Document of The World Bank Report No: ICR2798 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-H2800) ON A GRANT IN THE AMOUNT OF SDR 26.9 MILLION (US$40 MILLION EQUIVALENT) TO THE REPUBLIC OF BURUNDI FOR A COMMUNITY AND SOCIAL DEVELOPMENT PROJECT June 26, 2013 Post-Conflict & Social Development Practice Group(AFTCS) Eastern AFR 1Department Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective May 15, 2013) Currency Unit=Burundi Francs (BIF) 1.00 BIF = US$0. 643 US$1.00 = 1.556 BIF FISCAL YEAR ABBREVIATIONS AND ACRONYMS ABUTIP Burundian Public Works Agency CAS Country Assistance Strategy CDC Communal Development Committee CDD Community Driven Development CDP Communal Development Plan CMU Country Management Unit DGMAG Directorate General of Decentralization and Mobilization of Self-development EA Environmental Assessment ESMF Environmental and Social Management Framework EU European Union FA Facilitating Agency FM Financial Management GIZ German Development Cooperation GoB Government of Burundi ICR Implementation Completion and Results Report IDP Internally Displaced Person IEC Information, Education, and Communication IFAD International Fund for Agricultural Development IFR Interim Financial Report IRC International Rescue Committee ISN Interim Strategy Note ISR Implementation Status Report M&E Monitoring and Evaluation MIPS Ministry of Interior and Public Security MoH Ministry of Health MTR Mid-term Review NOSC National Orientation Steering Committee PAD Project Appraisal Document PDO Project Development Objective PIU Project Implementation Unit PRADECS Community and Social Development Project PRSP Poverty Reduction Strategy Paper QALP Quality Assessment of Lending Portfolio RF Results Framework SIL Specific Investment Loan SMU Sector Management Unit TTL Task Team Leader USAID United States Agency for International Development ii Vice President: MakhtarDiop Country Director: Philippe Dongier Sector Manager: Ian Bannon Project Team Leader: PiaPeeters ICR Team Leader: Leanne Bayer iii REPUBLIC OF BURUNDI COMMUNITY AND SOCIAL DEVELOPMENT PROJECT CONTENTS Data Sheet A. Basic Information…………………………………………………………………………v B. Key Dates………………………………………………………………………………….v C. Ratings Summary……………………………………………………………………….....v D. Sector and Theme Codes………………………………………………………………….vi E. Bank Staff…………………………………………………………………………………vi F. Results Framework Analysis……………………………………………………………..vii G. Ratings of Project Performance in ISRs…………………………………………………..xi H. Restructuring ……………………………………………………………………………..xi I. Disbursement Graph………………………………………………………………………xi 1. Project Context, Development Objectives and Design ........................................................... 1 2. Key Factors Affecting Implementation and Outcomes ........................................................... 4 3. Assessment of Outcomes ........................................................................................................ 9 4. Assessment of Risk to Development Outcome ..................................................................... 15 5. Assessment of Bank and Borrower Performance.................................................................. 16 6. Lessons Learned.................................................................................................................... 18 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners ....................... 19 Annex 1. Project Costs and Financing ...................................................................................... 20 Annex 2. Outputs by Component .............................................................................................. 21 Annex 3. Economic and Financial Analysis ............................................................................. 25 Annex 4. Bank Lending and Implementation Support/Supervision Processes ......................... 27 Annex 5. Beneficiary Survey Results ....................................................................................... 29 Annex 6. Stakeholder Workshop Report and Results ............................................................... 30 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR ................................. 31 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders ................................... 40 Annex 9. List of Supporting Documents .................................................................................. 41 MAP iv A. Basic Information Community and Social Country: Burundi Project Name: Development Project Project ID: P095211 L/C/TF Number(s): IDA-H2800 ICR Date: 06/11/2013 ICR Type: Core ICR MINISTRY OF Lending Instrument: SIL Borrower: FINANCE Original Total XDR 26.90M Disbursed Amount: XDR 26.68M Commitment: Revised Amount: XDR 26.90M Environmental Category: B Implementing Agencies: Ministry of Communal Development Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 07/26/2005 Effectiveness: 08/28/2007 08/28/2007 Appraisal: 11/27/2006 Restructuring(s): Approval: 03/21/2007 Mid-term Review: 09/2009 Closing: 12/31/2012 12/31/2012 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Substantial Bank Performance: Satisfactory Borrower Performance: Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance: Performance: v C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry No None at any time (Yes/No): (QEA): Problem Project at any Quality of No None time (Yes/No): Supervision (QSA): DO rating before Satisfactory Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General education sector 16 16 General public administration sector 15 15 General transportation sector 8 8 Health 8 8 Other social services 53 53 Theme Code (as % of total Bank financing) Conflict prevention and post-conflict reconstruction 33 33 Decentralization 17 17 Participation and civic engagement 16 16 Social Inclusion 17 17 Vulnerability assessment and monitoring 17 17 E. Bank Staff Positions At ICR At Approval Vice President: MakhtarDiop Hartwig Schafer (Acting) Country Director: Philippe Dongier Pedro Alba Sector Manager: Ian Bannon Francois G. Le Gall Project Team Leader: PiaPeeters Valerie Marie Helene Layrol ICR Team Leader: Leanne Michelle Bayer ICR Primary Author: Leanne Michelle Bayer F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) As per the Financing Agreement, the overall objective of the project is to establish and operate a decentralized, transparent, and participatory financing mechanism that empowers local vi governments and Communities for purposes of improved and more equitable local service delivery. ISR reports include the following PDO, which shows small differences, based upon advice from AFTRL during the first year of project implementation: The project development objective is to promote better and more equitable local service delivery. The project will achieve its objective by establishing and operationalizing a decentralized participatory and transparent financing mechanism that empowers local governments and communities. Revised Project Development Objectives (as approved by original approving authority) The original PDO did not change. (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 At least 15% of targeted communes have obtained funding for Communal Indicator 1 : Development Plan activities from outside the project. Value quantitative or 0 15% 100% Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments All communes have obtained at least one additional financing to their Communal (incl. % Development Plan from outside the project. achievement) At least 70% of infrastructure financed under subprojects is maintained and Indicator 2 : operational two years after completion of the respective Subproject. Value quantitative or 0 70% yearly 95% Qualitative) Date achieved 09/13/2007 12/31/2012 Comments 95% of sub-projects finalized are operational, based upon subprojects finalized (incl. % and in use. Maintenance measures and committees are in place in each commune. achievement) At least 70% of targeted communes show an increase in access to and use of primary health, primary education, and water services, since time of conclusion Indicator 3 : of as compared with that at the time of conclusion of the respective subproject Grant Agreement. Value quantitative or 0 70% yearly 97% Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments 97% of infrastructure related to health, education and increased access to water (incl. % are functional, accessible, and in use. Break down: primary health (74% achievement) achieved), primary education (99%), water (100%) At least 50% of the targeted Communes have integrated the concerns of Indicator 4 : Vulnerable Groups in their Communal Development Plans. Value 0 50% yearly 100% quantitative or vii Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 All targeted communes included the concerns of vulnerable groups in the Comments Communal Development Plans - both in terms of ensuring the participation of (incl. % vulnerable groups in meetings and identified projects within the plans that target achievement) vulnerable groups. (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 : People in project areas with access to Improved Water Resources Not applicable, Value subproject (quantitative 0 347,800 people selection is or Qualitative) demand driven. Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments 740 water sources finalized. Each source serves on average 94 households with (incl. % an average of 5 people per household. achievement) Indicator 2 : Direct project beneficiaries. Value (quantitative 0 Not applicable. 3,485,042 or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments (incl. % 1,707,671 male beneficiaries and 1,777,371 female beneficiaries. achievement) Indicator 3 : Latrines built Not applicable, Value subproject (quantitative 0 3131 selection is or Qualitative) demand driven. Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments 526 for primary schools, 482 communal colleges, 342 health centers, 1701 in (incl. % housing for vulnerable group members, 80 public. achievement) Indicator 4 : Improved community water points constructed or rehabilitated under the project. Not applicable, Value subproject (quantitative o 740 selection is or Qualitative) demand driven. Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments (incl. % 662 water sources and 118 water points rehabilitated and functional. achievement) viii Indicator 5 : Roads rehabilitated, rural. Not applicable, Value subproject (quantitative 0 102 kilometers selection is or Qualitative) demand driven. Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments (incl. % 20 rural roads rehabilitated, totaling 102 kilometers. achievement) Indicator 6 : Health facilities constructed, renovated, and / or equipped (number). Not applicable, Value subproject (quantitative 0 74 selection is or Qualitative) demand driven. Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments (incl. % 74 health centers constructed, 55 functional. achievement) Each audit report is unqualified following the audit stipulated by the project Indicator 7 : Financing Agreement. Value (quantitative N/A or Qualitative) Date achieved 08/28/2007 Comments (incl. % Audit reports submitted to the Bank unqualified. achievement) A communication strategy targeting all Project stakeholders has been developed Indicator 8 : and implemented. Value (quantitative N/A or Qualitative) Date achieved Comments 100% strategy completed. Assumption that 30% completed means framework (incl. % was adopted and conceived, and 100% means completely implemented. Project achievement) also added an additional communications strategy for project close-out. Indicator 9 : Value (quantitative or Qualitative) Date achieved Comments (incl. % achievement) At least 60% of targeted communes that provide robust monitoring and Indicator 10 : evaluation information in an appropriate timeframe. Value 0 60% 100% (quantitative ix or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments (incl. % 34 out of 34 communes provide regular reporting information. achievement) Indicator 11 : At least 70 subprojects targeting vulnerable groups have been completed. Value (quantitative 0 70 249 or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments A total of 249 subprojects projects targeting vulnerable groups have been (incl. % completed. achievement) Indicator 12 : At least 430 community and commune level subprojects completed. Value (quantitative 0 430 812 or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments 812 finalized under component 2a. While 818 subprojects were signed, 812 were (incl. % completed. achievement) Training plan of DGDMAD (the direction general for decentralization within the Indicator 13 : line Ministry) has been finalized and implemented satisfactorily. Value (quantitative 0 100% 100% or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 Comments Full training satisfactorily completed. Assuming that 30% adopted means that the (incl. % framework has been conceived, and 100% means the training has been achievement) completely implemented. The planned training was entirely completed. At least 90% of targeted communes that have developed or updated their Indicator 14 : Communal Development Plan on a participatory basis. Value (quantitative 0 90% 100% or Qualitative) Date achieved 09/13/2007 12/31/2012 12/31/2012 All targeted communes developed their Communal Development Plans at the Comments start of the project. Towards the end of the project the communes were (incl. % developing their second five-year CDP's and the project assisted 37 communes achievement) across six provinces. G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 10/12/2007 Satisfactory Satisfactory 2.83 x 2 01/30/2008 Satisfactory Satisfactory 3.12 3 07/29/2008 Moderately Satisfactory Moderately Satisfactory 3.76 4 10/20/2008 Satisfactory Satisfactory 4.59 5 04/21/2009 Moderately Satisfactory Moderately Satisfactory 6.58 6 12/23/2009 Moderately Satisfactory Moderately Satisfactory 11.96 7 05/27/2010 Moderately Satisfactory Moderately Satisfactory 18.96 8 03/27/2011 Moderately Satisfactory Moderately Satisfactory 29.64 9 11/01/2011 Moderately Satisfactory Satisfactory 33.09 10 04/11/2012 Moderately Satisfactory Satisfactory 35.19 11 09/22/2012 Satisfactory Satisfactory 38.08 12 04/24/2013 Satisfactory Satisfactory 41.11 H. Restructuring (if any) The Financing Agreement was amended four times. The first amendment of February 11, 2008 reallocated grant proceeds. The second amendment of December 23, 2010 changed the name of the line Ministry from the Ministry of Interior and Public Security to the Ministry of Planning and Communal Development, revised budget ceilings for subprojects, and lowered the community and communal contributions for subprojects. The third amendment of August 1, 2011 specified that the Second Vice President (as opposed to the First Vice President in the original FA) would chair the NOSC. The final amendment of November 26, 2012 again revised the name of the line Ministry to Ministry of Communal Development and completed a reallocation of proceeds. I. Disbursement Profile xi 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1. The Community and Social Development Project (PRADECS, known by its French acronym in Burundi) was developed following civil conflict in the country and in the midst of a major political transition. In 2005 the newly elected Government revised the electoral code prior to communal elections, and later approved a new Communal Law. The decentralization process initiated by the Government aimed to empower communities to improve local governance and social cohesion, and to improve access to socioeconomic infrastructure. PRADECS was designed as a Community Driven Development (CDD) project with a two-pronged approach to support the new national decentralization strategy and to ensure improved and more equitable access to local socioeconomic infrastructure and service delivery in rural areas where over 90% of the Burundian population live. 2. Country context. At the time of project appraisal, Burundi was emerging from more than a decade of civil conflict that started with the 1993 coup d’état against its first elected government. A decade of civil war resulted in the killing of an estimated 300,000 people and the displacement of 1.2 million (16% of the population) during the conflict. The Arusha Peace Agreement of August 2000 initiated a peace process that was successfully concluded with the promulgation of a new constitution in February 2005 and the installation of a democratically elected government in August 2005. 3. Burundi was, and still is, one of the poorest countries in the world due to the devastating effects of the war on its economy. GDP per capita had fallen by almost 40% from US$180 in 1994 to US$110 in 2003. The percentage of people living on less than one dollar a day had nearly doubled, from 35% overall to 68% in 2005. The country’s social indicators were among the weakest in Sub-Saharan Africa. 4. Sector context. At appraisal, the new Government was embarking on an ambitious decentralization process. In 2005 the Government of Burundi (GoB) adopted a new administrative structure through a Communal Law and a new electoral system at the communal level. Burundi’s 129 communes (116 of which are in rural areas) had been plagued by low capacity and a lack of transparency in public resource management. It was within this context that GoB launched a decentralization process to empower communes to improve local governance and social cohesion, and improve access to socio- economic infrastructure and services. 5. The conflict had gradually destroyed human, physical, and social capital, and created a growing population of vulnerable groups. There were numerous sources of insecurity and instability at the time of appraisal: lack of physical, food, social, and environmental security; extreme land scarcity; unequal access to services; return of refugees and internally displaced people (IDPs); high population growth (2.7% annually in 2003); and unequal access to resources related to gender, regions, and vulnerable groups. The Project Appraisal Document (PAD) noted that exclusion and inequity reinforce poverty and that rebuilding social capital at the community level was critical to restore adequate delivery of social services and consolidate the peace process. 