Document of The World Bank Report No: 19484 BU PROJECT APPRAISAL DOCUMENT ONA PROPOSED CREDIT IN THE AMOUNT OF SDR 9 MILLION (US$ 12 MILLION EQUIVALENT) TO THE REPUBLIC OF BURUNDI FOR THE SECOND SOCIAL ACTION PROJECT July 7, 1999 Human Development 4 Country Department 9 Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective July 7, 1999) Currency Unit Burundi Franc (FBu) FBu 1 = US$ 0.00182 US$ 1 = FBu 550 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AfDB African Development Bank AGETIP Agence Generale d'ExEcution de Travaux d'Interet Public BURSAP I First Burundi Social Action Project BURSAP II Second Burundi Social Action Project CAS Country Assistance Strategy ECD Early Childhood Development ICB International Competitive Bidding IDA International Development Association IEC Information, Education, and Communication Programs IFAD International Fund for Agricultural Development ISTEEBU Burundi histitute of Statistics and Economic Studies (nstitut de Statistiques et d 'Etudes Economiques du Burundi) LACI Loan Administration Change Initiative LCS Least Cost Selection NCB National Competitive Bidding NGO Non-Governmental Organization NS National Shopping PMR Project Management Report PMU Project Management Unit QBCS Quality-Based and Cost Selection SOE Statement of Expenditures UJNDP United Nations Development Program UNHCR United Nations High Commissioner for Refagees UNICEF United Nations Children's Fund WHO World Health Organization Vice President: Mr. Callisto Madavo Country Director: Mr. Emmanuel Mbi Sector Manager: Mr. Arvil Van Adams Team Leader: Mr. Menahem Prywes Burundi Second Social Action Project TABLE OF CONTENTS A. Project Development Objective I 1. Project development objective 1 2. Key perfonnance indicators 1 B. Strategic Context 2 1. Sector-related Country Assistance Strategy (CAS) goal supported by the project 2 2. Main sector issues and Government strategy 3 3. Sector issues to be addressed by the project and strategic choices 5 C. Project Description Summary 7 1. Project components 7 2. Key policy and institutional reforms supported by the project 7 3. Benefits and target population 7 4. Institutional and implementation arrangements 8 D. Project Rationale 9 1. Project alternatives considered and reasons for rejection 9 2. Major related projects financed by the Bank and other development agencies 10 3. Lessons learned and reflected in the project design 10 4. Indications of borrower commitment and ownership 11 5. Value added of Bank support in this project 11 E. Summary Project Analysis 1l 1. Economic 11 2. Financial 11 3. Technical 12 4. Institutional 12 5. Social 13 6. Environmental assessment 14 7. Participatory approach 14 F. Sustainability and Risks 15 1. Sustainability 15 2. Critical Risks 15 3. Possible Controversial Aspects 16 G. Main Loan Conditions 17 1. Effectiveness Conditions 17 2. Other 17 H. Readiness for Implementation 17 I. Compliance with Bank Policies 18 Annexes Annex 1. Project Design Summary Annex 2. Detailed Project Description Annex 3. Estimated Project Costs Annex 4. Cost-Effectiveness Analysis Summary Annex 5. Financial Summary Annex 6. Procurement and Disbursement Arrangements Table A. Project Costs by Procurement Arrangements Table B. Thresholds for Procurement Methods and Prior Review Table C. Allocation of Loan Proceeds Annex 7. Project Processing Budget and Schedule Annex 8. Documents in Project File Annex 9. Statement of Loans and Credits Annex 10. Country at a Glance Map: Burundi (IBRD 30118) Burundi Second Social Action Project Project Appraisal Document Africa Regional Office Country Departnent 9 Date: July 7, 1999 Team Leader: Mr. Menahem Prywes Country Director: Mr. Emmanuel Mbi Sector Manager: Mr. Arvil Van Adams Project ID: 64510 Sector: SA - Social Funds & Social Assistance Lending Instrument: Theme(s): Community Development, Social Protection, and Post-Conflict Specific Investment Credit Stabilization. Poverty Targeted Intervention: [X] Yes [] No Project Financing Data [ I Loan [X] Credit [I Grant [ ] Guarantee [ ] Oiler [Specify] For Loans/Credits/Others: Amount (US$ m): IDA US$ 12 million Proposed terms: [ ] To be defmed [ ] Multi currency [] Single currency [X] Standard Variable [ Fixed [ LIBOR-based Grace period (years): 10 Years to maturity: 40 Commitment fee: 0.5% Service charge: 0.75% Front-end fee on Bank loan: N/A Government 0.45 0.5 IDA 7.00 5.00 12.00 Beneficiary contribution 0.75 0.75 Total: 8.20 5.00 13.20 Borrower: Government of Burundi Guarantor: N/A Responsible agency: Twitezimbere (a not-for-profit organization) Estimated disbursements of IDA funds(Bank FY/US$M): Annual 1.3 3.0 2.8 2.7 2.2 Cumulative 1.3 4.3 7.1 9.8 12.0 Project implementation period: 4 years Expected effectiveness date: January 1, 2000 Expected closing date:. December 31, 2003 Implementing agency: Twitezimbere Contact person: Mr. Pontien Bikebako, Director Address: B.P. 2415, Bujumbura, Burundi Tel: (257) 22 18 04 Fax: (257) 22 84 51 E-mail: none; Project Appraisal Document Page 1 Bumndi - Second Social Action Project A: Project Development Objectives 1. Project development objectives: (see Annex 1) The development objectives are that: 1. Diverse community groups work together to select, partly finance, monitor and evaluate, use, and maintain their priority community investments. II. The poorest and most vulnerable groups benefit from improved social and economic services and improved social protection -- through transitory wages and social services that limit risks of poverty. To best achieve the objectives, the project will target the communes with the poorest and most vulnerable populations. Therefore, the project would focus some resources on communes attempting to reintegrate relatively high numbers of refugees and internally displaced persons. The reasoning is that these groups are often relatively poor and that targeting the corflmunes of return would promote community cooperation and reintegration. Within each commune, some activities would be targeted to pre-school children and mothers because they are among the poorest and most vulnerable groups. In the communes selected, the project would achieve the objectives by supporting the central community-based activity component, and two supporting components: 1. Community investments. * Mobilization and strengthening of community capacity (and in particular, of representative development committees) to participate in selection, financing, monitoring and evaluation, and the continued operation of the infrastructure sub-projects, * Reconstruction and construction of small-scale community infrastructure, such as classrooms, health centers, rural water supply, short rural roads, public marlcets, swamp drainage, and planting of forests; and * Improvement of early child development: through investments in the health and nutrition of young children and mothers (information, education and communication, de-worming, and growth monitoring and promotion). Communities will select additional investments, for example, in, nutrition education, micro-nutrient supplements, vegetable gardening, child- development training, trauma counseling; and 2. Strengthening of statistical capacity for monitoring the evolution of poverty, and finally 3. Administration. 2. Key performance indicators: (see Annex 1). The key indicators of realization of the project's objectives are: For Objective I: => Qualitative assessment of cooperation of diverse groups in selected sub-projects. => Percentage of women on the communal Development Committees reflects male-female ratio in the commune. z> Percentage of projects that reflect community priorities. Project Appraisal Document Page 2 Burundi - Second Social Action Project => Percentage contribution of the community to the project, in material, labor, and money. > Percentage of Development Committees that are still functioning one year after completion of the- infrastructure sub-project. = Percentage of infrastructure adequately maintained one year after completion. For Objective II: r Days of employment created. z Number of beneficiaries of infrastructure by type: Education, Health, Water Supply, Economic Infrastructure (markets, roads, forestry). => Capacity utilization rate of the infrastructure one year after construction. = Prevalence of wasting, stunting, and underweight of pre-school aged children. = Qualitative assessment of adult-child interaction. B: Strategic Context 1. Sector-related Country Assistance Strategy (CAS) goal supported by the project(see Annex 1) Document number: 14442-BU Date of latest CAS discussion: 05/02/1995 The CAS produced in 1995 no longer applies because of the international embargo and economic deterioration that followed the military coup in 1996 and because the suspension of disbursements from 10/1/1996 to 5/6/1997 interrupted the execution of Bank-financed projects. After the resumption of disbursements the Bank reopened pre-existing projects but did not follow the earlier strategy and did not prepare new credits. In 1998, the Government began peace discussions with the political parties represented in the elected national parliament and with armed groups based in neighboring countries and has made progress in reaching agreements. As the peace process progressed, the boycott was suspended, and the Bank agreed to review its program of assistance. The Country Team then prepared an Interim Strategy of Assistance for discussion with the Government. Once Burundi is more fully stabilized the Country Team would work with the Government to develop a Country Assistance Strategy with a view to long- term development. The broad objectives of the Interim Strategy are to (a) support the on-going peace process, (b) provide urgent support for recovery of poor communities and of the rural economy, and to (c) stabilize the macro-economy. The Interim Strategy starts with the BURSAP II and follows with an AGETIP that would reconstruct or build larger scale infrastructure, and with other actions. Inportantly, the Interim Strategy would provide for a macro-economic credit to permit freer import of goods and services for critical sectors. Finally, it would support a private investment guarantee scheme. The project is part of the Strategy because it fits well with the objectives. It provides employment and wage income in poor rural communities, restores community infrastructure, and expands delivery of social services. The project further bolsters recovery by increasing community capacity to participate in the selection, financing, and operation of local investments. Moreover, the project helps build a social basis for peace at the local level by bringing people together to seleclt and contribute to a project for their mutual benefit. The resulting improvement in local economic conditions and organizational capacity should then underpin and reinforce the peace process. Project Appraisal Document Page 3 Burundi - Second Social Action Project 2. Main sector issues and Government strategy: Sources. The assessment of issues is based on the Bank's recent sector work on Burundi: theBurundi Poverty Note, Reversing the slide into crisis: prospects for social protection in a period of great uncertainty, January 1999, and the draft Economic Update of late 1997. Poverty. The profile of poverty, measured in terms of household expenditure, has worsened rapidly since the beginning of the current crisis in 1993. Based on survey evidence, the incidence of rural poverty is estimated to have increased by 80% since 1993, with a doubling of urban poverty incidence. Poverty depth is estimated to be among the highest in sub-Saharan Africa. Social indicators also fell, reversing the steady improvement of the years preceding the crisis. Before the crisis of 1993, Burundi had undertaken a major program of infrastructure for health, surpassing the WHO norm of one health center per 20,000 inhabitants. In 1997, this had fallen to one center per 25,000 in rural areas because of the complete or partial destruction of 60 centers. Large numbers of medical personnel have also been lost to death or exile since the start of the 1993 crisis. The number of centers with at least.the WHO minimum of three trained auxiliaries dropped from 59% before the crisis to 41% in 1997. Reported cases of major endemic diseases have increased, and reported case of Malaria in 1995 and 1996 were four times higher than in 1988 and 1989. Respiratory infections are 2.5% higher and diarrhea 50% higher. Furthermore, HTV prevalence is rising rapidly, and the infection rate is estimated at 21% in urban areas and 6% in rural areas. Poverty among refugees and displaced persons. Conflict related population movements have led to poverty that is not adequately measured by conventional household surveys. Approximately 10% of the population were intemally displaced at any one time in 1997, with at least 14.4% suffering one or more displacements during the year. This follows previous periods of high population movement in 1993-96. Agricultural output is affected by the time spent in camps: households lose the savings necessary to buy inputs and their homes and farms are vulnerable to looting and destruction during their absence. In addition cultures such as coffee, which have not been maintained