations costs. * Capacity to innovate and adapt Most component designs were prepared with limited participation. Many PAPSCA NGOs did adapt original blueprints to the realities on the ground, but some stuck to the original design. Comparative littitons. * Limited replicability PAPSCA could be replicated, although the high level of support it required makes replication difficult, especially for poorly endowed local governments. * Limited self-sustainability PAPSCA's sustainability is limited, especially for humanitarian aid, primarily because few income- generating activities were created at the community level. Some of the infrastructure can be sustained by interested community groups (water supplies and health clinics, for example). * Limited technical capacity Some PAPSCA NGOs, especially those with international links, have good technical capabilities. Other smaller local NGOs do not. * Lack of broad programming content PAPSCA NGOs were not involved in strategic programming. In fact the question can be raised if NGOs should be responsible for strategic programming at the regional or national level. Another case study of NGO involvement in the PAPSCA program yielded a number of recommendations (box 6.2).53 The audit confirms these findings. Especially important are the need to establish proper selection mechanisms for NGOs, decentralize procedural authority, and ensure that the target population participates in the project from the beginning. 53. Hino, Toshiko, "NGO-World Bank Partnerships: A Tale of Two Projects," World Bank Human Capital Development Working Paper (Washington, D.C., 1996). 34 Box 6.2: Working with NGOs in Bank-Supported Projects i. Ensure that NGOs fully understand the practical implications of working with the government and the Bank. ii. Define the purpose ofNGO involvement, especially its complementary to the government's project, as clearly as possible. iii. Evaluate the capacity of the NGOs early in the design stage before the selection of suitable collaborative mechanisms. iv. Ensure consistency ofsupervision through the succession of task managers and, ideally, continuity. v. Encourage NGOs to promote the participation of the target population by providing flexible and responsive mechanisms during the supervision. vi. Ensure that adequate administrative capacity exists in the relevant government agencies. Findings 6.8 Participation. Contrary to popular belief, involving NGOs in projects does not automatically ensure participation, and it certainly did not mean that the PAPSCA project was participatory. In fact, PAPSCA's management style was top down from the beginning; the NGOs were not oriented toward participation. Plans were made rapidly, and planners simply assumed that communities would agree to the self-help concept. The communities were consulted only late in the project, and massive motivational efforts and sensitization training were required to make the participatory approach work. 6.9 Overhead. A substantial portion (up to 40 percent) of the project's resources was spent on overhead and supporting services. Under the war widow component, a staff apartment, office, and health clinic were built in Luwero. Were these facilities needed? Many widows did not think so. They were afraid of higher medical fees and would have preferred better water supplies. Again contrary to popular belief, the intermediation of NGOs was not cheap. 6.10 Selecting NGOs. There was no bidding process to select NGOs. Project preparation was based on the proposals presented, and the capabilities of the individual NGOs were not appraised. As it turned out, many NGOs proved quite capable. But the Bank should not assume that NGOs are always capable of executing its projects and should consider a more open and transparent selection process (possibly including a form of bidding) for selecting NGOs. 6.11 Competition. NGOs are frequently in competition with local government structures, and some have established community structures at the expense of local governments. Some local NGO offices are much better staffed and have more resources than government agencies. This situation creates a dilemma. NGOs are clearly often better prepared to implement projects than government agencies, but NGOs should not be used at the expense of local government capacity building. Are NGO activities and decentralization of government on a collision course? Should NGO activities be integrated with emerging district administrations? There is a lot to learn about the appropriate role of local government and NGOs. Alternatives to the PAPSCA Model 6.12 Despite PAPSCA's problems, the project's experience proves the value of community- based development and the advantages of working with NGOs. However, the management 35 arrangements were too centralized, complex, and cumbersome. Would a different approach have worked better? 