6. At appraisal, communities lacked basic social services (education, health, and water). In health, years of conflict had destroyed the health care system due to lack of resources and adequate health care personnel. As stability and peace were being restored in the country the new health authorities were confronted with the challenge of responding to emergency needs, including those of returning refugees and IDPs, while ensuring the sustainable development of the system. For water, sanitation services in rural areas were limited. Only 22% of the population had access to functional facilities, and 90% of these facilities were traditional pit latrines. In rural areas, only 43% of the population had access to potable water. The education sector was being squeezed by a highly popular policy of the new Government to 1 eliminate primary school fees from September 2005, which led to an immediate and dramatic increase in first grade enrollment. The primary school enrollment rate had surpassed the pre-crisis level, with a gross enrollment close to 100% in 2005-06. Basic infrastructure severely deteriorated during the conflict, and in 2003-04 only 40% of the schools in Burundi were built of hard or semi-hard materials with adequate roofs. Electricity and potable water were almost non-existent in schools. 7. Project context. Burundi’s 2005 Interim Strategy Note (ISN) set out a road map for Bank support for the period FY06-07. The ISN had two main objectives. First was to consolidate the achievements of the past three-year transition period and, in particular, to ensure that communities and the population had access to basic services and were supportive of the peace process and the economic reform agenda. Second was to lay the foundation for sustained economic recovery and growth. The ISN was fully aligned with the priorities set forth in Burundi’s 2006 Poverty Reduction Strategy Paper (PRSP). 8. The Community and Social Development Project came at a critical moment to address the weakened socio-economic service delivery. The project fit within the parameters of the ISN and the PRSP, leveraging three PRSP pillars: (i) improving governance and security, (ii) promoting sustainable equitable growth, and (iii) developing human capital. The project supported the first objective of the ISN by increasing communities’ access to basic socio-economic services, as well as enhancing social stability. At appraisal, other donors had also engaged with CDD projects, and PRADECS was seen as a mechanism to assist GoB to better coordinate donor interventions at the communal level and lay the foundation for longer-term communal development. 1.2 Original Project Development Objectives (PDO) and Key Indicators 9. As per the Financing Agreement, the overall objective of the project is to establish and operate a decentralized, transparent, and participatory financing mechanism that empowers local governments and Communities for purposes of improved and more equitable local service delivery. ISR reports include the following PDO, which shows small differences, based on advice from AFTRL during the first year of implementation: The project development objective is to promote better and more equitable local service delivery. The project will achieve its objective by establishing and operationalizing a decentralized participatory and transparent financing mechanism that empowers local governments and communities. 10. Key performance indicators at the PDO level include at least: (i) 70% of the targeted Communes show an increase in access to and use of primary health, primary education, and water services, as compared with that at the time of conclusion of the respective Subproject Grant Agreement; (ii) 70% of the infrastructure financed under Subprojects is maintained and operational two years after completion of the respective Subproject; (iii) 15% of the targeted Communes have obtained funding for Communal Development Plan activities from outside the Project; and (iv) 50% of the targeted Communes have integrated the concerns of Vulnerable Groups in their Communal Development Plans. 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification The PDO was not revised. 1.4 Main Beneficiaries 11. The main beneficiaries were located within government, at multiple levels, and within communities. At appraisal, the PAD captured well the vertical relationship among the government actors. The Ministry of Interior and Public Security (MIPS) was the original line ministry, with the Directorate General of Decentralization and Mobilization for Self-development (DGMAG). However, this portfolio moved several times between Ministries, finally settling at the Ministry of Communal Development, the last line Ministry for PRADECS and acknowledged as the principle partner in the final amendment to the 2 Financing Agreement. Beneficiaries of the capacity building component were government officials from three levels: communities, communes, and line ministries (at the central and decentralized level). The main beneficiaries of subprojects for social services were the 116 rural communes in Burundi, and the communities serviced by these communes. Vulnerable groups, including destitute women, youth, disabled, people living with HIV/AIDS, orphans and the Batwa (an ethnic minority in Burundi), were specifically targeted as beneficiaries to promote social cohesion and social inclusion. 1.5 Original Components 12. The project as approved consisted of three components: 1) Capacity Building; 2) Financing Subprojects for Local Development; and 3) Project Management, Monitoring and Evaluation, and Communication. Both components 1 and 2 were further broken into two sub-components to reflect specific support to vulnerable groups. 13. Component 1: Capacity building aimed to empower stakeholders for decentralization, focusing on strengthening the capacities of communes and communities to facilitate local development, and augment social cohesion and inclusion. The performance indicators are, at least: (i) 90% of the targeted communes have developed or updated their Communal Development Plans on a participatory basis, and (ii) 50% of targeted communes have developed a strategy to ensure the participation of vulnerable groups in their decision-making process, notably with respect to Communal Development Plans. 14. Sub-component 1a: Capacity building for local development (US$4.8 million) aimed to assist communities and communes undertake participatory planning, as well as initiating, preparing, implementing, managing, and monitoring subprojects. Through the recruitment of Facilitating Agencies (FAs), field-based facilitation teams strengthened local technical and fiduciary skills to implement local development activities. Institutional support was also provided at the national level to the corresponding line Ministry responsible for decentralization. 15. Sub-component 1b: Capacity building for social inclusion and cohesion (US$1.2 million) was designed to strengthen social inclusion and cohesion by targeting vulnerable groups, focusing on information, education, and communication (IEC) activities for vulnerable groups to benefit: (i) local government actors, (ii) organizations that represent or support vulnerable groups, and (iii) associations of vulnerable groups. 16. Sub-component 2a: Financing public subprojects (US$27 million) aimed to provide financial support to communes and communities to implement demand-driven subprojects in: health, education, transportation, water and sanitation, culture and sports, and collective economic infrastructure (markets, slaughterhouses, bus stations). Financial support was provided through a grant and complemented by a contribution from the beneficiaries. The selection of subprojects was carried-out under a participatory mechanism and approved by the respective Communal Councils. The performance indicator is at least 430 communal subprojects have been completed. 17. Sub-component 2b: Financing subprojects to strengthen social inclusion and cohesion (US$3 million) aimed to provide financing for subprojects targeting the special needs of vulnerable groups but unlike Component 2a, beneficiary contributions were not required. Potential subprojects included schooling, housing, professional and vocational training, and alphabetization programs. It also funded community services for counseling, particularly legal aid for vulnerable groups. The performance indicator is at least 70 subprojects targeting vulnerable groups have been completed. 18. Component 3: Project management, monitoring and evaluation (M&E), and communications (US$3.4 million) directly supported the technical, administrative, and financial 3 management of the project. An integrated M&E system was established and a communication strategy was developed and executed. The M&E performance indicator is at least 60% of the targeted municipalities provide robust monitoring and evaluation information in an appropriate timeframe. 1.6 Revised Components 19. Components were not revised throughout the duration of the project. 1.7 Other significant changes 20. The main programmatic changes during the project were related to the shift in project coverage between what was referred to as low- and high-intensity provinces. High-intensity intervention provinces would benefit from capacity building and subprojects, but low-intensity intervention provinces only from subprojects since other donors were providing similar capacity building support as envisaged in PRADECS. During implementation, the coverage of high and low intensity provinces shifted to address the exit or start of other donor programs. Annex 2 includes coverage by province. 21. The thematic working group on decentralization brought together the main actors supporting decentralization (namely GIZ, Swiss Cooperation, the EU, Belgium, the IRC operating with USAID funds, and PRADECS) and information flowed among the working group and the Ministry of Communal Development to ensure coverage across the country. PRADECS was viewed as flexible enough to respond quickly to changes on the ground, especially donor coverage. 22. The Financing Agreement was amended four times. The first amendment (February 11, 2008) reallocated grant proceeds. The second (December 23, 2010) changed the name of the line Ministry from the Ministry of Interior and Public Security to the Ministry of Planning and Communal Development, revised subproject budget ceilings, and lowered community and communal contributions. The third amendment (August 1, 2011) specified that the Second Vice President (as opposed to the First Vice President originally) would chair the Steering Committee. The final amendment (November 26), 2012 revised the name of the line Ministry to Ministry of Communal Development and reallocated proceeds. 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 23. Project preparation. At the time of project preparation, Burundi’s social sector needs were enormous. Project preparation took place during the immediate post-election period of 2005 when the country was optimistic about its future and development prospects. The new Government had moved quickly and by February 2007 had completed the national policy on decentralization and community development, which was followed by a process for all communes in the country to prepare their own Communal Development Plans for the first time. In essence, project preparation was being done in parallel with the Government’s roll out of its decentralization strategy within the context of the first PRSP. 24. As reflected in the PAD, a strong analytical base underpinned project design. Analysis was undertaken to draw from and situate the project vis-à-vis other Bank projects (Burundi Social Action, Agricultural Rehabilitation and Support, and Emergency Demobilization, Reinsertion and Reintegration). An analysis was also done to review donor interventions in Burundi in terms of scope and geographic coverage. This analysis led to the adoption of eight low-intensity and eight high-intensity provinces in project design. The Ministry of Interior also completed an extensive 80-page study on decentralization and CDD to inform the preparation of PRADECS. 25. Design. Project components were unchanged throughout the project. A Quality Assessment of Lending Portfolio (QALP) carried out in May 2010 rated the project’s overall design as Moderately 4 Unsatisfactory. QALP criticized the design of the project as overly complex for Burundi’s conditions, track record and absorptive capacity, and argued that the project should have a better balance between building local capacity, community participatory approaches, and service delivery. Support to decentralization should have been more gradual, using a transitional approach where the capacity of communes would first be strengthened, before the gradual transfer of responsibilities. Prior to the mid- term review (MTR), the tension between the two main components – capacity building and subprojects – became evident. While in essence PRADECS was conceived as a capacity building initiative, and the project spent its first year of implementation dedicated mostly to this task, the perception remained that PRADECS was an infrastructure project to finance construction of schools and health clinics. Capacity building was thus seen as a pre-requisite for receiving subproject funding, instead of supporting the overall decentralization process in the country. 26. Although the project designed framework implementation and supervision by Government was complicated, it worked well and remained in place throughout the project. The structure comprised the NOSC, a Technical Execution Monitoring Committee, the Executive Secretariat (Project Implementation Unit, PIU), and interactions with provincial and communal authorities. In fact, this framework appears heavy on paper but was operationalized within the project context and proved effective in keeping the project on track. This supervision structure also mitigated many of the high risks identified at project entry. The concern centered on the management of funds when creating a decentralized financing mechanism in a post-conflict, fragile context. The supervision framework mitigated the risk of funds mismanagement. During the ICR mission officials from the Ministry of Communal Development noted the funds flow mechanism and the tight control the project maintained on financial oversight. 27. The detailed flow of funds mechanism is a source of contention however for local actors. Considering the risks identified at entry it was necessary to put in place a detailed system for tracking and accounting of project resources being transferred to local communes. Feedback from communal Administrators and their Technical Advisors however points to the complexity of the flow of funds system and a lack of understanding from communal officials on the mechanism, steps, and timing of fund disbursements to local subproject accounts. The appraisal summary in the project document states: “The guiding premise of the project is that there is considerable latent technical and managerial capacity at the local level that could be released through a participatory strategic planning process, followed by a predictable investment envelope to finance activities.� Communal officials had difficulties, especially at the start of the project, in managing procurement and the flow of funds. It seems that the starting premise of the level of capacity existing among communal officials was overestimated. 28. This point feeds into a final comment on project design. It was deemed that enough capacity existed in Burundi among local actors to manage the design of the training curriculum. This capacity seems to have been overstated. Certainly training and facilitation would need to be done in the local language and with knowledge of the local context. However, local organizations had limited prior experience in decentralization and capacity development of local government officials. They were strong on conflict resolution and facilitation, very important aspects during the start of the project in a post- conflict context, but the project could have benefitted from more professional curriculum design, standardization, and training of trainers. This issue was eventually corrected after the MTR. 2.2 Implementation 29. During the five-year project implementation period, Burundi went through significant changes in the political and institutional setting that directly affected the project. The Ministry of Interior and Public Security was initially the line Ministry for the project, but this changed a year later with the creation of a separate Ministry for Decentralization and Communal Development. This reflected the increasing importance of decentralization. Following the change of line Ministry, the national steering committee 5 switched from the First to the Second Vice-Presidency, since the Ministry of Interior reported to the First Vice-Presidency, while the Ministry of Decentralization and Communal Development reported to the Second Vice-Presidency. 30. Overall, the project was implemented successfully. Three factors contributed to this success. First, Government project “champions� emerged early and remained involved. Second, the PIU was staffed by a competent team with strong technical skills. Third, PRADECS underwent a shift in focus and many corrective measures were taken following an early MTR. 31. Government ownership. PRADECS benefitted from high level Government ownership throughout the project. The Director General from the Directorate General of Decentralization and Mobilization for Self-development (DGMAG) from the Ministry of Interior who was involved with the project since its inception, moved with the decentralization office to the Ministry of Communal Development and remained the technical contact point for both the Bank and the PIU teams during project implementation. In addition, NOSC supervised the project closely. The Second Vice President chaired the Committee, bringing relevant Cabinet Ministers together for field visits to review the work of PRADECS and address specific problems. For example, when local health clinics that were non- operational due to lack of trained personnel and medical equipment, the Minister of Health was requested to attend the NOSC meeting and participate in a field visit with the Second Vice President to address the problem. Officials from the Ministry of Environment would tour randomly selected projects at least twice a year with the PIU Safeguards Specialist to monitor environmental compliance of subprojects. PRADECS was a highly visible project that was closely supervised and managed by the Government. 32. Technically strong PIU. The PIU was staffed with competent and skilled personnel and largely avoided political appointments. While the PIU was technically strong, the project struggled with management in 2009 and 2010, and in 2011 the Ministry replaced the Executive Secretary. The change in PIU management led to a more open, technically sound, and productive collaboration between the project and IDA. Overall, the technical strength of the PIU, combined with Government ownership, was a key factor in successful project implementation. 