during the absence of the population, require more labor to produce equivalent yield on their retum. In addition, many refugees and displaced persons contract health and nutrition problems that make it difficult for them to work at full capacity. According to the UNICEF survey of November 1997, 84% of the displaced showed symptoms of conjunctivitis, 92% had body lice, 94% had scabies, 96% had diarrhea, and 96% suffered from worms. Poverty among children. Infant mortality rates per 1,000 live births have increased from an already high 110 in 1992 to 136 in 1996. Immunization rates are down from 83% in 1993 to 54% in 1996. Malnutrition, measured by wasting among children under 5 is estimated to have increased from 6 to 20% since 1993. Moreover, gross primary school enrollment has fallen from 70% to 40% since the start of the crisis, in part because about 480 schools required total or partial reconstruction by 1995-96. National social indicators hide an even worse picture at the regional level: four provinces had primary school enrollment below 30% in 1996-97, reaching a nadir of 9% in one of the most-violence stricken areas. Social protection. Given the extent of poverty, there is a near-term need for social protection measures to preserve Burundi's human capital and to prevent a further deterioration of social indicators. The most effective way to provide social protection is often through labor-intensive public Project Appraisal Document Page 4 Bumrndi - Second Social Action Project works. A further tactic is to strengthen household incomes through small and micro-credit. Government can deliver social protection directly to particularly vulnerable groups, such as refugees, displaced persons, and children and young mothers. The recognized good practice in community reintegration is to avoid introducing conflict by assisting all groups in the community, rather than just returnees. The project could also promote reintegration through peace education and cooperative projects. Young children and mothers will receive some protection against future poverty by information and education programs to promote early childhood development and nutrition. Public expenditures on health and education. A long-term resolution of the sources of poverty would require lasting civil peace and security and redirection of the Government budget. Government choices for the allocation of public funds have not been strongly oriented to basic services and poverty reduction. Between 1992 and 1997, the share of military expenditures in GDP rose significantly. And between 1992 and 1997, education expenditure declined from 4.6 to 2.6% of GDP and health expenditures declined from 1.2 to 0.5% of GDP. The withdrawal of donor support explains part of the decline, but part of the decline is also related to Government efforts to retain personnel in all sectors. Agriculture. A further reason for increased poverty is that the crisis exacerbated structural problems in the agricultural sector. For, the limits on land available for development, declining soil fertility, low use of modem inputs, and adverse incentives for investment in the state-controlled cash crop sector (especially coffee) had already set in motion a slide in yields. The crisis and embargo have worsened these problems, through looting and destruction of household goods and livestock, population displacement, and collapse of distribution channels for agricultural inputs. Populations unsure whether they will be displaced are planting crops which are fast growing and easy to cultivate, but of less nutritional value than other crops requiring greater upkeep. Distribution of inputs and seasonal credit to restart agriculture might help in the short-term and provide some social protection. In the longer term, public investment in agriculture and in particular on extension and rural roads would help. Moreover agriculture would benefit from decreased reliance on export taxation, which would improve producer prices of coffee, and from accelerated liberalization in the pricing and marketing of cash crops. Participation. The meaning of 'participation,' in practice, is not always clear. In general, participation in the context of a social fund means ceding control and responsibility for some aspects of project activities to local residents. The First Social Action Project (BURSAP I) has succeeded in ceding responsibility for part of the financing of its investments. Communities finance at least 5%, and on average 15%, of project costs, in terms of local government funds and household contributions of labor and building materials. The BURSAP I also attempts to cede selection of investments to local populations by requiring that applications originate with local groups, usually with the sponsorship of a NGO or a local government official. In practice, applications seem to have been generated by NGOs, in areas where they happen to be active, and by some local government officials, and to reflect their own programs rather than local choice. The policy issue is what extent of community participation is achievable and acceptable to the Government in the context of a centrally administered system that has experienced civil strife. The Ministry of Plan and the Ministry of Community Development have declared their interest in finding ways to improve participation. Moreover, the feasibility of expanding community participation was successfully tested by Twitezimbere, in the context of the BURSAP I. Community mobilizers were able to work with existing local development committees, modify committee membership to assure that it is representative, and obtain the support of local government officials. Project Appraisal Document Page 5 Burundi - Second Social Action Project Poverty monitoring. The final issue is how to assure continued monitoring cf the evolution of poverty so that poverty reduction efforts can be targeted effectively and to judge tfLe success of development efforts. The BURSAP I supported a national survey of household expenditures, which can form the statistical basis for more frequent, smaller surveys. 3. Sector issues to be addressed by the project and strategic choices: Social protection. The project would provide social protection to some of the poorest and most vulnerable Burundians and to some resettled people, over the short-term, by investing in infrastructure sub-projects that generate wage incomes in the communes where they live and through activities to promote young children's health, nutrition and social development. Health and education. The project would support an increase in basic health and education services by financing the construction of health centers, primary and secondary schools, and rural water supply, and also simple technologies to improve the health and nutrition of pre-school children and mothers. Moreover, the project would promote the continued flow of services by organizing community groups that would maintain the projects and by providing relevant training. Public expenditures on health and education. The project would not address the relatively low share of public expenditures on the social sector in the Govemment budget. Inst.ad, the Country Team has agreed to address the issue in the context of an Emergency Economic Recor,struction Credit. Agriculture. The project would not directly target agriculture. It would have a positive impact on agricultural output and productivity since communities can decide to use project funds to build agricultural infrastructure such as market places, slaughterhouses, small irrigation canals, or seeds for vegetable gardens. Moreover, the wages generated by project works should, initially, spur local agricultural production by driving up the prices of local food crops and livestock products. This should also help promote the reintegration of returnees. Main strategic choices. The proposed project would resemble the BURSAP I. The major change would be an improvement in the substance of community participation and in the targeting of communities. Participation. In the proposed project, communities would participate more fully by choosing the infrastructure sub-project, or combination of sub-projects, that the project would finance. To launch this process in each commune, Twitezimbere would conduct or contract an assessment of needs applying participatory rural appraisal techniques. This would lead to production of a short list of priority projects for the commune, and formation (or reformation) of a Development Committee that is representative and capable of participating in supervision and on-going maintenance of the sub- project. The project would then finance the community projects, so long as they are qualifying social or economic infrastructure, up to the limit of US$ 75,000 per commune. Moreover, communities would contribute at least 10% of project costs compared to a minimum of 5% in the BURSAP I. This should assure that the infrastructure sub-projects provide the greatest benefit to poor communities, among alternative uses of the project funds, and provide a greater incentive for communities to contribute to the maintenance and operation of the infrastructure. During the participatory assessment, the project would also provide information on early child development. Groups of adults, usually of mothers, would also be formed at the level of thxolline to carry out early childhood programs. These would be subject to the appioval and supervision of the Project Appraisal Document Page 6 Burundi - Second Social Action Project Development Committee and would have a lower cost of supervision than the infrastrmcture component. Improved participation could eventually extend to the areas of project implementation and infrastructure management. All these improvements cannot be achieved at once in Burundi because of limited project management capacity in rural communities, because of the hierarchical tradition of centralized administration, because of the post-conflict context, and also because the attempt would delay delivery of urgently needed social infrastructure, services, and incomes. The improvement in participation in the proposed project should then be viewed as part of a process of dialogue and of mutual learning through experience. Targeting. This implies a change from the practice, in the BURSAP I, of letting NGOs implicitly select the areas for project investments. So instead, Twitezimbere would propose, annually, a list of communes for project investment according to the following criteria: (a) Poverty index. The BURSAP II is targeted to the poorest communes, where poverty of a commune would be measured by an index calculated from information on access to social and economic services collected by the ISTEEBU. The poverty index will be constructed from data on use and proximity to potable water, health services, primary schooling, and to a public market. (b) Return of external refugees and internally displaced persons. Refugees and displaced persons are often extraordinarily poor. Moreover, support for areas where they return supports the peace process by helping to stabilize local economies. For this reason the BURSAP II will also target communes that have a relatively high proportion of extemal refugees and internally displaced persons that have retumed since 1996. The UNHCR will provide statistical information to identify areas of retum. The BURSAP II would target all residents in the commune and not only retumees, in line with good practice for local peace building. Finally, the BURSAP II will not invest in refugee or displaced persons camps. (c) Communes already receiving similar assistance. The UTNDP agreed to organize an informal committee for, among other purposes, the exchange of information, between donors, on geographic areas of activity of their projects. Twitezimbere would contribute information to the committee and, to avoid duplication, would target areas not receiving similar infrastructure investments from other donors. (d) Security. The BURSAP II would not serve communes where the staff of Twitezimbere cannot circulate safely and where they cannot work effectively. The technical process of selection of communes would work as follows. First, the Director of Twitezimbere would estimate the number of communes that can be covered in the annual work program. Then, he would select the first three-quarters of the total number of communes for that year from the poverty (access to infrastructure) index, starting with the poorest. He would then select the final quarter of the communes from the index of returnees, starting with those with the highest proportion of returnees in the population. If any of the same communes appears in both lists, or if other donors are active in those communes, or if any of the communes are insecure, the Director would repeat this process until he has selected enough communes to fulfill the project's annual capacity, The Director of Twitezimbere would then submit the list of communes for project investments for approval by its Executive Committee, with the non-objection of IDA. The Government will have a role in this decision since the Ministries of Finance, Plan, and Agriculture are represented on the Executive Committee, along with several NGOs, and other representatives of civil society. Project Appraisal Document Page 7 Burundi - Second Social Action Project Poverty monitoring. Another strategic choice is for smaller scale statistical monitoring. The BURSAP I is financing a relatively large-scale survey that will yield detailed baseline information on poverty. In contrast, the proposed project would finance less expensive surveys that can be repeated to provide up-to-date follow-up information on the evolution of poverty. Furthermore, these surveys would be oriented toward action to reduce poverty rather than simply gathering analytical infor:mation about poverty. To that end, the surveys would provide the information on the conditions of roads, water supply, health and education services, and local institutions that is necessary for planning reconstruction programs, and provide information on health and nutrition status. The relevant indicators would be selected through a survey of statistics users in the Government, donor, and NGO sectors. C: Project Description Summary 1. Project components: (see Annex 2 for a detailed description and Annex 3 for a detailed cost breakdown) The following are total costs of the project in current US dollars and include contributions from the Central Govemment and from beneficiaries. 