6.13 The midterm review in late 1992 considered ways to improve management and implementation. It listed as the alternatives retaining the PCMU as the centralized management unit; decentralizing management to individual implementing agencies; incorporating components in regular line ministry programs; or closing the project down altogether. In the end, the PCMU was retained, largely because reappraisal would have been costly and disruptive to project execution, especially to the implementing agencies. However, in retaining the PCMU, the Bank recommended strengthening the unit with additional staff, a financial controller, higher procurement thresholds (through prudent shopping), and more government and Bank supervision. Some of these recommendations were carried out and did in fact improve the PCMU, but in general the unit remained ineffective. 6.14 Why was PAPSCA so unsuccessful? Contrasting it to successful social funds is revealing.54 Unlike social funds, which generally have strong political support, PAPSCA did not have full government support. It also did not have the autonomy most social funds enjoy, and it did not provide the necessary training to its implementing agencies that is a feature of social funds. Social funds are by nature participatory, involving beneficiaries in every aspect of project preparation, planning, and implementation. Because they are participatory, social funds are also relatively transparent, as the PCMU that administered PAPSCA was not. PAPSCA employed a bureaucratic, top-down management approach that did not make participation a centerpiece of the program until it was too late. Its staff was inexperienced and not fully dedicated.55 6.15 Would a social fund approach have worked better? Like the design of a social fund, PAPSCA's design was based on participatory community development. But unlike social funds, PAPSCA did not attempt to fund large numbers of small-scale projects directly; rather, it dealt with clusters of small projects through intermediaries. Many of its programs were recurrent, providing direct support to orphans and widows. Also, at the time PAPSCA was prepared, Uganda was just emerging from years of civil unrest, and the government was unlikely to agree to full autonomy for the program. The government may also have seen a relatively prescriptive approach as more appropriate than a decentralized model. In retrospect, the social fund model as it was emerging from Latin America and West Africa would not have been applicable for the PAPSCA program. 54. Although the Bank's experience with social funds is somewhat limited, at the time PAPSCA was being prepared, the Bolivia emergency social fund had been operational since 1987 and Senegal's AGETIP was also in the preparation phase. 55. See, for example, the Performance Audit Report on Senegal-AGETIP, Report 16516, 1997 (para. 5.9): AGETIP provides a new and innovative solution to public works contract management without creating bureaucracy and without corruption. What explains its success? First, it was set up as an NGO with a "private sector" philosophy. Second, it has benefited from strong political support. Third, it is accountable to the government and donors and therefore pursues high standards of efficiency and transparency. Fourth, it has a highly motivated and dedicated staff that follows strict ethical and professional standards (independence, impartiality, cost-consciousness, and efficiency). Fifth, it has complete financial autonomy in decision-making within the context of it statutes, bylaws, and project agreements with the government. Sixth, it provides substantial training to its partners (enterprises, consultants, and municipal managers). This training increases general awareness of AGETIP's approach and improves the quality of project execution. And finally, it has a results-oriented management culture. 36 6.16 In the debate over whether the Bank can work effectively with NGOs in its projects, the PAPSCA experience will be very useful. Having NGOs execute community development programs in defined geographical areas was a sound plan. But doing so within a rigorously centralized management framework was a mistake. Management responsibilities for the components (including procurement, disbursement, accounting, auditing, and M&E) should have been decentralized to the implementing agencies. As social funds have shown, decentralization can occur only within a transparent framework of agreed objectives and procedures that are clearly spelled out in manuals and written agreements. Under such a scenario, the PCMU would have resembled a secretariat with oversight responsibilities. In the end, both the Bank and governments may have to accept that maintaining NGOs' independence, flexibility, and ability to remain innovative will require more effective delegation of authority. 