33. MTR refocusing. The MTR was initially scheduled for early 2010, but it was advanced to October 2009 as the Task Team had identified several short-comings, particularly that the project had strayed from its original design. The MTR was a turning point in terms of bringing the project back on track. At the point of the MTR, disbursements were low (15.9%), and there were concerns over attaining some of the performance indicators. 34. In mid-2009 the Task Team had concluded that the project had lost its focus on supporting the decentralization process and social development. The project had become a fund primarily for larger municipal infrastructure investments (with pressure to deliver larger health and education infrastructure), to the detriment of capacity building to support decentralization, and smaller subproject at the community level. In addition, while the project had been successful in promoting social inclusion, the cohesion aspect through advocacy and social marketing remained almost neglected in program implementation. 35. As per MTR recommendations, an in-depth analysis of capacity building activities, including an assessment of FAs as well as a training needs assessment, was completed. This resulted in a comprehensive capacity building strategy which was implemented throughout the remainder of the project. Capacity building at the central level was also strengthened. 36. Social cohesion, particularly in terms of capacity building, seemed to have been interpreted as a form of “social inclusion�, indicating the underlying assumption that promoting the participation of 6 vulnerable groups in the local decision making process would be sufficient to address social cohesion. While PRADECS was very successful in ensuring representation and inclusion of vulnerable groups in preparation of Communal Development Plans, the MTR recommended strengthening of IEC activities for vulnerable groups to enhance social cohesion at all levels, as well as include subprojects to establish local mechanisms for provision of community services, such as psychosocial counseling and legal assistance. 37. The MTR also re-focused the selection of subprojects, emphasizing smaller community infrastructure projects, as well as diversification from health and education to other sectors such as transport and water and sanitation. A key issue identified during the MTR was the sustainability of health centers—only 2 out of 10 health centers were functional, largely due to lack of equipment and personnel. As per MTR recommendations, the Minister of Health became directly involved with PRADECS, and MoH prioritized provision of equipment and personnel for health centers built by PRADECS. In addition, the MTR recommended not supporting new health centers unless specifically approved by MoH at the central level, instead of at the provincial level as previously required. Finally, the MTR emphasized the need to respect the subproject selection criteria, to establish a committee for management of the proposed subproject, and put in place arrangements, including financial and human resources, for maintenance of the subproject. Recognizing the potential for political capture during upcoming elections in 2010 the team recommended a freeze on new subprojects during the campaign period. 38. Although significant issues were raised during the MTR, the Task Team and the GoB agreed not to restructure the project as the PDO could be met if the recommended changes were implemented. This was especially relevant for capacity building, where significant re-focusing and strengthening of the process was needed. 39. In early 2010 the project underwent a Quality Assessment of the Lending Portfolio (QALP) review which mirrored many of the MTR concerns. Several recommended actions, and identified solutions by the Task Team, were addressed during the remaining two years of the project, e.g., lowering community and communal contributions, ensuring sustainable maintenance of infrastructure, and ensuring that health clinics became operational. The QALP review rated Bank supervision as Moderately Satisfactory, and noted that the significant staff turnover and managers (four TTLs at MTR since project preparation) did not seem to have significantly affected the quality of supervision or client support. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 40. Design. The M&E system was based on an extensive results framework that captured the data to monitor progress toward achieving the PDO. The PAD states that the M&E system intended to: (i) inform stakeholders of the progress made by the project in achieving its objectives; (ii) provide information to evaluate the efficiency of implementation; and (iii) build M&E capacity within the line Ministry. To achieve this, a baseline study was to be completed before effectiveness to establish the value of the main indicators and to design a system of data collection at the national, provincial, and communal level. 41. Implementation and utilization. At project outset and as stipulated in the Financing Agreement, an M&E manual was prepared and approved in mid-2007 and updated in September 2008 for use at the decentralized project levels. The baseline study was completed at the beginning of the project. The data collection system was put in place and followed, including communes submitting monthly reports to PRADECS. One difficulty was the lack of electricity in several communes who relied on paper reports. At the decentralized level it was impossible to institute a database. The MTR Aide Memoire notes however that the quality of reporting from the communes improved following training and capacity reinforcement. One project indicator states that at least 60% of targeted communes provide robust M&E in an appropriate timeframe. This was 100% achieved as all communes began sending regular reports of increasingly better quality. The project M&E Specialist and the final report submitted by the Government 7 (Annex 7) note that the M&E system improved following a dedicated M&E mission in early 2010. Until the MTR, no regular quarterly reports were delivered to the Bank, but since then, quarterly reports were consistently submitted on time and with complete data on indicators. Progress toward achievement of the PDO is easily tracked through these quarterly reports and all ISRs. 42. ICR mission interviews with donors supporting decentralization confirmed that many of the studies and reports generated by PRADECS were of use for the broader decentralization thematic working group. German Development Cooperation (GIZ) noted that they did not know of another project that was as closely supervised and evaluated. Also, having these reports and studies shared among the working group was acknowledged and appreciated. The M&E reports cited as salient for both the project and other actors include a review of the capacity building Components as recommended by the MTR, a beneficiary survey, an impact evaluation on the ability of subprojects to augment service delivery, an infrastructure audit of all subprojects that ultimately led to a new draft law on the maintenance of infrastructure, and a national mapping exercise to identify social infrastructure supported by the project. 2.4 Safeguard and Fiduciary Compliance 43. Safeguards. The project was classified as Category B and triggered three safeguard policies: Environmental Assessment (OP 4.01), Indigenous Peoples (OP 4.10), and Involuntary Resettlement (OP 4.12). The project developed an Environmental and Social Management Framework (ESMF), a Plan concerning the Batwa population to address OP 4.10, and a Framework for Involuntary Resettlement and Compensation to address OP 4.12. All subprojects underwent an environmental assessment (EA) and completed an environmental and social evaluation form. The EA and environmental and social screening were designed to be done by provincial level authorities, but due to lack of capacity and training the project Safeguards Specialist completed all of the work with officials from the Ministry of Environment. The Ministry collaborated closely with PRADECS and benefitted from capacity building as community environmental committees were established to monitor subprojects and an expert consultant worked with the Ministry and PRADECS to develop an environmental screening system. Specific attention to the Batwa population was strengthened in the project in 2009 through the recruitment of a consultant to specifically address issues related to the Batwa population, resulting among others, in support for housing of vulnerable Batwa. Safeguards compliance is rated as satisfactory. 44. Financial management. The overall financial management aspects (budgeting, accounting, internal control, funds flow arrangement, reporting and external auditing) improved throughout project implementation. The implementing entity was compliant with the Bank’s financial management requirements. The project external auditor issued a clean opinion on the financial statement for the last years of implementation and the management letter did not raise any major issues. Unaudited Interim Financial Reports (IFRs) for the project were submitted on time, reviewed and found to be satisfactory; there are no overdue audit reports. 45. The project used the accounting software TOMPRO which helped to maintain proper books of accounts; and the financial management manual details key internal control procedures from transaction initiation, review, approval recording and reporting. The manual was updated to take into consideration changes during implementation. There was a clear separation of duties within the fiduciary unit, all fiduciary functions were well staffed with one financial and accounting officer, one chief accountant, one accountant and one procurement specialist; they were all trained on the use of Bank fiduciary procedures. 46. Although the flow of funds to communities was complicated, with support from the Fiduciary Agency of the project and regular support from the PIU to the communes when needed, the system performed well. The Final Independent Evaluation of the project noted that the incidence of corruption is 8 not present in the project. The consultant further noted that this was significant on a US$40 million project channeling funds to communal bank accounts for community subprojects. 47. With regards to funds flow, the overall disbursement rate as of December 31, 2012 was 99.08%. Some categories such as Consultant, Training and Operating Costs have been overdrawn due to the high cost of FAs which was underestimated. The total undisbursed amount of US$370,812.94 will be canceled. 48. Based on the above analysis, the overall financial management system for the entire project implementation period is rated as satisfactory. 49. Procurement. Procurement activities were conducted at implementation unit level (using Bank procurement and consultants guidelines) and commune/community levels with support from FAs (using community-based procurement guidelines). Despite the high procurement risk identified during appraisal, the adoption of mitigation measures resulted in a smooth implementation of procurement activities, for example the hiring of a procurement specialist at the PIU and intensive procurement training at commune/community levels. The procurement risk was moderate during the last three years of the project. 50. With regards to Bank support, four post procurement reviews were carried out in April 2010, May 2011, May 2012 and November 2012. Findings were reported in Aide Mémoires of supervision missions and pointed out some weaknesses in procurement planning, processing and record keeping, which the project gradually corrected. As a result, no major procurement shortcomings have been identified at the closing date and all contracts were already completed as communicated by the client during the last Bank supervision mission of November 2012. 51. Community-based activities require strong knowledge and familiarity with communities and local context. For this reason the project elected to renew the annual contracts of the FAs recruited after effectiveness. This could have posed a potential reputational risk in not opening up the process to competition. However, procurement training delivered by FAs to communes and communities was very well appreciated as reported in the independent final project evaluation. 52. Based on the above analysis, the overall procurement system and process for the entire project implementation period is rated as satisfactory. 2.5 Post-completion Operation/Next Phase 53. While in a letter dated June 2012, the Minister of Finance requested follow-on support for decentralization and community development from the Bank, the discussions held during the preparation of the new CAS led to the decision to not consider a follow-up project, given that other donors remain engaged in decentralization. However, the decision was made to focus on two areas of related support. First, Economic and Sector Work on fiscal decentralization is under way. Second, Pillar 2 of the CAS reinforces improved access to social services and a program that addresses social protection is being developed. The Bank is therefore not disengaging, but re-focusing through the current CAS. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Rating: Satisfactory 54. The PDOs, which were drawn from Burundi’s 2005 ISN and aligned with the subsequent CAS and the first PRSP, were and remain highly relevant to Burundi’s conditions. The project spanned a period of two CAS documents and both maintained Objective Two: Improve Access to Social Services and Consolidate Social Stability. Furthermore, the PDOs maintained their relevance during the PRSP-II 9 process concluded in January 2012, specifically a focus on improving access and quality in basic social services. The relevance of the PDOs is also deemed satisfactory due to their alignment to the on-going decentralization process that in the last year developed a new three-year strategy, adopted a national capacity building plan, and outlined a strategy to provide quality social services at the decentralized level. 55. The relevance of original project was substantial. The project addressed the need to provide funding to support economic and social recovery in Burundi and to rebuild social capital after more than a decade of war. The project design supported the newly established institutional framework around decentralization put in place by the new Government, identified a five year timeframe in order to be realistic on the time required to build local capacity in a fragile context, and identified activities to target the provision of social infrastructure and simultaneously build capacity and support to vulnerable groups. 56. The relevance of project implementation was substantial. As stated earlier, prior to the MTR in late 2009 there were concerns of achieving at least two PDO level indicators. Measures taken at this point in the project have already been covered, but the link here is to acknowledge that the successful implementation of the project, particularly in the final two years, led to the attainment of all indicators. 3.2 Achievement of Project Development Objectives Rating: Satisfactory 57. The Results Framework (RF) shows that the PDOs were achieved. Details are provided below. Component 1a: Capacity building for local development 58. This sub-component lagged behind others, receiving a rating of moderately satisfactory until the final year. The assumption at appraisal was that latent talent existed within local communities and among communal officials to be able to operationalize the new decentralization policy. Two problems emerged that were eventually overcome. First, mid-way through the project over 85% of communal level officials changed. During the 2010 elections a major shake-up occurred among locally elected officials, essentially removing thousands of officials and communal staff who had been trained by the project and who were managing on-going subprojects. Second, the quality of the capacity building provided by the FAs varied significantly, each using different training manuals. 59. The 2010 elections, which coincided with the MTR period, re-focused this component to analyze the training curriculum and standardize all modules across all FAs and re-assess areas of low and high intensity to better address shifting donor interventions. For example, the Province of Bujumbura Rural was originally a province of low intensity, but after the 2010 elections most of the newly elected commune officials in the province had never worked in local government. The donor originally covering this province had ceased operations. This is only one example of how the component re-focused to pragmatically analyze coverage, but also the content of support provided. By the end of the project, 10 different training modules were produced in French and Kirundi and absorbed by the Ministry of Communal Development in their development of a National Capacity Building Plan. 60. This Component successfully supported the Government in the process of developing Communal Development Plans for all communes across the country. These plans then acted as the local framework for the selection of subprojects. The related project indicator states that at least 90% of targeted communes would develop or update their Communal Development Plans on a participatory basis. Early in the project 100% of targeted communes had completed their plans, and at project close a second generation of Communal Development Plans was being initiated. Furthermore, training was conducted on financial management and budgeting, procurement, project design, and monitoring. During the ICR mission several communal Administrators attested that they were able to obtain funding for other communal projects through their training and experience with PRADECS. Both GIZ and Swiss 10 Cooperation confirmed to the mission that in developing greater fiscal decentralization with the communes it is apparent that the management of the subproject funds, combined with technical support, has been the only concrete experience of communal officials in managing projects with regards to procurement procedures and managing funds. One of the related PDO level indicators is that at least 15% of targeted communes have obtained funding for Communal Development Plan activities from outside of the project. Before the end of the project, all communes (100% attainment) had obtained funding outside of the project to finance the Communal Development Plans. 