1. Community investments Poverty 10.7 81.4 10.0 93.0 Of which alleviation a. Community participation c;1.00 7.6 1.0 100.0 b. Infrastructure rehab & construction 7.75 58.7 7.0 90.3 c. Investmentinyoung children& 2.00 15.1 2.0 100.0 mothers. 2. Poverty monitoring 1.20 9.1 1.0 83.3 3. Administration 1.25 9.5 1.0 80.0 Total 13.20 100.0 12.0 90.9 Note: Includes a government contribution of US$ 0.45 million, distributed to components 2 and 3, and community contributions of $ 0.75 million to component lb. 2. Key policy and institutional reforms supported by the project: The key policy reform would be to increase the target population's participation in selection and financing of infrastmcture sub-projects. (See section B3 above). 3. Benefits and target population: Benefits: The target populations would benefit, in the short-term, from wages generated through intensive employment on the infrastructure sub-projects. Young children and mothers would benefit from better nutrition and health. Returnees would benefit from improved social stability from mobilization of the community around a project that they selected and that are in the common interest of diverse groups. The inflow of wages and greater social stability would provide poor communes with some social protection during the transition. Some secondary short-term benefits to local agricultural production, household investment, and commerce will probably arise from local spencding of project wages. Project Appraisal Document Page 8 Burundi - Second Social Action Project The long-term benefits would be an improvement in the health, education, and economic productivity of the target population. The improvement would result from the improvement in social and economic services that flow from the infrastructure, information, education, and other training activities, and the strengthening of the communal Development Committees, which would become more able to maintain the infrastructure and manage social services. Target population. The target population would be the inhabitants of communes that are unusually poor, that are areas of return for refugees and internally displaced persons, and that have an unusually high number of children and other dependents, along with poor child nutrition. (See section B3 above for details). 4. Institutional and implementation arrangements: Implementing agency. A local non-profit and non-governmental organization Twitezimbere, has been executing the BURSAP I. Twitezimbere specializes in the execution of projects for donors, including the African Development Bank and the UNHCR. Twitezimbere works with the Government under an agreement (Convention) of April 20, 1993 and is recognized by the Ministry of the Interior as a non profit organization on the Ordonnance Ministerielle no. 205.01/082/93DU of February 25, 1993 Its capacity was somewhat weak through mid-1997, and then improved steadily and is now satisfactory. It has received formal opinions from its financial auditors for 1997 and 1998 that were expressed without reserves, indicating greatly improved financial management. Execution of civil works has also accelerated markedly, and disbursements have increased in parallel. Judging from the technical audits for 1997 and 1998-99, the quality of most of the work is satisfactory. Moreover, a survey of beneficiaries that was conducted alongside the technical audit suggested that the project is serving their interests and has won their support. Therefore the BURSAP II would make use of the capacity and experience that is already in place. Twitezimbere would implement the community infrastructure, community mobilization, and administration components of the proposed project. It would apply the procedures detailed in the Manuals of Procedures. Manuals of Procedures. The Manual of Technical Procedures is based on the one used by the BURSAP I, with some modifications to increase community participation, protect the environment and to set procedures for child health, nutrition, and development interventions. The project will also apply a new Manual of Financial Procedures to improve financial control, financial reporting, disbursement, and procurement. Early childhood development. Twitezimbere will contract and train intermediaries to implement the early child development activities. Early child development will be discussed with groups in the identified communes during the same community participation process as the community infrastructure activities. After the communities choose among the available options, the local intermediaries will develop tailored packages of interventions, including manuals to guide the implementation of the ECD activities. Project performance monitoring. Twitezimbere would recruit a local consulting company or contract with the ISTEEBU to monitor project perfonnance. Poverty monitoring. The national statistical institute (theISTEEBU would implement monitoring of national poverty. Project Appraisal Document Page 9 Bumndi - Second Social Action Project Consistency with sector standards and guidelines. To assure consistency with national development plans for education and health, the project would continue the practice of the BURSAP I by only building education and health infrastructure when they appear on the ministry national maps of desired infrastructure. To assure technical consistency, the project woul.d follow the practice of the BURSAP I by using the standard ministry-approved designs for education and health infrastructure. D: Project Rationale 1. Project alternatives considered and reasons for rejection: The Govermment of Burundi formally asked the Bank to launch a second phase of the BURSAP I. TIhe Country Team decided to launch the project because of the request and because of the urgent need to contribute to the stabilization and development of poor rural communities. A further consideration was that the methods and execution unit needed to provide quick assistance had been established in the context of the BURSAP I. The BURSAP I. The BURSAP I became effective in 1994 but execution did not accelerate until the end of the suspension of disbursement in mid-1997. Furthermore, the execution unit was strengthened at around the same time. Since then the project has enjoyed increased support from the Government and from civil society and a wide array of group. Moreover, the project has financed construction of an impressive and ever increasing volume of social infrastructure. The number of BURSAP I-financed infrastructure sub-projects completed rose from zero during the suspension to 23 in the second half of 1997 to 67 in the first three quarters of 1998. The number of sub-projects begun accelerated even faster from zero during the suspension to 98 during the second half of 1997 and 268 in the first three quarters of 1998. Furthermore, the quality of the sub-projects appears to be good, judging from the most recent financial and technical audits. AGETIP. There was also some discussion of launching an AGETIP that would promote employment and generate wage incomes through rehabilitation and construction of larger scale infrastructure. The advantages are that an AGETIP could provide roads necessary to improve access to markets and lower transport costs, and also agriculture infrastructure (such as water supply) that could improve productivity. The reason for the decision to proceed with the BURSAP first were the time necessary to prepare an AGETIP and the need to create a fumctioning execution unit Micro-credit. The Country Team considered a micro or agricultural credit component, but this was rejected because of the difficulty experienced by similar social funds in delivering such credit. The credit component of the social fund in Rwanda achieved a reimbursement rate of only around 50% and the African Development Bank-funded project in Burundi, executed by Twitezimbere, has not been particularly successful. Furthermore, the management of Twitezimbere, which would execute the proposed project, said that it would not be able to manage additional micro-credit. The Country Team concluded that financing of credit would be best handled through an agriculture or financial sector operation rather than the BURSAP II. Extent of community participation. The country team also discussed launching a project with nearly complete management by elected community groups -including procurement. The team's view was that rapid progress in decentralization and community participation might not prove possible and that it would be more effective to improve participation in steps, through a process of dialogue with the Government and other groups, piloting of participation, and mutual learLing. Moreover, the team felt that the extensive work needed to prepare a completely new system of participation would delay delivery of urgently needed development assistance to the poorest communities. Project Appraisal Document Page 10 Burundi - Second Social Action Project Early child development. There is some support for execution of early child development activities on the national level and through the formal system of health centers. However experience has shown that_ ECD efforts are most effective when the delivery structure is as close as possible to the community targeted. The main reason for choosing execution through the BURSAP II is that it facilitates this approach to ECD: it relies on the same process of community mobilization as the infrastructure sub- component. Groups of caregivers will be trained and encouraged to organize to find ways to improve the health of young children. They will select specific ECD activities from a menu of options in order to realize their identified objectives. Like the infrastructure sub-component, this could also serve to bring disparate sub-groups together and therefore reinforce the theme of peace education. Neither of these roles would be well served through execution by civil servants in the formal health system. 2. Major related projects financed by the Bank and/or other development agencies: Implementation Development Progress (IP) Objective (DO) Bank-financed Labor intensive public works, ABURTIP N/A N/A especially rural roads (planned for 2000) Budgetary support (planned for the Emergency Economic N/A N/A second half of 1999) Reconstruction Credit Refugee reintegration and resettlement Post-Conflict Fund N/A N/A (approved) grant __ Health (closes 6/30/2001) Health/Population II S S Infrastructure (to be extended) Water Supply S S Poverty alleviation (closes 12/31/1999) Social Action Project S S (BURSAP I) Other development agencies _ . _ g Poverty alleviation (closes 12/31/2000) African Development Bank (executed through _____________________________Twitezimbere) .,j .; Poverty alleviation / Agriculture IFAD (effective end-June 1999) Poverty Alleviation / Umbrella UTNDP (to be funded Program (effective once adequate through a Trust Fund) funds are raised).,V~ Belgian-Burundian Fund for Belgium 4. ' z Employment (ongoing) Bg , IP/DO Ratings: HS (Highly Satisfactory), S (Satisfactory), U (Unsatisfactory), HU (Highly Unsatisfactory) 3. Lessons learned and reflected in the project design: Simplicity. The most important lesson from the BURSAP I and from the Rwanda Food Security & Social Action Project is the importance of simplicity of design for delivering urgently needed support quickly in a post-conflict country with limited local capacity for project execution. Therefore, the proposed project focuses on the relatively simple activity of financing community infrastructure. Project Appraisal Document Page 11 Burundi - Second Social Action Project Community participation. A second lesson, drawn from numerous Bank studies, is the importance af true community participation in assuring the maintenance and continued operation of Bank-financed infrastructure. Therefore the project seeks to improve community participation in the selection, financing, and management of the infrastructure sub-projects and in delivery of services that benefit the health and nutrition of young children and mothers. Develop and retain project execution capacity. A third lesson, drawn from the health project in Burundi and the social funds in Rwanda is the importance of retaining skilled personnel to achieve quick and efficient execution of projects. A similar project in Rwanda has been delayed in part, by the closing of the operating project unit, and the difficulty of forming a good quality execution unit. 4. Indications of borrower commitment and ownership: The most substantive evidence of Government commitment is its support for the BURSAP I since the resumption of disbursement in 1997. For example, the Ministry of Plan supported the appointment of a project director and accountant, because the use of IDA funds and financial management skill of the staff of Twitezimbere had been weak. This led to improved financial management of the unit and to the first financial audit without reserves. With Government urging, the staff rapidly accelerated use of IDA funds and construction of community infrastructure. Furthermore, the Minister of Plan wrote the Bank to request the credit and the President requested that the Bank increase its activity in the social sector. The BURSAP II has the support of the ministries of Finance, Community Development, and Resettlement and Reintegration of Refugees and Displaced Persons. 