Document of The World Bank Report No: ICR00002988 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-H4090 IDA-H6020 TF-93124) ON IDA GRANTS IN THE AMOUNT OF SDR 16.4 MILLION (US$ 25.9 MILLION EQUIVALENT) AND ON A CO-FINANCING GRANT FROM THE FOOD PRICE CRISIS RESPONSE TRUST FUND IN THE AMOUNT OF US$7.0 MILLION TO THE REPUBLIC OF TOGO FOR A Community Development Project March 26, 2014 Social Protection Sector Unit (West/Central) Country Department AFCF2 Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective September 30, 2013) Currency Unit = CFA Franc US$ 1.00 = CFA Franc 486 FISCAL YEAR: January 1 - December 31 ABBREVIATIONS AND ACRONYMS AGAIB Agency for support to grass roots initiatives (Agence d'Appui aux Initiatives de Base) CDD Community-Driven Development CDP Community Development Project CRW Crisis Response Window CVD Village Development Association (Comit Villageois de Diveloppement) DP Development Partner EDF European Development Fund EPPR Emergency Program for Poverty Reduction ESMF Environmental and Social Management Framework EU European Union GDP Gross Domestic Product GFRP Global Food Crisis Response Program GPE Global Partnership for Education ICRR Implementation Completion and Results Report IDA International Development Association IFAD International Fund for Agricultural Development IFR Interim Financial Report I-PRSP Interim Poverty Reduction Strategy Paper ISN Interim Strategy Note LICUS Low-Income Countries Under Stress LIPW Labor-Intensive Public Works M&E Monitoring and Evaluation MTR Mid Term Review NGO Non-Governmental Organization O&M Operation and Maintenance ODEF Organisation de Diveloppement des Eaux et Forets PAD Project Appraisal Document PDCPlus Community Development and Safety Nets Project PCU Project Coordination Unit PDO Project Development Objective PERI Projet Education et Renforcement Institutionnel PIM Project Implementation Manual PTA Parent/Teacher Association QEA Quality at Entry Assessment QSA Quality of Supervision Assessment QUIBB 2006 Welfare Survey (Questionnaire des Indicateurs de Base du Bien- Etre) RESEN Rapport d'etat du systeme educative national RPF Resettlement Policy Framework TF Trust Fund THIMO Travaux a haute intensite de main d'wuvre TOR Terms of Reference TS Technical Secretariat TTL Task Team Leader UNDP United Nations Development Programme Vice President: Makhtar Diop Country Director: Ousmane Diagana Sector Manager: Stefano Paternostro Project Team Leader: John Van Dyck ICR Primary Author: Christopher S. Ward TOGO Community Development Project CONTENTS D ata Sheet ....................................................................................................................................... i A. Basic Information ........................................... .......... i B. Key Dates.......................................................1 C. Ratings Summary..................................................1 D. Sector and Theme Codes .............................................. E. Bank Staff........................................................ F. Results Framework Analysis........................ ..................iii G. Ratings of Project Performance in ISRs..................................ix H. Restructuring (if any)..............................................ix . Disbursement Profile................................................ x Implementation Completion and Results Report...............................1 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.............................22 5. Assessment of Bank and Borrower Performance ....................... .....23 6. Lessons Learned opltnd..............ml...................................................................... 26 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners...........27 Annexes................................................................................................ 28 Annex 1. Project Costs and Financing............................ .........28 Annex 2. Outputs by Component .........................................29 Annex 3. Economic and Financial Analysis ....................................36 Annex 4. Bank Lending and Implementation Support/Supervision Processes...........37 Annex 5. Beneficiary Survey Results..................................................................................... 39 Annex 6. Stakeholder Workshop Report and Results .................... ......40 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR..................41 Annex 8. Comments of Financiers and Other Partners/Stakeholders................42 Annex 9. List of Supporting Documents .............................. .....43 Data Sheet A. Basic Information Country: Togo Project Name: TG-Community Dev. Project ERL (FY08) IDA-H4090,IDA- Project ID: P110943 L/C/TF Number(s): H6020,TF-93124 ICR Date: 03/07/2014 ICR Type: Core ICR Lending Instrument: ERL Borrower: REPUBLIC OF TOGO Original Total Commitment: USD 17.20M Disbursed Amount: USD 32.33M Revised Amount: USD 32.90M Environmental Category: B Implementing Agencies: Technical Secretariat Community Development Project Coinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 02/27/2008 Effectiveness: 09/10/2008 09/10/2008 10/17/2008 Appraisal: 04/18/2008 Restructuring(s): 06/29/2010 Approval: 06/26/2008 Mid-term Review: 01/22/2011 02/01/2011 Closing: 06/30/2012 09/30/2013 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: Satisfactory Government: Satisfactory Implementing Quality of Supervision: Satisfactory . Satisfactory Agency/Agencies: Overall Bank Overall Borrower Satisfactory Satisfactory Performance C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments (if P PIndicato"rs Rain Performance any) Potential Problem Project Quality at Entry at any time (Yes/No): Y_(QEA):on Problem Project at any time YsQuality of Supervision on Yes one (Yes/No): (QSA): DO rating before Closing/Inactive status: D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General agriculture, fishing and forestry sector 25 25 General public administration sector 10 10 Health 18 18 Other social services 17 17 Primary education 30 30 Theme Code (as % of total Bank financing) Education for all 17 17 Health system performance 17 17 Participation and civic engagement 33 33 Rural services and infrastructure 16 16 Social Inclusion 17 17 E. Bank Staff Positions At ICR At Approval Vice President: Makhtar Diop Obiageli Katryn Ezekwesili Country Director: Ousmane Diagana Madani M. Tall Sector Manager: Stefano Paternostro Eva Jarawan Project Team Leader: John Van Dyck Giuseppe Zampaglione ICR Team Leader: John Van Dyck ICR Primary Author: Christopher S. Ward 11 F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) To provide poor communities with improved basic socio-economic infrastructures and income generating activities by financing at least 350 subprojects to be identified and implemented directly by communities. Revised Project Development Objectives (as approved by original approving authority) To provide selected poor communities with improved basic socio-economic infrastructures, income-generating and labor intensive activities and access to food, by financing, among others, sub-projects to be identified and implemented directly by communities. (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Achieved at Indicator Baseline Value appro Revised cmplet approval TreVaus Completion or documents) Target Years Indicator 1 :Direct Project Beneficiaries (number), of which female (%) Value quantitative or 0 NA 199,500 (50%) 2819(4 Qualitative) Date achieved 06/03/2008 09/30/2013 ]09/30/2013 09/30/2013 Comments (incl. % Percentage of women does not include agricultural inputs component, for which achievement) no gender disaggregated data was available Indicator 1(a): Beneficiaries of public works program Value quantitative or 0 NA 25,000 26,623 Qualitative) _06/03/2008 ___________09 30/201 Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 1(b): Students receiving at least a meal per day Value quantitative or 0 NA 57,000 60,934 Qualitative) Date achieved 06/03/2008 _09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 1(c): Farmers with access to fertilizers/seeds Value quantitative or 0 NA 14,000 13,774 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) 111 Indicator 1(d): People directly involved in new or improved income generating activities Value quantitative or 0 NA 1,500 5,981 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 1(e): Students enrolled in constructed/rehabilitated schools Value quantitative or 0 NA 72,000 97,367 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % The numbers presented above are cumulative. 31,699 students attended class in achievement) project classrooms in 2013. Indicator 1(f): People with access to an improved water source Value quantitative or 0 NA 30,000 63,440 Qualitative) I Date achieved 06/03/2008 _09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 2: Area under improved land management (ha) Value quantitative or 0 NA 2,200 2,032 Qualitative) I Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Repartition of financing of subprojects consistent with regional targeting criteria Indicator 3: as estimated by the QUIBB 2006 (Variance between QUIBB 2006 repartition and actual allocatios by region) Value or 0 Less than 10% Less than 10% . quantitative orvariation variation 3.5% variation Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 (. The indicator is a measure of targeting effectiveness. It measures the variation c ment between each region's contribution to poverty (from the QUIBB of 2006) and the actual allocations under the project to each region. Indicator 4: Subprojects functioning one year after completion Value quantitative or 261 90% 96% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (T. arget revised from a number (261) to a percentage (90 %) in June 2010. achievement) 1V (b) Intermediate Outcome Indicator(s) Original Target Formally Actual Value Values (from Achieved at Indicator Baseline Value appro Revised Target cmplet approval Values Completion or documents) Target Years Basic socio-economic infrastructure subprojects completed (number) Value (quantitative or 200 200 290 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 2 : Classrooms built or rehabilitated (number) Value (quantitative or 0 A 360 621 Qualitative) Date achieved 06/03/2008 109/30/2013 109/30/2013 Comments (incl. % achievement) Indicator 3 : Health facilities constructed renovated and/or equipped (number) Value (quantitative or 0 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 4 :Improved community water points constructed or rehabilitated under the project (number) Value (quantitative or 0 Qualitative) Date achieved 06/03/2008 09/30/2013 109/30/2013 Comments (incl. % achievement) Indicator 5 : Income-generating