61. Institutional support was provided to the Ministry of Communal Development and open communication seems to have prevailed between the PIU, the Ministry, and NOSC. Although much of the success in implementation from a technical perspective is due to the strength of the PIU team, the Ministry navigated the broad decentralization process and was pragmatic in managing the project. Ministry staff was exposed to a training program, with one project indicator being that the training plan be finalized and implemented satisfactorily. This was 100% achieved. 62. The early ratings in the ISR of moderately satisfactory are directly tied to the view that the outputs being produced (training of officials, but not generating increased capacity, or existence of modules, but no standardization or quality assurance of training materials) did not create a causal link to the intended outcome. The Government and the PIU team took adopted a more strategic approach and moved away from a series of scattered trainings toward technical support and coordinated to support the ongoing decentralization process. Component 1b: Capacity building for social inclusion and cohesion 63. This component was designed to strengthen social inclusion and cohesion by targeting vulnerable groups. This sub-component created a link between national policies addressing vulnerable groups and legal support to these groups after the post-MTR corrective measures. The project also created a link between Components 1b and 2b, meaning capacity building around social cohesion was augmented by financing subprojects that supported vulnerable populations. Examples include national workshops on the legal rights of vulnerable groups and financing legal costs for 1,250 orphans to be registered in the civil registry. A national association for albinos was supported and later homes for the most vulnerable albinos were constructed. Specific attention was paid in the preparation of Communal Development Plans supported by PRADECS to ensure that vulnerable groups were included in the consultations and their concerns addressed. Specific set-asides were made in subproject allocations for activities to support these groups, surpassing the target of 50%. PRADECS adopted this approach with the Ministry of Communal Development, and through the thematic working group on decentralization all development partners financing the Communal Development Planning process emphasize inclusion of vulnerable groups. Component 2a: Financing public subprojects 64. This sub-component aimed to provide financial support to communes and communities to implement demand-driven subprojects. The target was to complete at least 430 community and commune level subprojects—the project was able to complete 818 subprojects. Table 1 – Breakdown of subprojects by sector Sector Signed Sector (%) Functional Functional (%) Education 333 41 327 99 Health 74 9 55 74 Water 56 7 56 100 Roads 69 8 69 100 Other 286 35 286 100 Total 818 100 793 97 11 65. The more salient analysis is the review of increased access to quality social services – with the emphasis on quality, meaning functional infrastructure. The related project development indicator aimed to achieve 70% of targeted communes showing a significant increase in access to and use of primary health, primary education, and water services. Achievement was 97%. The indicator was broken down by sector. The project struggled early in health, but the Bank worked closely with the PIU and MoH to ensure health centers were operational. The project reached the 70% target, but with 18 of the 74 health centers constructed not operational, the Government still has work to do. 66. All subprojects financed under the project were included in the Communal Development Plans. With the assistance of FAs and the PIU team, communal Administrations undertook public consultations to identify priority projects for financing. A significant achievement under this Component was that following technical support from a Bank disability expert, since 2010 all building designs were modified to ensure accessibility for disabled people. The Government is proud of this accomplishment. Component 2b: Financing subprojects to strengthen social inclusion and cohesion 67. This sub-component targeted the same beneficiaries as component 1b, aiming to provide financing for subprojects targeting the special needs of vulnerable groups. The target for an intermediate results indicator at the outset of the project was to complete at least 70 subprojects targeting vulnerable groups. The project was able to complete 289 subprojects. 68. This component worked in synergy with component 1b, and in fact the staff member on the PIU team responsible for both components remained on the project from start to finish. There was an evident evolution over the years toward support to local associations advocating for vulnerable groups, and the financing of subprojects to leverage the advocacy work. Examples include the funding for national Batwa and albino organizations and the construction of houses for the most vulnerable Batwa and albinos. Small enterprise projects were also developed for these groups, e.g., a local flour mill run by albinos. 69. The financing of subprojects made more visible issues affecting vulnerable groups and provided these groups leverage within the local community. The Bank team often recommended caution in terms of potential resentment that could emerge around the selection of beneficiaries, but the project was able to address this concern by partnering with reputable national associations and working with the FAs that had developed strong relationships with local communities. 3.3 Efficiency 70. At appraisal, a cost effectiveness and financial analysis was not carried out as the project did not lend itself to such analysis for three reasons: (i) all investments would be demand-driven, and their nature cannot be known beforehand; (ii) all eligible subprojects would be non-productive; and (iii) the economic benefits of capacity-building are difficult to quantify. The project aimed to ensure cost efficiency through robust financial control by the PIU and a Fiduciary Agency that verified the communal accounts financing of all subprojects. Also, the adoption of competitive procurement procedures at the local level helped communes to adopt the least cost approach in the selection of contractors, and it was assumed community and beneficiary involvement would be a powerful accountability mechanism to ensure transparency and cost efficiency. 71. A comparative analysis of unit costs of subprojects implemented by PRADECS suggests the project was cost effective in the Burundi context (details in Annex 3). Primary schools, rural roads and water subprojects were compared to projects funded by other partners or Bank projects:  The average cost of building a primary school under PRADECS was US$59,610, compared to US$88,254 under another Bank project working with ABUTIP, Agence Burundaise pour les Travaux 12 d’Interet Publique, (Burundian Agency for Public Works). For PRADECS the cost per primary school classroom was US$8,516 and for ABUTIP US$14,709.  Both PRADECS and the International Fund for Agricultural Development (IFAD) rehabilitated rural roads. The cost under PRADECS for the rehabilitation of one kilometer was US$16,461 while IFAD’s average cost was US$23,458.  PRADECS and the Ministry of Water, Energy, and Mines, rehabilitated water sources to bring potable water to population centers in rural areas. The cost of rehabilitating one kilometer of water under PRADECS was US$6,334 while the average for the Ministry was US$9,640. 72. All cost comparisons for sub-component 2a point toward PRADECS being more efficient than other partners, for similar projects during the same time period. An infrastructure audit was done on PRADECS in 2011 and overall the quality of construction was confirmed as good. 73. Three reasons explain greater cost efficiency by PRADECS.  PRADECS targeted smaller local firms. Bidding documents encouraged local firms to apply and achieved this by putting a lower threshold for the business turn-over. This attracted small companies to apply for the construction of subprojects. These smaller companies had lower over heads and administrative costs and were able to provide the same quality for less cost. ABUTIP and the GoB seem to have worked with the larger, national construction firms for similar subprojects, resulting in higher costs.  PRADECS elected to hire one consultant per province as an Engineer/Supervisor to supervise all subprojects in each commune within the province, lowering the cost of supervision as other projects opted to recruit consultancy firms for each subproject at much higher cost.  Project design was such that communal Administrations were directly involved in the planning, procurement, supervision, and maintenance of subprojects. Local community committees were also established for supervision and maintenance. This accountability mechanism created greater local ownership over the subprojects and encouraged on-time delivery, the use of quality materials, and the rapid correction of faults. 74. A final comparator is that of the cost of developing the Communal Development Plans. Significant resources were allocated to capacity building. The average cost per commune of developing a Communal Development Plan under PRADECS was US$17,352. GIZ provided data showing that their average cost per commune was approximately US$20,000. 3.4 Justification of Overall Outcome Rating Rating: Satisfactory 75. The overall outcome rating of the project is satisfactory. This rating is based on the analysis of the on-going relevance of the project, both in terms of the relevance of the PDOs and its design, as well as the achievement of all PDOs. Although efficiency is more difficult to assess under this type of operation, no deficiencies were identified. A qualitative and quantitative assessment of outputs by component (Annex 2) confirms the completion of all anticipated outputs within the original timeframe and budget of the project, a significant achievement for a CDD project in a post-conflict environment. 76. The institutional setting in Burundi has matured around decentralization, with the Communal Law and the National Policy firmly rooted. The Government’s PRSP-II maintains a focus on community development and access to social services. Likewise the project maintained direct relevance to the ISN at appraisal, the CAS of FY08-12 (Pillar 2), and most recent CAS for FY13-16 (Pillars 2 and 3). Feedback from Government officials and donor counterparts during the ICR mission confirms the strong relevance of PRADECS today. PRADECS was the largest project in the sector and the intervention directly 13 supported the Ministry of Communal Development and local communes in 16 of Burundi’s 17 provinces. A widely held view emerged from interviews on the relevance of the project in supporting the decentralization agenda in Burundi, and collaboration with other development partners to support decentralization. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Gender Aspects, Poverty Impacts and Social Development 77. Gender aspects. The project was effective in addressing gender throughout all components and the outcome was strong participation by women. Burundi’s constitution sets minimum gender thresholds for women in elected office. This rationale translated into the project, to the same level of 30% participation of women targeted in several activities. Specifically, in the development of the Communal Development Plans women were particularly targeted and the needs of women in development planning were addressed by the FAs during their local facilitation between members of the community and local authorities. It is estimated that women participated upwards of 40% during the community meetings for communal development planning. The structure of Communal Development Committees (CDCs) established by the project in all target communes comprised a minimum of 30% female members. This allowed women at the local level to engage in decision-making. Activities that targeted vulnerable groups provided support to local associations for advocacy and income generating projects. The project targeted vulnerable women and more than 40% of the members of the associations were women. A specific target group was identified as female victims of violence and assistance was provided in the form of legal aid, psychosocial assistance, and support for vocational training. The number of direct project beneficiaries is recorded as 3,485,042, with more than half (1,777,371) women. 78. Poverty impacts. Burundi is one of the poorest countries in the world. Prior to the conflict, the percentage of people living below the national poverty line was 36.4% (1990). During the conflict, this had increased to 68% in 1998. Prior to the start of the project in 2006 the rate was still high at 67%. Anecdotal evidence from interviews during the ICR mission notes a profound impact on the lives of rural Burundians. Project data confirms this with a detailed break-down included in quarterly reports. With a population of approximately 9 million, the direct project beneficiaries are over 3.4 million people. The Final Independent Evaluation of the project estimates indirect beneficiaries around 6 million people. For example, the project calculated that each rehabilitated water source (total of 740) impacts 94 households, with an average of five members per household. This translates into 347,800 people in rural areas have better access to safe water. A study on the impact of subprojects on communities found that the average distance to access health centers and primary schools was reduced from 5-10 km to less than 5 km. (b) Institutional Change/Strengthening 79. Through PRADECS, the communal administrations were able, for the first time, to take direct responsibility over funds for subprojects. This included the development of the first Communal Development Plans, the identification of priority subprojects through participatory community meetings, managing the procurement process, monitoring subprojects, and finally guaranteeing the maintenance of newly built communal infrastructure. This comment was provided by donors (GIZ and Swiss Cooperation), by communal administrators and technical staff, and cited in the Final Evaluation Report of the project. In fact, GIZ reiterated that the only practical experience that communal authorities have in managing procurement is through PRADECS. It seems that this practical approach to programming strengthened the capacity of local officials to manage Communal Development Plans. (c) Other Unintended Outcomes and Impacts (positive or negative) Not applicable. 14 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 80. Stakeholder workshop. A formal beneficiary survey was not carried out, nor was a stakeholder workshop held for this ICR. However, a member of the ICR team participated in a stakeholder workshop organized by the project in December 2012 just a few weeks before the close of the project. The workshop brought together over 200 communal Authorities from all communes targeted by the project. Four key messages from project stakeholders emerged from the workshop: (i) the 1,067 subprojects realized through the project had a direct impact on improving access to social services in rural areas where 90% of the population live, (ii) the needs remain great and stakeholders argued for continued support through PRADECS to accompany the communes in the on-going decentralization process, (iii) the impact of social services, particularly schools and health clinics was of urgent priority at the start of the project, but the diversification of projects for more water, transport, and local economic initiatives had tremendous gains for the local population, and (iv) communal officials expressed that their capacity to manage local public administration was augmented due to the training provided by the project. 81. Participants of the stakeholder workshop, as well as interviews conducted in four communes during the ICR mission (two low-intensity and two high-intensity communes) elaborated on point 3 above – the impact of subprojects on local economic development. For example, the communal Administrator of Makebuko, in the Province of Gitega, explained how a subproject to provide electricity to the commune office and surrounding area sparked economic growth. Where two years earlier there had been little activity, now buildings were going up and small businesses were emerging. A hair salon, a flour mill, and a small motel had recently been established down the road from the communal administration headquarters. The electricity connection also provided lighting that has augmented security for the local community and allowed for economic activity to continue into the evening hours. A communal Administrator from Bujumbura Rural spoke of the impact of a new rural market area. Before, vendors displayed products on small wooden tables and mats on the ground. The new market, with a cement floor and roof sheeting to protect against the elements, has increased the number of vendors on market days. Communal officials explained the necessity of continuing to improve social infrastructure where needed, while simultaneously financing subprojects that target local economic development, particularly in areas that have been historically more isolated. Communal Administrators indicated that these subprojects financed by PRADECS had the double effect of improving the livelihoods of citizens, while bringing in revenues to the commune through local taxation. The line Ministry for PRADECS echoed these same views and during the last year has created a second Directorate General for Local Economic Development. 4. Assessment of Risk to Development Outcome Rating: Substantial 82. Risks identified during project preparation were appropriate, as they reflected the post-conflict environment of Burundi and the Government’s own limited capacity at that time, as well as the significant fiduciary risks involved in adopting a decentralized financing mechanism. Four risks were rated high, two as significant, and three as moderate. The overall risk rating was high. The identified risks played out over the life of the project with much focus during implementation spent on fiduciary accountability. 