5. Value added of Bank support in this project: The Bank will bring its experience of social funds in more than 30 countries, as well as country- specific knowledge gains through the implementation of the BURSAP I Moreover the Bank teams has accumulated experience with social funds and community development in neighboring Rwanda. E: Summary Project Analysis: (detailed assessments are in the project file, see Annex 8) 1. Economic: (supported by Annex 4) [ ]Cost-Benefit Analysis: NPV= N/A; ERR= N/A IXj Cost Effectiveness Analysis See annex 4. [ ] Other (specify): Beneficiary assessment to determine needs, and fiscal. analysis 2. Financial: (see Annex 5) Cost sharing and cost recovery Communities will be expected to contribute at least 10% of the cost of infrastructure sub-projects in cash, construction materials, and/or labor to provide an incentive for communities to maintain and operate infrastructure. This has worked well in the BURSAP I, where the average contribution is 15%. In the BURSAP II, the project would work with local administrations and with national and international NGOs to form community organizations that would maintain the infrastructure and assure its continued operation to produce social and economic services. Project Appraisal Document Page 12 Burundi - Second Social Action Project Financial controls Financial System and Audit. There will be a single special account for all components of the project that will be established and managed by Twitezimbere. The cash advance for three months activities will be for use by Twitezimbere, as the executing agency. The financial service at Twitezimbere will maintain financial records in compliance with accepted accounting principles and will prepare financial statements. The financial officer will make the necessary arrangements to conduct the annual audit of the project. The Government will ensure that the external auditors are independent and approved by the World Bank. The financial service will not be required to adopt the financial management report-based disbursement system, but will improve its financial management and reporting and conduct a mid-term review of its capacity to shift to the system. To these ends, the Manual of Financial Procedures includes a detailed chapter on accounting practices and reports and on financial controls. 3. Technical: The technical quality of infrastructure financed by the BURSAP I was evaluated through technical audits in 1997 and 1998 and found to be sound. The project's technical design is based on the use of simple, appropriate, and environmentally sound technologies, corresponding to the populations' needs and capacity for maintenance and operation. The design of education and health facilities is that of the government ministries. Labor intensive work is preferred wherever it is proven as technically and economically efficient and insofar as it generates employment opportunities. 4. Institutional: a. Executing agencies: Twitezimbere The overall management, coordination, and execution of the project will be entrusted to a not for profit organization, Twitezimbere. This is a light structure that is already executing the BURSAP I under an agreement (Convention) with the Government. Twitezimbere will also be responsible for contracting of intermediaries, liaison, supervision, monitoring, training, and long-term planning and policy support. In addition, it will assure overall financial management and accounting. To facilitate the programming, coordination and monitoring of BURSAP II, the Government will assure that Twitezimbere will continue to be administered by an Executive Committee (Conseil executij) where the Govenmment and civil society are represented. The Committee will convene regularly to review the progress of the project, ensure complementary and an efficient coordination of the activities of the different project components, and consolidate programs, budgets, as well as quarterly and annual progress reports. Twitezimbere serves as a secretariat to the Committee, and will therefore be in charge of preparing biannual activity programs and related annual budget proposals and the annual program. The budget and program will be submitted to the Committee for approval. Project oversight and orientation will be the responsibility of the Executive Committee. The Project Appraisal Document Page 13 Burundi - Second Social Action Project Committee will organize at least one annual meeting with Bilateral Donors, Line Ministries concerned, NGOs, communities, etc. The purpose of the meeting would be to (i) review the proposcd annual work program, (ii) review the implementation status and progress toward achievement of the project's objectives, (iii) decide on necessary corrective actions relative to project implementation, and (iv) coordinate the various projects in different areas. b. Project management: Accounting. Accounting and Internal Control Systems are set out in detail in the Manual of Financial Procedures and are satisfactory for providing reasonable assurances that accounts are properly recorded and resources safeguarded. Accounting Systems for the Project Management Unit for the BURSAP I have already been computerized using TOMPRO software, and the systems for the BIJRSAP II will continue to be computerized. The Chart of Accounts for the project is based on the National Chart of Account and in conformity with the Intemational Accounting Standard satisfactory to IDA. Financial Reporting. The annual Financial Statements of the project will be prepared in accordance with accounting principles acceptable to IDA. The Financial Statement will include at least a statement of sources and uses of funds, a statement of reconciliation of the special account and a balance sheet. The Financial Statements will be submitted to IDA no later than six months after the end of the fiscal year. The project will not be required to adopt the financial management report-based disbursement procedure at first. However a mid-term review will examine whether the project can move to the system. Auditing. An Independent Auditor acceptable to IDA will audit the financial statement of the project. The selection of the auditor is a condition of effectiveness. The auditor will provide an opinion on: a) Adequacy of Accounting and Intemal Control System b) Financial Management System c) Project Financial Statements d) Special Accounts and Statement of Expenditures e) Project Account Staffing. A Finance and Administration Director will be appointed in Twitezimbere in accordance with detailed terms of reference that have been developed. An accountant will be appointed in Twitezimbere to keep the detailed accounting records. Twitezimbere will contract for the additional technical expertise necessary to execute ECD activities. 5. Social: The project is expected to support reconciliation by mobilizing diverse groups to work together for their common benefit. The aim is not only ethnic reconciliation, but also to create a supportive environment for the reintegration of retuming extemal refugees and intemally displaced persons. The project staff is planning to reinforce this aspect of the project by developing and including early child development services and peace education into the community mobilization process. Project Appraisal Document Page 14 Burundi - Second Social Action Project 6. Environmental assessment: Environmental Category [ A [X] B [ ] C An environmental analysis was conducted for this project in April 1999. The report identifies potential positive environmental and social impacts of future sub-projects (rural water supply, short rural roads, public markets, health and child care centers, classrooms, swamp drainage, and planting of forests). These include employment opportunities and income generation, use of local technologies, increased availability of social services, and stabilization of soils through forest planting. Potential negative environmental impacts, largely as a result of construction activities, include soil erosion, deforestation, habitat destruction, as well as water contamination and sedimentation. The production of bricks by traditional methods inevitably leads to cutting of trees, since the bricks are baked using local charcoal. The project will pilot the use of the non-energy consuming mechanical press technique for producing bricks, to reduce the impact of the project on forests. The report presents mitigation measures for the various types of sub-projects. Furthermore, the report recommends that Twitezimbere strengthen its envirommental analysis capacity to ensure effective review, clearance, and implementation of the sub-projects. The Manual of Technical Procedures will include environmental guidelines for sub-projects that do not require such analysis, as well as procedures for sub-projects requiring environmental analysis. 7. Participatory approach: a. Primary beneficiaries and other affected groups: The primary beneficiaries are the residents of unusually poor communes and communes of return of refugees and displaced persons. (See section B3 above for details). The main change in the proposed project, relative to the BURSAP I, is therefore to empower residents of communities to choose and maintain their own project investments. The specific changes are to: o Use participatory assessments to establish local needs and investment priorities in each commune; o Modify the committees, if necessary, and through consultation with the public, to assure that they are representative. o Allocate a financial envelope of US$ 75,000 to the residents of each commune to finance their priority infrastructure sub-projects (but not including ECD activities); o Increase the choice available to communities by allowing them to chose a mix of infrastructure investments, within the financial envelope. b. Other key stakeholders: At the Central Govemnment level, the key stakeholders are the ministries that work with donors engaged in communal development: the Ministries of Finance, Plan, Reintegration and Repatriation of Displaced Persons and Refugees, and Communal Development. The key sector ministries involved in the project are the Ministries of Health and Education. At the local Government level, the key stakeholders are the Provincial Govemors4dministrateurs de commune, and Chefs de secteur. The cooperation of these officials, in partnership with the local population, is necessary for successful execution of sub-projects in the communes. The key multi-lateral partners are the African Development Bank (AfDB), the UNHCR, and the UNDP. The AfDB is executing a social fund through Twitezimbere (the BURSAP execution unit) using the BURSAP I Manual of Procedures -virtually co-financing with the Bank. The Bank and the Project Appraisal Docwnent Page 15 Burundi - Second Social Action Project AfDB reached an agreement to exchange information and collaborate on supervision of the project. The United Nations WHgh Commissioner for Refugees (UNHCR) is aL key partner because of their activity in resettling refugees, because of their collaboration with the Bank in execution of a Post- Conflict Fund-financed grant, and because they will provide Twitezimbere with information on areas of resettlement of refugees and displaced persons. The UNDP is a key partner because it can provide an effective forum for an exchange of information and cooperation among the donors involved in community and rural development Among the bilateral donors, Belgium is important because of its similar and moderate sized community reconstruction project, le Fonds Belgo-Burundais d 'emploi. The Agence fran,aise de developpement has opened discussions with Twitezimbere, with the agreement of the Govemnment, on financing of activities similar to the BURSAP II and also sent a representative to participate in the appraisal mission. They are now co.nsidering a grant of Euro 1 to 3 million (about US$ 1.1 to 3.2 million). The grant may be made directly to Twitezimbere, and executed in cooperation with the BURSAP II. The key non-governmental partners are the domestic and foreign NGOs that execute sub-projects financed by the BURSAP I. Action Aid may become a more important partner because they run a school in Burundi to train nationals in participation. F: Sustainability and Risks 1. Sustainability: The project relies mainly on intensified community participation arLd capacity building to sustain operations of the project (See section on participation). 