activities subprojects (number) Value (quantitative or 0 0___150 ___ 150_ _______294 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Basic socio-economic infrastructures assessed as having satisfactory technical quality (%) Value (quantitative or 1 _________0________160_____________ 90% 93% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (ic. % Target was revised from a number (160) to a percentage (90%) in June 2010. achievement) V Indicator 7 : Meals provided to primary school students (in thousands) (annual) Value (quantitative or 2352 each year NA12,100 17,013 Qualitative) (3,500 in two years) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 8 : Schools successfully implementing the School Feeding Program (number) Value (quantitative or 84 NA 180 256 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 9 : Schools providing primary school children with one fruit per day (%) Value (quantitative or 0 Qualitative) Date achieved 06/03/2008 109/30/2013 109/30/2013 Comments (incl. % ot measured achievement) Indicator 10: Person days of work provided (number) Value (quantitative or 0 11139979 Qualitative) Date achieved 06/03/2008 109/30/2013 109/30/2013 Comments (incl. % achievement) Indicator 11: Participants below the poverty line in public works (%) Value (quantitative or 0 A5A Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (N.ot able to be measured because impact evaluation was cancelled achievement) Indicator 12: Female participation (%) Value (quantitative or 0 NA 50% 39.6% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (ic. % Refers to public works achievement) Indicator 13 : Young people (aged less than 35) employed (%) Value (quantitative or 0 58% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (ic. % Refers to public works achievement) vi Indicator 14 : Total cost allocated to wages (%) Value (quantitative or 0 A 70% 70.1% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (ic. % Refers to public works achievement) Indicator 15: Public works scheme completed with satisfactory technical quality (%) Value (quantitative or 0 NA 75% 80% Qualitative) Date achieved 06/08/2008 09/30/2013 09/30/2013 (. For reforestation only. Figures excludes a reforested area damaged by fire in the c ment Savanes region which was replanted, but too late to be included in the technical achievement)audit. Annual net increase in cash income of targeted participants compared to control Indicator 16 :. group in public works (%) Value (quantitative or 0 A0A 1 A30% NA _ __ _ Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (N.ot able to be measured because impact evaluation was cancelled achievement) t CVD are inclusive, well representative, and well trained: (i) at least 25% of the members are women Value (quantitative or 1 0____90%____ 90% 6.6% Qualitative) 0 % 9676/ Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) CVD are inclusive, well representative, and well trained: (ii) at least one woman in a leadership post Value (quantitative or 0 90% 90% 96.7% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) CVD are inclusive, well representative, and well trained: (iii) number of CVD members successfully trained Value (quantitative or 0 2,100 11663 _________ 0_____2,100____________ 2,100 1,663 ______ Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 20 : AGAIB Project Management expenses as a % of disbursements for subprojects Value (quantitative or 1 0____15%____ 15% 1.1% Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) vii IFR submitted by AGAIB to the TS on a quarterly basis, within 15 days Indicator 21: following the end of the quarter (frequency of IFR transmission changed from monthly to quarterly) (number) Value (quantitative or Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) IFR submitted by TS to the Bank within 45 days from the end of each Quarter Indicator 22: (nme):_______________ (number) Value (quantitative or 1 1 Qualitative) 01 81 Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 23: Fertilizer procured and distributed to farmers (tons) Value (quantitative or 0 A 4,275 4,275 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Seeds procured and distributed to farmers (tons) (includes both commercial and foundation seeds) Value (quantitative or 0 A 370 858 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 25: (i) commercial seeds distributed Value (quantitative or 0 NA 356 844 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) Indicator 26 : (ii) foundation seeds distributed Value (quantitative or 0'A 14 14 Qualitative) Date achieved 06/03/2008 09/30/2013 09/30/2013 Comments (incl. % achievement) viii G. Ratings of Project Performance in ISRs Date ISR Actual Disbursements Archived (USD millions) 1 10/30/2008 Satisfactory Satisfactory 1.54 2 03/27/2009 Satisfactory Moderately Satisfactory 1.60 3 04/16/2009 Satisfactory Moderately Satisfactory 1.64 4 07/28/2009 Moderately Satisfactory Moderately Satisfactory 1.93 5 10/30/2009 Moderately Unsatisfactory Moderately Unsatisfactory 3.39 6 02/25/2010 Moderately Satisfactory Moderately Satisfactory 5.53 7 06/10/2010 Satisfactory Satisfactory 7.85 8 12/22/2010 Satisfactory Satisfactory 11.53 9 05/31/2011 Satisfactory Satisfactory 14.38 10 12/27/2011 Satisfactory Satisfactory 19.51 11 06/26/2012 Satisfactory Satisfactory 21.38 12 12/17/2012 Satisfactory Satisfactory 22.79 13 06/25/2013 Satisfactory Satisfactory 25.22 H. Restructuring (if any) Board ISR Ratings at Amount Restructuring Restructuring Disbursed at Reason for Restructuring & Key Date(s) Approved PDO Restructuring Changes Made Change DO I in USD millions To respond to the global food and financial crisis, the restructuring created a new sub-component for 10/17/2008 Y 0.00 School Feeding and a new component for Enhanced Cereal Production. The PDO was revised to take the new activities into account. The project was restructured to create a new Labor Intensive Public Works Sub-component to 06/29/2010 Y S 5 12.95 respond to the global food and financial crisis. The PDO and results indicators were also restructured, and the project was extended by 12 months. 