83. The risk to the project’s development outcome is rated as substantial for three reasons. First, a lasting impact of the project is the existence of newly built social infrastructure at the commune level that provides access to social services for citizens. The likely sustainability of this infrastructure is enhanced because communal administrations have taken over their management. A significant level of Government and community ownership was developed through the project’s implementation modalities. The project was successful in supporting the development of a new draft law for the maintenance of communal infrastructure. However, limited communal budgets will remain a challenge. In August 2007, a national decree was passed which foresaw that approximately 15% of the national budget transferred to the 15 communes directly. Since the approval of this project, no actual transfer mechanism is functional. The reforms cited are substantial, but without financial means to support the new laws it will be difficult for the Government at the local level to ensure the sustainability of social infrastructure. 84. Second, although capacity at the commune level needs to be further strengthened, Government counterparts confirmed the technical skills learned and adopted (specifically on procurement and budgeting) through the training delivered under the capacity building component was absorbed by communal administration officials. Other donors also noted that the communes who were targeted as high intensity communes are ahead of counterparts in low intensity communes in terms of budget and project management. On the other hand, upcoming elections in 2015 could result once again in a high turn-over of currently trained local government representatives. The capacity of local public administration officials remains at risk due to an overly politicized culture where bureaucratic stability is not yet the norm. 85. Finally, the risk to the project’s development outcome is mitigated by a stronger institutional structure that is in place for decentralization in Burundi. The political risk referenced earlier in the text remains in place, but has been mitigated somewhat by developing clearer roles and responsibilities for Ministries involved in the decentralization process. However, the political-institutional tension within the context remains and the Ministry of Communal Development could benefit from a political champion from senior Government advocating for the decentralization process. 5. Assessment of Bank and Grantee Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 86. The Bank team had the appropriate background and technical expertise to prepare and appraise the project. The preparation phase was sufficient in length, and included numerous missions. The necessary studies and analyses were conducted to develop a solid framework for the operation; the development objective and indicators were properly identified; and the appropriate components were devised to achieve the project’s goal. The operation was properly aligned with the country’s ISN and PRSP, and responded to the Government’s expressed development goals. The Bank team put in place mitigation measures that addressed the project overall high risk rating, focusing on the recruitment of a strong technical team for the PIU, the establishment of a sound financial management system, including the recruitment of external auditors, and the establishment of a baseline for the M&E system. No standardized training curriculum was put in place at the outset. QALP rated Bank performance in ensuring quality at entry as moderately unsatisfactorily, and found that project design was overly complex for Burundi’s conditions. However, during project implementation, these challenges were overcome. Thus, the Bank’s performance in ensuring quality at entry is rated Moderately Satisfactory. (b) Quality of Supervision Rating: Satisfactory 87. Overall the Bank carried out 12 full supervision missions, plus an MTR mission, and more specific missions such as an M&E mission. In each case the missions were comprised of team specialists and complemented by the FM and Procurement Specialists. In its final report (Annex 7) the Government made special mention of the frequency and quality of supervision support to the project. However, there were four project TTLs which the Government admitted was complicated during the first half of the project. From the MTR phase onwards one TTL was in place for the remainder of the project. Additionally, the project benefited throughout implementation from field-based technical specialists based in the Burundi Country Office which assisted in fostering regular contact with the PIU and line 16 Ministry. It also allowed the Bank to more effectively engage in the thematic working group on decentralization. TTL stability and field presence enhanced the Bank’s capacity and quality of supervision. 88. Regular reporting of Aide Memoires and ISRs maintained a clear view of project implementation, particularly the progression of PDO indicators. ISR ratings and write-ups also indicate the evolution of the capacity building component, which required a concerted effort from the Bank to re-focus. The Bank team also undertook regular field visits as part of formal missions and as specific issues emerged. Many of the communes were visited in order to assess the capacity building component and to inspect social infrastructure. Bank staff often participated in project workshops, including a national workshop for all communal Administrators in the country, and a workshop on environmental and social safeguards. 89. QALP rated Bank supervision as moderately satisfactory, and noted that the significant staff turnover and managers did not seem to have affected the quality of supervision or client support. A key issue of concern for QALP was the need for the CMU and SMU to determine whether fundamental changes and or restructuring would be needed. The Task Team, SMU and CMU decided after the MTR that no Level 1 restructuring was needed to address implementation issues. The corrective measures put in place after the MTR, as well as before and after the 2010 election period, and close supervision of the project until closing, resulted in the project achieving all its development objectives. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory 90. Although some deficiencies were identified in quality at entry, these were corrected during project implementation and Bank supervision was strong and responsive. On balance, Bank performance is rated as satisfactory. 5.2 Grantee Performance (a) Government Performance Rating: Satisfactory 91. The Government was actively involved in the preparation of the project, and maintained ownership, particularly at the technical level. The Ministry of Communal Development (and the Director General who followed changing Ministries and Vice-Ministries on decentralization throughout the years) actively engaged in project management and closely monitored the work of the PIU. This involvement was stronger during the last two years of project implementation. Likewise, NOSC maintained an oversight function over the project. When problems arose (e.g., switching of the project coordinator, a review of communal contributions on subprojects) open communication between the Bank and the Government proved effective for taking corrective action. The Government moved quickly and transparently in most cases. Public reports were disseminated following NOSC meetings and information regarding PRADECS was widely available. Although strong Government performance on PRADECS was demonstrated, as of today there is no operational mechanism for financial transfers to the communes – this reality weakens the overall ability of local administrations. (b) Implementing Agency or Agencies Performance Rating: Satisfactory 92. The PIU performed at a high technical level. FM and Procurement reviews consistently confirmed the performance of the team according to fiduciary guidelines. Team coordination was strong, particularly when a staffing change was made around mid-way through the project. The reports prepared and disseminated by the PIU were of a high standard. The PIU team participated in the thematic working group on decentralization and often led specific committees. In short, the team managing PRADECS was an enormous resource to the Government and proved highly effective at implementing a complex project. 17 The core team remained with the project from start to finish, ensuring a high level of institutional memory and the development of working relationships with communal officials and local communities. 93. In terms of implementing partners, the first Fiduciary Agency did not perform and the contract was not renewed. However, this was corrected and a new Fiduciary Agency completed the work to the end of the project under stronger leadership and more clear Terms of Reference. The FAs performed well overall. The critique mentioned earlier in terms of their ability to design and develop training modules was addressed by the recruitment of expert consultants to revise all training modules. (c) Justification of Rating for Overall Grantee Performance Rating: Satisfactory 94. On the basis of the performance ratings of both the Government and the main implementing agency (PIU), and the development objectives achieved under the project, the overall rating for Grantee performance is satisfactory. 6. Lessons Learned 95. A SIL operation on its own is not a sufficient vehicle to promote broad policy and institutional reforms needed for effective decentralization. PRADECS contributed significantly to supporting the young decentralization process in Burundi. However, to generate the required policy momentum in a newly launched decentralization process, it would have been more effective to combine a SIL operation with other mechanisms to promote policy changes, such as DPLs. The decentralization process is still challenged by lack of coherence on fiscal decentralization, including the mechanism for fiscal transfers and local tax collection, and policy stagnation on devising clear roles and responsibilities for decentralized services such as health and education. The limited transfer of funds from the national to the local level, although required by law, severely constrains local government to carry out their new responsibilities. A CDD project, combined with a mechanism to support policy and institutional reforms could have better supported the development of a fully effective decentralization process. 96. In a low capacity environment, CDD projects should use a gradual, transitional approach. CDD approaches have proven to be effective in low capacity, post-conflict environments. They transfer financial resources directly to communities to address their most important needs, rebuild social infrastructure, make government presence visible at the local level and contribute to peace building. However, sufficient time needs to be given to develop effective mechanisms at the community level and build the capacity of local authorities. Only in a second phase should responsibilities be transferred, including developing and implementing construction for social infrastructures. Otherwise, capacity building will be seen only as a pre-requisite to obtain funding for infrastructure subprojects, instead of as a key objective on its own to support sustainable local development. 97. A comprehensive assessment of actual capacity among local officials in a fragile context is essential to avoid over-optimism and an unrealistic design at entry. The success of a CDD operation partially hangs on the capacity and ability of local officials to manage resources and provide technical inputs. Over estimating the ability of local officials does more harm (do no harm principle) and will ultimately waste time and resources at the start of a project while modifications are made to address capacity constraints. 98. CDD projects need to be flexible during periods of elections in post-conflict settings. CDD projects are especially vulnerable to political interference and corruption during election periods, even more so in post-conflict countries. Bank and government teams need to take pro-active actions to avoid these potentially negative impacts on project performance, such as putting on hold selection of 18 infrastructure projects at the community level during a significant period before elections, and foreseeing additional capacity building for newly elected officials at all levels following elections. 99. Bank presence in-country, particularly in a post-conflict environment, is important. This lesson has been repeated in several recent Burundi ICRs, and no doubt from other countries. Complex operations, such as a large-scale CDD project, require closer guidance and mentoring by Bank staff than more straightforward projects or when working in a more stable context. In-country staff allows real-time technical support. 7. Comments on Issues Raised by Grantee/Implementing Agencies/Partners (a) Grantee/implementing agencies 100. The Government confirmed by letter their satisfaction with the ICR report and had no comments to provide. Additionally,the Government provided the ICR team its completion report which is included in Annex 7. (b) Cofinanciers Not applicable. (c) Other partners and stakeholders 101. The ICR mission consulted partners in the thematic working group on decentralization. GIZ and Swiss Cooperation, who have taken turns chairing the group, are familiar with PRADECS from its role supporting the Ministry of Communal Development in the working group, as well as its role financing subprojects in communes where GIZ and Swiss Cooperation manage capacity building activities. 102. GIZ specifically focused on the support role that PRADECS provided the Ministry of Communal Development, stating that members of the PIU often led working groups on specific topics – for example, developing the national capacity building plan, or devising a draft law on the maintenance of communal infrastructure. GIZ stated that they saw the activities of PRADECS as tightly aligned with the policies and work of the Ministry and the welcomed the project’s role as technical advisor. 103. The Head of Swiss Cooperation was very direct in stating that PRADECS was an opportunity for the other donors working in the sector, and that the resources and approach used in the project was a “force de frappe�. Additionally, it was acknowledge that PRADECS followed principles of the Paris Declaration, in terms of Government ownership, in a genuine way by having the NOSC operationalized and involved in practical project supervision, as well as being established in a manner that provided guidance to the project. 19 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Actual/Latest Appraisal Estimate Percentage of Components Estimate (USD (USD millions) Appraisal millions) Capacity building 6.00 5.50 92.0 Financing subprojects for local 30.00 30.00 100.0 development Project management, M&E, and 3.40 3.80 112.0 communication Total Baseline Cost 39.40 39.30 99.8 Physical Contingencies 0.14 0.00 0 .0 Price Contingencies 0.16 0.00 0.0 Total Project Costs 39.70 39.30 Front-end fee PPF 0.30 0.30 100.0 Front-end fee IBRD 0.00 0.00 0.00 Total Financing Required 40.00 39.60 99.0 (b) Financing Appraisal Actual/Latest Type of Percentage of Source of Funds Estimate Estimate Cofinancing Appraisal (USD millions) (USD millions) Grantee 0.00 0.00 0.0 IDA Grant 40.00 39.60 99.0 20 Annex 2. Outputs by Component The project has three main components, each with sub-components. Sub-component 1a: Capacity building for local development. This sub-component aimed to assist communities and communes undertake participatory planning, as well as initiating, preparing, implementing, managing, and monitoring subprojects. Through the recruitment of Facilitating Agencies (FAs), field-based facilitation teams strengthened local technical and fiduciary skills to implement local development activities. Institutional support was also provided at the national level to the corresponding line Ministry responsible for decentralization. The major outputs of this component include:  The development of Communal Development Plans in 34 communes in provinces of high- intensity intervention within the first year of the project.  A further 37 Communal Development Plans were prepared in 2012 as part of the second generation of the five year Communal Development Plans.  10 training modules on capacity building were designed and disseminated to the Ministry of Communal Development to feed into the National Capacity Building Plan.  The drafting of a new law on the maintenance of infrastructure has been completed. PRADECS identified a consultant to work with the Ministry of Communal Development to prepare this new law. The draft law will now be considered for legislation. Sub-component 1b: Capacity building for social inclusion and cohesion. Through targeting vulnerable population groups, this sub-component was designed to strengthen social inclusion and cohesion. The focus was on information, education, and communication (IEC) activities with respect to vulnerable groups for the benefit of: (i) local government actors, (ii) organizations that represent or support vulnerable groups, and (iii) associations of vulnerable groups. The major outputs of this component include:  The project has provided legal assistance to 106 individuals, most of whom are orphaned children, as well as securing the registration of 1250 orphans and other vulnerable people in the civil registry. An additional 600 people have been trained at the commune level on the legal rights of vulnerable people, bringing the total of those trained to 1370.  The project has supported 136 associations with either technical assistance and/or by providing materials to support productive activities. Examples include assistance to draft legal texts of the association, assistance to register the association with the authorities, and provision of materials to leverage productive activities such as mills, soap-making ateliers, and bakeries.  The project has constructed homes for vulnerable people, including the Batwa and the most vulnerable members of the albino community. Sub-component 2a: Financing public subprojects. This component aimed to provide financial support to communes and communities to implement demand-driven public subprojects. The following sectors were identified: health, education, transportation, water and sanitation, culture and sports, and collective economic infrastructure (markets, slaughterhouses, bus stations). Financial support was provided through a grant and complemented by a contribution from the beneficiaries. The selection of subprojects was carried out under a participatory mechanism, and approved by the respective Communal Councils. The performance indicator is: At least 430 Communal level subprojects have been completed. The Table below provides the detail of subprojects completed by the project. The total number of subprojects signed is 818, and 812 were completed. 21 Table 2 – Breakdown of subprojects sub-component 2a NUMBER OF NUMBER OF NUMBER OF SUBPROJECTS SUBPROJECTS SUBPROJECTS TYPES OF SUBPROJECT COMPLETED SIGNED BY NOT BY CATEGORY COMPLETED CATEGORY Communal Colleges 86 82 4 Primary Schools 106 104 2 Pre-school 1 1 0 Technical Secondary School 1 1 0 Health Centers 35 35 0 Living Quarter for Health Center 1 1 0 Rural Roads 14 14 0 Communes of High-intensity Intervention Bridges 29 29 0 Water connections for Drinking Water 9 9 0 Water Source Rehabilitation 12 12 0 Non-medical Equipment for Health Centers 28 28 0 Equipment for Primary Schools 50 50 0 Equipment for Communal Colleges 43 43 0 Equipment for Secondary Schools 28 28 0 Equipment Vocational Training Center 2 2 0 Vocational Training Centers 2 2 0 Latrines 14 14 0 Warehouses 2 2 0 Markets 1 0 0 SCEP 2 0 0 Slaughterhouse 1 0 0 Water Reservoir 1 0 0 Sport Facility - Stadium 2 0 0 Sports Facility - Field 10 10 0 SUB-TOTAL 480 474 6 Communal Colleges 53 53 0 Technical Secondary School 2 2 0 Communes of Low-intensity Primary Schools 83 83 0 Pre-school 1 1 0 Intervention Health Centers 39 39 0 Rural Roads 8 8 0 Bridges 18 18 0 Water connections for Drinking Water 10 10 0 Water Source Rehabilitation 7 7 0 Living Quarters for Teachers 3 3 0 Rural Electrification 9 9 0 22 Sport Facility - Stadium 3 3 0 Markets 1 1 0 Warehouses 2 2 0 Wall Construction 1 1 0 Sports Facility - Field 9 9 0 Equipment EQPEP 29 29 0 Equipment Communal Colleges 19 19 0 Equipment Primary School 9 9 0 Non-medical Equipment Health Centers 17 17 0 Equipment Vocational Training Center 2 2 0 Slaughterhouse 1 1 0 Latrines 12 12 0 SUB-TOTAL 338 339 0 TOTAL 818 812 6 Sub-component 2b: Financing subprojects to strengthen social inclusion and cohesion. This component targeted the same beneficiaries as in component 1b. The aim was to provide financing for subprojects targeting the special needs of vulnerable groups but unlike Component 2a, no contribution was required by the beneficiaries. Identified potential subprojects included schooling, housing, professional and vocational training, and alphabetization programs. It also provided financing for the provision of community services for counseling and assistance, particularly legal assistance for vulnerable groups. The performance indicator is: At least 70 subprojects targeting Vulnerable Groups have been completed. The Table below provides the detail of subprojects completed by the project. The total number of subprojects signed and completed is 249. Table 3 – Breakdown subprojects sub-component 2b TYPES OF SUBPROJECT NUMBER OF SUBPROJECTS SIGNED BY CATEGORY Construction of socio-cultural centers 5 Communes of High-intensity Intervention Lodging for vulnerable groups 58 Support for school fees for vulnerable groups 18 Support to persons living with HIV/AIDS 1 Youth vocational skills training 14 Construction of sports fields 8 Support to sports teams 10 Support to IDPs and returnees 2 Support to vulnerable children in schools 3 Education support to Batwa children 11 Legal assistance to vulnerable groups 4 SUB-TOTAL 134 23 Construction of socio-cultural centers 5 Lodging for vulnerable groups 50 Communes of Low-intensity Education support for vulnerable students 26 Youth vocational skills training 15 Intervention Support to sports teams 4 Construction of sports fields 1 Education support to Batwa children 5 Computer equipment 1 Support to street children 1 Legal assistance to vulnerable groups 7 SUB-TOTAL 115 TOTAL 249 24 Annex 3. Economic and Financial Analysis A comparative analysis of unit costs of subprojects implemented by PRADECS and that of other projects underscores some cost efficiency. Table 4 below outlines in detail a cost comparison for the unit costs of specific subprojects under PRADECS compared to those of other donors. 1 Table 4 – Cost comparison between PRADECS and other donors on subprojects Type of COST COST TOTAL # OF COST PER Infrastr PROVINCE COMMUNE PROJECT SUBPROJECT SUPERVI WORKS COST CLASSROOMS CLASSROOM ucture SION Primary Schools Extension Ndava GITEGA BURAZA PRADECS 7 7,700 Primary School 51,992 1,911 53,903 Extension RUYIGI NYABITSINDA PRADECS Murama Primary 7 9,331 63,834 1,484 65,317 School AVERAGE COST PRADECS PER PRIMARY SCHOOL 59,610 AVERAGE COST PRADECS PER CLASSEROOM 8,516 Kabuye Primary KAYANZA KAYANZA ABUTIP 6 13,769 School 78,114 4,499 82,612 Ceyerezi Primary KAYANZA MURUTA ABUTIP 6 13,321 School 75,430 4,499 79,928 Mihigo Primary KAYANZA KAYANZA ABUTIP 6 16,767 School 96,103 4,499 100,601 Ryamukona KAYANZA KABARORE ABUTIP 6 15,187 Primary School 86,621 4,499 91,119 Rugazi Primary KAYANZA KABARORE ABUTIP 6 14,501 School 82,509 4,499 87,008 AVERAGE COST ABUTIP PER PRIMARY SCHOOL 88,254 AVERAGE COST ABUTIP PER CLASSEROOM 14,709 Health Centers Nyabikenke KIRUNDO VUMBI PRADECS Health Center 55,277 3,728 59,005 Buheka Health MAKAMBA NYANZA-LAC PRADECS Center 59,200 2,576 61,776 AVERAGE COST PRADECS PER HEALTH CENTER 60,391 AVERAGE COST GOVERNMENT PER HEALTH CENTER 112,468 AVERAGE COST ABUTIP PER HEALTH CENTER 160,668 Rural Roads Rehabilitation Rural Road RN CIBITOKE MUGINA PRADECS 10-Mugina ( 1 15,035 695 15,729 km) Rehabilitation Rural Road KARUSI GIHOGAZI PRADECS Ruvubu- 16,027 1,167 17,193 Ruganira ( 1 km ) AVERAGE COST PRADECS PER KM REHABILITATED 16,461 AVERAGE COST IFAD PER KM REHABILITATED 23,458 Water Rehabilitation 1 All costs in Annex 3 are only for subproject or activity costs, and do not include costs of the PIU. Comparative costs for other donors also exclude project management costs and are therefore comparable. 25 Water rehabilitation BUBANZA MPANDA PRADECS Butembe- 57,187 1,621 58,808 Kanenga (9 km) Water rehabilitation CIBITOKE MUGINA PRADECS Rushima- 22,249 1,285 23,535 Rubirizi ( 4km) AVERAGE COST PRADECS PER KM REHABILITATED 6,334 AVERAGE COST GOVERNMENT PER KM REHABILITATED 9,640 The table elaborates the costs of primary schools, health centers, rural roads rehabilitated and water sources rehabilitated. For all four types of subprojects, PRADECS was more efficient than the other donors.  The average cost of building a primary school under PRADECS was about US$59,610, compared to US$ 88,254 under another World Bank project working with an implementing body called ABUTIP, Agence Burundaise pour les Travaux d’Interet Publique, (Burundian Agency for Public Works). The breakdown of cost per classroom is also available. For PRADECS the cost per classroom for a primary school is US$8,516 and for ABUTIP US$14,709. Compared to ABUTIP, PRADECS is approximately 42% less costly.  Both PRADECS and the International Fund for Agricultural Development (IFAD) rehabilitated rural roads. The cost under PRADECS for the rehabilitation of one kilometer was US$16,461. IFAD rehabilitated one kilometer of rural road for an average cost of US$23,458. Compared to IFAD, PRADECS is approximately 30% less costly.  PRADECS and the GoB, through the Ministry of Water, Energy, and Mines, rehabilitated water sources to bring potable water to population centers in rural areas. The cost of rehabilitating one kilometer of water under PRADECS was US$6,334. The GoB rehabilitated the same distance for US$9,640. Compared to The GoB, PRADECS is approximately 34% less costly. All cost comparisons for sub-component 2a point towards PRADECS being more efficient than other partners, for similar projects during the same time period. An infrastructure audit was done on PRADECS in 2011 and overall the quality of construction was confirmed as good. Sub-component 2b, financing subprojects to strengthen social inclusion and cohesion, also provides cost data in terms of a light economic analysis to show how the investments into small economic initiatives lead to real change for beneficiaries. Two examples, for illustrative purposes, provide this analysis. One subproject trained a group of 60 youth in skills training (tailoring) for a cost of US$170/person. The revenues per month from graduates fluctuate depending on rainy season (a low of US$12/month) to the dry season (high of US$26/month). On average, if a graduate makes US$19/month then the initial investment is recovered in 9 months. A second project financed a mill for an association of vulnerable albinos. The cost to finance the start-up of the mill was US$2,585. The mill started producing in February 2013, again with lower revenues during the rainy season (low of US$129) and higher revenues during the dry season (high of US$193). After 16 months the association is able to generate enough revenues to cover the initial cost. A final comparator is that of the cost of developing the Communal Development Plans. Significant resources were used in the project for component 1, capacity building. The average cost per commune of developing a Communal Development Plan under PRADECS was US$17,352. GIZ provided data that their average cost for producing the same plan per commune was approximately US$20,000. The cost for capacity building under PRADECS to that of GIZ, a partner implementing complementary capacity building activities in provinces of low-intervention, is therefore similar. 26 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Pin Foon K. F. Ah-Kee Procurement Analyst AFTA1 Procurement Mohamed Arbi Ben-Achour Consultant AFTN2 Social Safeguards Yvette Laure Djachechi Social Development Specialist AFTCS Social Safeguards Sameena Dost Counsel LEGES Legal SoulemaneFofana Rural Development Specialist AFTA1 Michael P. Fowler Finance Officer CTRDM Disbursement MathurinGbetibouo Consultant LCCHT PamphileKantabaze Operations Officer AFTHE Toni NtagandaKayonga Consultant AFTCS Joseph KizitoMubiru Sr Financial Management Specialist LCSFM Financial Mgmt Prosper Nindorera Procurement Specialist LCSPT Procurement Africa EshogbaOlojoba Environmental Specialist MNSEN Environ. Safeguards Korotimi Sylvie Traore Program Assistant MNAFM Team Assistance Supervision/ICR Andrew OseiAsibey Monitoring & Evaluation AFTDE M&E OtienoAyany Financial Management Specialist AFTME Financial Mgmt Jean-Claude Balcet Consultant LCSTR Pamela OrdingBeecroft Operations Analyst AFTCS Marcelo Jorge Fabre Sr Operations Officer OPSFC Team Leadership Toni NtagandaKayonga Consultant AFTCS MelanceNdikumasabo Procurement Specialist AFTPE Procurement Marco Nicoli Sr Knowledge Management Officer LEGVP Prosper Nindorera Procurement Specialist LCSPT Procurement Deo-Marcel Niyungeko Sr Municipal Engineer AFTU1 Marie-Claire Nzeyimana Communications Associate AFRSC PiaPeeters Sr Social Development Spec AFTCS Team Leadership Dominique Puthod Operations Officer LCSHE Sylvain AugusteRambeloson Senior Procurement Specialist AFTPE Procurement Thomas Jeffrey Ramin Sr Operations Officer HDGPE Chantal Rigaud Communications Associate HDGPE Fanny Robert Consultant AFTCS ClaretteRwagatore Program Assistant AFMBI Team Assistance Cheikh A. T. Sagna Sr Social Development Specialist AFTCS Social Safeguards AuroreSimbananiye Program Assistant AFTN3 Team Assistance Emmanuel Sinzohagera Consultant AFMBI Team Leadership (Co- Leanne Michelle Bayer Sr Social Development Specialist AFTCS TTL) 27 (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY06 37.95 248.14 FY07 39.68 319.84 Total: 77.63 567.98 Supervision/ICR FY07 9.1 31,042.00 FY08 26.9 122,074.00 FY09 25.16 115,984.00 FY10 35.36 139,004.00 FY11 26.54 147,511.00 FY12 21.29 110,108.00 FY13 14.85 67,901.00 Total: 159.20 733,624.00 28 Annex 5. Beneficiary Survey Results N/A 29 Annex 6. Stakeholder Workshop Report and Results N/A 30 Annex 7. Summary of Grantee's ICR and/or Comments on Draft ICR The GoB submitted to the Bank a final report on PRADECS to include in this ICR report. Below, a summary of the Government’s report is included. I. Evaluation of the objective of project operation and design 1.1. Consistency of the objective with national priorities The project’s objective was to help reduce poverty through the setting up and the operationalization of a transparent, participatory and decentralized funding mechanism that allowed capacity building for decentralized local communities in order to improve access to basic socio-economic services. The PRSP- I, which coincides with the project period, established the country development priorities in four areas: (i) improvement of governance and security, (ii) promotion of sustainable and equitable growth, (iii) development of human capital, and (iv) the fight against HIV/AIDS. The first area gives importance to capacity building at all levels, particularly at community level to strengthen their participatory process in the local development process. For the third, the government has set the objective of restoring pre-conflict performance and achieving the millennium goals by 2015 in the areas of education, health and drinking water, hygiene and sanitation. The same area establishes as government priority the action of putting in place mechanisms that will take into account the needs of conflict affected vulnerable populations. Furthermore, the letter of national decentralization policy and Community Development, adopted by the Government in February 2007, as well as the national decentralization policy paper, adopted in 2009, establishes decentralization as a pathway toward development in that it allows the transfer of powers, authority, responsibilities and resources from the central government to the decentralized administrations in order to promote community development. 1.2. Logic of a differentiated intervention per province The project design had planned that the project would intervene in a strong way in eight provinces and in a lighter way eight others. This logic was relevant in itself, considering the complementarity approach advocated by the project with other donors. These donors include the Gutwara Neza Program deployed in the eight provinces of the project’s lighter intervention. The complementarity approach was also considered with the Swiss cooperation in Ngozi province, with the Dutch cooperation in Muyinga province, with the Belgian Cooperation in Cibitoke province and with the UNDP / UNCDF project in Rutana province. In reality, this differentiated intervention has been detrimental to the provinces of lighter intervention. On the one hand, the Gutwara Neza Program withdrew from 4 provinces when PRADECS was at midterm; on the other hand some TPFs like Gutwara Neza did not fund subprojects and, no other partner besides PRADECS took into account the inclusion and social cohesion component. Complementarity based on geographical distribution of interventions cannot be fair if it is solely based on a harmonization of the size of funding and the content of programs. If the exclusion of Bujumbura Town is justified by the district’s centralized structure, it could also be revisited. In fact, access to basic social services is almost as problematic in this district as in the other provinces, if one would refer to classroom occupancy rate in public schools, for instance. 1.3. Institutional implementation arrangement The institutional arrangements favored the implementation of the project by a project management unit instead of implementation by the Ministry. It is possible that this choice explains the level of results achieved by the project. This choice does not pose the problem of ownership because the viability of interventions was provided by the effective participation of senior officials from the Ministry of Communal Development in the project activities, the connection of the implementation unit with sectorial ministries, the interface role played by the Provincial Office of the Cooperative and Associative Movement which are a provincial branches of the MDC, as well as the participation of beneficiaries in the 31 choice and funding of subprojects. On the other hand, the strategy of involving project implementation service providers helped not to burden the management unit and to gain in terms of cost. The services of facilitation agencies cost a total of 3.184 billion Burundi francs. It is certain that this cost would have been exceeded if the Executive Secretariat had positioned their own staff with the accompanying logistics. 1.4. The choice of funding instrument The project received a total of $ 40 million funding for a 5-year period. This instrument was preferred to the alternative of a long-term adaptable program (Adaptable Program Loan). Decentralization is a long- term process that could have justified the choice of the second form of funding. However, compared to the type of funding chosen, the duration and amount of funding are relevant to the project objectives. A question arises, however, regarding the amount of funding allocated to vulnerable groups. Given the proportion of the Burundian population likely to fall within this category, which alone could justify a large income creating project, it may appear that the $ 4.2 million allocated to social cohesion and inclusion, all components included, was insufficient. II. Evaluation of Results with Regard to Key Project Indicators The project development objective evolved around three main areas, namely, the improvement of access to basic social services, capacity building of local actors and of communities to make them able to be responsible for their development, and the consideration of vulnerable groups for inclusion for the sake of social cohesion. The achievement of this objective would be measured by a set of indicators defined in the Grant Agreement. As shown in the results developed below, the indicators have been fully achieved, with some even exceeding the targeted figures. Indicator N°1: At least 70% of the communes targeted see an improvement in access and use of primary health care, primary education and water, this in relation to the signing of a subproject funding agreement. This indicator was redefined in the mid-term review for the sake of clarity and was reformulated as "The targeted communes see an improvement in access and use of primary medical care (a), primary education (b) and water (c), this in relation to the rate of access and use at the signing time of a subproject funding Agreement." Moreover, the evaluation takes as reference 2010, which is the year in which the first subprojects were made functional. a) Improving access to primary health care To reflect the achievement of this indicator, Table 5 shows the progress of health centers built by PRADECS in number of infrastructures, in functionality and number of beneficiaries. Table 5 – Progress of primary health indicators from March 2010 to December 2012 Subproject Status March-10 Dec-11 Dec-12 Subprojects registered 67 69 74 Completed and functional Subprojects 8 31 55 Functionality rate 12% 45% 74% Number of direct beneficiaries 67 543 422 237 550 791 Table 5 shows the progress of the indicators at three levels: the construction of 74 new health centers which have allowed to reduce access distances, the progressive opening services of these infrastructures to reach a 74% functionality rate, the gradual progress of direct beneficiaries increases, hence going from 67,543 direct beneficiaries in March 2010 to 550,791 direct beneficiaries in December 2012. b) Improving access to primary education Achievements in the field of primary education are reflected in Table 6 which shows the number of schools built and their functionality. 32 Table 6 – Progress of primary education indicators from March 2010 to December 2012 Subproject status March-10 Dec-11 Dec-12 Subproject registered 103 154 189 Completed and functional Subprojects 18 95 184 Functionality rate (%) 17% 62% 97% Number of direct beneficiaries 148 654 662 705 1 170 179 As shown in Table 6, 189 primary schools to be built with the financial support of PRADECS were selected. The number of functional schools went from 18 schools in March 2010 to 184 primary schools in December 2012, which corresponds to a 97% functionality rate in December 2012. Moreover, with 184 primary schools newly built or rehabilitated, it was possible to produce 712 new classes that are now all occupied; which corresponds to a 14.4% contribution by PRADECS compared to the national need for 4,932 class rooms estimated at the start of the project in 2007. Then, the number of direct beneficiaries is increasingly growing because it went from 148,654 people in March 2010 to 1,170,179 in December 2012. c) Improving access to drinking water This indicator is measured by the improved or rehabilitated water points by PRADECS as well as by the beneficiaries of these developments as reflected in Table 3 below. Table 7 – Progress of access to drinking water: indicators from March 2010 to December 2012 Subproject status March-10 Dec-11 Dec-12 Subprojects registered 368 755 831 Completed and functional subprojects 95 496 812 Functionality rate (%) 26% 66% 98% Number of direct beneficiaries 45 321 233 120 382 580 Out of a total of 831 water points expected (664 managed sources and 167 tap-stands), 812 water points (98%) could be made functional in December 2012 (655 managed sources and 157 tap-stands), which made it possible to reach 382,580 people served with drinking water in December 2012. Indicator 2: At least 70% of infrastructures funded under Subprojects are maintained and are functional two years after completion. Table 8 below summarizes the local development subprojects implemented at the communal and hillside levels as well as at the level of their operationalization. Table 8 - Infrastructure built and their status on 31 December 2012 Total Subprojects Subprojects Type of subproject % functional Number completed functional Education 333 328 300 91% Health 74 74 55 74% Drinking water 56 56 54 96% Tracks 69 67 67 100% Other 286 286 280 98% Total 818 811 756 93% At the end of December 2012, 93% of infrastructures built or rehabilitated are maintained and functional, even if maintenance problem is not yet an issue since the infrastructure is still new. It should also be noted that almost all of the infrastructure and other works are immediately functional as soon as construction is completed and the construction deadline is usually between 2 and 6 months. There were some exceptions 33 where health centers were not immediately functional after construction, mainly because not only the technical equipment that should be provided by the Ministry of Health and Action against AIDS were slow to come, but also the medical staff was not quickly deployed. Note that here also, 74% of health centers are maintained and functional after completion and efforts are underway to make them all functional. Indicator N° 3: At least 15% of the targeted communes received funding outside the Project for activities in their Communal Community Development Plan (PCDC).The PCDC development process was accompanied by stakeholder and community leader sensitization on the use of PCDC as an instrument of advocacy and fundraising outside PRADECS. From 2009, systematic observations in the communes demonstrated that all communes (100%) were able to obtain funds outside PRADECS in order to fund activities under their PCDC. In its area of intervention, PRADECS funded provincial planning and evaluation meetings of their development program with the participation of other TFPs, with results on PCDC funding. Indicator N°4: At least 50% of the targeted communes have included the concerns of vulnerable groups in the Communal Community Development Plan. The results achieved in relation to this indicator are linked to a series of awareness actions by PRADECS, including technical notes on taking into account the concerns of vulnerable groups, the action plan for vulnerable groups, the setting up of thematic groups specialized on issues of vulnerable groups as well as relevant training activities. The result is that all 34 PCDCs supported by PRADECS integrated subprojects for vulnerable groups and that 289 subprojects from these PCDCs were funded by the project. Indicator N°5: At least 430 hillside and commune subprojects were completed under the Project. Table 9 below shows the progress of subproject infrastructures over 3 years from 2010 to 2012 to provide an idea of comparison between the target value for the entire project period and the actual value achieved by the Project. Table 9 – Evolution of subprojects Subproject Status March-10 Dec-11 Dec-12 Recorded subprojects 182 594 818 Completed subprojects 70 393 782 % of subprojects completed against 16% 91% 182% the target indicator (430 MP) At the end of December 2011, the number of subprojects already recorded (MP completed and MP underway) had already exceeded the number of 430 subprojects targeted for the entire project period. At the end of 2012, the percentage of subprojects completed had already reached 182% with regard to the target value of 430 subprojects. Indicator N°6: Not less than 90% of the targeted communes developed or updated their Communal Community Development Plan in a participatory manner. The initial target was of 50 communes of major intervention provinces but, the Swiss Cooperation supported the communes of Ngozi and the Dutch Cooperation through IRC, an international NGO, supported the communes of Muyinga. The development of PCDC followed an inclusive, participatory and bottom-to-top approach, a process that combined all the layers of the population, regardless of their socio-economic status and vulnerability. Indicator N°7: At least 80% of the targeted communes have set up a strategy to ensure the participation of marginalized groups in the decision-making process, particularly in relation to the Communal Community Development Plan. It should be noted that 100% of the targeted communes have set up a strategy to ensure the participation of marginalized groups in the decision-making process, particularly in relation to PCDC and for the following reasons: (i) activities related to local planning of the PCDCs have 34 consistently started in the census hillsides to give opportunity to all vulnerable people to be present and participate, (ii) techniques and strategies to give voice to the floor to the weak were used, (iii) the community development structures put in place during the process systematically included representatives of vulnerable groups, (iv) an actual awareness and training campaign for vulnerable groups was conducted before and after the production of PCDCs to raise their awareness on their own development, inclusion and social cohesion. Indicator N°8: At least 70 subprojects for vulnerable groups have been completed under the Project. With regard to the targeted 70 subprojects for the entire project period, the figure of 289 subprojects was reached over the five years of the Project, which corresponds to an excess of 219 subprojects (413%). Indicator N° 9: At least 60% of the communes targeted provide adequate data for monitoring and evaluation within the time required in the implementation of their development programs. With the systematic support of facilitation agencies, consulting engineers, financial controllers of the Trust Agency, provincial offices of cooperative and association movement, it should be noted that although the beginning was difficult for some communes, in the end, 100 % of communes have managed to provide adequate data for monitoring and evaluation in a timely manner. Monthly, quarterly and annual project reports were made based on data from the communes. Indicator N°10: A communication strategy taking into account all stakeholders has been developed and implemented. From a study conducted at the start of the project, a communication strategy and an action plan on communication have been developed. Then, the implementation of this plan involved major communication agencies in the country on the basis of a contract between the project and media agencies specialized in the production of programs, magazines, advertisements and newspaper articles. Periodic assessments of the strategy took place to improve the performance of each communication sector and adjustments were made whenever needed. III. Evaluation of Government performance PRADECS acted as the operationalization of the government's commitment to support grassroots communities in decentralization and the provision of both social and socio-economic community infrastructures, as reflected in the February 2007 national decentralization policy and community development paper. It thus played a leading role both in the preparation and the implementation of the project. Conduct of project preparation activities: The government has put in place an appropriate institutional framework to ensure a design that is consistent with national priorities and sectorial needs of the project scope. The process was under the coordination of a preparation unit whose team was set up competitively, supported by a preparatory committee representing the various sectorial ministries. Various studies were conducted to secure the implementation framework of the project, including (i) a study on the social impact to identify the socio-economic variables to be taken into consideration in the preparation of project (ii) an institutional assessment to determine the appropriate institutional framework and the strengths and weaknesses to be taken into account in the definition of a capacity building program, (iii) a study of the decentralized financial management systems to provide the operational mechanisms for the transfer of resources, and (iv) an environmental study leading to the design of an Environmental and Social Management Framework (ESMF). Setting up the conditions for project implementation: While negotiations on the Grant Agreement took place on 8 February 2007, the government took steps to fulfill the implementation of the project in set timeframe. A project management unit was set up in February 2007 and the national supervisory bodies were established proactively to the project implementation, namely the Order No. 04/VP1 of 22/01/2007 of the First Vice-President of the Republic by establishing the National Project Steering Committee and the Ministerial Order 558/530 of 18/5/2007 for the Technical Monitoring of Project Implementation. 35 These bodies, as well as the Monitoring Committee of the IDA portfolio, accompanied the project implementation. Key strategic government intervention during project implementation: The government took action both at the institutional and operational levels to ensure proper implementation of the project. Without being exhaustive, the following may be mentioned: - Reinforcement and operationalization of the Provincial Offices of Cooperative and Associative Movement (BPMCA). The institutional arrangement of PRADECS chose not to create branches of the Executive Secretariat at the provincial level, but rather rely on BPMCA as a decentralized structure of the Ministry of Communal Development. From the start of the project, the government redeployed staff to put the best people in the provinces of strong PRADECS intervention. These structures played an important role in supporting the communes in all project related activities, in the vertical communication between the ES and the communes as well as in the monitoring and evaluation of Project activities. - Coordination of actions through the Thematic Group "Decentralization and Community Development." The government coordinated the Technical and Financial Partners working in the area of decentralization through monthly thematic meetings. These meetings allowed the implementation of the complementary approach between the project and other TFPs. Under this approach, the provincial advisors of the Gutwara Neza Program played the role of facilitation agency for PRADECS in the provinces of minor intervention where these advisors were present. - Mobilization of communal contributions. Initial previsions had set the contribution of communes and communities to funding local development subprojects at US$1,465,000 against US$27.7 million for IDA, i.e. a total contribution of 5.3%. The achievements are such that the actual contributions totaled 1,554,908,945 BIF against 36,997,439,980 BIF, representing 4.4% of participation. These figures show that the government has far exceeded its expectations in both absolute and relative terms, taking into account the fact that rates have been halved in the process. In many cases where communes were not able to pay their contribution, the government supplemented their efforts by calling upon FONIC to make available their financial contribution. It turned out that the communes could not keep up after participation in the financing of two to three subprojects, and the financial support of FONIC served to give them a boost. - Collaboration agreement with sectoral ministries: the decentralized sector ministries (education, public works, water, transport, youth, national solidarity, health and others) have played an important role in providing technical advice to communes with regard to compliance with quality standards and national standards. Cooperation agreements were signed in the early days of the project between the project and those Ministries. Under these agreements all electrification subprojects and all drinking water subprojects were respectively carried out under the management of REGIDESO and AHR. Moreover, despite the difficulties, the Ministry of Public Health continued to equip health centers built under PRADECS to the extent that, at the end of the project, 74% of health facilities were functional. - Achievements of quantitative and qualitative objectives of the project: The quantitative objectives for the project were achieved in all its components. The "capacity building" component, reached 69,934 men/training days on 21 topics that can be classified into two groups: the reinforcement of commune project management and the reinforcement of civic consciousness. For the "subproject" component, while the targets were respectively 430 commune and hillside subprojects and 70 subprojects for vulnerable groups, the achievements reached 818 subprojects in the first category and 289 in the second, respectively, double and quadruple of the expected number. In connection with the "management, M&E and communication" component, the first major indicator of government performance is the funding consumption rate that reached 99%; the second is the quality 36 of management as evidenced by the unqualified opinion of the auditors on all financial years of the project. In conclusion, the Government expresses its satisfaction with the role played by mandated actors in the preparation and implementation of the project. A key lesson learned is the level of representativeness of the government in the project steering bodies. Indeed, the National Steering Committee (CNOP) was chaired at the Office of the Vice-President of the Republic and was made of members from ministerial level. This high representation is an advantage in that it can follow up the project at the highest policy level, but at the same time a disadvantage on the regularity of statutory meetings. IV. Evaluation of IDA performances Support for project preparation: IDA has made available to the Government an advance of US$579,420, as well as US$265,000 from the Japanese Grant. This funding helped to set up a project preparation unit, to carry out the main studies in order to better define the project concept, to prepare the implementation manual, the manual of administrative, financial and budgetary procedures and the monitoring and evaluation manual, all three conditions for project implementation and to recruit key project personnel. Therefore, the provision of an advance for project preparation (PPF) was a factor promoting the respect of key dates in the project implementation process. With the technical support of IDA, the work plan for the first eighteen months of the project was developed, hence providing an introduction to monitoring project implementation by the team put in place. It should be noted that the Japanese grant was poorly executed. Indeed, out of an amount of US$265,000, disbursement was limited to US$75,000. Regular supervision in the form of support and advice: Project implementation was facilitated by the support and advice of IDA project managers in the form of supervision missions conducted every six months, after which an action plan for the next 6 months was developed. Various forms of communication (e-mail, teleconference, etc.) punctuated this interval. This ongoing communication between the project management and IDA supervision helped solve problems as they arose, but also and more so, to act proactively in anticipation of problems. Occasional technical support missions supplemented regular IDA missions. These include a support mission for the improvement of M&E indicators of the project and a mission that strengthened the inclusion of vulnerable groups in the construction of community facilities (access ramps, wider doors, appropriate latrines, etc.). Support in resolving crises: during the project life cycle, three events are identified as having been major factors affecting the success of the project. Bank Team contributed in finding solutions to critical situations. - Communal and Community Contributions: Halfway through the project, the communes were unable to mobilize their contributions in cash and in kind for the communities, especially when they were in the second or third subprojects. The Grant Agreement was amended to reduce the rate of these contributions. This measure had a positive impact on the result of subprojects, so mobilization of communal and community contributions were not a problem. - Institutional dysfunction: The IDA team which initiated the modification of the Grant Agreement also amended it in response to institutional dysfunction when the change in government structure placed the project under the Ministry of Communal Development instead of the Ministry of Interior and Public Security, at the same time blocking the functioning of project organs. - Crisis of Trust Agency: The IDA team invested in finding a temporary solution to restart disbursements that had stopped for two months following the non-renewal of contract of the Trust Agency for incompetence. During the recruiting process of another agency, an individual consultant was hired to act as a Trust Agency for urgent disbursements. 