2. Critical Risks: (reflecting assumptions in the fourth column of Annex 1) From Outputs to Objective A breakdown in the peace process would probably H Use of a participatory approach that reduce community participation by reducing contributes to reconciliation of diverse cooperation between diverse groups. groups by bringing them together for work that is in their common interest. The Govemnment does not transfer complementary M Country team will work with the Gov. (recurrent) budget resources for operation of health on a Public Expenditure Review. and education infrastructure. Negotiate minimum adequate spending levels in the context of an emergency rehabilitation credit. The Communities are not able to raise local resources M Reinforcement and training of to partly compensate for insufficient Govemment deve:lopment committees during the funding. participatory assessment. Community development committees do not conduct M Same as above routine maintenance. The Government does not increase training and does M Provide training and incentives for key not improve incentives for teachers, nurses, doctors, staff in the context of a public and other skilled staff of social infrastructure. expenditure review, preparation of any education and health credits, and through coordination with other donors. Poverty monitoring system is not used by policy N Use Df the data in economic and sector makers to target sub-projects and to design better workl with the Govemnment and in policies to reduce poverty. donor coordination. Project Appraisal Document Page 16 Burundi - Second Social Action Project From Components to Outputs If the peace process fails, increased physical H World Bank and other donor support insecurity would limit outputs. for the peace process through investment in the human resources of the poorest and most vulnerable groups, through stabilization and then expansion of the rural economy, and design of projects in ways that promote reconciliation. Unavailability of a sufficient number of people M The project will support an extensive trained in participatory assessments, financial training program in participatory management, construction, and management of small methods. businesses and NGOs would slow the outputs. Most Central and Local Government officials will M Central and local Government officials cooperate with the participatory approach. will be consulted regarding the Manuals of Procedures. No infrastructure sub- projects will be launched without written support from the Administrateurs de commune. Government pays its counterpart funds. N hifrastructure sub-projects component, which the Government supports, cannot be executed without counterpart funds. Moreover, the planned Emergency Reconstruction Credit should increase the availability of funds. Overall Risk Rating S Risk Rating - H (High Risk), S (Substantial Risk), M (Modest Risk), N (Negligible or Low Risk) 3. Possible Controversial Aspects: Project Appraisal Document Page 17 Burundi - Second Social Action Project G: Main Loan Conditions 1. Effectiveness Conditions: a) Twitezimbere has adopted the Manuals of Procedures, in form and substance satisfactory to the Association; b) Twitezimbere has established an accounting and financial managemnent system satisfactory to the Association; c) Twitezimbere has appointed the independent auditors in accordance with the Association's Consultant Guidelines; d) a Project Account has been opened and an initial deposit in the amount of $15,000 has been deposited therein; e) a Convention has been executed on behalf of the Borrower and Twitezimbere; f) standard conditions of effectiveness related to legal opinions. 2. Other: Assurances obtained at negotiations: a) The Government will contribute counterpart funds to the Project. Account that are adequate to execute the project's annual work program. b) The Government will provide adequate skilled staff (teachers, nurses, doctors) and recurrent budget to operate community infrastructure financed by the project. a) Twitezimbere will carry out a work program to improve financial reporting. H: Readiness for Implementation [X] 1. a) The engineering design documents for the first year's activities are complete and ready for the start of project implementation. [ ] 1.b) Not applicable. [ ] 2. The procurement documents for the first year's activities are complete and ready for the start of project implementation. [X] 3. The Project Implementation Plan has been appraised and found to be realistic and of satisfactory quality. [ ] 4. The following items are lacking and are discussed under loan conditions (Section G): Project Appraisal Document Page 18 Burundi - Second Social Action Project I: Compliance with Bank Policies [X] 1. This project complies with all applicable Bank policies. [ ] 2. The following exceptions to Bank policies are recommended for approval. The project complies with all other applicable Bank policies. Team Leader: Mr. M nahem Prywes Secto Manaker: Mr. Arvil Van Adams Country Director Mr. Emmanuel Mbi V Project Appraisal Document Annex I Bumndi - Second Social Action Project Page 1 of 4 Annex 1: Project Design Summary Burundi: Second Social Action Project Sector-related CAS Goal: Sector Indicators: Sector /Country Reports: (froma Goal to Bank I1. Support the peace . Percentage of poor in total a Household surveys of Mission) process. population. consumer expenditt-res and Continued willingness of 2. Deliver urgent support * Social indicators. social conditions. the Government to support for recovery of poor * Education and healih poverty alleviation efforts. communities and of statistics. the rural economy. * Public Expenditure Reviews. * Updates of the Poverty Note. Project Appraisal Document Annex 1 Bumndi - Second Social Action Project Page 2 of 4 Project Development Outcome I lmpact Indicators: Project Reports: (from Objective to Goal) Objective: I1. * Technical Audit reports * Improved mianagement 1. Diverse community * Qualitative assessment of prepared by independent of the governiment groups work together cooperation of diverse engineers. budget and irmproved to select, partly groups in selected sub- * Participatory assessments price stability. finance, construct, use, projects. of the project by * Improved donor and maintain * Percentage of projects that independent auditors support and therefore community reflect community (attached to the technical Improved availability investments. priorities. audit). of foreign exchange for 2. The poorest and most * Percentage contribution of vital imports. vulnerable groups the community to the * The embargo is not re- benefit from improved project, in material, labor, imposed. social and economic and money. services and improved * Percentage of Development social protection -- Committees that are still warough transitory functioning one year after wage incomes and completion of the social services that infrastructure sub-project. limit risks of poverty. Percentage of infrastructure adequately maintained one year after completion. 2. * Days of employment created. * Number of beneficiaries of infrastructure by type: Education, Health, Water Supply, Economic Infrastructure (markets, roads, forestry). * Capacity utilization rate of the infrastructure one year after construction. * Prevalence of wasting, stunting, and underweight of pre-school aged children, in communes where to project is active. * Qualitative assessment of adult-child interaction. Project Appraisal Document Annex 1 Burundi - Second Social Action Project Page 3 of 4 S~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ = @ Output from each Output Indicators: Project Reports: (from Outputs to Objective) component: la. . Quarterly and Annual . The peace process la. * Participatory rural progress reports prepared advances and physical . Comm-unities mobilized assessments conducted. by Twitezimnbere. security is adequate in for decision making and * Participatory assessment of . Technical Audit reports most provinces. delivery of community whether diverse groups prepared by independent . The Government contributions. work together. engineers transfers . Diverse groups within lb. . Participatory assessments complementary the commune work . Number of infrastructure of the project by recurrent budget toward a common goal. sub-projects started, by independent auditors resources for operation lb. type. (attached to the technical of health and education . Access to infrastructure * Quality of the audited audit). infrastructure. improved to meet infrastructure. . Bank supervision mirssion . Communities are able priority needs. . Number of agreements reports. to raise local resources * Improved capacity of signed with NGOs. a Financial Audit reports to compensate for some private sector: local * Number of infrastructure prepared by indeper dent of any shortfall in contractors, consultants sub-projects completed, by, auditors. Government funding. and NGOs. type. * Mid-term review and ICR . Community • A targeted and . Days of employment prepared by Twitezimbere Development transitional increase in created. and the Bank. Committees conduct wage incomes. . Training programs routine maintenance. lc. conducted for NGOs and * The Government . Improvement in health Smal and Medium increases training and and nutrition of young Enterprises. imnproves incentives for children and mothers. Ic. teachers, nurses, 2. . Households reached by doctors, and other . Recent & relevant data infonnation, education, and skilled staff of social made available for communication activities. infrastructure. policymaking and * Persons trained to improve * Poverty monitoring progress monitoring. the health and nutrition of system is actually used . Recent & relevant data young children. by policy makers to on progress of the 2. design better policies to project is available. * Statistical surveys produced reduce poverty and 3. by the national statistics target interventions. * Project effectively institute (ISTEEBU) and managed and other local institutions implemented. trained by the project 3. * Cost of administration as a percentage of expenditures on infrastructure sub- projects and (plus) community mobilization. * Average time for sub- project cycle (from identification to hand over). Project Appraisal Document Annex 1 Burndi - Second Social Action Project Page 4 of 4 Project ComponentslSub- Iiiputs: (IDA) Project Reports'. (from Components to components: 1. US$ 10.0 miillion a Quarterly and Annual Outputs) 1. Community la. US$ 1.0 miillion progress reports prepared . If the peace process investments: lb. US$ 7.0 mnillion by Twitezimbere. fails, increased a. Community ic. US$ 2.0 million * Financial Audit reports physical insecurity participation & 2. US$ 1.0 million prepared by independent would limit outputs. capacity building. 