37 Annex A Basic Data Sheet UGANDA: PROGRAM FOR ALLEVIATION OF POVERTY AND SOCIAL COST OF ADJUSTMENT PROJECT (PAPSCA) (CREDIT 2088-UG) Key Project Data (amounts in USS million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 37.1 37.2 100 Credit Amounta) 28.0 31.1 111 Cofinancingb) 3.1 2.3 75 Date physical components completed 06/30/94 06/30/95 123c) Economic rate of returnd) a) Increase due to favorable exchange rate fluctuation between SDR and US$. b) Not included in appraisal financing plan, but donors were expected to separately fund parts of the program. SIDA later provided US$3.1 million under trust fund arrangements managed by IDA. c) Periods calculated from date of signing. d) No economic rates of return were calculated at appraisal or at completion. Cumulative Estimated and Actual Disbursements FY90 FY91 FY92 FY93 FY94 FY95 FY96 Appraisal estimate (US$M) 3.50 8.00 14.00 21.50 27.50 28.00 - Actual (US$M) 0.98 4.67 13.53 20.78 27.30 29.63 31.07 Actual as % of appraisal 28 58 97 97 99 106 - Date of final disbursement: 09/11/96 Project Dates Original Actual Identification 10/06/87a) Preparation 02/89b) IEPS 04/06/89 FEPS 05/10/89 Appraisal 06/89 Negotiations 11/20/89 Board Approval 02/01/90 Signing 02/08/90 Effectiveness 06/29/90 Closing date 09/30/94 09/30/95 a) First file reference to a possible social adjustment credit. b) Although regarded as an identification mission, it was in fact the preparation mission. Annex A 38 Staff Inputs (staff weeks) FY89 FY90 FY91 FY92 FY93 FY94 FY95 FY96 Total Preappraisal 12.1 12.1 Appraisal 14.6 29.2 43.8 Negotiations 15.5 15.5 Supervision 20.8 12.6 12.7 29.0 15.5 11.4 19.1 121.1 TF Supervision - 6.1 4.1 10.2 ICR 4.1 4.1 Total 26.7 65.5 12.6 12.7 29.0 21.6 15.5 23.2 206.8 Mission Dataa) Date No. of (month/year) persons Staffdays in field Specializationb) Performance c) Identification/ 02/89 8 135 E, PS, H, E, CD, Preparation NGO, ILO, ILO Appraisal d) 06/89 12 200 PS, CO, F, M, ED, E, CD, CD, P, H, E, E Supervision le) 02/90 1 7 PS - Supervision 2 06/90 1 4 PS 2, 1 Supervision 3 05/91 2 25 OP, CO Supervision 4g) 08/91 1 5 IS Supervision 5 11/91 2 20 IS, D Supervision 6 03/92 1 20 IS - Midterm Review h) 09/92 3 45 IS, H, E 2,2i) Supervision 7 02/93 1 1 IS 2,2 Supervision 8 07/93 1 4 IS - Supervision 9 09/93 2 20 IS, E 2,2 Supervision 10 02/94 1 3 is 1,l Supervision 11 09/94 3 24 IS, NGO, RM S,S Supervision 12k) 02/95 2 6 iS, PO _1) Completionm) 10/95 5 75 TM, IS, RM, AC S,S a) The ICR had omitted to include this table. Hence it had to be prepared fully by the audit mission, largely on the basis of file research. b) E=economist, PS=public sector specialist, H=health expert, CD=community development specialist, NGO=NGO specialist, ILO=expert from ILO, CO=country officer, F=financial analyst, M=management specialist, ED=education specialist, P=political scientist, OP=operations officer, D=disbursement officer, IS=implementation specialist, RM=resident mission OP, PO=project officer, TM=task manager, AC=accounting and auditing specialist. 39 Annex A c) Ratings of performance are respectively for overall status of implementation and achieving project development objectives. Initially they were rated on a scale of 1, 2, 3 but subsequently on an alphabetic scale: Satisfactory, Unsatisfactory. It should be noted that only very few Form 590's are included in the Bank's files and that with the reorganizations the divisional black books have also disappeared. d) Includes staff from SIDA and ODA. e) In addition to the "normal" supervision of the project, the SDA component was supervised by separate missions, increasingly and more intensely later in the project period, and the water and sanitation components were also covered by separate supervision missions, especially in the beginning of the project period. f) Supervision had been delegated to the resident representative for the period June 1990-May 1991. g) Special mission to clear up procurement backlog. No other records on file other than a letter to PCMU confirming certain decisions. h) This mission also assessed the feasibility of including VAP in the PAPSCA program. i) As noted, there are few Form 590's in the files. These are not really the ratings of the midterm review but the end of fiscal year Form 590 ratings prepared 06/30/92. At that time a number of subratings were assessed at the 3 level: availability of funds, procurement progress, studies progress. j) This mission was done in conjunction with the preparation of a case study of NGO-World Bank partnerships. k) A special mission by a senior financial management specialist took place in June 1995. 