1x If PDO and/or Key Outcome Targets were formally revised (approved by the original approving body) enter ratings below: Outcome Ratings Against Original PDO/Targets Satisfactory Against Formally Revised PDO/Targets Satisfactory Overall (weighted) rating Satisfactory I. Disbursement Profile - Original ---- Formally Revised -- Actual 40- 0 Un S10- CY C CY CY C CY CY C C CY CY C C CY CY CY C 0 I ll ll ll ll ll CD CD C CD 0 CD CD C' C rq N r N 1 N N4 N I N N N NN N N N N N x Implementation Completion and Results Report Togo Community Development Project 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal During a prolonged governance and economic crisis (1990-2005), the people of Togo fell into a spiral of poverty and destitution. GDP per capita dropped at an average rate of 1% per annum, health services deteriorated, and only 17% of 11 year old children were enrolled in school. In 2006, 62 percent of the population lived below the poverty line. During this period, international aid was reduced and then suspended. Following a Presidential election in 2005 and legislative elections in 2007, Togo engaged in a dialogue with the international community which led to the re-establishment of full economic and financial relations. Between 2005 and 2008 Bank involvement focused on analytic work and on a LICUS TF funded Emergency Program for Poverty Reduction (EPPR) based on a community driven development (CDD) approach. Government and Bank strategies were set out in a Government Interim Poverty Reduction Strategy Paper (I-PRSP) and a Bank Interim Strategy Note (ISN), both of which targeted improved governance and transparency, return to private sector-led growth, and provision for urgent social needs. The Community Development Project (CDP) was a centerpiece of the Bank ISN. The project was the main instrument for delivering on the ISN's third pillar of addressing urgent social needs. Building on the positive CDD experience of the EPPR, the project was to help the Government deliver visible results to the population through improving access to, and supply of, basic social services for the most vulnerable groups. The CDD approach was seen to be particularly appropriate given the low capacity, fragile context and the need to stimulate participation at the grassroots level. The design included support for construction/rehabilitation of primary schools, health posts, water points, rural roads, and other socio-economic infrastructure, as well as support for community-driven income generating activities. Support was later added for school feeding, labor intensive public works, and agricultural inputs as part of the Bank's response to the global food, fuel, and financial crisis. 1.2 Original Project Development Objectives (PDO) and Key Indicators Initially financed by a grant of USD 17.2 million (IDA grant H4090), the Togo Community Development Project was approved by the Bank Board in June 2008 as an Emergency Operation. The original Project Development Objective (PDO) was to provide poor communities with improved basic socio-economic infrastructures and income generating activities by financing at least 350 subprojects to be identified and implemented directly by communities. The original key indicators were as follows: 1 + Repartition of financing of sub-projects consistent with regional targeting criteria by the 2006 Welfare Survey ('QUIBB') + Number of new sub-projects functioning one year after completion + Number of communities that, following sensitization and training, have presented a proposal which was approved by the regional support agency (AGAIB) approval committee + Minimum number of subprojects which reflect the priorities of targeted communities and beneficiaries at mid-term review (MTR) and at end of project 1.3 Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification In response to escalating food prices, the Bank provided additional financing of US$7.0 million (TF-91324) for the PDC in October 2008 from the Food Price Crisis Response Trust Fund, four months after Board approval of the initial financing. As a response to rising food prices, the additional financing supported an additional sub-component for school feeding and a new component for enhanced cereals production. The support for school feeding was meant to respond to the impact of the food price crisis on children in rural and poor areas of the country. The rationale for the enhanced cereals production component was to counteract rising cereals prices by boosting productivity in the sector through provision of seeds and fertilizer, which were in limited use in the country. The PDO was formally amended as follows (changes marked): 'To provide poor communities with