3. US$ 1.0 million auditors. . Unavailability of a b. Rehabilitation & . Technical Audit reports sufficient number of construction. prepared by independent people trained in c. Social services for engineers. participatory young children and . Participatory assessments assessments, financial mothers. of the project by management, 2. Poverty and project independent auditors construction, and performance (attached to the technical management of small monitoring. audit). businesses and NGOs, 3. Administration. . Bank supervision mission and in early childhood reports. development would slow the outputs. a Most Central and Local Govemment officials will cooperate with the participatory approach. * The Govemment of Burundi pays its counterpart funds. Project Appraisal Document Annex 2 Burundi - Second Social Action Project Page 1 of 3 Annex 2: Project Description Burundi: Second Social Action Projiect By Component: Project Component I - US$ 10.75 million Community investments The community investment component consists of: o Mobilization and strengthening of community capacity (and in particular, of representative Development Committees) to participate in selection, financing, and the continued operation of the infrastructure; o Contracting for the reconstruction and constr iction of small-scale community infrastructure, such as classrooms, health centers, rural water supply, short roads, public markets, swamp drainage, and planting of forests. o Investments in young children and mothers. The small-scale community infrastructure sub-component will resemble that in the BURSAP I. The main change aims at making community participation more substantive. This means ceding control and responsibility for some aspects of project activities to local residents. The BURSAP I succeeded in ceding responsibility for part of the financing of its investments: local residents contributed 15 percent, on average, of costs in tenus of materials, labor, and cash. The BURSAP I also attempted to cede selection of investments to local populations by requiring that applications originate with local groups, usually with the sponsorshLip of a NGO or a local government official. In practice, applications seem to have been generated by NGOs, in areas where they happen to be active, and to reflect their own programs rather than local choice. To deepen participation, the project will work with the local population to assure that they actually choose the infrastructure sub-projects that the project will finance, subject to a financial envelope. Moreover the project will reinforce community institutions that will maintain the infrastructure. The project will also apply the participatory approach to making the investments in pre-school children and mothers. The specific changes in the community activities of the BURSAF' II, relative to the BURSAP I, will then be to: > Launch the sub-project cycle through participatory assessments to establish local needs and investment priorities in each commune; > Allocate a financial envelope of US$ 75,000 to the residents of each commune to finance their priority infrastructure sub-projects; > Allow any commune that completes an infrastructure sub-project successfully to apply for a furither sub-project of up to US$75,000. > Increase the choice available to communes by allowing them to choose a mix of investments, within the financial envelope. > Provide a sub-set of targeted communes with the opportunity to strengthen grassroots capability for basic disease prevention, nutrition and other early child development activities according to priorities established by the community. Project Appraisal Document Annex 2 Burundi - Second Social Action Project Page 2 of 3 To this end, multi-disciplinary teams will work with the communities to determine their choice of priorities. The teams that conduct the participatory assessments could be composed of personnel from Twitezimbere, or qualified Agences d'encadrement (NGOs, consulting companies, or individual consultants). Otherwise, the roles of Twitezimbere and theAgences d'encad'-ement will remain the same as in the BURSAP I: that is, either can prepare the technical and financial proposals, procure construction of civil works following the Manual of Technical Procedures, supervise construction of the works, and implement the ECD activities. The teams will apply a methodology, in their work with communities that combines community meetings with use of graphics, thematic maps, calendars of the seasons, and activity profiles. This procedure will be used to identify the poorest people, in consultation with key sources, such as the 'Bashingantahe' (elders) and other social or religious leaders. To assure that the Development Committees are representative, the teams will add members after consulting with the beneficiaries and the local administration. This procedure, judging from the experience of a pilot, will lead to a better representation of women, young people, and other groups. Specifically, at least 30% of the members of each Development Committee should be women. The teams will work with each Development Committee to produce a report on the participatory assessment. The report will include (i) a brief description of the community and the social and economic conditions in the commune, (ii) a map of public infrastructure and of concentrations of population and economic activities in the commune, (iii) a list of priority investments with a brief explanation and justification, and (iv) the responsibilities of the committee and the community during and after construction. The report of the team must receive the approval of theAdministrateur de commune. Twitezimbere, through its personnel or through contractors, will be responsible for the supervision of the participatory process, using a standard form. Moreover, Twitezimbere will confirm that the infrastructures investments identified as community priorities are consistent with provincial development plans. The manual includes a detailed description of the participatory process to be followed. This kind of participatory approach implies that NGOs and local governments will no longer decide where the project will become active. The Director of Twitezimbere and the Executive Committee (where the Government and civil society are represented) and will decide, each year, in which communes the project will finance activities using the criteria below, using the criteria and procedures described in section B3, pages 5-7, of the main document. To accelerate execution of the project, Twitezimbere will reinforce its capacity by limited recruitment and by training personnel. A specialist in participation will be recruited to reinforce and advance participation activities. Moreover, Twitezimbere will launch a training program in participation for its staff and for the staff of the ministries concerned, NGO partners, UN organizations, and consultants. The third and final sub-component will start by delivering a basic package of information, education, and communication to promote early child development during the participatory assessment process for a sub-set of the communes with infrastructure sub-projects. Allcollines from each of these communes will be eligible to participate. The participatory process will produce a brief assessment, by the community, of the needs of young children and mothers, ranked by order of priority. Also as a result of the participatory process, there will be a simple development plan reflecting the ranking of ECD sub-projects by order of preference. Options that communes could choose for subsequent Project Appraisal Document Annex 2 Burundi - Second Social Action Project Page 3 of 3 activities to meet the identified priorities could include micro-nutrition., gardens to provide nutritio, monitoring of the growth of infants and children, and trauma counseling. The identified activities will be prepared for each commune and implemented with support from contracted Agences d'encadrement. Training of community mobilizers, health workers and educators will back these activities. Twitezimbere will provide monitoring and evaluation. Project Component 2 - US$ 1.20 million Strengthening of statistical capacityfor monitoring the evolution ofpoverty. Like the BURSAP I, the new project will finance strengthening of the capacity of the ISTEEBU to monitor the evolution of poverty. To strengthen capacity, the project will support a systematic program of retraining of statisticians and will supply some equipment. With regard to monitoring, the BURSAP I financed a relatively lar.ge-scale survey that produced detailed baseline information on poverty. In contrast, the BURSAP II will finance less expensive surveys that can be repeated to provide up-to-date follow-up information on the evolution of poverty. Furthermore, these surveys will be oriented toward aetion to reduce poverty rather than simply gathering analytical information about poverty. To that end, the surveys will provide the information on the conditions of roads, water supply, health and education services, and local institutions that is necessary for planning reconstruction programs. The relevant indicators will be selected through a survey of statistics users in the Government, donor, and NGO sectors. Project Component 2 - US$ 1.25 million Administration. The project will be executed by Twitezimbere, which is a non-governmental organization that executes the BURSAP I under an agreement (Convention) with the Government. A board that represents a cross-section of Burundian society governs Twitezimbere: NGOs, individual members, and the Ministries of Finance, Plan, and Agriculture. Project Appraisal Document Annex 3 Burundi - Second Social Action Project Page I of 1 Annex 3: Estimated Project Costs Burundi Second Social Action Project ~~~ ~ ~ 0" *.9013 OUNUM 1. Community investments 5.77 3.53 9.30 a. Community participation 0.54 0.33 0.87 b. Infrastructure rehab & construction 4.16 2.55 6.71 c. Investment in young children and mothers 1.07 0.65 1.72 2. Poverty monitoring 0.65 0.39 1.04 3. Administration 0.67 0.41 1.08 Total Baseline Cost 7.09 4.33 11.42 Physical Contingencies (10%) 0.82 0.50 1.32 Price Contingencies (3.5%) 0.29 0.17 0.46 Total Project Costs 8.20 5.00 13.20 Project Appraisal Document Annex 4 Burumdi - Second Social Action Project Page 1 of 2 Annex 4: Cost Effectiveness Analysis Summary Burundi Second Social Action Project The analysis of costs shows that the current social action project, the BURSAP I, has generally achieved unit costs of construction and lower percentage costs of administration than comparable projects. The table below shows that the BURSAP I achieved lower unit costs of construction, per square meter, of primary schools, health centers, and secondary schools, than other projects. A cost-benefit analysis is not feasible because of the difficulty of measuring benefits in a rural economy where data are scarce and where goods, services, and labor are often not traded and therefore are difficult to value in money terms. The BURSAP II will use the same execution unit and sirmilar methods of execution, and especially procurement. This suggests that the BURSAP II will be a highly cost efficient project. Comparison of Unit Costs of Construction of Community Infrastructures: BURSAP l-Twitezimbere versus other Projects Implementing Type of Net Costs (net Local r c Surface Built Unit Costs FBu Unit costs Agency hIfrastructure of supervision) Participation (sq m.) (sq m.) US$ (sq m.) UNHCR Primary school 24,000,000 0 24,000,000 403.00 58,824 130.72 ARP Primary school 23,437,106 0 23,437,106 505.76 46,340 102.98 (AustrianNGO) Health center 13,598,382 0 13,598,382 186.03 73,100 162.44 Secondary school (first level) 21,710,674 0 21,710,674 466.34 46,555 103.46 BPE Primary school 25,000,000 0 25,000,000 432.00 57,870 128.60 (Public agency) Communal school 18,000,000 0 18,000,000 328.00 54,878 121.95 CEPBU Primary school 35,000,000 3,850,000 38,850,000 542.00 71,679 159.29 (Religious NGO) Health center 74,000,000 8,140,000 82,140,000 786.00 104,504 232.23 Twitezimbere* Primary school 10,566,050 1,943,069 12,509,119 432.00 28,956 64.35 Health center 12,401,794 1,445,867 13,847,661 192.00 72,123 160.27 Secondary school (first level) 11,270,234 2,144,048 13,414,282 520.00 25,797 57.33 The cost comparison assumes the each type of infrastructure is homogeneous and does not vary with the executing agency. Dollar values have been calculated using the exchange rate FBu 450=$US I for the period covered by the table. * Net costs and total costs are the average costs of construction cited in the technical audit. for the social infrastructure mentioned above. Project Appraisal Document Annex 4 Burundi - Second Social Action Project Page 2 of 2 The weighted-average cost of administration (AC) of the BURSAP I, over 1995-97, as a percentage of project disbursements, was about 8% (see table below). This is significantly lower than the 13% cost_ in Rwanda, the 17% cost in Eritrea and the 30% cost in Zambia. The cost of administration is probably also below the comparable costs in the social funds in Burkina Faso, Angola, and Mauritania. Senegal and Malawi appear to have achieved even lower costs of administration. Some difficulties in comparability arise from differences in what is included in the cost of administration. In the case of Zambia, the administrative costs include costs for training, capacity building, micro-project launch workshops, and field appraisals. A better measure of the cost- effectiveness of the project is obtained when the ratio of administrative expenditures to project disbursements is computed as the average over the life of the project. A weighted-average has been calculated in order to take into account the variable level of administrative costs according to the different phases of the project. Comparison of Administrative Costs (AC) of Several Social Funds in Africa Senegal Malawi Zambia Burundi Rwanda Burkna Niger Angola Eiitrea Mauritania Mali Agetip Agetip Nigetip Amextipe 96-99 97-98 92-98 95-97 95-98 92-97 91-97 96-98 93-98 92-98 AC/Total costs, % 8.53 4.84 10.4 ACIDisbursement, % 4.40 4.50 29.89 8.89 6.32 18-20 17.15 7.7 AC/Investment costs, % Average, % 14.21 WeightedAverage, % 8.04 12.57 AC: Admninistrative costs (project management costs) Please note that the definition of administrative costs varies between proj ects. Senegal: the ratio was obtained using estimated disbursements for 1999. Malawi: Figures were only available for 1997 and 1998. Zambia: Administrative costs are high because they include capacity building, additional staff for centralized responsibilities, and task assistants for regional officers. Bunmdi and Rwanda: Two types of ratios were calculated here. Ihe simple average of administrative / disbursement ratios for consecutive years and a weighted average across consecutive years to account for the different volume of adminirstrative costs. Niger. The expenses of the Nigetip have been used as an estimate of operating costs. Total