1) The end FY95 Form 590 rates the project: S,S. m) The Cofinancier was invited but did not participate. But one of the international NGOs actively assisted the mission. A separate participatory beneficiary assessment was also commissioned and executed by a local consultant firm. Other Project Data Borrower/Executing Agency: PAPSCA Project Coordination and Monitoring Unit Follow-on operations: none for PAPSCA, but primary school rehabilitation is continued under the Northern Reconstruction Project (Credit 2362) and the Primary Education and Teacher Development Project (Credit 2493).  41 Annex B List of Main Documents Armstrong, Jill. 1995. "Uganda's AIDS Crisis: Its Implications for Development." World Bank Discussion Paper 298. Washington, D.C. Bamberger, Michael, Abdullahi M. Yahie and George Matovic, eds. 1996. "The Design and Management of Poverty Reduction Programs and Projects in Anglophone Africa." EDI Learning Resource Series (in particular chapter 8 on PAPSCA). Washington, D.C. Cernea, Michael M. 1988. "Nongovernmental Organizations and Local Development." World Bank Discussion Paper 40. Washington, D.C. Colleta, Nat J., Markus Koster and Ingo Wiederhofer. 1996. "Case Studies in War-to-Peace Transition." World Bank Discussion Paper 331. (section n Uganda: "Consolidating Peace"). Washington, D.C. 1996b. The Transition from War to Peace in Sub-Saharan Africa. World Bank Directions in Development Series. Washington, D.C. Danish International Development Agency (DANIDA). 1996. "Demobilization of the Ugandan Army: Lessons Learned and Strategy Options for Reintegration of Veteran Families." Denmark. Gopal, Gita. 1995. "Procurement and Disbursement Manual for Projects with Community Participation." World Bank Discussion Paper 312. Washington, D.C. Gopal, Gita and Mark Alexandre. 1994. "World Bank-Financed Projects with Community Participation-Procurement and Disbursement Issues." World Bank Discussion Paper 265. Washington, D.C. Government of Uganda. 1989. "Program for the Alleviation of Poverty and the Social Costs of Adjustment." Final Report. Uganda. 1996. "Poverty Monitoring Sectoral Working Group Report." Hino, Toshiko. 1996. "NGO-World Bank Partnerships: A Tale of Two Projects." World Bank Human Capital Development Working Paper. Washington, D.C. Midland Consulting Group. 1996. "A Draft Report on the Participatory Evaluation for PAPSCA." Uganda. Ofcansky, Thomas P. 1996. Uganda: Tarnsihed Pearl ofAfrica. Boulder: Westview Press. Opio, Fred. 1996. "The Impact of Structural Adjustment Programme on Poverty and Income Distribution in Uganda." Draft Report. Kampala: Economic Policy Research Centre. Annex B 42 PAPSCA/PCMU. 1995. "Program to Alleviate Poverty and Social Costs of Adjustment." Final Evaluation Report. UVAB (Uganda Veterans Assistance Board). 1996. Information Digest. Government of Uganda. World Bank. 1990a. "Government of Uganda, Alleviation of Poverty and Social Costs of Adjustment Project." President's Report P-5174-UG. Washington, D.C. 1990b. "The Republic of Uganda, Alleviation of Poverty and the Social Cost of Adjustment Project." Staff Appraisal Report 8179-UG. Washington, D.C. 1992. "Republic of Uganda, Northern Reconstruction Project." Staff Appraisal Report 0411 -UG (for reference to the Community Action Program). Washington, D.C. 1993a. "Demobilization and Reintegration of Military Personnel in Africa: the Evidence from Seven Country Case Studies." Discussion Paper IDP-130. Washington D.C. 1993b. "Demobilization and Reintegration of Military Personnel in Africa: Eight Country Case Studies." Working Paper. Washington D.C. 1993c. "Uganda, Growing Out of Poverty." A World Bank Country Study. Washington, D.C. 1994. "Republic of Uganda: The Role of Nongovernmental Organizations and Community-Based Groups in Poverty Alleviation." Report 12262-UG. Washington, D.C. 1996a. "Uganda: Program for Alleviation of Poverty and Social Costs of Adjustment Project." Implementation Completion Report 15726. Washington, D.C. 1996b. "Uganda: The Challenge of Growth and Poverty Reduction." World Bank Country Study. Washington, D.C. SUDAN ,Moy KonUM 'UGANDA ) ARUA 2 , KTU ® DISTRICT CAPITALS KITGUM 4ARUA KOTI® NATIONAL CAPITAL GULU DISTRICT BOUNDARIES* ZAIRE GULU RIVERS NEBBQ; NEBBI k - INTERNATIONAL BOUNDARIES 4APAC Y' LIRA 0 4 24 tMOROTO Distnct n as coin-ide with District Capita!, with one I~/J 4~1.exception Fort Porto) is the District Capitsas äoaorote District. --S IND2 S O RO T I 2 - - enoSOROTI -1 U KUMIIeom no o os .1APCHOR A, , KUMI KAÖCd R,, on KiBOGA PALILtSA - s KAMULI PALLISA E MALEo 4 KiB .E K BROG LU (ERO \KAF5UUlg i,. ,MBALE L LUWERO y *BÄL K BLE , TORORO,, BUNDtl'UGYO 9ì NOORO.* FORMUBNDE MUK JINJAi GANGA,0 KILOMETER$ 0 50 100 50 PORTAL KABAROLE MUBENDE'i N_ _ AABAROKMPALA JINJA IGANGA iS00 MPIGI M G LA KENYA KASESC -M- MASAKA Edwrd ENYI MBARARA KALANGALA UGANDA RUK NGII-h INTUNGAMO BANZANIA- RWAND2% ~ __ ___ ___ ____ ___ _..____ ___ ___ w....'. __ ___ ____ __ ___ ____ ___ ___ 't IMAGING Report No.: i05I Type: PPAR