37 Points of disagreement: Overall, the project implementation benefited from the consistency of the views of IDA and government staff. However, these points have not converged on the eligibility of homes for health personnel. While IDA argued that these homes are not eligible for funding as homes for officials, the government continues to maintain that the homes are integral part of the health infrastructure and essential to their functioning. In conclusion, the government expresses its satisfaction with the role of IDA both during the preparation and implementation of the project. Two main lessons are to be learned from the performance of the World Bank. On the one hand, the Government recognizes the benefits of a decentralized supervision unit put in place at the local office of the World Bank. Such close supervision can act on problems in real time and even anticipate them together with the project management unit; it allows the World Bank to be present in the reflections of the decentralization thematic group and be in close contact with other TFPs involved in the sector. V. CONCLUSIONS 5.1. Key lessons Successful operation despite the high degree of identified risks: The preparation of the project described as "substantial" the inherent risk in its implementation. Among them was the risk of seeing resources monopolized by communes with more active leaders than others, political interference in the project implementation, the low technical and financial capacity of the communes, the risk of exclusion of vulnerable groups, the risk of communal tension between the allocation of resources and the risk of misuse of resources. The project has established control mechanisms of these risks to the extent that no major event has come to hinder the smooth running of the project. Transparent implementation and apolitical project management is a key factor in its success. In addition, transparency and objectivity that characterized the allocation of resources, both at the initial allocation and the level of allocation of bonus helped ease tensions on the allocation. Special incentives for considering the needs of vulnerable groups have been initiated. These incentives include the extension of a specific action plan for vulnerable groups, the establishment of a thematic groups specialized on problems of vulnerable groups and the determination of prior allocation to activities related to this category of beneficiaries. A pioneering experimental funding approach for other donors: of all identified risks, the accountability of communes in the management of funds was the most critical, both in terms of management capacity and security of funds. This accountability is also a source of very exciting satisfaction on the part of local actors. For the first time, they have experienced the true role of managing a local community outside of routine administrative acts. Trial and error has been observed at the outset in terms of abiding by the procurement and financial management reference principles (transparency, efficiency, economy and fairness), but the training and supervision of the Executive Secretariat allowed the maturing of the system. The PRADECS is a successful experiment of this accountability that could inspire and strengthen other donors. Important lessons could be learned by evaluating PRADECS and PADLPC interventions that took place in the same province of Cibitoke, with different approaches. Subprojects with visible economic impacts and with domino effects that can serve as a model for the promotion of the local economy: one of the impacts directly measurable of the project is the creation of employment and revenues induced by the implementation of subprojects. Of the US$37 million dollars for the implementation of subprojects, about US$26 million were injected in the local economy in the form of labor and purchase of materials. Some community facilities such as markets and electrical connections have led to the emergence of income-generating activities such as workshops, welding and carpentry, hairdressing salons, mills, as well as entertainment activities such as film projection. From this perspective, the project has completed a dual purpose of providing basic services to the population, while serving as a model that can inspire interventions promoting the local economy. However, substantial 38 results in terms of promoting the local economy could only be achieved with an approach focused on strengthening local entrepreneurship, an approach that was outside the field defined for the project. 5.2. Key challenges A drop of water in an ocean of needs: The PRSP-II recognizes a significant advance in terms of peace building and progress in the restoration of basic social services with regard to the context of the PRSP I. The contribution of PRADECS in the progress of certain social indicators is remarkable as evidenced by the results presented in Section 3. However, these indicators remain below normal standards: infant mortality is 59 deaths per 1,000 live births, the infant and child mortality rate at 96 deaths per 1,000 live births and maternal mortality at 499 maternal deaths per 100,000 live births (PRSP-II). The completion rate of school enrollment remains 47.7% and makes it impossible to achieve the millennium goals in this area in 2015 (PRSP-II). Access rate to drinking water is limited to around 50% with the goal of reaching 70% in 2015 (PRSP-II). As for the implementation of decentralization, the new law on the transfer of powers from the state to communes raises new challenges in terms of empowerment of local authorities and communities to take charge of their development. The other side of participatory planning: participatory planning is an approach that empowers the hillside population in prioritizing their development needs through CDP’s. The hillside consultations also create at the same time expectations of a population that awaits positive returns through the implementation of subprojects originally included in the CDP’s and the satisfaction expressed needs. If substantial funding does not follow the ongoing planning process, these people will end up five years later with the same needs and the same state of vulnerability, to the extent that the CDP’s will eventually be seen more as a tool for propaganda rather than a development tool. Maintenance of infrastructure as the best indicator of ownership: the project has established mechanisms for the appropriation of achievements such as contributions in kind and cash to finance subprojects and monitoring of implementation of subprojects by beneficiary committees. However, it appeared that the best indicator of ownership and the best guarantee of the sustainability of infrastructure is the development of a culture of maintenance and repair. The fact is that despite the awareness raising and training on the subject, there is lack of little gestures like cleanliness and small repairs for which omission may cause more material damage. In addition to the continuation of awareness raising and training, the government is preparing to issue an order establishing the obligations of communes and communities in terms of infrastructure maintenance. 5.3. Conclusion PRADECS is an operation that has achieved remarkable success in all aspects of project management: its efficient and effective implementation, the achievement of its development goal and the materiality of its impact, the high degree of satisfaction of beneficiaries. The government hopes that the World Bank continues its support in the area of decentralization in order to maintain and consolidate project achievements. A request to this effect has been sent to IDA. 39 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders Comments by local officials from a stakeholder workshop: A member of the ICR team participated in a stakeholder workshop organized by the project in December 2012 just a few weeks before the close of the project. The workshop brought together over 200 communal authorities from all communes targeted by the project. Four key messages from project stakeholders emerged from the workshop: (i) the 1,067 subprojects realized through the project had a direct impact on improving access to social services in rural areas where 90% of the population live (2008 census), (ii) the needs remain great and stakeholders argued for continued support through PRADECS to accompany the communes in the on-going decentralization process, (iii) the impact of social services, particularly schools and health clinics was of urgent priority at the start of the project, but the diversification of projects for more water, transport, and local economic initiatives had tremendous gains for the local population, and (iv) communal officials expressed that their capacity to manage local public administration was augmented due to the training provided by the project. Comments by GIZ: GIZ specifically focused on the support role that PRADECS provided the Ministry of Communal Development, stating that members of the PIU often led working groups on specific topics – for example, developing the national capacity building plan, or devising a draft law on the maintenance of communal infrastructure. GIZ stated that they saw that the activities of PRADECS were tightly aligned with the policies and work of the Ministry and without the role of PRADECS as technical advisor there will now be a void. Comments by Swiss Cooperation: The Head of Swiss Cooperation was very direct in stating that PRADECS was an opportunity for the other donors working in the sector, and that the resources and approach used in the project was a “force de frappe�. Additionally, it was acknowledge that PRADECS followed principles of the Paris Declaration, in terms of Government ownership, in a genuine way by having the NOSC operationalized and involved in practical project supervision, as well as being established in a manner that it provided guidance to the project. 40 Annex 9. List of Supporting Documents A. Project Documents Date/period Title Project Appraisal Document February 2007 Financing Agreement b/w GoB and IDA May 22, 2007 Aide Mémoire – Mid Term Review Mission October 13–25, 2009 Aide Mémoire – ISM April 29 – May 12, 2011 Aide Mémoire – ISM December 12 – 20, 2011 Aide Mémoire - ISM May 21 – 25, 2012 Aide Mémoire – ISM November 26 – December 5, 2012 Implementation Status and Results Report # 1 10/12/2007 Implementation Status and Results Report # 2 01/30/2008 Implementation Status and Results Report # 3 07/29/2008 Implementation Status and Results Report # 4 10/20/2008 Implementation Status and Results Report # 5 04/21/2009 Implementation Status and Results Report # 6 12/23/2009 Implementation Status and Results Report # 7 05/27/2010 Implementation Status and Results Report # 8 03/27/2011 Implementation Status and Results Report # 9 11/01/2011 Implementation Status and Results Report # 10 04/11/2012 Implementation Status and Results Report # 11 09/22/2012 Implementation Status and Results Report # 12 04/24/2013 B. World Bank documents (analytical and operational) IEG 2011.Preparing High-Quality Implementation Completion and Results Reports.Perspectives from the Independent Evaluation Group.Handouts from the IEG Workshop. Washington DC: The World Bank. OPCS. 2006. Implementation Completion and Results Report Guidelines. Last updated on 10/05/2011. Washington DC: The World Bank. OPCS & IEG. 2006. Harmonized Evaluation Criteria for Bank Operations. Washington DC: The Word Bank. World Bank. 2013. Designing Community Driven Development Operations in Fragile and Conflict Situations: Lessons from a Stocktake, Pre-review Draft. Washington DC: World Bank. World Bank (2012) Country Assistance Strategy for the Republic of Burundi for the Period FY13-16, Report No. 72334-BI, Washington DC: the World Bank. World Bank, (2008) Country Assistance Strategy for the Republic of Burundi for the Period FY09-12. World Bank, (2007) Interim Strategy Note for the Republic of Burundi for the Period FY07-08. 41 C. Government of Burundi documents République du Burundi, (2012), Enquête auprès des Bénéficiaires du Projet d’Appui au Développement Communautaire et Social, PRADECS, CURDES, Bujumbura: Burundi. République du Burundi, (2012), Evaluation Finale du Project, PRADECS, Egis Bdpa, Bujumbura: Burundi. République du Burundi, (2010), Programme et Plan d’action opérationnel de renforcement des capacités des acteurs en charge de la mise en œuvre de la politique de la décentralisation et du développement communautaire, PRADECS , Consultants Gérard Baltissen et Elias Sentamba, Bujumbura: Burundi. République du Burundi, (2006), Cadre Stratégique de Croissance et de Lutte Contre la Pauvreté, CSLP II, Bujumbura: Burundi. République du Burundi, Ministère de l’Intérieure et de la Sécurité Publique (2006), Analyse de l’Etat des Lieux des Projets de Développement Décentralisé dans le cadre de la Préparation du PRADECS, Bujumbura: Burundi. 42 Annex 10. List of People Met During the ICR Mission The ICR Mission took place during May 12-17, 2013. The mission met with key members of the PRADECS PIU, the Ministry of Communal Development, and other development partners in the decentralization thematic working group. A field trip was undertaken to visit four communes to discuss with communal administrators, communal technical advisors, and chief of hillsides where subprojects and project activities were implemented. The four communes were broken down to visit two communes of high-intensity intervention and two communes of low-intensity intervention. The list of people met during the ICR mission and their respective affiliation is provided below. Name Title Ministry / Department / Agency Government of Burundi Son Excellence Jean Claude Minister Minister of Communal Ndihokubwayo Development Son Excellence Martin Ex-Minister Minister of Communal Nivyabandi Development Monsieur Jean Marie Director General Ministry of Communal Ntihirageza Development Monsieur Serges Karikurubu Chief of Zone Commune of Rutegama Monsieur Diomède Muzigira Chief of Hillside Hillside of Munanira II, Commune of Rutegama Monsieur FrancoisNinzigamye Chief of Hillside Hillside of Camunandu, Commune of Rutegama Monsieur Thomas Bukuru Chief of Hillside Hillside of Gashingura, Community Development Committee Member Monsieur Célestin Ntirandekura Chief of Hillside Commune of Rutegama Monsieur Tite Nsengimana Chief Hillside Hillside of Bubanda, Community Development Committee Member Monsieur Sylvestre Bukobero Chief of Hillside Commune of Nyarunazi Monsieur Roger Nijimbere Chief of Hillside Commune of Nyarukere Monsieur Diomède Member of the Communal Commune of Rutegama Ntibanezerwa Development Committee Monsieur Bonaventure Chief of Hillside Hillside of Munanira I Nduwimana Monsieur Nestor Nshimirimana Director of Primary School Commune of Rutegama, Communal Development Committee Monsieur Emmanuel Agronomist Community Development Ndabambarire Committee Member Monsieur Vital Nsengiyumva Technical Advisor in Commune of Giheta Charge of Development Name Title Ministry/Department/Agency Monsieur Stanislas Niyondezi President of the Communal Commune of Giheta Water Utility Monsieur Cyriaque Birushe Technical Advisor in Commune of Giheta Charge of Social Affairs and 43 Administration Monsieur Balthazar Nihenda Director of the Communal Commune of Mutaho College Monsieur Fidèle Nizigiyimana Technical Advisor in Commune of Giheta Charge of Development Monsieur Emery Placide Director of the Vocational Commune of Giheta Ntakarutimana Training Center Monsieur Jean Marie Vianney Communal Administrator Commune of Mutimbuzi Mutabazi Monsieur Nathanael Manirakiza Technical Advisor in Commune of Mutimbuzi Charge of Development Monsieur Joel Habiyambere Technical Advisor in Commune of Mutimbuzi Charge of Social Affairs and Administration Monsieur Jérémie Hakizimana Community Development Commune of Mutimbuzi Committee Member Monsieur Onesphore Communal Administrator Commune of Makebuko Ntakiyinanira Project Implementation Unit (PIU) PRADECS Monsieur Thomas Minani Executive Secretary PRADECS Monsieur Guido Infrastructure Specialist PRADECS Ndayihanzamaso Monsieur Bernard Ndikumana Procurement Specialist PRADECS Monsieur Cyrille Sindahabaye M&E Specialist PRADECS Monsieur Tharcisse Social and PRADECS Barakamfitiye Environmental Specialist Monsieur Emile Gahungere Capacity Building PRADECS Specialist Development Partners Monsieur Claudio Tognola Head of Cooperation Swiss Cooperation Madame Deputy Head of Office Swiss Cooperation AnnonciataNdikumasabo Monsieur Philip Kusch Technical Advisor for GIZ Decentralization Monsieur Léopold Bapfutwabo Program Officer on GIZ Governance 44 IBRD 33380 BURUNDI SELECTED CITIES AND TOWNS MAIN ROADS PROVINCE CAPITALS PROVINCE BOUNDARIES NATIONAL CAPITAL INTERNATIONAL BOUNDARIES RIVERS 29°E 30°E 31°E Lake This map was produced by the Map Design Unit of The World Bank. To Kigali Kivu The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any To endorsement or acceptance of such boundaries. Gitarama Kagera Lake Lake Rweru Cohoha R WANDA KIRUNDO To Cyangugu Kirundo To Butare To Rulenge u ar y u an uv uv K R CIBITOKE MUYINGA NGOZI Muyinga To Cibitoke Nyakanura Ngozi Kayanza Rusiba 3°S Musada Ruvuvu 3°S Buhiga To Bubanza AY AYA N Z A K AYA Karuzi si Kakonko M w eru BUBANZA izi KARUZI CANKUZO Rus vu Cankuzo vu Muramvya V YA Ru R AM MU To Uvira BUJUMBURA L uvironza Gitega WA R O M WA RA Ruyiga DEM. REP. Mwaro RUYIGI BU OF CONGO GITEGA To Kibondo M gu JU p un Mt. Heha U Rum (2,670 m) Bukirasazi B Matana TA NZA NIA BURURI Mutangaro R U TA N A Bururi Most distant Rutana Rumonge headwater of the Nile River 4°S 4°S Makamba BURUNDI MAKAMBA z i ara Mabanda r ag Mu Lake Nyanza-Lac Tanganyika To Kasulu 0 10 20 30 40 Kilometers 0 10 20 30 Miles 29°E 30°E 31°E SEPTEMBER 2004