project costs includes non-IDA contributions. Eritrea:The Community Development Fund Program is a multi-donor project and its admrinistrative costs were fundedby the Govemment of Eritrea There is no indication of the administrative costs linked to IDA funds Project Appraisal Document Annex 5 Burnmdi - Second Social Action Project Page 1 of 1 Annex 5: Financial Summary Burundi: Second Social Action Project Years Ending December 31 _~~~~~~Ya I ea Year 3 l Year 4 Total Financing Required Project Costs Investment Costs 3.00 3.00 3.00 3.00 Recurrent Costs 0.30 0.30 0.30 0.30 Total Project Costs 3.30 3.30 3.30 3.30 Interest duning construction Front-end Fee Total Financing 3.30 3.30 3.30 3.30 Financing IBRD/IDA 3.09 3.05 2.93 2.93 Govemment Central 0.03 0.06 0.18 0.18 Provincial Co-financiers User Fees/Beneficiaries 0.18 0.19 0.19 0.19 Total Project Financing 3.30 3.30 3.30 3.30 Project Appraisal Document Anmex 6 Burundi - Second Social Action Project Page 1 of 6 Annex 6: Procurement and Disbursement Arrangements Burundi: Second Social Action Project A. Procurement. Procurement of goods and works required for the project and to be financed out of the proceeds of the credit shall be govemed by the Guidelines for Procurement under IBRD Loans and IDA Credit published in January 1995, and revised in January and August 1996, September 1997, and January 1999. Procurement of consultants' services shall be governed by theGuidelines for the Selection and Employment of Consultants by World Bank Borrowers published by the Bank in January, revised in September 1997, and January 1999. The following paragraphs describe ihe procurement arrangements under the project. Capacity Assessment of Twitezimbere for Procurement Twitezimbere will be the executing agency for this project: it has been executing the Burundi Social Action Project (BURSAP I, Cr. 2494-Bu) since 1994. This local NGO took some time to accelerate execution to a reasonably rapid pace. Since 1996 and furthernore since the second half of 1997, Twitezimbere has demonstrated impressive progress in the number of sub-projects financed as well as in the quality of the execution. Technical audits took place in 1997 and 1998 and thoroughly reviewed selected sub-projects, including the procurement procedures. According to these audits the executing agencies (Twitezimbere and other NGOs) carried out procurement in accordance with the provisions of the Manual of Procedures (of the BURSAP I). The Manual of Technical Procedures for the BURSAP II was built up from the Manual for the BURSAP 1, and was, in particular, revised to strengthen community participation. In this respect the following two main changes will be introduced: (i) community participation will be allowed for smaller works and, (ii) standard bidding documents for civil works costing less than US$ 30,000 equivalent per contract will be incorporated in the Manual. These latter documents will be made available in Kirundi, the national language of Burundi. Procurement Methods for Civil Works and Goods (see Table A) Civil works for infrastructure sub-projects involving the construction, repair, and rehabilitation of community infrastructure, including primary and secondary schools, health centers, water supply and sanitation, feeder roads, bridges, limited markets facilities, etc. costing US$ 30,000 or more but less than US$ 75,000 equivalent per contract up to an aggregate amount of US$ 3 million will be procured through NCB procedures described in the Manual of Technical Procedures of Twitezimbere. Civil works contracts for sub-projects costing US$ 7,500 equivalent per contract or more but less than US$ 30,000, up to an aggregate amount of US$ 3.5 million will be procured through obtaining at least three bids from small local contractors as described in the Manual of Technical Procedures. Smaller works costing less than US$ 7,500 equivalent per contract involving full community participation for an aggregate amount of US$ 0.7 million will be carried out in accordance with procedures acceptable to the Bank. The procedures are described in details in the Manual of Technical Procedures. Goods for infrastructure sub-project. Furniture and equipment for community sub-projects costing US$ 30,000 or more but less than US$ 75,000 equivalent per contract up to an aggregate amount of US$ 200,000 will be procured through NCB procedures described in the Manual of Technical Procedures. Contracts costing less than US$30,000, equivalent, up to an aggregate amount of US$ Project Appraisal Document Annex 6 Bunndi - Second Social Action Project Page 2 of 6 350,000 will be obtained through national shopping procedures as described in the Manual of Technical Procedures. Goods for Twitezimbere. Vehicles, computer and office equipment for Twitezimbere offices costing more than US$ 30,000 equivalent but less than US$ 100,000 per contract, up to an aggregate amount not to exceed US$ 200,000 may be procured through NCB procedures described in the Manual of Technical Procedures. Office supplies, equipment and furniture for Twitezimbere estimated to cost less than US$ 30,000 equivalent per contract will be procured through national shopping based on price quotations obtained from three reliable suppliers provided that the aggregate amount does not exceed US$ 100,000. Prior review thresholds (Table B) IDA financed contracts above the threshold of US$ 75,000 for works and goods will be subject to IDA's prior review. All other goods and works contracts will be subject to post review. However, a technical audit will be performed during each year of project implementation. About 10 percent of the on-going or completed sub-projects (selection will be on a random basis) will be subject to an in-depth review as part of the annual technical audit. This review will look in detail into compliance with the procurement procedures utilized to ensure conformity with Twitezimbere's Manual of Technical Procedures. Consultants (Technical Assistance and Training) Consultants Services for infrastructure and capacity building sub-projects. Consultant services will be procured in accordance with the Guidelines for Selection and Employment of Consultants by World Bank Borrowers of January 1997, revised September 1997 and January 1999. Except for the annual audits (financial and technical) most of the contracts will be procured locally due to the nature and scope of consultancy services which will be required. The types of services concerned are the following: (a) participatory rural assessment; (b) design and supervision of infrastructure sub-projects; (c) training of beneficiaries in basic skills; and (d) training of Twitezimbere, ISTEEBU and NGO staff. Competition will be encouraged to the extent possible, mainly through Least Cost Selection (LCS) for consultants services related to sub-projects and Quality-Based and Cost Selection (QBCS) for studies or consultants services provided to Twitezimbere and other partners in the implementation of the project. Single Source Selection will be used for an NGO when it appears that its local involvement, experience, knowledge and approach offer the best guarantee for the success of the sub- projects. Prior Review for Consultants IDA prior review for consultants' services will required for: (a) all contracts (for any amount) for Twitezimbere high level personnel, and (b) contracts with individuals with an estimated value of US$ 20,000 or more and (c) contracts with consulting firms with an estimated value of US$ 50,000 or more. The technical evaluation report for recruitment of consulting firms will be subject to prior review for contracts exceeding US$ 50,000 each and amendments to contracts with consulting firms lead to substantial modifications of the Terms of Reference and the terms of employment. All contracts not subject to the above thresholds will be subject to post review. Project Appraisal Document Annex 6 Burundi - Second Social Action Project Page 3 of 6 TABLE A: PROJECT COSTS BY PROCUREMENTARRANGEMENTS (in US$ million equivalent) 1. Works for sub-projects - 3.00 3.50 0.7 7.20 - (3.00) (3.50) 0.70 (7.20) 2. Goods - 0.40 0.45 - 0.85 for sub-projects and equipment, vehicles, - (0.40) (0.45) - (0.85) futniture, materials for ECD and PIU 2. Consultant's Services - - - 2.20 2.20 - - - (2.20) (2.20) 3. Training - - - 0.65 0.65 - - - (0.65) (0.65) 4. Operating Costs - - - 1.55 1.55 - - - (1.10) (1.10) Total - 3.40 3.95 5.10 12.45 (3.40) (3.95) (4.65) (12.00) Note: The amount in the upper part of each box is the cost of the activity including Government contributions, but not community contributions, which are made in kind. The amount in parentheses is the cost attributed to the IDA credit. 'National Competitive Bidding or IAPSO Z At least 3 Price quotations (Works) or National Shopping (Goods). Figures in parentlhesis are amounts to be financed by IDA Credit. Project Appraisal Document Annex 6 Bunmdi - Second Social Action Project Page 4 of 6 Table B: Thresholds for Procurement Methods and Prior Review 1. Works =/>US$30,000 NCB >US$75,000/US$ 0.6 million for sub-projects =/>US$7,500US$30,000 NCB Post Review (samples) b/ equipment, vehicles, furniture, and materials 3. Services FiTms QCBS, LCS >US$50,000/US$ 0.2 million Individuals Short List >US$20,000/US$ 0.2 million NGOs Single Source All / US$ 0.5 million Twitezimbere High level Short List All / $ 0.2 million Staff LCS All / US$ 0.2 million Technical Audits LCS All / US$ 0.2 million Financial Audits 4. Miscellaneous N/A N/A N/A Total Value of Contracts subject to prior review: US$2.10 or 17.5% of total costs. Project Appraisal Document Annex 6 Burundi - Second Social Action Project Page 5 of 6 Table C: Allocation of Loan Proceeds I. Works 6,340000 100% 2. Vehicles, Equipment, 730,000 100% furniture, goods 3. Consultant Services 2,070,000 100% 4. Training 600,000 100% 5. Operating Costs 1,060,000 90% until 12/3 1/00 80% /util 12/3 1/01 40% thereafter Unallocated 1,200,000 __ Total 12,000,000 B. Disbursement Reporting and LACI. The Bank is introducing an initiative to change loan administration in Bank Group projects, the Loan Administration Change Initiative (LACI). This initiative assists projects to put in place sound Financial Management, Procurement and Output Monitoring Systems. Where appropriate, if these systems are in place, disbursements may be made on the basis of agreed quarterly Project Management Reports (PMR) rather than on the basis of individual invoices or statements of expenditures. An assessment has been made that the project management unit will not apply LACI for BURSAP II at first. Instead, the project will begin by establishing a monitoring and evaluation process that includes output-monitoring reports consistent with the PMR requirements. The financial, procurement and output monitoring systems will be reviewed as part of the mid-tierm review at the end of the second year to assess whether it will be appropriate to move to PMR based disbursement at that time. The quality of general management by Twitezimbere, judging by the cost of execution, is relatively strong at a bit under 9 percent of project expenditures, which is relatively low for a social fund. The weakness of the BURSAP I had been in the area of financial management. However, with the appointment of a new Director and a new accountant, the quality of financial management has improved steadily, and this is reflected in the financial audits for 1996 and 1997. The market for people with technical skills to manage a project is tight in Burundi. The difficulty of finding qualified employees places a limit on the expansion of the project staff and on expansion of project activities. It is therefore unlikely that Twitezimbere could execute more than US$ 4 to 5 million in high quality activities per year (slightly higher than the current rate). Project Appraisal Document Annex 6 Burundi - Second Social Action Project Page 6 of 6 Special Account. The proposed allocation of the credit is shown in Table C. The IDA Credit will be disbursed over a period of 4 years from 2000 to 2003, with a closing date of December 31, 2003. The estimated disbursement schedule is shown in Annex 5. All applications to withdraw proceeds from the Credit will be fully documented for contracts subject to prior review by IDA (contracts valued for civil works not less than US$ 75,000, for equipment and vehicles not less than US$ 75,000, and for consultant services not less than US$ 20,000 for individual consultants and US$ 50,000 for firms. For the contracts not subject to prior review by IDA, disbursements will be made against Statements of Expenditures (SOEs) certified by the Director of the Project Management Unit (PMU). Supporting documentation will be retained by Twitezimbere and will be available for review as requested by IDA supervision missions and program auditors. The accountant of Twitezimbere under the guidance of its Administrative and Financial Director will provide technical assistance and monitoring. To expedite disbursements and ensure that project funds will be available when needed, the Government will open a Special Account in the Central Bank. The authorized balance of the Special Account will start at US$ 500,000 and then rise to US$ 800,000 once disbursements