improved basic socio-economic infrastructures and income generating activities by financing, among others, at least 350 subprojects to be identified and implemented directly by communities.' With the new focus on child nutrition and food production, two additional results indicators were added to the results framework. The additional project outcome indicators tracked the number of beneficiaries of school feeding activities, the quantity of cereals delivered, and number of farmers benefiting. The Bank again made use of the flexibility of the project implementation mechanism with a second additional financing of $8.7 million (Grant H6020) in June 2010 to respond to the global financial and economic crisis. The second additional financing, with resources from the Pilot Crisis Response Window (CRW), scaled up school feeding and introduced a new sub-component for labor-intensive public works as safety net mechanisms to help blunt the impact of the global crisis. The PDO was formally amended for a second time, as follows (changes marked): 'To provide selected poor communities with improved basic socio-economic infrastructures, income- generating and labor intensive activities and access to food, by financing, among others, at-least 350 sub-projects to be identified and implemented directly by communities.' At the same time, the outcome indicators were revised to reflect the additional sub-components and to strengthen the relevance of the indicators to the PDO, particularly to measure improved access. Two indicators were added, and two were dropped so as to streamline the number of indicators being tracked and reported on, and one indicator was amended. 2 Table 1: Original and Revised Outcome Indicators Original outcome indicators Outcome indicators as revised in June 2010 Number of direct project beneficiaries, of which: - Female - Beneficiaries of public works programs - Students receiving at least one meal a day - Farmers with access to fertilizers/seeds - People directly involved in new or improved income- generating activities - Students enrolled in constructed/rehabilitated schools - People with access to an improved water source (new indicator and sub-indicators added in June 2010) Area under improved land management (hectares) (new indicator added in June 2010) Repartition of financing of sub-projects Repartition of financing of sub-projects consistent with consistent with regional targeting criteria by regional targeting criteria by the 2006 Welfare Survey the 2006 Welfare Survey ('QUIBB') ('QUIBB') (unchanged) Number of new sub-projects functioning one Percentage of new sub-projects functioning one year after year after completion completion (amended in June 2010) Number of communities that, following (dropped in June 2010) sensitization and training, have presented a proposal which was approved by the regional support agency (AGAIB) approval committee Minimum number of subprojects which reflect (dropped in June 2010) the priorities of targeted communities and beneficiaries at mid-term review (MTR) and at end of project 1.4 Main Beneficiaries The primary target groups of the project were poor communities in all five regions of Togo, as identified through the 2006 Welfare Survey (QUIBB). 1.5 Original Components The original components were: (1) community sub-projects (USD 12.5 million) to improve access to education, health, water and sanitation and other socio-economic infrastructures, and to increase income generating opportunities; (2) training (USD 1.7 million) at community, regional and national levels; and (3) management, operating costs and audits (USD 3.0 million). Total project cost was USD 14.2 million, plus an imputed community contribution of USD 0.63 million. 1.6 Revised Components The restructuring to accommodate the additional resources made available to the project in September 2008 and June 2010 resulted in the following project structure: 3 Component 1: Community Development (USD 23.2 million): 1.1 Community Development Sub-Projects (USD 12.5 million, financed by IDA H4090) a) Socio-economic infrastructure b) Income-generating activities 1.2 Emergency School Feeding (USD 6.3 million, of which USD 4.3 million from the Pilot Crisis Response Window (H6020) was added to USD 2.0 million from the 2008 GFRP TF-91324) 1.3 Labor-intensive Public Works ('LIPW', USD 4.4 million from the Pilot Crisis Response Window, H6020) Component 2: CDP Training (USD 1.7 million financed by IDA H4090) Component 3: CDP Management and Operating Costs (USD 3.0 million, financed by IDA H4090) Component 4: Enhanced Cereals Production (USD 5.0 million, financed by GFRP TF-93124) 1.7 Other significant changes None 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry The original operation was an emergency project prepared under Operational Policy 8.0 (Rapid Response to Crises and Emergencies) designed to consolidate the Bank's re-engagement with Togo. The project was designed for rapid impact within the shared goals of the Government's I- PRSP and the Bank ISN, particularly poverty reduction, satisfaction of urgent social needs, and the strengthening of local and regional capacity for sustainable development. The flexible and modular design provided a framework within which further activities and financing could be added. Involvement of government agencies, central, regional and local civil society institutions and other Development Partner (DPs) in preparation ensured that project design reflected objectives and experience of all stakeholders and created ownership of the resulting project. The identification of the specific problems to be addressed - poverty, health, institutional and educational status - was accurate, and the resulting design responded to these problems by providing income generation, health and educational services provision and access, and nutrition for the most vulnerable group, children. The project was designed on the basis of past experiences both worldwide and in Togo (particularly the EPPR). Lessons incorporated included: (1) simplicity of design (at least initially); (2) participatory processes that strengthened community ownership; (3) a planning and 4 implementation process that combined strong leadership from the center with effective regional planning and implementation support and, at community level, well-tested bottom up institutional mechanisms , notably the village development associations ('CVDs'), of which 1,700 already existed across the country; and (4) clear and comprehensive implementation procedures set out in manuals. The comprehensive 2006 Welfare Survey (QUIBB) provided an effective basis for targeting. The reliance on the existing regional Grassroots Development Support Agencies (Agences d'Appui aux Initiatives de Base-AGAIB) and implementation procedures that had already proved effective meant that the project was ready for implementation. Risks and mitigating measures, most of them common to all CDD projects, were assessed as follows: Table 2: Risks and Mitigating Measures Risk Mitigating measure Excessive control by central authorities Design, reflected in the implementation manuals, gave decision-making and spending authority to communities and regional support structures. Lack of counterpart contribution for the (very poor) The contribution level was set at just 5% to allow communities affordability whilst ensuring ownership Lack of sub-project sustainability because staff and Regional approval committees and the central operating budget might not be available from line Project Steering Committee included all relevant ministries that were to operate facilities (schools, ministries, and line ministries signed off on sub- clinics etc.) projects in their sector Weak capacity on procurement, financial Intensive training at all levels. management and M&E at both local and national levels. Weak sustainability and poor targeting of income Rigorous screening, selection of organizations with generating activities. past experience, and focus on agriculture with its lower risks and greater experience. These risks were correctly identified (and the mitigation measures proved appropriate and largely effective). A further risk assessment was conducted for the September 2008 additions to the project (School Feeding and Enhanced Cereals Production). A hygiene risk was identified for School Feeding and was appropriately mitigated by measures to ensure hand washing and cleanliness. The risks identified for Enhanced Cereals Production - further escalation in global food prices, tension in communities concerning the selection of beneficiaries, and negative health and environmental effects of fertilizer and pesticide use - were formally correct, but did not capture many of the main risks of a subsidized input distribution program (such as undermining of nascent commercial markets, rent-seeking behavior, low cost recovery, and absence of sustainability or development impact). 5 Risks assessed in the Project Paper presented to the Board for the LIPW program added in 2010 were limited to possible resettlement issues and tensions regarding the selection of beneficiaries and did not touch on other common risks attached to LIPW projects - poor quality of assets created and difficulty of ensuring sustainability of those assets. 2.2 Implementation Despite the implementation risks identified at appraisal (see 2.1 above), implementation was largely satisfactory, even though extra activities were successively added to the project. The project was rated satisfactory or moderately satisfactory throughout implementation except for a period between October 2009-February 2010 when the project was rated moderately unsatisfactory because of implementation delays, especially in regard to Income Generating Activities and community training. The delay arose because of poor performance of the FM and M&E specialists and procurement delays in purchasing vehicles. The team agreed with the Government on corrective measures, such as replacing staff and standardizing community training, which resulted in the project returning to satisfactory performance in four months. Much of the overall smooth implementation is attributable to the energy and quality of the staff of the project coordination unit ('the PCU') and regional implementing agencies ('AGAIB') staff and to the practice of contracting implementation support to locally based NGOs and consultants. Management of the PCU and the five AGAIB was generally sound and professional. Staffing was adequate, although some posts were slow to be filled (M&E specialist, for example, see 2.3 below). The decision to trim ten posts from the AGAIB in 2012 reduced the AGAIB's supervision and capacity building performance, and the elimination of the PCU procurement post at the same time slowed up remaining procurement. Based on the village development associations (CVDs) and a light touch participatory planning process, the delivery of Community Subprojects, the heart of the project, was on the whole efficiently implemented. There were only minor execution delays and procurement problems despite the risks in devolving functions to community level. Technical quality was good (96% of infrastructure of satisfactory standard), and costs were lower than for comparable infrastructure under other programs in Togo. Implementation of the Income Generating Activities was contracted out to NGOs which had varying capacity to identify, screen and support the activities. Two common implementation problems encountered were that the financial resources available to the NGOs (10% of the project cost) were inadequate to support implementation, and the two year duration of the support was too short to ensure the activities were properly established and stakeholders fully trained. Nonetheless, the activity was fully implemented and reached four times the number of people targeted. The School Feeding program, an implementation challenge because of the dispersed nature of beneficiary schools and other factors, benefitted from good cooperation between the regional implementing agencies (the AGAIB) and the local branches of the ministries responsible for education and social services. Implementation through local NGOs with ongoing programs and experience in each region and the provisions for close follow up and inspection contributed to 6 success. The main implementation challenges were the wide geographic distribution of beneficiary schools, very high supervision requirement to ensure proper management of funds and meal delivery, and the cumbersome system of payments. Despite these constraints, the component satisfactorily delivered 50% in excess of targets (17 million meals in 256 schools, against a target of 12.1 million meals in 180 schools). Despite the inherent and particular complexities attached to the implementation of labor- intensive public works (LIPW), implementation encountered only moderate problems which were eventually resolved. Challenges involved institutional arrangements, application of the criteria for targeting of beneficiaries, and execution across 128 forestry sites and 123 other sites. Contractual arrangements between the AGAIB and the National Agency for Forestry (ODEF) took time to be concluded and put into practice, which led to delays. Arrangements for paying workers through the postal system proved very slow and also contributed to delays before being finally resolved after the post office increased staffing for the payment processing. Nonetheless, the activity was completed and delivered 12% more work days than targeted. The Enhanced Cereals Production component required cooperation between a number of agencies. Implementation problems occurred with the seeds supply activity due to the limited experience of the Ministry of Agriculture in assessing seed requirements and handling seed stocks. The component nonetheless delivered on its (limited) targets. Training was delivered by NGOs and was generally of satisfactory quality. Some deficiencies occurred in the training of village development associations (CDVs) and local groups carried out through NGOs which, although committed and possessing good local knowledge and relations with communities, sometimes lacked technical knowledge and also lacked the in-house resources for follow up once the two year training contracts were completed. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization The M&E system was designed to capture the outcomes and outputs of project activities across several dimensions: the quantity and quality of physical infrastructure, the manner in which project resources were targeted geographically, inclusion and gender representativity, and the respect of project procedures. The M&E system relied on two main components: (i) consolidation at the central level by the Technical Secretariat of results monitoring data produced by the AGAIBs, for which a computerized management information system was to be developed, and (ii) a series of evaluations of quality and results, including two technical audits, two evaluations of income generating activities, an evaluation of school feeding, an environmental and social audit, and annual financial audits. Initially the project did not attempt to assign targets for the number of sub-projects or beneficiaries in each of the respective sectors in which the project operated (education, health, water, etc.) as the choice of sectors was left to communities and unknown a priori. To conform to new Bank-wide mandatory core indicators, the June 2010 additional financing and restructuring added outcome indicators on numbers of beneficiaries overall and across different types of activities and sectors. The estimation of numbers of beneficiaries and sub- projects in each sectors was aided by the experience of the project up to that point, specifically 7 the distribution of community priorities across sectors. At the same time, in order to prevent the number of outcome indicators from being too large, two of the original outcome indicators measuring community empowerment were dropped. In retrospect, it might have been preferable to retain at least one of these indicators as a measure of the aspect of the PDO relating to community empowerment. The June 2010 project paper also planned for a randomized control trial impact evaluation on the labor intensive public works component. Funds were not allocated under the Bank financing to fund this costly impact evaluation with the expectation that trust fund resources could later be mobilized. At the beginning of project implementation, M&E was slow to get underway. The first M&E specialist had health issues resulting in lengthy medical leave. AGAIBs produced data on the progress of project activities but consolidation at the central level was slow. Certain indicators, such as those concerning the number of students attending schools, and the number of beneficiaries of water points, were found to be very labor intensive to collect and as a result were infrequently updated. In 2011 a new project M&E specialist was recruited and an international consultant brought in to set up and computerize the M&E system, which finally became operational in 2012 at both the AGAIB and central level. Towards the end of the project, the Bank and Technical Secretariat took stock of the indicators. To consolidate efforts and reduce the burden of AGAIBs, it was agreed to finance a large end-of- project beneficiary survey to collect data that not been conducive to regular collection for financial and logistical reasons. TORs were developed covering indicators such as the numbers of students attending schools, the number of beneficiaries of water points and health posts, the representation of women in community groups, the quality of community training, and beneficiary satisfaction. The survey was completed in September 2013 and results summarized in Annex 5. The Bank task team's application for trust fund resources to finance the randomized control trial impact evaluation on the labor intensive public works component was not approved and no other resources could be identified, rendering this aspect of the M&E efforts not possible. 2.4 Safeguard and Fiduciary Compliance Environmental, social and resettlement impacts were assessed at appraisal as limited and the environmental category was B. An ESMF and RPF were prepared and focal points in each agency were responsible for supporting their application. Training was conducted at all levels, use of checklists was systematic in all applications for project financing, and all works contracts contained provisions for environmental and social protection. Instances of non-compliance with requirements occurred but these were corrected through further training and did not affect results. An end-of-project audit found only minor anomalies and the Togolese Ministry of Environment issued a certificate in October 2013 to confirm that the project had complied with environmental and social protection requirements. Procurement and financial management complied with Bank procedures and no major problems were identified during either statutory audit or verification by World Bank procurement and financial management specialists. A fiduciary issue did arise in 2011 when the AGAIB salaries were raised without prior approval of the Bank. This led to temporary downgrading of the 8 overall project rating (to Moderately Satisfactory), but was corrected when staff repaid the unauthorized increases. 2.5 Post-completion Operation/Next Phase Arrangements to ensure sustainability post-completion are as follows: Community Subprojects: At the completion of implementation, responsibility for operation and maintenance (O&M) lies with the village development associations (CVDs) and - in the case of education and health facilities - with the local branches of ministries. Training and support were provided during implementation to develop the capacity of the CVDs and to help them develop sources of funds for O&M - for example, revenue raising plantations of cash crops or trees on common or public land. Income Generating Activities: Responsibility for O&M of assets and for business management lies with the producer association party to the contract. Training and support were provided during the implementation period. School Feeding: As the poorest communities generally lack the resources to fund school feeding operations on their own, continuation of the program requires sustained donor and government financing. The school feeding operations under the PDC succeeded in convincing Government to begin providing financing for school feeding out of domestic resources beginning in 2014. Together with resources under the PDC's successor project (PDCplus), continued operation of school feeding is assured for the next few years. Over the longer-term, efforts will need to be made to ensure predictable and adequate Government funding for this activity. LIPW: For reforestation works, the Ministry of Environment and Forest Resources has taken responsibility and an agreement to this effect was signed between the Ministry and the PDC. CDVs are responsible for other LIPW-created assets. Agricultural inputs: As a short-term emergency measure, the component supporting agricultural inputs focused primarily on meeting the immediate need for inputs while supplying modest capacity-building support for advancing with the design of a more sustainable system for provision of agricultural inputs. The project was not meant to take a larger role in the dialogue on agriculture sector policy, which instead was left to the Bank's agriculture task team. