Document of The World Bank Report No: ICR00001152 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA H076 ­ TF 053794) ON AN IDA GRANT IN THE AMOUNT OF SDR 22 MILLION (US$ 33 MILLION EQUIVALENT) AND A TRUST FUNDED GRANT IN THE AMOUNT OF SDR 28 MILLION (US$ 41.8 MILLION EQUIVALENT) TO THE REPUBLIC OF BURUNDI FOR AN EMERGENCY DEMOBILIZATION, REINSERTION AND REINTEGRATION PROJECT June 23, 2009 Sustainable Development Department Fragile States, Conflict and Social Development Unit Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective May 2009) Currency Unit = Burundi Francs US$1 = BIF 1,241 SDR1 = US$ 1.33 FISCAL YEAR July 1 ­ June 30 ABBREVIATIONS AND ACRONYMS AMIB African Peacekeeping Mission in Burundi ASPEN Africa Region Social and Environmental Safeguards Unit (World Bank) APPMs Armed political parties and movements BINUB United Nations Integrated Mission to Burundi BNP Burundi National Police CAS Country Assistance Strategy CNDD-FDD Conseil National Pour la Défense de la Démocratie ­ Forces de la Défense de la Démocratie ­ National Council for the Defense of Democracy ­ Democratic Defense Forces D&R Demobilization and Reintegration DRRP Demobilization, Reinsertion and Reintegration Program ESMF Environmental and Social Management Framework FAB Forces Armées Burundaises ­ Burundi Armed Forces FNL Forces Nationales de la Libération ­ National Forces of Liberation FNL-D Forces Nationales de la Libération ­ Dissidents ­ National Forces of Liberation - Dissidents FTA Forces Technical Agreement GoB Government of Burundi HIV/AIDS Human Immunodeficiency Virus / Acquired Immune Deficiency Syndrome IBRD International Bank for Reconstruction and Development IDA International Development Association ITS Indemnité Transitoire de Subsistance (French translation of TSA ­ see below) JCC Joint Cease-Fire Commission JOP Joint Operations Plan JVMC Joint Verification and Monitoring Mechanism MDRP Multi-Country Demobilization and Reintegration Program MDTF Multi-Donor Trust Fund M&E Monitoring & Evaluation MIS Management Information System NCDRR National Commission for Demobilization, Reinsertion and Reintegration NDF National Defense Force NGO Non-Governmental organization OP/BP Operational Policy/Bank Policy PDO Pre-discharge orientation PDO Project Development Objective PIM Program implementation manual PNDRR Programme National pour la Démobilisation, Réinsertion et Réintégration ­ National Program for Demobilization, Reinsertion and Reintegration PRSP Poverty Reduction Strategy Paper TSA Transitional subsistence allowance UN United Nations UNDP United Nations Development Program UNICEF United Nations Children Fund VCT Voluntary Counseling and Testing Vice President: Obiageli Katryn Ezekwesili Country Director: John McIntire Country Manager: Mercy Myiang Tembon Sector Manager: Ian Bannon Project Team Leader: Marcelo Jorge Fabre ICR Team Leader: Bernard Harborne REPUBLIC OF BURUNDI Emergency Demobilization, Reinsertion and Reintegration Project CONTENTS 1. Project Context, Development Objectives and Design ....................................................................... 1 2. Key Factors Affecting Implementation and Outcomes ....................................................................... 6 3. Assessment of Outcomes ................................................................................................................... 11 4. Assessment of Risk to Development Outcome ................................................................................. 18 5. Assessment of Bank and Borrower Performance .............................................................................. 18 6. Lessons Learned ................................................................................................................................ 21 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners ................................... 24 Annex 1. Project Costs and Financing................................................................................................ 25 Annex 2. Outputs by Component ....................................................................................................... 26 Annex 3. Economic and Financial Analysis ....................................................................................... 27 Annex 4. Bank Lending and Implementation Support/Supervision Processes .................................. 29 Annex 5. Beneficiary Survey Results (if any) .................................................................................... 31 Annex 6. Summary of Borrower's Comments on Draft ICR ............................................................. 34 Annex 7. Comments of Cofinanciers and Other Partners/Stakeholders............................................. 36 Annex 8. List of Supporting Documents ............................................................................................ 37 Annex 9. List of key informants ......................................................................................................... 39 [MAP] ....................................................................................................................................................... 40 A. Basic Information Burundi Emergency Demobilization, Country: Burundi Project Name: Reinsertion and Reintegration Program Project ID: P081964 L/C/TF Number(s): IDA-H0760,TF-53794 ICR Date: 06/26/2009 ICR Type: Core ICR GOVERNMENT OF Lending Instrument: ERL Borrower: BURUNDI Original Total USD 33.0M Disbursed Amount: USD 29.3M Commitment: Revised Amount: USD 0.0M Environmental Category: B Implementing Agencies: National Commission for Demobilization, Reinsertion and Reintegration Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 01/15/2003 Effectiveness: 09/10/2004 Appraisal: 10/20/2003 Restructuring(s): Approval: 03/18/2004 Mid-term Review: 06/15/2007 06/15/2007 Closing: 12/31/2008 12/31/2008 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: Substantial Bank Performance: Moderately Satisfactory Borrower Performance: Moderately Unsatisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Satisfactory Government: Not Applicable Implementing Quality of Supervision: Moderately Satisfactory Not Applicable Agency/Agencies: Overall Bank Overall Borrower Moderately Moderately Satisfactory Performance: Performance: Unsatisfactory i C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry Yes None at any time (Yes/No): (QEA): Problem Project at any Quality of Yes None time (Yes/No): Supervision (QSA): DO rating before Moderately Closing/Inactive status: Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Health 5 5 Other social services 95 95 Theme Code (as % of total Bank financing) Conflict prevention and post-conflict reconstruction 29 29 Gender 14 14 HIV/AIDS 14 14 Other accountability/anti-corruption 14 14 Social safety nets 29 29 E. Bank Staff Positions At ICR At Approval Vice President: Obiageli Katryn Ezekwesili Calisto Madavo Country Director: John McIntire Emmanuel Mbi Sector Manager: Ian Bannon Joseph Baah-Dwomoh Project Team Leader: Marcelo Jorge Fabre Ingo Wiederhofer ICR Team Leader: Bernard Harborne ICR Primary Author: Fanny Robert F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The overall goal of the DRRP would be to help consolidate peace in Burundi and the Great Lakes region through the implementation of relevant articles of the Arusha Peace Accord and subsequent cease-fire agreements. The three principal objectives of the Program would be to: ii (i) demobilize an estimated 55,000 ex-combatants from the FAB, members of Armed Political Parties and Movements (APPMs), and subsequently the NDF, as well as support their reintegration into civilian life; (ii) provide reinsertion assistance to 20,000 Gardien de la Paix and 10,000 combattants militants militia; and (iii) contribute to the reallocation of Government expenditure from defense to social and economic sectors. Revised Project Development Objectives (as approved by original approving authority) (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years Up to a maximum of 55,000 ex-combatants from FAB and APPMs have been Indicator 1 : demobilized and reintegrated into civilian life. Up to 30,000 militias provided reinsertion support 26,283 ex- combatants up to 55,000 ex- demobilized. combatants Value 28,383 Gardiens de up to 30,000 quantitative or 0% la Paix and Gardien de la Paix Qualitative) Combattants and Combattants Militants have Milittants received reinsertion support. Date achieved 09/10/2004 12/31/2008 12/31/2008 100% of eligible combatants were demobilized.The discrepancy between initial Comments and actual target numbers is due to an overestimation of the original size and (incl. % composition of the different fighting forces, and to delays in the negotiations achievement) with FNL-PALIPEHUTU At least 40% of ex-combatants are engaged in sustainable income-generating Indicator 2 : activities within 18 months of demobilization. Over 87% of ex- 40% are engaged combatants Value in sustainable reported being quantitative or 0% income-generating employed in 2008 Qualitative) activities survey on reintegration. Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments (incl. % achievement) Recurrent expenditures on defense fall from FBU49.4 billion in 2005, and fall to Indicator 3 : FBU 23.8 billion in 2007. Expenditures on health & education combined increase by FBU25.7 over same period iii 2008 revised FBU 62.7 billion for budget: FBU 132.3 Value defense Projected 2009- billion for health & quantitative or FBU 51.3 billion for 2.2% education Qualitative) social expenditures combined; FBU 57.6 for defense. Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments Demobilization has reduced defense expenditures between 2005 and 2009, but (incl. % the increase in the police force has more than offset the savings achievement) (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years Indicator 1 : At least 14,000 ex-combattants are demobilized annually. year 1:15,170 Value demobilized; year (quantitative 0% N/A 2: 2,181; year 3: or Qualitative) 2,735; year 4: 1,987 Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments The discrepancy between initial and actual target numbers is due to an (incl. % overestimation of the original size and composition of the different fighting achievement) forces, and to delays in the negotiations with FNL-PALIPEHUTU Each demobilized has received his/her Transitional Subsistence Allowance Indicator 2 : payment within 30 days of scheduled payment dates Value 23,022 have (quantitative 0% 95% received their TSA or Qualitative) Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments Initially scheduled tranches payment was 3, 6 and 9 months after demobilization. (incl. % Average payments were made 3.01, 6.03 and 9.01 months after demobilization. achievement) Over 98% on average have received their TSA on time. Each disabled ex-combatant has received medical rehabilitation assistance within Indicator 3 : 18 months of demobilization 1,660 disabled ex- Value combatants were (quantitative 0% 100% provided medical or Qualitative) rehabilitation assistance Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments 100% of disabled ex-combatants have received basic medical assistance in (incl. % demobilization camps. 40% of severely and very severely disabled ex- achievement) combatants have received medical rehabilitation assistance starting July 2008. 100% of registering ex-combatants have access to reintegration assistance within Indicator 4 : 9 months of their demobilization iv 21,996 ex- Value combatants have (quantitative 0% 100% received or Qualitative) reintegration assistance. Date achieved 09/10/2004 12/31/2008 12/31/2008 Comments The average wait time between demobilization and the beginning of the (incl. % reintegration assistance was 1 year throughout the project. achievement) Up to 20,000 Gardiens de la Paix and 10,000 Combattants are disarmed and each Indicator 5 : receive FBU 100,000 service allowances 18,709 Gardiens de la paix and 9,674 Value Combattants were (quantitative 0% 100% disarmed and or Qualitative) received FBU100,000 service allowance. Date achieved 09/10/2004 12/31/2005 12/31/2008 Comments (incl. % 94% of original target achievement) Up to 20,000 Gardiens de la Paix and 10,000 Combattants are disarmed and each Indicator 6 : receive FBU 100,000 service allowances Value (quantitative 0% 100% or Qualitative) Date achieved 09/10/2004 12/31/2005 Comments (incl. % achievement) G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 09/29/2004 Satisfactory Satisfactory 0.00 2 04/19/2005 Satisfactory Satisfactory 6.05 Moderately Moderately 3 06/30/2005 11.17 Unsatisfactory Unsatisfactory Moderately Moderately 4 09/16/2005 11.17 Unsatisfactory Unsatisfactory Moderately 5 10/25/2005 Unsatisfactory 11.37 Unsatisfactory 6 12/02/2005 Moderately Satisfactory Moderately Satisfactory 11.37 7 03/27/2006 Moderately Satisfactory Moderately Satisfactory 12.04 8 06/30/2006 Moderately Satisfactory Moderately Satisfactory 12.82 v 9 12/22/2006 Moderately Satisfactory Moderately Satisfactory 22.81 Moderately 10 05/30/2007 Moderately Satisfactory 22.81 Unsatisfactory Moderately Moderately 11 08/13/2007 22.81 Unsatisfactory Unsatisfactory Moderately Moderately 12 03/17/2008 25.00 Unsatisfactory Unsatisfactory 13 06/26/2008 Moderately Satisfactory Moderately Satisfactory 25.81 14 12/19/2008 Moderately Satisfactory Moderately Satisfactory 26.54 H. Restructuring (if any) Not Applicable I. Disbursement Profile vi 1. PROJECT CONTEXT, DEVELOPMENT OBJECTIVES AND DESIGN Note: Per OPCS guidelines, this ICR will focus on rating the outcomes of the project as well as identifying lessons learned from this operation. Further, some additional detail is provided on the relevant socio-political context due the difficult operating environment in which this project was launched. 1.1 Context at Appraisal 1. In 2000, emerging from the cycle of recurrent politico-ethnic conflict that had plagued the country for several years, Burundi reached a milestone in the peace process. The Arusha Peace and Reconciliation Agreement on August 28, 2000 laid the groundwork for constitutional and institutional reform to foster national reconciliation between the nine co-signatory parties. A power sharing transitional Government was successfully inaugurated on November 1, 2001. However, large-scale violence continued to prevent the complete stabilization of the country. Two major armed political parties and movements (APPMs), the Conseil National Pour la Défense de la Démocratie - Forces Pour la Défense de la Démocratie (CNDD-FDD) and the Forces Nationales Pour la Libération - PALIPEHUTU (FNL-PALIPEHUTU) continued to fight the transitional Government and refused to participate in the Arusha process. Only in November 2003 did the CNDD-FDD, the larger of the two armed groups, finally reach a comprehensive cease-fire agreement with the Government. 2. At the time of project launch, in 2003, the situation was marked by rising poverty: the proportion of people living below the poverty line rose from 35% in 1992 to more than 60%. Burundi was presenting some of the worst socio-economic indicators in the world, including high infant mortality and acute malnutrition. Of an estimated population of 8 million persons, over 800,000 were living in refugee camps in neighboring Tanzania, in addition to 300,000 internally-displaced civilians still in country.1 Protocol IV of the Arusha Agreement had dedicated an entire portion of its final text to the issue of refugees and IDPs by planning for an independent legal framework and the establishment of a national commission for the rehabilitation of war-affected populations. In 2002, in the midst of ongoing violence and with little preparation, refugee camps in Tanzania started to be dismantled by the Burundian Army because they were considered a recruitment base for the armed groups. The ensuing migratory situation was later identified as a variable in the reintegration of ex-combatants into civil society.2 3. In turn, the economic situation was marked by falling output and incomes (GDP declined by nearly 20% in real terms over the period 1993- 2002); high levels of inflation; diminished public revenues; the withdrawal of international financial assistance (international aid transfers fell from an annual average of nearly US$300 million to less than an annual average of US$ 100 million between 1992 and 2002); lower investment (the investment rate fell from an annual average of more than 15% to 9% between 1992 and 2002); and the accumulation of domestic and external deficits and external payment arrears (which stood at US$148.5 million in 2002). 4. On August 28, 2003 and in direct response to the renewed efforts to seek peace and stability, the Government established a National Commission for Demobilization, Reinsertion and Reintegration 1 International Crisis Group briefing, [December, 2003]. 2 Peter Uvin [May, 2007], Ex-combatants in Burundi: why they joined, why they left, how they fared. 1 (NCDRR) to provide political and strategic guidance to the DDR process. An Executive Secretariat was put in place to implement a national DDR program, as well as to coordinate the contributions of international partners. Within the framework of the Multi-Country Demobilization and Reintegration Program (MDRP) for the greater Great Lakes region, the Government asked the World Bank for financial and technical assistance to support a Demobilization, Reinsertion and Reintegration Project (DRRP). Efforts towards the demobilization and reintegration of ex-combatants in Burundi were supported by representatives of several MDRP donors and partners, the African Union Mission in Burundi (AMIB) and the Joint Cease-fire Commission (JCC) chaired by the United Nations. In this context, the Bank worked with the Government and other key stakeholders to facilitate the preparation of a Joint Operations Plan (JOP) outlining the roles and responsibilities of all actors engaged in the disarmament and demobilization process. 5. The Government's request for technical and financial assistance for the DRRP had come at a critical time in the peace process. IDA's involvement was justified by its considerable experience with such programs, the scale of resources required for the Program, the focus on socio-economic reintegration of ex-combatants, and the need for IDA to be responsive to a client request to lead and coordinate international support for a critical peace-building operation. IDA support was intended to complement the activities of other international stakeholders in Burundi and the region. The establishment of a multi-donor trust fund under the umbrella of the MDRP program allowed for donor coordination and flexibility in the funding available. 1.2 Original Project Development Objectives (PDO) and Key Indicators [as approved]: Development objectives: 6. The overall goal of the DRRP was to help consolidate peace in Burundi and the Great Lakes region through the implementation of relevant articles of the Arusha Peace Accord and subsequent cease-fire agreements. The three main objectives of the Program were to: (i) demobilize an estimated 55,000 ex-combatants from the Forces Armées Burundaises (FAB), members of Armed Political Parties and Movements (APPMs), and subsequently the National Defense Force (NDF), as well as support their reintegration into civilian life; (ii) provide reinsertion assistance to 20,000 Gardiens de la Paix and 10,000 combattants militants militia; and (iii) contribute to the reallocation of Government expenditures from defense to social and economic sectors. Letter of Demobilization Policy: 7. In February 2004, the Government addressed to the President of the World Bank a document containing the Government commitment to implement DDR activities in the country. This "Letter of Demobilization Policy," outlined the legal framework for DDR activities, the target numbers and a schedule for implementation. This key information was embedded into the project design and drove the establishment of estimates and targets for the project. The strategy described in the Letter also outlines a gradual decrease in the size of the National Defense Force (NDF) over a period of four years, after its creation out of personnel selected from the former APPMs and Burundian Armed Forces (FAB). In summary, the Letter establishes a two-phase program including: 2 a. Phase 1 (year 1): The Joint Operations Plan establishes criteria for combatant status and eligibility for integration in the regular forces. Ineligible combatants (for the army) would be demobilized (ca. 14,000) b. Phase 2 (years 2- 4): The NDF is gradually downsized with demobilization of combatants (ca. 14,000 in year 2; 14,000 in year 3 and 13,000 in year 4). Key indicators: 1. At least 14,000 Ex-Combatants are demobilized annually; 2. Each demobilized Ex-Combatant has received his / her TSA payment within thirty (30) days of scheduled payment dates; 3. Ex-Combatants have access to reintegration assistance within nine months of their demobilization; 4. Average income levels of Ex-Combatants are similar to average income levels in their communities of settlement within 24 months of the Ex-Combatants' demobilization; 5. An external evaluation of the implementation of the Project is carried out annually and includes an analysis of environmental and social monitoring indicators; and 6. Recurrent expenditures on defense are a maximum of FBU 49.4 billion in 2005, FBU 36.1 billion in 2006 and FBU 23.8 billion in 2007. At the same time, expenditures on health and education combined increase by FBU 6.7 billion in 2006 and FBU 19 billion in 2007. 1.3 Revised PDO (as approved by original approving authority) and key indicators, and reasons/justification: N/A 1.4 Main Beneficiaries: 8. Below are the original tables from the technical annex detailing the target groups for the two phases of the DRRP. Stage 1 for the first year and stage 2 for years 2-4. Figure 1. Original tables of estimated target groups 3 1.5 Original components [as approved]: 9. Component 1 ­ Demobilization. STAGE 1: The demobilization of 9,000 ex-combatants from the APPMs and 5,000 soldiers of the FAB would be completed in a period of approximately six months during stage I of the DRRP. This component includes verification of combatant status, registration and distribution of ID cards, socio-economic data gathering, pre-demobilization orientation, medical screening and voluntary counseling and testing for HIV/AIDS and transport to communities of return. STAGE 2: Stage II demobilization would be undertaken during years 2-4. During which 41,000 personnel of the NDF would be demobilized. This process would be more gradual during stage II, and would be undertaken from a single Demobilization Centre. The procedures would be identical to those in stage I. 4 10. Component 2 ­ Reinsertion. All ex-combatants receive a transitional subsistence allowances (TSA) based on salary levels of FAB soldiers and providing for basic resettlement needs of ex- combatants over a period of 18 months. 11. Component 3 ­ Reintegration. Reintegration assistance was provided to help ex-combatants return and/or settle in a community and establish sustainable livelihoods. This included: - information and referral - counseling - advocacy from the implementing agency to local communities - training (vocational, on-the-job training, and apprenticeship) - micro-projects (small-scale agriculture, livestock keeping, fisheries, handicraft, construction, services or trade) and micro-finance - access to land: facilitation for ex-combatants' requests according to established procedures (but not amounting to the acquisition of land) - social reintegration 12. Component 4 - Customized reintegration assistance for special groups. Special groups included children ex-combatants (as defined by the Cape Town Principles3), who were initially assisted via an MDRP special project implemented by the United Nations Children Fund (UNICEF); female ex- combatants and disabled ex-combatants. The customized assistance entailed among others separate facilities for the encampment of women and children ex-combatants, medical assistance and rehabilitation for disabled ex-combatants, counseling activities at the community level with partners and women associated with ex-combatants, and gender awareness for the implementing agency and implementing partners. 1.6 Revised Components: N/A 1.7 Other significant changes: 13. Over the course of the project, negotiations with the FNL did not succeed in time to incorporate the ex-combatants from this group in the Demobilization and Reintegration activities. Furthermore, the verification process for the status of combatant and registrations showed that the original size and composition of the different fighting forces in Burundi had been largely overestimated. This resulted in a much smaller number of eligible beneficiaries to be assisted. Overall, the project has supported nearly 100% of combatants eligible and opting for reincorporation into civilian life. This translates to a number of ca. 26,000 beneficiaries. 3 Cape Town Principles ­ UNICEF [1997] Cape Town Principles and Best Practices on the Recruitment of Children into the Armed Forces and on Demobilization and Social Reintegration of Child Soldiers in Africa. 5 2. KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOMES 2.1 Project Preparation, Design, and Quality at Entry: 14. The Government had made significant progress towards peace and stabilization in 2003 after the signature of the 2000 Arusha peace accords and the November 2003 cease-fire agreement with the CNDD-FDD. This provided the overall framework for subsequent negotiations and agreements in Burundi. A proactive role of the Government and the political momentum behind the peace process were positive factors for the preparation, design and quality of the project at its inception. They were especially instrumental to the integration process whereby ex-combatants from APPMs integrated the national forces. Only strong political will from the Government could guarantee the buy-in from the FAB, then the national army, and confidence-building amongst all armed groups set to integrate the national army. Although outside the scope of the project, these were determining factors throughout inception and implementation. 15. The dire economic situation was a key factor affecting implementation of the reintegration component. After signing peace agreements in the early 2000s, the country did not experience a bounce in GDP growth as is generally the case when some degree of `post-conflict' stability resumes. With a population of about 8 million distributed over 27,834 square kilometers, Burundi has one of the highest population densities in Africa and is ranked the third poorest country in the world according to the latest World Development Indicators.4 The country is landlocked and thus depends on the extremely degraded road and rail infrastructure of its neighbors to conduct international trade. 63 percent of the population suffers from food insecurity. Agriculture, which provides 94 percent of all employment, 80 percent of export revenues and 40 percent of the aggregate GDP, is marked by a high dependency on coffee exports and low levels of investment from the private sector.5 Poverty is deeper and more pervasive in rural areas, and substantial disparity exists between regions. Almost 97 percent of Burundi's poor live in rural areas, which have a poverty incidence of 69 percent, compared to 34 percent in urban areas. Against this challenging background, socio-economic reintegration activities were intended to support ex-combatants in establishing a livelihood and sustain themselves and their dependents. However the weakness of the private sector and infrastructure limited the opportunities to generate wealth. It must be noted that reintegration was meant to act as an equalizer and not an opportunity for higher incomes. One of the key indicators for reintegration was therefore that average income levels of ex-combatants are similar to average community incomes in their communities of settlement within 24 months of their demobilization. 16. Project design benefited from lessons learned from past Bank operations such as the Burundi Community Rehabilitation project financed by the Post Conflict Fund in 2001, but also past reintegration projects in Rwanda and Ethiopia. The World Bank had gained experience in working in a fragile environment and established a level of confidence with the Government instrumental to national ownership. Among the lessons learned were : 4 World Development Indicators Database, World Bank, revised April 24 2009. 5 World Bank working paper No. 147 [May, 2008], Breaking the cycle, A Strategy for Conflict-Sensitive Rural Growth in Burundi. 6 a. Demobilization. The assembly of ex-combatants during the demobilization phase provides a good opportunity for HIV/AIDS voluntary counseling and testing (VCT), as well as the dissemination of information about program benefits and civilian life. b. Reinsertion. During this period, ex-combatants require a transitional safety net to cover their families' basic material needs. c. Economic reintegration. Targeted support is advisable to help ex-combatants establish sustainable livelihoods. In order to be useful and cost-effective, such assistance should be tailored to their socio-economic profile and adjusted to the economic environment. Female ex-combatants often require specialized support to re-establish sustainable livelihoods. In general, ex-combatants should receive no more support than is necessary to help them attain the general standard of living of the communities into which they reintegrate. Furthermore, reintegration assistance should be limited in time to avoid the creation of a dependency syndrome. Information, counseling and referral services can assist ex-combatants to assess their livelihood options and find employment. Apprenticeships in the informal sector combine the benefits of vocational training and employment and are thus preferable to training only. Micro-projects can help ex- combatants secure their economic independence through self-employment. Microprojects are, however, complex interventions that require significant administrative capacity and a stable environment. d. Social reintegration. Ex-combatants should be encouraged to reintegrate in the vicinity of their extended families. Informal networks of ex-combatants can be important elements to successful economic and social reintegration. e. Child ex-combatants should be reunified with their families wherever possible, and assisted to resume their education or enter into gainful enterprise. f. Disabled ex-combatants commonly require specialized medical and assistance in overcoming the constraints of the disability. g. Institutional issues. Program coordination by a single civilian agency with overall program responsibility, balanced by decentralization of implementation to regions and communities makes for a powerful institutional arrangement. The employment of ex- combatants as counselors and field staff can facilitate reintegration. In peacekeeping environments such as in Burundi, close operational linkages between the national implementation body and the peacekeeping force are essential. 17. The project was designed with original key features that reflected the specificity of the D&R operations and allowed for greater flexibility of the World Bank model for this kind of program. In particular, the 2-phase approach was a response to the Burundi political context and the need to demobilize members of armed groups first before progressively decreasing the size of the national forces. Moreover, the project was designed over a 5-year period as opposed to the traditional 3 year lifespan of regular Bank operations. Finally, the overall structure of the MDRP as a regional program with trust funds available for financing of operations created an unprecedented level of donor coordination on the ground as well as flexible and adapted financing mechanisms. 18. Risks were accurately identified in the project's technical annex. Yet, some of the measures set up to mitigate these risks proved to be insufficient. For example, program-specific risks which were identified, such as the limited capacity or the repartition of Government expenditures, were not adequately addressed in the mitigation measures proposed. Indeed mitigations measures for low capacity 7 such as careful monitoring and technical assistance to the Executive Secretariat throughout implementation were put in place but their scope was too limited, allowing for inadequate management and important delays in implementation, particularly for the reintegration component. 19. Reintegration was planned as a short-term assistance to be linked with longer-term recovery projects such as a planned UNDP community-based development project. This intentionally short-term view was chosen for two main reasons. From lessons learned from other similar projects, the objective was to avoid perceptions amongst the population that perpetrators of violence were being rewarded. Short-term reintegration assistance was intended to provide just the sufficient amount of initial support for ex-combatants to return to their communities. The second reason pertained to the synergies planned with longer-term, community-based recovery programs that would assist the entire community and allow for reintegration in the process. The UNDP had notably started preparation of such a project at the time of design, but broad-based community recovery efforts never materialized due to several reasons, mainly associated with continued insecurity in rural areas caused by the FNL-PALIPEHUTU forces. 2.2 Implementation: 20. The implementation was characterized by an efficient demobilization and reinsertion process but also by important delays in the reintegration component. The main reasons behind these delays were the overall political process and low implementation capacity. The political process. Three key processes had a direct impact on project implementation: - The peace process became stalled after the FNL-PALIPEHUTU refused to enter the demobilization process along with the other armed groups identified in the Letter of Demobilization Policy issued by the Government. This had a significant impact on the activities carried out by the implementing agency but more importantly on the overall security context. Negotiations were ongoing after the 2006 cease-fire agreement between the newly elected Government and the FNL-PALIPEHUTU but were not finalized in time to allow for this armed group to participate in the project. - The Forces Technical Agreement (FTA), negotiated between the transitional Government and the CNDD, was an essential component of the preparation of demobilization and reintegration activities, spelling out rank harmonization arrangements upon which reinsertion benefits are based. The political process leading up to the FTA experienced major delays and resulted in the first year of project effectiveness seeing very few activities on the ground. - Finally, the 2005 elections had a major impact upon the Executive Secretariat, as former trained- up staff left and new members were recruited at the behest of the new Ministerial team, undermining the training and technical assistance components. 21. Mid-term review and Institutional Capacity. The Bank's mid-term project review in June 2007 along with the four independent payment verification surveys highlighted the efficiency of the implementation of the reinsertion component in the project. After receiving a first installment of the reinsertion support on discharge, and within three months of their return to communities, beneficiaries received a second tranche of the transitional subsistence allowance to cover their basic needs and start income generating activities. The schedule of payments was respected throughout the project: a third tranche was received after six months and a last tranche after nine months. All beneficiaries opened an 8 account in a local bank, facilitating the transaction and 98.89% claimed to have received their payments on-time and in the right amount. Earlier, in 2005, upon the request of the Bank task team, an assessment on the institutional capacity of the Executive Secretariat had been conducted. This highlighted structural inefficiencies with the Secretariat and the need for more technical assistance and supervision for the National Commission. In particular, the assessment recommended further decentralization at the provincial levels and clarification of the role of the financial department6. Following this report, this department was restructured to encompass logistics and procurement as well. However, most of the assessment recommendations were not applied, such as the recruitment of a technical assistant to support the project coordinator, and hence the 2007 mid-term review noted continuing capacity weaknesses. The review revealed management issues at the level of the Executive Secretariat causing delays in the contracting of implementing partners for the reintegration component. The review led to the downgrade of the project's progress to achievement of PDOs to moderately unsatisfactory. In the Aide Memoire of the mid-term review the mission "strongly urged the SE/CNDRR to reduce unit costs per ex-combatant and the overall management costs of the program, which are both found too high."7 It also insisted that the CNDRR itself would be much more actively involved in the oversight of the program which would also be increasingly decentralized. The issue of supervision from the CNDRR over to the Executive Secretariat remained ongoing throughout the project. As identified in the mid-term review, the political and strategic guidance that were supposed to be provided by the CNDRR were largely lacking after the successful 2005 elections. This caused the Executive Secretariat to freewheel for the second phase of demobilization and most importantly, for the implementation of the reintegration component. It is only later in the process, in 2007, that a new leadership within the Commission made the program a priority again for the Government, which was reflected at the level of the Secretariat by the contracting of implementing partners for the reintegration component and the vulnerable groups component, along with the further demobilization of members of the NDF. 2.3 Monitoring and Evaluation: 22. M&E Design. The monitoring and evaluation of the project was supported by the establishment of a Management Information System (MIS) that would record and keep all data related to the ex- combatants as they entered the demobilization camps and continued through the reinsertion and reintegration process, as well as data on service providers and benefits. Monitoring was designed at all levels: central, provincial and in the communes: sample surveys were planned bi-annually to verify that benefits had been delivered to the beneficiaries and that reintegration activities were ongoing. Independent evaluations were designed to include statistics on the reintegration of ex-combatants along with a beneficiary assessment in order to capture qualitative data. A robust MIS system was installed which comprised a considerable amount of socio-economic data. This data was to inform the design of reintegration activities in response to the specific profiles of the beneficiary group. 23. M&E implementation. The MIS was an efficient management tool as it allowed tracking of the number of beneficiaries entering the program along with data on their age and gender. It functioned with a well-designed, efficient system involving digital fingerprint scanners and cameras but also a fully automated payment system for the delivery of reinsertion support. Four verification surveys were conducted throughout the project with a substantial sample size (an average of 10% of the total number 6 BINZEGGER [September, 2005] Institutional capacity assessment of the NCDRR Executive Secretariat, ACTS. 7 Aide-memoire of the mid term review for the Burundi Emergency Demobilization, Reinsertion and Reintegration Project [June, 2007] 9 of beneficiaries) and yielded a wealth of data on the situation of beneficiaries after demobilization. Independent evaluations were also carried out by consultants competitively selected by the implementing agency. 24. M&E utilization. The MIS and verification surveys were used to track beneficiaries once they were back in their communities and had started the reintegration activities. A qualitative study financed by the program and conducted by international expert Peter Uvin exposed trends and dynamics amongst beneficiaries, notably the difference between reintegration in rural and in urban settings. However informative this study is, it must be noted that the size of the sample interviewed by Mr. Uvin is inferior to 100, and bears no statistical value. 2.4 Safeguard and Fiduciary Compliance: 25. Safeguards. The project was classified as a category B for its income generating activities component. Although the program had minimal social and environmental impacts due to the widely dispersed and small-scale reintegration components, detailed social and environmental safeguard arrangements were made at the inception stage. An environmental and social management framework was going to be developed by the MDRP in collaboration with the Africa Region's Social and Environmental Safeguards unit (ASPEN), although the operation did not require one since it was processed under OP 8.50, the operational policy for Emergency Recovery Assistance. It must be noted that such a framework failed to be produced. Social mitigation measures included ensuring: (i) no involuntary resettlement was taking place and (ii) targeted assistance and equal access to benefits for vulnerable populations such as Batwa ex-combatants, female ex-combatants and child soldiers. Environmental mitigation measures included systematic screening of reintegration sub-project and adequate disposal of medical waste. The systematic screening of sub-projects came as a mitigation measure to compensate for the lack of a framework. 26. Financial management. The initial risk analysis rated this project as high risk in the absence of adequately trained personnel and therefore required a close monitoring from the Bank for activities related to the financial management. Six safeguards had been built into the program to ensure IDA and trust funds were utilized for the purposes intended. Among them for instance was the conditional replenishment of the quick disbursement fund according to project performance. As concluded in the final independent evaluation commissioned by the Government, the 18-month-long absence of an Administration and Finance Director resulted in project mismanagement issues. The lack of internal audits was also highlighted as a major issue. In October 2008 an investigation from a Governmental commission was triggered leading to the firing of the head of the Executive Secretariat. 27. Procurement. An assessment of the procurement capacity of the Executive Secretariat was carried out during appraisal and concluded that the project was high risk. As a result, an international procurement specialist was hired to help the implementing agency set up a detailed procurement plan. However, the procurement service of the Executive Secretariat failed to appropriately implement procurement procedures agreed upon with the World Bank especially with regards to publication of calls for bids. This was highlighted in the final independent evaluation commissioned by the Government.8 The major issue related to procurement throughout the project was the lack of bidders with appropriate 8 ACT Consultants and Case d'Afrique [December, 2008] Final evaluation of the National Demobilization, Reinsertion and Reintegration Program, Analysis of Procurement procedures p. 56. 10 capacity to carry out the activities advertised. This lead to a high number of unconventional procedures such as sole source selection which in turn created longer turnaround times to award contracts. Insufficient supervision for ongoing contracts is a shared responsibility between the World Bank and the Government, both of which were unable to provide appropriate guidance when conflicts on the contracts arose between implementing partners and the Executive Secretariat. 2007 and 2008 were the busiest years for procurement with contracting of implementing partners for the reintegration component as well as the assistance for disabled ex-combatants. The project would have greatly benefited from more supervision and technical assistance especially at that time. 2.5 Post-completion Operation/Next Phase: 28. The project supported the demobilization, reinsertion and reintegration of over 26,000 ex- combatants over the course of 5 years which represents nearly 100% of eligible beneficiaries. As of December 31, 2008 (closing date of the project) a small portion (estimated around 1,100) of beneficiaries had not completed the program and were still awaiting reintegration assistance. This is partly due to implementation delays but also to the late demobilization of elements in the national forces. In parallel to the winding down of D&R activities under the first phase of the program, negotiations progressed between Government and the FNL-PALIPEHUTU, which had abandoned the ethnic reference in its name and split into two main groups, FNL-Rwasa and FNL-Dissidents. In April 2009 an agreement was reached regarding the numbers of eligible FNL ex-combatants for either integration in the national forces or demobilization and reintegration. The Government has therefore been maintaining a response capacity since the closing of the project in December 2008 in anticipation of new D&R support to the last remaining armed group of the FNL, and has requested assistance from the World Bank. A new project to address the limited pending caseload from the past project and the demobilization and reintegration of the FNL is currently under preparation in the Africa region Fragile States, Conflict and Social Development unit (AFTCS) as a stand-alone project. 3. ASSESSMENT OF OUTCOMES 3.1 Relevance of Objectives, Design and Implementation: 29. Relevance is rated highly substantial because D&R remains highly relevant in Burundi. The project was established based on the Arusha peace agreement and the subsequent request from the Government for assistance in demobilization and reintegration of ex-combatants. D&R is a short-term, vital phase of the transition from conflict to stability and in the case of Burundi has led to the establishment of social peace and of a unified national army. The project continues to be in line with Bank assistance strategy as highlighted in the 2008 Country Assistance Strategy (CAS) for Burundi outcome: 2.1 Improved reintegration of ex-combatants and vulnerable groups9. A follow-up project is currently under preparation to continue to support Burundi in its efforts to demobilize and reintegrate ex-combatants remaining from the caseload of the previous project and ex-combatants from the FNL. 3.2 Achievement of Project Development Objectives: 9 Report No 44193-BI, IDA and IFC Country Assistance Strategy for the Republic of Burundi for the period FY09-FY12, [July 8, 2008]. 11 PDO 1: Demobilize an estimated 55,000 ex-combatants from the FAB, members of APPMs, and subsequently the NDF, and support their reintegration into civilian life. 30. This PDO is rated as moderately satisfactory ­ only moderate due to the weakness of the reintegration component and the program management issues (but not due to the discrepancy in initial targets and final output). 31. There is a difference between the first PDO and actual program outcomes. The PDO stated that services for D&R would be available for all eligible combatants up to a ceiling of 55,000. Only slightly over 26,000 have been supported in the project. While this may be interpreted as a failure to achieve the objective, the number supported represents 100% of the eligible combatants for demobilization in Burundi and as such, the project has achieved this goal, albeit below the full utilization of the available capacity. The discrepancy between the initial target numbers and the number of beneficiaries assisted is due to (i) an overestimation of the original size and composition of the different fighting forces in Burundi, and (ii) the delays in the political negotiations with the last rebel force in the country, the FNL- PALIPEHUTU. The table below is extracted from the final independent evaluation commissioned by the Government and shows the progression of the demobilization activities throughout the project. Figure 2. Number of individuals demobilized each year by the ES/NCDRR 32. Linkages between the main outcomes of this project and the outputs are evaluated in the table below. Figure 3. Table on outcomes and outputs Outcome Linkage with output 1. All armed groups except the FNL are Strong: demobilization through the demobilization demobilized centers operated by the implementing agency. 12 2. Ex-combatants not selected for Modest: the reinsertion grant allowed ex-combatants to integration in the army return to civilian cover their basic needs and start an income generating life activity but their successful reintegration in the community depends on many factors including a preexisting social/family network. Evidence has shown for instance that ex-combatants in rural areas had better chances of social reintegration than ex-combatants in urban settings.10 If one considers the varying quality of reintegration activities, it becomes difficult to clearly establish a causal relation between reintegration outputs and outcomes. 33. Transitional Subsistence Allowance (TSA) grants were successfully transferred to beneficiaries in a timely and efficient manner. The 4th payment verification survey on 3,463 demobilized ex- combatants illustrates how the reinsertion allowance was spent. On average, 22.32% was used for the satisfaction of basic needs, 15.58% to start income-generating activities, 15.04% to build a house and 14.81% for the purchase of live stock11. 34. Targeted assistance for vulnerable groups was provided by implementing partners. Child soldiers were separated from adults at the early stage of the process and supported thereafter by an MDRP special project implemented by UNICEF. Disabled ex-combatants received targeted medical assistance including rehabilitation treatment. Their assistance suffered from significant delays due to the late contracting of implementing partners and delays in purchasing of necessary prosthetic material (some of the prosthetic material was only made available to the implementing partner in the fall of 2008). 35. Outcomes from the reintegration component are mixed. Qualitative data gathered in verification surveys and independent studies suggest that ex-combatants successfully transitioned to civilian life and have been widely accepted in their communities. However, the success of economic reintegration and establishment of a sustainable livelihood is uneven. The reintegration component suffered from weak implementation throughout the project. Other external factors also affected results, such as the failure of international partners to launch planned development projects to bridge the gap between the short-term assistance provided by the project and more sustainable reintegration. At the planning stage, the World Bank task team was instrumental in setting up a strategy for reintegration. A workshop was organized in 2005 following the publication of a report written by an international consultant setting out the reintegration strategy for the project. This plan was not effectively followed. In turn, reintegration activities and contracts for implementing partners were delayed during the first year of implementation when the Executive Secretariat was organizing demobilization. Once bids were made to recruit implementing partners to carry out reintegration activities, very few qualified bidders materialized to implement the activities. Lack of oversight from the NCDRR allowed ex-combatants to obtain funding for "micro-projects" that were in fact simple monetary exchanges. Indeed, many ex-combatants would use their reintegration grant to purchase grain, goats or beer and resell in order to pocket the cash. Although it is debatable whether this constitutes failed reintegration, it is clear that the reintegration component as it was planned was poorly executed. Communication and sensitization activities were also 10 Findings from Uvin [2007]. 11 Butoyi [July, 2008]. Quatrième Enquête de Vérification des Paiements des Allocations de Réinsertion : Rapport d'Enquête 13 insufficient to address the ongoing issue of high expectations from the ex-combatants regarding reintegration activities. Finally, linkages with broader-recovery projects were weak due to the very few number of development programs set up during the lifetime of the project. Key indicators for the reintegration component were: 1. Ex-combatants have access to reintegration assistance within nine months of their demobilization. 2. At least 40% of ex-combatants are engaged in sustainable income-generating activities within 18 months of demobilization. 3. Average income levels of ex-combatants are similar to average community incomes in their communities of settlement within 24 months of their demobilization. 36. Another key issue lies in the delay between the return to the community and the beginning of the reintegration activities. Indicator 1 allowed for a maximum of nine months between the two but a survey realized by the Burundi National Institute for Statistics in 2008 showed the average delay between demobilization and receipt of grant was one year. Indicator 2 has been attained: preliminary data indicates that 55% of beneficiaries are engaged in income-generating activities. Indicator 3 is likely to be attained although longer-term surveys will be needed to assess this. 37. A second important issue was the lack of communication and sensitization activities around the reintegration component. Whereas reintegration activities were intended to be provided by implementing partners in the form of trainings and funding for sustainable micro-projects, they were widely perceived by beneficiaries as further monetary compensation. Anecdotal evidence suggests that ex-combatants were aware of the amount budgeted per capita for their reintegration by the implementing partners (FBU 600,000). This resulted in the misconception that ex-combatants expected to receive this amount in cash instead of receiving it in the form of training and other activities. Furthermore, a lack of an aggressive communication strategy on the part of the Government led to widely held misperceptions on the project amongst the population. If bad news was relayed extensively in the media, good news was never communicated. For instance the publication of the survey on the impact of reintegration in 2008 showed substantial results and could have been used as a communication tool. As one of the key informants for this ICR puts it: "The timid communication arrangements put the project on the defensive, thus further reducing ownership by the Government." PDO 2: Support the reinsertion of an estimate 20,000 Gardiens de la paix militia of Government and approximately 10,000 Combattants militants militia of the APPMs. 38. This PDO is rated highly satisfactory. Each eligible beneficiary from the Gardiens de la paix or combattants militants received one-time payment for reinsertion of 100,000 FBU (~ 80 USD). A total of 18,709 Gardiens de la paix and 9,674 Combattants militants received the grant. Although external to the project, it must be noted that the reinsertion process was not without protests among the population and issues related to the lists of beneficiaries from these two groups. The outcome for this PDO is similar to the outcome listed in PDO1: demobilization and reintegration into civilian life. However, the objective for this group was strictly limited to distribution of the "recognition-of-service allowance". 14 PDO 3: Contribute to the reallocation of Government expenditure from defense to social and economic sectors over a period of five years. 39. This PDO is rated moderately unsatisfactory. While education and health expenditures continuously increased and defense expenditures decreased, the overall security sector expenditures grew overtime due to increased military, police and interior security expenditures. The original performance indicator for this PDO: Figure 4. Table on PDO 3 2005 2006 2007 Defense FBU 49.4 billion FBU 36.1 billion FBU 23.8 billion expenditures max. max. max. Health & + FBU 6.7 billion + FBU 19 billion Education combined 40. The actual evolution of Burundi public budget expenditures is detailed in the 2008 Burundi CAS: "The Government has increased public expenditure allocations to priority sectors over the past six years. Demobilization has reduced defense expenditures, but the increase in the police force has more than offset the savings." Figure 5. Total public expenditure (domestically financed, current FBu billion) 2001 2002 2003 2004 2005 2006 2007 2008 actual actual actual actual actual actual revised revised expenditure expenditure expenditure expenditure expenditure expenditure budget budget FBu billion Education 17.0 22.3 27.9 32.8 43.5 66.4 87.5 97.8 Health 3.7 3.8 4.0 4.7 5.2 9.9 16.0 34.5 Defense 34.0 38.4 40.0 42.7 43.3 47.9 53.6 57.6 Agriculture 1.9 2.0 1.9 2.1 2.2 3.2 6.4 15.8 Energy and Mines 0.5 0.3 0.4 0.6 0.6 2.1 4.7 5.2 Transport and Telecom 0.5 0.6 0.3 0.7 0.8 0.6 1.5 2.3 Public works 1.1 1.5 1.5 2.1 2.2 4.5 8.5 15.3 Other 19.5 25.4 30.8 35.3 51.2 73.7 95.2 138.2 Total 78.2 94.3 106.8 121.0 148.9 208.3 273.4 366.9 Source: Ministry of Finance, SIGEFI; excludes interest payments and amortization 41. It is possible to establish linkages between the project's outputs (downsizing of the army) and the outcome (decrease in defense expenditures). However, the overall increase in security sector spending indicates that the risk of savings from the downsizing not being used to increase health and education spending, identified in the technical annex, has not been completely averted. It must be noted that this risk had been identified as external and as such, the available mitigation measures were limited. 15 3.3 Efficiency: 42. Efficiency is rated as moderately unsatisfactory. Although the number of beneficiaries assisted was less than half of the maximum number planned for at inception stage with a little over 23,000 adult ex-combatants demobilized, 76% of the budget for reintegration was utilized as of December 2008. The Executive Secretariat allowed significant cost overruns while failing to provide timely reintegration assistance and efficiently award contracts to implementing partners. Financial management was downgraded in the last ISR of the project to moderately unsatisfactory to reflect poor management of funds and potential cases of ineligible expenses. As of today, approximately 30% of the project's budget (ca. US$20 million) remains undisbursed. 3.4 Justification of Overall Outcome Rating: Rating: moderately satisfactory 43. The overall outcome rating is moderately satisfactory based on highly substantial relevance, an average of moderately satisfactory PDOs, and moderately unsatisfactory efficiency. The main objective of the project was to contribute to peace and stability in Burundi by supporting the efforts of the Government to demobilize and reintegrate up to 55,000 ex-combatants from national forces and armed political parties and movements. With the notable exception of combatants from the FNL-Rwasa and FNL-Dissidents, which was beyond the control of the project, this objective has been widely achieved. A unified army is now functioning and considered by key informants as a source of stability for the country. Over 26,000 ex-combatants have been demobilized including over 3,000 child soldiers. Reinsertion grants have been delivered on schedule to close to 100% of beneficiaries. Reintegration outcomes, although mixed, have allowed for a peaceful return of ex-combatants to their communities. However the efficiency of the project was undermined by poor management from the implementing agency, the overall objective of decrease in the defense budget was not fully achieved and the reintegration component partly failed to provide ex-combatants with long-term development solutions. 3.5 Overarching Themes, Other Outcomes and Impacts: (a) Poverty Impacts, Gender Aspects, and Social Development 44. Female ex-combatants were identified in the project as part of those vulnerable groups requiring targeted assistance. This assistance included separate living areas in demobilization centers for female ex-combatants and their children, specialized medical screening in the demobilization centers, gender sensitization for implementing agency personnel, promotion of female associations and involvement of spouses during reintegration activities. However an ongoing issue remained eligibility since female individuals were rarely given the status of ex-combatant. The project attempted to circumvent this issue by including families and spouses in reintegration activities however this method only partially addressed the issue. Analytical work was conducted on youth in Burundi and the impact of the conflict on gender norms. (b) Institutional Change/Strengthening 16 45. Institutional strengthening was not a component of the project as D&R activities are transitional by nature. The institutions in charge of implementation are not intended to outlast the demobilization of the last remaining armed group. The follow-up phase intends to phase out and mainstream most of the longer-term activities related to pensions and special assistance to the relevant ministries. However, it is worth noting that this project is has allowed a national agency, in this case the CNDRR, to run a DDR program by itself, as opposed to many DDR projects implemented by international partners. Although the CNDRR will not be sustained after the demobilization of the last remaining armed group, the institutional experience gathered during this process along with the phasing out of activities to line ministries allow significant institutional strengthening within the GoB altogether. (c) Other Unintended Outcomes and Impacts 46. There are no unintended outcomes and impacts that are measured. The spread of HIV in the demobilization centers had been included as a risk and mitigation measures were adopted such as HIV/AIDS sensitization, voluntary counseling and testing in the centers. 47. Although there is no hard data supporting this, anecdotal evidence suggests that reintegration activities in the communities fostered the creation of ex-combatant associations. 48. One of the larger outcomes of the project is related to the overall peace process and security sector reform exercise. Indeed, the army integration process could not have gone forward without the "exit valve" provided by the project for those not eligible for integration. In this regard, the project acted as a "catalyst" for progress towards peace and stability in Burundi and was a key element of the peace process. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops: 49. Throughout the project, monitoring and evaluation activities produced a wealth of data on the status of beneficiaries of the program. A beneficiary survey was conducted in 2006 on the reintegration of children associated with armed forces, assisted in an MDRP special project with UNICEF. The results from this survey are briefly discussed although child soldiers were not beneficiaries of the national project. Although no formal beneficiary survey was conducted for the project, four extensive surveys have been conducted (August 2005, December 2005, June 2007 and July 2008) and these rendered key information on the social and economic status of beneficiaries. These four surveys initially aimed at verifying that payment had indeed been transmitted for the TSA and other assistance, but the quality of the data gathered and the size of the sample (in average 10% of target population) make them extremely informative beyond verification of payment. They indicate for instance how cash transfer or kits were used. In addition, an impact evaluation was conducted in 2008 on social economic reintegration and produced useful data such as whether skills acquired at training had been used since. For this evaluation 377 ex-combatants were selected from a population of 6,513 reintegrated ex-combatants through random sampling from 36 communities after stratification by regional bureau. 50. Unemployment: The third verification survey showed 69.1% of ex-combatants had set up an income generating activity. In the 2008 study on reintegration, 12.4% of ex-combatants were marked as unemployed (5% said they 17 were unemployed but had worked before and 2.9% said they were unemployed and had never worked before).12 55% of ex-combatants interviewed said they were able to satisfy at least one of the following basic needs: access to potable water, decent housing, adequate clothing, eating at least twice a day, children's education, and healthcare. 22% said they could cover two and 14% said they could cover three of these needs. Only 2% of those interviewed said they could satisfy all six of the basic needs listed.13 51. Comparison with the remainder of the community: In the 2008 survey, 42.8% of ex-combatants stated that their socio-economic situation was worse than that of their former colleagues or members of their community, but another 38.4% said it was the same. Relative housing perceptions varied widely by survey. 38.6% of child ex-combatants said their housing conditions were the same as those of others in the village, but 19.3% said their housing was much worse while only .7% said theirs was much better than other housing in the village. Child ex-combatants were also asked about their households' ownership of animals and land relative to their community. The mode for both responses was "less" (25.6% for animals and 40.4% for land) than the average for the community, but the next entry with the highest responses was "same" (10.2% and 21.8%, respectively). The same pattern followed for food ­ 46.0% of child ECs believed that they had less to eat than other people in their community, and 31.9% felt they had the same amount. 52. Relative poverty: One of the two surveys using the ladder diagram specifically labeled the 6 rungs as "very rich," "rich," "comfortable," "average (can get by)," "poor," and "very poor." Most responded "average (can get by)," with 37.3% of responses. 31.3% and 5.0% of respondents said they were "poor" and "very poor," respectively. Interviewees were also asked about their position just after demobilization. 41.3% of respondents indicated a higher level for their position at demobilization than at present, while 35.7% indicated a lower position. 22.9% indicated the same level for both periods. 4. ASSESSMENT OF RISK TO DEVELOPMENT OUTCOME Rating: Significant 53. The risk is rated significant for two main reasons: (i) the overall political context in Burundi remains uncertain and the upcoming 2010 elections could provide momentum for civil unrest and political instability; (ii) the long-term results of reintegration activities cannot be measured in the short- term and will be observed overtime. The follow-up phase for the demobilization and reintegration of the FNL will provide an opportunity to establish the occurrence of "recycling", when a previously demobilized beneficiary attempts to pose as an ex-combatant again. Registration and verification against pre-existing lists from the DRRP could provide the supporting data for this analysis. 5. ASSESSMENT OF BANK AND BORROWER PERFORMANCE 5.1 Bank Performance 12 ISTEEBU (2008). Evaluation de l'Impact de Réintégration Economique et Sociale. 13 ISTEEBU (2008). 18 (a) Bank Performance in Ensuring Quality at Entry: Rating: satisfactory 54. Bank performance at entry is rated as satisfactory. Project preparation was highly relevant to the country context and the Bank's assistance strategy. The financing arrangements put in place with the complementarity of IDA funds and the multi-donor trust fund allowed for a flexible and responsive assistance to the borrower's needs. Risks identified were appropriate especially with regards to financial management risk ratings and mitigation measures. The excellent turnover time for the preparation of this operation justifies the satisfactory rating as it was able to respond to a critically urgent need from the Government in a timely yet efficient manner. (b) Quality of Supervision: Rating: moderately satisfactory 55. The quality of supervision is rated as moderately satisfactory. The Bank team devoted significant resources for the supervision of this project, with extended missions at least 4 times a year from 2004 to 2008. Moreover, from 2005 to 2008 a full-time position was financed in the Bujumbura country office to supervise the project. Three TTLs succeeded each other at the head of the project during the 5 years of preparation and implementation. Although these transitions did not severely affect the project, the team was not fully able to maintain continuity in supervision and data collection. Notwithstanding the fact that missions from the task team were significantly more frequent than for any regular Bank operation, and that an MDRP representative was based in Bujumbura, even stronger supervision was needed, as was identified by all stakeholders interviewed for this ICR. 56. The original structure of the MDRP as an umbrella regional program provided the team with the flexibility to supervise, provide technical assistance and commission analytical work. However this original funding mechanism did not always allow for assistance to be fully targeted and adapted to country-specific parameters such as the low capacity. Donor partners have argued that the project would have benefited from a more targeted, country-specific program without specifying exactly why. 57. Financial management, although identified early as a high risk for the project, did not benefit from enough technical assistance and supervision. This lead to blockages in cash flow and a lack of transparency. This in turn resulted in delays for the selection of implementing partners, and a potential case of mismanagement currently under internal investigation by the Government (INT was alerted and is currently reviewing the case). (c) Justification of Rating for Overall Bank Performance: Rating: moderately satisfactory 58. Bank's performance is rated as moderately satisfactory. One important lesson learned is that the Bank's business model has to be further adapted to operations in fragile states to allow for enhanced supervision. A stronger presence on the ground has been widely identified by partners as a necessity for this type of project. 19 5.2 Borrower Performance: NOTE: Government and implementing agency being indistinguishable, rating and justification are only provided for overall borrower performance. Justification of Rating for Overall Borrower Performance: Rating: moderately unsatisfactory 59. Borrower's performance is rated as moderately unsatisfactory. It is important to note that this project was implemented in a transitional environment, until the elections of 2005, which did not allow for continuity within the Executive Secretariat. The generally weak capacity must also be factored in the rating. Mismanagement issues and delays in implementation have nonetheless hindered the efficiency of the project and justify this rating. 60. At the inception stage, the Government, particularly the NCDRR and Ministry of Defense, showed great political will and pro-activeness in preparing the project and working with the task team. However after successful elections in 2005 priorities at the Government level shifted and the project team noted an increasing lack of oversight from the NCDRR over the Executive Secretariat. This lack of oversight impacted upon the implementation of the project especially in the reintegration component. 61. The 2005 institutional capacity assessment highlighted numerous structural issues that were only partially addressed. The heavily hierarchical structure of the Executive Secretariat created significant delays in implementation. The level of "counselors" was identified as superfluous for instance, and was later removed. Moreover, lack of transparency in the division of labor and roles and functions of each position hindered efficacy. Human resources, procurement and financial management were key areas where the consultants identified a need for further technical assistance and restructuring. The decentralization of decision-making and application of operations plans at the provincial and local levels was encouraged by the authors of the report. However these recommendations were only superficially followed. 62. A potential financial mismanagement case is currently under investigation by a special commission of the Government. This case has been transmitted to the World Bank's Department of Institutional Integrity and is currently being reviewed. 20 6. LESSONS LEARNED Most lessons learned identified below have been included in the preparation of the next phase for demobilization and reintegration14. Demobilization Improved demobilization center counseling and sensitization. Counseling must be better tailored to the profile of each ex-combatant, and classes be more practical than theoretical. Enhanced security around demobilization camps Demobilized identification card retrieved upon departure from the camp to allow for better social reintegration Reinsertion Incentive mechanism to encourage beneficiaries to use interim transitional support in a sustainable manner. Parallel "Indemnité Transitoire de Subsistance" (ITS) and training, followed immediately by reintegration. Although not initially conceived this way, ITS payments were paid out and then training and reintegration occurred later, sometimes much later. To make ends meet in the interim, many ex-combatants, particularly in urban settings, went into debt, and were forced to sell their reintegration activity inputs to settle their accounts. Ensuring the correct phasing of ITS and reintegration activities can mitigate this problem, and allow for a more logical progression of support, learning and implementation. Reintegration Synergies with wider development activities. The short-term approach to reintegration in the last phase fell short when other long-term development initiatives did not come on line as expected. To ensure maximum exploitation of all possible synergies the following options can be considered: (i) continuous mapping of wider national development activities that exist or are coming on line; (ii) continuous outreach to relevant organizations on possibilities for referral or ways to integrate reintegration activities into their overall programming; and (iii) a referral mechanism to assist ex-combatants to exploit such opportunities in their communities. From the beneficiary perspective, dovetailing the program's transitional reintegration support with long-term global development assistance will multiply activity gains and the likelihood of sustainability. From the programmatic perspective, this will ultimately allow for 14 Emergency Project Paper for the proposed Emergency Demobilization and Transitional Reintegration project in Burundi, Annex 8 "Lessons learned from EDRRP". 21 mainstreaming of ex-combatant reintegration activities, allowing for a phase-out of such activities through the national D&R program. Quality of training and follow-up. For reasons of parity with past beneficiaries, material reintegration entitlements cannot be changed. However, the quality and intensity of both training and follow-up should be reinforced, in the acknowledgement that wider development efforts are still limited in Burundi and economic growth is weak. For some ex-combatants, D&R reintegration support may comprise a major part of their overall development assistance. An effort should also be made to ensure that training content is better adapted to the target groups in question. Some combatants have had little or no schooling and need more basic skills and a different kind of follow-up to increase their chances of success at their chosen reintegration activity. Additionally, the training should be differentiated by age and gender. Reintegration activities adapted to Burundi's job market and opportunities. Promote labor-intensive public works for reintegration activities and establish monitoring mechanisms at the implementing partners' level to continuously assess the economic opportunities in the community of resettlement. Strong encouragement of spouse and family participation. Spouses and families should be encouraged to participate at some or all stages of reintegration support (activity choice, training, and implementation) by all implementing partners. The involvement of spouses and families in reintegration activities has been shown to increase the chances of their sustainability, but spousal involvement in Burundi has been low. This issue must be addressed in a more coherent manner. Possibilities for community involvement/support should also be identified. Encouragement of ex-combatant participation in associations. Associations can be an important means to social reintegration and economic security. They are also another possible means of reinforcing sustainability and mainstreaming reintegration support, since associations can also apply for support via other development programs and initiatives. Establishment of "peer exchange" mechanisms for knowledge transfer among implementing partners and among ex-combatants. Exchange of experiences among implementing partners would enhance coherence of approach (largely lacking before) and the transfer of operational lessons. A peer-learning mechanism for ex-combatants could be an important means of increasing the chances for success in reintegration projects. Adapted assistance for children associated with the armed forces that have children themselves. Reintegration support for children should be adapted to take into account those older "children" and/or almost adults, especially those with their own families. Greater consideration and inclusion of the psycho-social aspect. Psychosocial treatment capacity in Burundi is extremely low, and there are few (if any) talking therapy options for the general population. Psychosocial support in the 22 last program depended on the capacity and discretion of individual implementing partners, and referrals largely resulted in vouchers for medication at the one psychiatric center in Burundi. Given the many psychological traumas observed among ex-combatants, this issue should be discussed with the GoB and partners to see if any better solutions can be found, including how to address traumas related to gender-based violence for female ex-combatants. Institutional mechanisms and processes Project implementation by a small technical coordination team. The TCT should be staffed with competitively recruited technical personnel. This would counter past institutional weaknesses associated with its political nature interfering with technical capacity. Highly qualified financial and procurement staff, reinforced by an independent international consultant to verify financial and procurement tasks. Such an arrangement would ensure improved financial and procurement procedures following irregularities in project financing and procurement observed in the last project. Enhanced monitoring & evaluation. Proper and systematic registration of all beneficiaries (biometric or other). Systematic data collection for tracking purposes after demobilization. A single-purpose multi-donor trust fund, with an associated steering committee including GoB and donors. This would help preserve the same harmonization of approach hitherto carried out through the MDRP framework. An active role for the NCDRR. The NCDRR did not fulfill its policy role in any active sense until relatively recently when a delegate was appointed to act in the President's stead. This active guidance has proved crucial to resolving challenges at many levels, and it must continue. Other Communication and sensitization activities must be proactive, not reactive. There was virtually no pre-emptive strategic communications effort during the last program. Many negative reactions among beneficiaries could have been at least minimized had there been adequate and constant targeted messaging to various groups. The new project should ensure effective communication and sensitization at different levels, including beneficiaries, the communities of return, the general population as well as the donor community and other partners such as the UN agencies. Sensitization campaigns. Events on the national stage should be organized for the population and public officials to educate on the scope and objectives of the project. 23 7. COMMENTS ON ISSUES RAISED BY BORROWER/IMPLEMENTING AGENCIES/PARTNERS (a) Borrower/implementing agencies: See Annex 6 (b) Co-financiers: See Annex 7 (c) Other partners and stakeholders (e.g. NGOs/private sector/civil society): See Annex 7 24 ANNEXES ANNEX 1. PROJECT COSTS AND FINANCING (a) Project Cost by Component (in US$ million equivalent) Actual /Latest Estimate Percentage of Appraisal Appraisal Estimate Components 31/12/2008 (US$ million) (US$ million) 1. Demobilization 4.8 2. 3 47.92 2. Reinsertion 31.5 16.8 53.33 3. Reintegration 20.9 16.0 76.56 4. Technical assistance 2.0 2.4 115.00 5. Program management 8.8 9.6 107.95 6. Special target groups 12.4 5.4 43.55 Total Baseline Cost 80.4 52.3 65.05 Physical Contingencies 4 Price Contingencies 0 Total Project Costs 84.4 Project Preparation Facility (PPF) 0.59 0.59 100 Front-end fee (IBRD only) 0 Total Financing Required 84.4 (b) Financing Appraisal Actual/Latest Percentage of Source of Funds Type of Financing Estimate Estimate Appraisal (US$ million) (US$ million) [Government] Borrower counterpart 3.0 0 0% financing [IBRD/IDA or GEF] IDA 33.0 33.0 100% [MDRP regional TF] WB-administered 41.8 41.8 100% MDTF Germany: German Bilateral 6.0 0 0% Technical Assistance Corporation (GTZ) 25 ANNEX 2. OUTPUTS BY COMPONENT Components Outputs 1. Demobilization 26,283 combatants demobilized (22,502 adult men, 516 adult women, 3,121 children assisted through special project). 26,283 ex-combatants received medical screening. 40% of beneficiaries get tested for HIV/AIDS (voluntary). 26,283 ex-combatants received pre-discharge orientation. 18,709 Gardiens de la Paix dismantled. 9,674 Militants Combattants dismantled . 2. Reinsertion 98.89% of ex-combatants opened a bank account. 99.15% of ex-combatants demobilized received at least the first tranche of transitional subsistence allowance. 23,022 (over 98%) have received all three tranches of reinsertion. 18,709 Gardiens de la paix received their single reinsertion payment. 9,674 Militants Combattants received their single reinsertion payment. 3. Reintegration 21,996 (83%) have received reintegration assistance: - 21,456 received funds and training for income- generating activities - 469 received vocational training - 71 enrolled in formal education 4. Customized 1,907 disabled ex-combatants have received medical reintegration assistance support. for special groups 1,195 disabled ex-combatants have received specialized socio-economic reintegration support. 3,261 child soldiers have been released from the armed forces, 3,259 reunited with their families (program run by UNICEF in parallel MDRP special project). 516 female ex-combatants have received targeted medical assistance and socio-economic reintegration assistance. 26 ANNEX 3. ECONOMIC AND FINANCIAL ANALYSIS At its inception, the project performed an analysis of defense expenditure in relation to key economic and financial data in order to assess the potential long-term impact of military downsizing on social and economic development. Four main sources were used: (i) GDP and other macroeconomic data from the IMF for 2000 to 2003, with projections for 2004 - 2006; (ii) the FAB salary structure; (iii) data on expenditure from the Ministry of Finance; and (iv) data provided by Government in the Letter of Demobilization Policy. Analysis of defense expenditure. From 1992 to 2001, defense expenditure increased from 3.5% of GDP to 8.0%. Nominal defense expenditure increased from FBU 30.5 billion in 1999 to FBU 45.6.3 billion in 2002. Impact of Demobilization. The analysis of the impact of demobilization is based on several assumptions, and is therefore only indicative15. Although 55,000 combatants would be demobilized under the DREW, the calculation of the Program's fiscal impact is based on the planned net reduction in force of the NDF of 20,000 from the current size of the FAB. With demobilization completed during 2007, the net reduction in force of 20,000 NDF soldiers would reduce defense expenditure by approximately FBU 22 billion (US$19.9 million) annually over the 2003 budget. The share of defense expenditure in total current expenditure would decrease from 31.7% in 2003 to 13.5% in 2007. The share of defense expenditure in GDP would decrease from 7.1% to 2.6% over the same period. Concomitantly, social sector expenditures would almost double. Analysis of alternatives. The Government's proposal, and the IDA appraisal of it, have taken into consideration various alternatives in the design of the DRRP. The final design is based on best practices and responds to the specific requirements of the situation in Burundi. It pursues a gradual approach to 15 The following assumptions are used: (i) integration of APPMs into the national defense force, increasing its size from 45,000 in 2003 to 66,000 in 2004; (ii)the demobilization of 5,000 FAB soldiers in 2004,14,000 NDF soldiers in 2005, 14,000 NDF soldiers in 2006, and 13,000 NDF soldiers in 2007; and (iii)the reallocation of 100% of budgetary savings to social sectors. 27 demobilization and reintegration as compared to a simple one-off discharge. The political and security requirements stipulated in the Arusha Agreement (August 2000) and the Forces Technical Agreement (November 2003), the stabilization of the country before and during elections, the need for a structured defense and security sector reform process, and the difficulty of economic reintegration in the immediate post-war period were judged to justify the related short-term defense expenditure costs. The Program tries to support the reintegration of ex-combatants into civilian life by providing a transitional safety net to facilitate the re-establishment of households and to avoid social tensions. The Program's design attempts to minimize interference with market forces and to maximize the choice of the individual and synergies with ongoing interventions. Furthermore, economic reintegration measures would be based on the socio-economic profile of the ex-combatants, their own efforts and aspirations, and the specific opportunity structure in their communities of return. Cost-effectiveness analysis. The Program design seeks to minimize costs without compromising quality. Whenever possible, existing structures would be used rather than parallel systems created. For example, with respect to the disbursement mechanisms, transitional safety net allowance payments would be effected directly by the ES/NCDRR through participating banks. With overall per capita costs of US$1,325 equivalent, the costs of the DRRP are within the range established by the MDRP framework. Cost-benefit analysis. It is instructive to compare the total cost of the Program US$84.4 million with the direct savings in recurrent cost to the Government's budget from the medium-term net reduction in force of 20,000 soldiers. Compared to the counterfactual16, the reduction in force of the NDF would save approximately US$40 million during the Program's duration and US$22 million annually thereafter. Fiscal savings from demobilization would thus exceed Program costs in 8 years. This calculation does not take into account long-term savings in overhead costs to the NDF, including training and maintenance of material. I t also does not take into account the budgetary impact of a planned increased in the size of the national police force. Finally, this simple comparison does not take into account productivity gains and benefits to the communities and households of the ex-combatants' return to civilian life. 16 The counterfactual would be that the share of military expenditure in Government current expenditure remains constant at 32%. 28 ANNEX 4. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION PROCESSES (a) Task Team members Names Title Unit Responsibility/ Specialty Lending (The system pulls from Task Team in PAD Data Sheet, if any.) Supervision (The system pulls from Task Team Members in all archived ISRs.) Senior Social Development Gregory Alex AFTCS Spec Financial Management Otieno Ayany AFTFM Specialis Pamela Ording Beecroft Operations Analyst AFTCS Nina Chee Sr Environmental Spec. MIGEP Sr Financial Management Bella Lelouma Diallo AFTFM Specia Senior Social Development Madjior Solness Dingamadji AFTCS Spec Sameena Dost Sr Counsel LEGAF Senior Social Development Marcelo Jorge Fabre AFTCS Spec Senior Social Development Harald Hugo Hinkel AFTCS Spec Ludovic Joseph Kabran Senior Operations Officer CSRRM Toni Ntaganda Kayonga Operations Officer AFTCS Senior Social Development Cornelis Arend Kingma AFTCS Spec Elisabeth Maier Consultant COCPO Estella Malayika Program Assistant MDO Natacha Meden Communications Officer AFTS3-HIS Sarah G. Michael Social Development Spec. AFTCS Sr Financial Management Joseph Kizito Mubiru LCSFM Specia 29 Daniel J. Murphy Senior Country Officer AFCC1 Senior Procurement Prosper Nindorera AFTPC Specialist Leoncie Niyonahabonye Office Manager AFTEN Marie-Claire Nzeyimana Communications Associate AFREX Senior Social Development Pia Peeters AFTCS Spec Fanny Robert Operations Analyst AFTCS Mamadou Sevede Consultant AFTH3 Aurore Simbananiye Team Assistant AFMBI Financial Management Emmanuel Tchoukou AFTFM Specialis Ingo Wiederhofer Senior Operations Officer EASSO (b) Staff Time and Cost (from SAP) (The system pulls data available for all fields) Fiscal year Source of fund No. of Staff weeks Total cost (USD Thousands) 2005 Bank budget 42.08 104.831 2006 Bank budget 29.46 75.184 Bank budget 46.58 130.851 2007 Trust fund 28.09 204.016 Bank budget 35.68 111.042 2008 Trust fund 94.36 620.889 Bank budget 14.44 60.002 2009 Trust fund 80.58 334.974 30 ANNEX 5. BENEFICIARY SURVEY RESULTS (IF ANY) Selected results from the beneficiary assessment of child soldiers (2006) and the impact evaluation of socio- economic reintegration (2008): Beneficiary Assessment of the Child Soldier (2006) B.3 What do you think of your housing in comparison to others in the village? Much worse 19.3% Worse 30.5% Same 38.6% Better 7.4% Much better .7% B.7 Do you think your household has more or less animals than others in the community? Much less 9.5% Less 25.6% Same 10.2% More 3.9% Much more .4% B.9 Do you think you have more or less land than other people in the community? Much less 14.0% Less 40.4% Same 21.8% More 5.3% Much more 1.4% B.25 Do you feel are isolated within your community? Often 10.5% Sometimes 23.9% Never 64.2% C.9 Do you think you have just as much access to education as your peers? Much less 2.1% Less 2.5% Same 28.8% More 3.5% Much more 2.5% N/A 58.6% D.6 Do you eat your fill? Yes 22.5% No 77.2% D.7 Do you think you have more/ less to eat than other people in your community? Much less 18.2% Less 46.0% Same 31.9% 31 More 3.5% D.8 Do you receive food aid? Yes 35.8% No 62.5% E.1 What are your sources of income? No income 56.5% Work for someone 11.6% Own boss 13.7% Association activities .4% Other 17.2% E.7 What do you think of your business compared to that of other youth in your community? Much worse .7% Worse 2.8% Same 6.3% Better 7.7% Much better 1.8% N/A 80.4% F.6 Where were you on this ladder a year ago? The top of the ladder is someone successful. [5-step ladder] Not there 3.9% 1 40.4% 2 29.1% 3 17.9% 4 6.7% 5 1.1% F.7 Where are you today [on the 5-step ladder]? Not there 6.3% 1 31.6% 2 29.8% 3 20.7% 4 8.4% 5 1.8% Impact evaluation of socio-economic reintegration (2008): CP10. Employment status of the interviewee Employed 87.6% Unemployed 12.4% EP01. What is your current employment status? Employed 90.6% Unemployed, has worked before 5% Unemployed, never worked before 2.9% Student 0.8% Housewife 0.5% Other 0.3% EP06. How do you see the progress in your current main business? Good 26.2% 32 Average 57.3% Bad 16.4% PB01. Are you currently able to satisfy your basic needs regarding:17 Healthcare 2% Children's education 8% Eating at least twice a day 9% Adequate clothing 14% Decent housing 12% Access to potable water 55% PB03. Here is a diagram representing a prosperity ladder with 6 levels. According to your own feelings, on which rung are you now? 1. Very Rich 0% 2. Rich 4.4% 3. Comfortable 21.9% 4. Average (can get by) 37.3% 5. Poor 31.3% 6. Very poor 5.0% PB04. How is your socio-economic situation in comparison to that of your former colleagues or members of your reintegration community? 1. Better than theirs 11.5% 2. Same as theirs 38.4% 3. Worse than theirs 42.8% 4. Can't tell 7.2% 17 The report calculates the number of respondents that can satisfy at least one of their basic necessities, from bottom to top. 33 ANNEX 6. SUMMARY OF BORROWER'S COMMENTS ON DRAFT ICR The Government provided the task team with its own ICR in the later stages of the ICR process. The report is provided as a separate document as it exceeds 10 pages. However, below is the ratings section for Bank and Borrower as translated from the Borrower's ICR: "The World Bank assumed a leadership position in financing of Demobilization, Reinsertion and Reintegration in Burundi. Other donors have followed the path created by the World Bank by contributing their own resources in addition to the World Bank's. We make a special mention of the MDRP which considerably contributed to the PNDRR with its substantial funding. Supervision missions from the World Bank allowed to build operational capacity amongst national structures especially with regards to planning, financial management, monitoring and evaluation et management in general." The Executive Secretariat has reviewed the first draft of this ICR and has provided the team with specific comments regarding the following issues. Component 1: Demobilization The Executive Secretariat wishes to stress the extensive assistance that was provided to beneficiaries upon arrival in the demobilization centers. Specific data was gathered upon registration and were key for the monitoring and evaluation of the project. In particular, data on disarmament (implemented before the project) but also qualitative data on the beneficiaries' expectations were collected. Finally, specific data on disabled ex-combatants were collected in the demobilization centers to tailor the assistance benefits. Upon departure from the demobilization centers, each beneficiary received 20,000 FBu to facilitate transportation to the community of resettlement. Lessons learned The Executive Secretariat identifies as a lesson learned the need for more preparation and information of the beneficiaries before they enter the program. Indeed, the Secretariat regrets the lack of preparation and available information amongst armed groups before the start of the demobilization process. Leaders of armed groups after negotiations should prepare potential candidates for demobilization by providing them with information on reintegration in civilian life. 2005 elections The Executive Secretariat does not identify the 2005 elections and changes in the Implementing Agency as a key factor affecting implementation. M&E 34 The Executive Secretariat wishes to specify the issue of unregistered digital prints. Scanners were found defective but not signaled in time to register all incoming beneficiaries. Digital prints taken also include the Gardiens de la Paix and the Militants Combattants. The lack of tracking tools identified in the ICR under "M&E utilization" is mitigated by actions from the implementing partners. The Executive Secretariat has signaled that although they were not integrated in the MIS, monitoring tools were used by implementing partners during the reintegration component. All reintegration benefits received were later integrated in the MIS. Follow-up phase The Executive Secretariat corrected the estimated number of pending caseload from the project from 1,100 to 1,587. 35 ANNEX 7. COMMENTS OF COFINANCIERS AND OTHER PARTNERS/STAKEHOLDERS Most of cofinanciers were interviewed during the ICR mission in Bujumbura. Below is a summary of their comments on the project. The overwhelming feeling amongst all co-financiers was that the project responded to an emergency situation and was an indispensable step towards the stabilization of the country. In this regard, most donors and members of the international community recognize that the project paved the way for other development projects to be developed in Burundi by guaranteeing a return to peace and stability. Demobilization is generally considered the most successful component of the project. Although the issue of target numbers and overestimation of beneficiaries' caseload is problematic from a World Bank point of view, it is very much overlooked by cofinanciers and other partners. Indeed, everybody understands the political context in which the project was implemented and the reason behind the overestimation of caseloads. The discrepancy between the target numbers and caseloads assisted is therefore not interpreted as a failure but rather a consequence of an overestimation at inception. With the exception of the FNL groups, cofinanciers consider all other ex-combatants to have been demobilized by the program. Reintegration was identified as the most problematic component of the project: there is a lot of criticism amongst cofinanciers regarding implementation of the reintegration component. The first criticism concerns significant delays between the return of ex-combatants to the communities and the beginning of the reintegration activities. There also seems to be misperceptions regarding the relevance of reintegration activities offered. Some cofinanciers expressed concern regarding the high percentage of individual projects involving small trade. Two schools of thought emerged amongst cofinanciers and partners interviewed during this ICR process on the subject of reintegration. The "minimalist" approach on reintegration was held by interlocutors who considered that an ex-combatant returning to his/her community is reintegrated as long as he/she will resettle peacefully among neighbors and have access to the same opportunities than the rest of the community. The "extensive" approach considers reintegration as an opportunity to create sustainable livelihoods, where ex-combatants not being able to create a sustainable income source for themselves is considered a failure of reintegration. Depending on interlocutors' conception of reintegration, the assessment of the project was rated satisfactory to failed. Cofinanciers mentioned the lack of communication as a key issue during implementation of the project. In particular, the weakness of the sensitization component and scarcity of sensitization campaigns was identified as a major obstacle for the program, leading to misperceptions that also affected some partners. Finally, all partners were unanimous in regretting the lack of tight supervision from the World Bank team on the project and the implementing agency. Ongoing allegations of mismanagement within the Secretariat, public demonstrations of ex-combatants as well as conflicts between implementing partners and the Executive Secretariat have convinced the cofinanciers and other partners that the World Bank should monitor more closely the implementation of the project. It is however unclear to most cofincancier how responsibilities and liabilities are split between the World Bank and the Borrower. 36 ANNEX 8. LIST OF SUPPORTING DOCUMENTS 1. Republic of Burundi - Burundi Minister of Finance, [February, 2004] Letter of Demobilization, Reinsertion and Reintegration Policy. - Case d'Afrique/ACT Consultants [December, 2008] Final independent evaluation of the National Demobilization, Reinsertion and Reintegration Program. - Executive Secretariat of the National Commission for Demobilization, Reinsertion and Reintegration of Ex-combatants [August, 2004] Implementation manual for the Demobilization, Reinsertion and Reintegration project. - Executive Secretariat of the National Commission for Demobilization, Reinsertion and Reintegration of Ex-combatants [September, 2007] Evaluation of income-generating activities implemented by international implementation partners. - Executive Secretariat of the National Commission for Demobilization, Reinsertion and Reintegration of Ex-combatants [July, 2007] Social and environmental monitoring & evaluation report - Institut de Statistiques et d'Etudes Economiques du Burundi (ISTEEBU) [August, 2008] Impact Evaluation of Socio Economic Reintegration on Beneficiaries and Host Communities. - Presidential Decrees o Institutional structure for the DDR program [August, 2003] o Rank harmonization [May, 2004] o Status of combatant [August, 2004] o Status of Gardiens de la Paix and Militants Combattants [August, 2004] o Verification mechanism for combatant status [November, 2006] o Demobilization of members of the NDF [February,2008] o Demobilization of members of the national police [March, 2008] o Demobilization of members of the NDF [March, 2008] - Sample verification surveys o BUTOYI [June, 2007]. Troisième Enquête de Vérification des Paiements des Allocations de Réinsertion: Rapport d'Enquête. o BUTOYI [July, 2008]. Quatrième Enquête de Vérification des Paiements des Allocations de Réinsertion : Rapport d'Enquête. - TAOUTI-CHERIF [June, 2006]. Beneficiary Assessment of the Social and Economic Status of the "Child Soldier" Special Project Beneficiaries in Burundi. 2. The World Bank - Aide-memoires/Back to office reports o December 2008 o February 2009 o Mid-term review [June, 2007] - BAGHDADLI, HARBORNE, RAJADEL [May, 2008], Breaking the cycle, A Strategy for Conflict-Sensitive Rural Growth in Burundi, World Bank working paper No. 147. 37 - Burundi Country Assistance Strategy [July, 2008] - Burundi Poverty Reduction Strategy Paper [September, 2006] - Development Grant Agreement [April, 2004] - Implementation Status and Results Reports - Memorandum of the President for the Burundi Emergency Demobilization, Reinsertion and Reintegration project [February, 2004] - Project Technical Annex [February, 2004] - VALENSART [October, 2006], Independent evaluation of reintegration support implemented by local implementation partners. 3. Other - BINZEGGER [September, 2005], Institutional capacity assessment of the NCDRR Executive Secretariat, ACTS. - The Economist Intelligence Unit [February, 2009], Burundi Country Report. - UVIN [May, 2007]. Ex-combatants in Burundi: Why they joined, why they left, how they fared, MDRP Working Paper No. 3. 38 ANNEX 9. LIST OF KEY INFORMANTS Below is a list of key informants who have been interviewed for this ICR in Bujumbura and Washington, DC. Many others have been contacted and provided significant input, such as then country manager Alassane Sow, but have not been formally interviewed due to time and logistical constraints. - Otieny Ayany ­ World Bank, Financial management specialist, AFTFM - Alain Darthenucq - Delegate of the European Commission in Burundi - Madjior Solness Dingamadji ­ World Bank, project coordinator in Burundi until May 2008 - Marcelo Fabre, World Bank, Task Team Leader, November 2007-present - Colonel Faye ­ BINUB, head of the Security Sector Reform/Small Arms Reduction service - Hunter Fraser ­ BINUB, Security Sector Reform/Small Arms Reduction service - Sue Hogwood ­ DFID representative in Burundi - Lieutenant Colonel Jaulin ­ French military attaché in Burundi - Cornelis Kingma, World Bank, Task Team Leader, November 2006 - November 2007 - Gilles Landsberg ­ Belgian embassy, First Secretary - Sarah Michael ­ World Bank, Task team member, 2006-2007 - Charles Muvira ­ Executive Secretariat, Coordinator for reintegration - Jeroboam Nzikobanyaka ­ Executive Secretariat, Project coordinator - Israel de la Piedra ­ IMF, Burundi country representative - Mercy Tembon ­ World Bank, Burundi country manager - Ingo Wiederhofer ­ World Bank, Task Team Leader, 2004-2006 39 [MAP] 40