have reached SDR 1,500,000. These amounts will represent IDA's anticipated share on the basis of monthly applications supported by full documentation of SOEs as required and verified by Twitezimbere's Administrative and Financial Director. Counterpart Funds - Project Account. The Government will open a Project Account in a commercial bank to cover its counterpart contribution to the project, to be used exclusively to meet expenditures under the program. The Borrower will deposit into the Project Account an initial amount of US$ 15,000 equivalent in Burundi Francs. The establishment of this account with an initial deposit of US$ 15,000 equivalent in Burundi Francs will be a condition of Credit Effectiveness. Accounting and Auditing. Twitezimbere will maintain accounting records for the project components that they manage directly. In addition to their financial statements, Twitezimbere agreed to submit a quarterly report, one month after the end of each quarter, presenting a summary of sources and uses of funds, procurement/ contracts schedule and statement and summary of expenditures by component and by category. Assurance has been obtained from the borrower that the Twitezimbere's special accounts, the financial statements, and the project account will be audited annually in accordance with international audit standards by experienced and recognized audit firms satisfactory to IDA. After the end of the fiscal year but no longer than six months after its completion, the Government will present Audit Reports of the Special Account and the Project Account with the special opinion on the use of the Special Account and the SOEs. The contracting of Auditors, for at least the first year, acceptable to IDA, and a certification by Auditors that Project Accounting System is operational, is a condition of Credit Effectiveness. Project Appraisal Document Annex 7 Burundi - Second Social Action Project Page 1 of 1 Annex 7: Project Processing Budget and Schedule! Burundi: Second Social Action Project Time taken to prepare the project (months) 10 _ First Bank mission (identification) 23102/1999 23/02/1999 Appraisal mission departure 05/17/l1999 05/l1911999 Negotiations 06/01/1999 06/10/1999 Planned Date of Effectiveness 01/01/2000 I /19 Prepared by: Ministries of Finance, Plan, Community Mobilization, and of Resettlement and Reinstallation or Refugees and Displaced Persons, and by Twitezimbere. Financial assistance for preparation was provided by the Norwegian Trust Fund for Early Child Development, Norwegian Trust Fund for Statistics, and the Spanish Trust Fund (for Intern). Bank staff who worked on the project included: Menahem Prywes, AFTH4 Economist and Task Team Leacler Essimi Menye, AFTI1I Statistician Markus Kostner, SDVPC Economist and Social Scientist (Participation) Siaka Bakayoko, AFTS2 Financial Management Specialist Bertrand de Chazal, AFTS2 Fina-ncial Management Specialist Abdul Haji, AFTS2 Financial Management Specialist Ren6 Welter, AFTS2 Procurement Specialist Agnes Albert-Loth, LOAAF Financial Management Specialist Pascale Helene Dubois, LEGAF Lawyer Mpoy-Kamulyi, LEGAF Lawyer Pamphile Kantabaze, Resident Mission Operations Officer Juvenal Nzambimana, Resident Mission Operations Officer Prosper Nindorera, Resident Mission Operations Officer Inmaculada Montero Luque, Intern Cost efficiency Susan Opper, AFTH2 Education Specialist (Early Child Development) Christine Pena, AFTH4 Economist/Early Child Development Anne Kielland, AFTH2 Child Development Specialist Nathalie Lopez-Diouf, AFTH4 Language Team Assistant Anne Anglio, AFTH4 Language Team Assistant (Foimatting) Project Appraisal Document Annex 8 Burundi - Second Social Action Project Page 1 of 1 Annex 8: Documents in the Project File BURUNDI SECOND SOCIAL ACTION PROJECT A. Project Implementation Plan Manual of Technical Procedures. Manual of Financial Procedures. B. Bank Staff Assessments C. Other Analyse environnementale des activits futures du Second Projet d 'Actions Sociales (Environmental Analysis of the Expected Activities of the Second Social Action Project), by Damas Nduwumwami, Consultant, April 1999. Project Appraisal Document Annex 9 Burundi - Second Social Action Project Page I of I Annex 9: Statement of Loans and Credits Status of Bank Group Operations in Burundi Operations Portfolio As of 31-May-99 Number of Closed Projects: 43 Active Projects BI-PE-216 1995 Govemrnment Health/Population II 0.00 21.30 0.00 14.51 9.26 0.00 BI-PE-227 1993 Government Social Sector 0.00 10.40 0.00 2.90 2.90 0.00 BI-PE-217 1992 Government Water Supply Sector 0.00 32.70 24.32 2.64 26.77 2.46 BI-PE-204 1990 Ministry of Public Health Transport Sector 0.00 43.20 6.59 2.41 6.94 -1.39 Total 0.00 107.60 30.91 22.46 45.87 1.07 Active Projects Closed Projects Total Total Disbursed (IBRD and IDA) : 54.50 539.58 594.08 of which has been repaid : 0.00 36.24 36.24 Totai now held by 1BRD and IDA 76.69 485-56 562.25 Amount sold 0.00 8.46 8.46 of which repaid 0.00 8.46 8.46 Total Undisbursed : 22.46 2.94 25.40 Actual disbursements to date minus intended disbursements to date as projected at appraisal. Note: Disbursement data is updated at the end of the first week of the month and is currently as of 30-Apr-99. Project Appraisal Document Annex 10 Burundi - Second Social Action Project Page 1 of 3 Burundi at a glance 10/1/98 Sub- POVERTY and SOCIAL Saharan Low- Burundi Africa Income Developfnfen: diarnond' 1997 Population, mid-year (millions) 6.6 614 2,048 Life expectancy GNP per capita (Atlas method, US$) 180 500 350 GNP (Atlas method, USS billions) 1.2 309 722 Average annual growth, 1991-97 Population (%) 2.6 2.7 2.1 X G Labor force (%) 2.6 2.6 2.3 GNP Groaa per - primary Most recent estimate (latest year available, 1991-97) capita enrollment Poverty (% ofpopulation belowrnationalpoverlyl/ne) Urban population (% of total population) 7 32 28 Life expectancy at birth (years) 49 52 59 Infant mortality (per 1,000 live births) 98 90 78 Child malnutrition (% ofchildren under5) 12 .. 61 Access to safe water Access to safe water (% of population) 37 44 71 Illiteracy (% of population age 15+) 65 43 47 Gross primary enrollment (% ofschool-age population) 43 75 91 -Burundi Male .. 82 100 -- Low-income group Female 40 67 81 - KEY ECONOMIC RATIOS and LONG-TERM TRENDS 1976 1986 1996 1997 EcononmIc ratlos GDP (US$ billions) 0.45 1.2 1.1 1.2 Gross domestic investment/GDP 9.1 11.6 8.8 4.4 Trade Exports of goods and services/GDP 13.7 11.7 4.4 8.1 Gross domestic savings/GDP 6.2 1.1 -0.9 -0.8 Gross national savings/GDP 4.2 4.9 2.1 2.5 Current account balance/GDP .. -6.7 -6.7 -2.0 Domestic i X Interest payments/GDP 0.1 1.0 1.1 1.1 SangsInvestment Total debt/GDP 5.0 45.7 1002 100.8 Total debt service/exports 4.5 24.5 97.2 49.5 Present value of debtGDP .. .. 47.1 Present value of debt/exports .. .. 947.4 Indebtedness 197646 1987-97 1996 1997 199842 (average annual growth) GDP 3.7 -1.4 -8.4 0.7 3.9 --Burundi GNP per capita 1.1 -3.9 -10.8 -1.5 0.9 --Lo-Income group Exports of goods and services 5.9 0.1 -49.3 27.1 7.3 A STRUCTURE of the ECONOMY 1976 1986 1996 1997 Growth raltes of output and Investment (%) (% of GDP) 20 Agriculture 64.8 58.5 67.5 46.1 Industry 12.8 13.5 16.9 11.5 o 0 Manufactunng 8.9 8.8 10.6 7.7 -20 Services 22.4 28.0 25.6 42.4 -40 - Private consumption 85.1 89.1 90.8 90.4 -eo General government consumption 8.7 9.8 10.0 10.5 GDI C GDD Imports of goods and services 16.7 22.3 14.1 13.3 G (average annual growth) 1976-46 1987-97 1996 1997 Growth rates of exports and imports (%) Agriculture 2.4 -1.3 1.5 0.5 40 - Industry 7.3 -4.4 -13.2 -7.9 20 ------ Manufacturing 5.8 -5.1 -16.4 -0.7 o' Services 6.6 0.1 -11.4 3.2 2 Private consumption 3.2 -0.2 -11.6 13.4 40 General government consumption -2.9 -0.1 -18.8 9.2 V Gross domestic investment 13.7 -7.9 0.0 -44.5 d Imports of goods and services 6.9 -1.9 -43.8 9.4 Exports lmports Gross national product 3.9 -1.3 -8.5 1.0 - Note: 1997 data are preliminary estimates. *The diamonds show four key indicators in the country (in bold) compared vvith its income-group average. If cata are missing, the diamond will be incomplete. Project Appraisal Document Annex 10 Burmndi - Second Social Action Project Page 2 of 3 Burundi PRICES and GOVERNMENT FINANCE 1976 1986 1996 1997 Inflation (%) Oomestic prices 3 (% change) 30_ _ Consumerprices 6.9 1.7 26.4 21.1 20 Implieit GDP deflator 8.1 -43 22.4 26.8 10 Government finance (% of GDP, includes cunent grants) a Current revenue 15.8 12.3 9.8 92 93 94 9S 96 97 Current budget balance 2.8 -0.8 -3.2 GDP deflator CPI Overall surplus/deficit -8.1 -8.2 -6.1 TRADE (US$ millions) 1976 1986 1996 1997 Export and import levels (USS millions) Total exports (fob) 129 40 93 300 Coffee 106 29 73 250 Tea .. 5 5 13 20 Manufactures 5 9 4 4 Total imports (cil 207 117 133 150 Food 16 18 23 100 Fuel and energy27 17 0 Capital goods63 3 37A AM N Export price index (1995=100) .. 122 76 125 91 92 93 94 95 9B 97 Import price index (1995=100) .. 71 103 97 H Exports a Imports Temisoftrade(1995=100) 172 73 129 BALANCE of PAYMENTS 1976 1986 1996 1997 Currentaccount balance to GDP ratio (%) (USS millions) Exports of goods and services 62 141 50 100 o Imports of goods and services 75 268 160 165 -2 Resource balance -13 -127 -110 -65 Net income 0 -21 -14 -11 Net current transfers 67 48 52 -6 Current account balance -81 -76 -24 - Financing items (net) 102 Is -7 -1- Changes in net reserves -15 -21 58 31 -12 Memo: Reserves including gold (US$ millions) 49 70 145 60 Conversion rate (DEC, lacalVUSS) 86.3 114.2 302.8 352.3 EXTERNAL DEBT and RESOURCE FLOWS 1976 1986 1996 1997 (USS millions) Compositon of total debt, 1997 (USS millions) Total debt outstanding and disbursed 22 549 1,139 1,247 IBRD 0 0 0 0 G:72 IDA 5 191 588 567 F: 1 I E: 169 _ Total debt service 3 35 55 53 IBRD 0 0 0 0 IDA .0 2 7 9 B:667 Composition of net resource flows Official grants 21 51 164 Official creditors 3 88 14 50 Private creditors 0 -6 -1 0 D: 40 Foreign direct investment 0 2 1 Portfolio equity 0 0 0 C: 30 World Bank program Commitments 8 32 0 0 A - 1RD E - Bilateral Disbursemends 1 44 17 12 8 - IDA D - Oter mutilateral F - Private Principal repayments 0 1 3 4 C - MF G - Short-term Net flows 1 43 14 7 1 X Interest payments 0 2 4 5 Net transfers 1 42 10 3 Development Economics 10/1/98 Project Appraisal Document Anmex 10 Bumndi - Second Social Action Project Page 3 of 3 Burundi PRICES and GOVERNMENT FINANCE 1976 1986 1996 1997 Inflation (%J Domesdc prices 3D (% change) .0 Consumer prices 6.9 1.7 26.4 21.1 20 Implicit GDP deflator 8.1 -4.3 22.4 25.8 10 Governmentfinance (% of GDP, includes cunrent grants) Current revenue .. 15.8 12.3 9.8 92 93 94 95 9D 97 Current budget balance .. 2.8 -0.8 -3.2 - GDP deflator *GCPI Overall surplus/deficit .. -8.1 -8.2 -6.1 TRADE 1976 1986 1996 1997 Export and import levels (US$ millions) (US$ millions) Total exports (fob) . 129 40 93 300 Coffee .. 106 29 73 250 Tea .. 5 5 U.3..0. Manufactures 9 4 4 Total imports (cif) .. 207 117 133 1 * .I * *IE I _ Food .. 16 18 23 lOt Fuel and energy 1 6 27 17 20 s l Capital goods .. 63 34 37 91 92 03 94 95 9s 97 Export prce index (1995=100) .. 122 76 125 Import price index (1995=100) 71 103 97 r3 Exports I Imports Terms oftrade (1995=100) .. 172 73 129 BALANCE of PAYMENTS (US$ millions) 1976 1986 1996 1997 currentaccountbalancetoGDP ratio (/) Exports of goods and services 62 141 50 100 0 Imports of goods and services 75 268 160 165 2 Resource balance -13 -127 -110 -65 Net income 0 -21 -14 -11 Net current transfers 67 48 52 -6 Current account balance .. -81 -76 -24 -8 Financing items (net) 102 is -7 -10 Changes in net reserves -15 -21 58 31 -12 Memo: - Reserves including gold (USS millions) 49 70 145 60 Conversion rate (DEC, local/US$) 86.3 114.2 302.8 352.3 EXTERNAL DEBT and RESOURCE FLOWS 1976 1986 1996 1997 (US$ millions) Compositior of total debt, 1997 (US$ millions) Total debt outstanding and disbursed 22 549 1,139 1,247 IBRD 0 0 0 0 F G.72 IDA 5 191 588 567 F: 1 Total debt service 3 35 55 53 IBRD 0 0 0 0 IDA 0 2 7 9 :5:67 Composition of net resource flowvs Official grants 21 51 164 - Official creditors 3 88 14 s0 Private creditors 0 -6 -1 0 D:400 :6 Foreign direct investment 0 2 1 Portfolio equity 0 0 0 . c: 30 World Bank program Commitments 8 32 0 0 A - IBRD E - Bilateral Disbursements 1 44 17 12 B - IDA D - Other multilateral F - Private Principal repayments 0 1 3 4 C-IMF G - Short-term Net flows 1 43 14 7 1 _ 1 Interest payments 0 2 4 5 Net transfers 1 42 10 3 Devellopment Economics 10/1/98 280 0e 300 310 A Lake W R N D Lae JRwru.. \o Lak) glvNt TOav c Cohoha ROADS BURUNDI, 6.x T a SELECTED TOWNS AND CITIES I i ® NATIONAL CAPITAL /' , > | ---"INTERNATIONAL BOUNDARIES i, ge7r( ~-.-.RIVERS ,'- 40~~~~~~~~~~~~~~~~~~T ¢ l^ \ Bu6anza >/ % DEMOCRATIC REPUBLIC -;wiwa 4 This map was produced by the / 8 ^ tX B R9 Map Design Unit of The World Bank. OF CON GO )- T akec nyikc The boundaries, colors, denominations g I1Rwnit ana anyoiher information sho-wnon Ithis .. map do not imply, on the port of The 0 25 50 K.ILOETR rn World Bank Group, any iudgment on /* az-N 25 50 KILMTR c the legolstatus of any territory,or any I I- ci 4 S 9 endorsement or acceptance of such jU;j| t025 50OMILES < boundaries. 1- . ' ' ' ToO ° 280 2 90 / wX / Kigoma 3 0°0< 31°0g '0, , , .. * . -, OAD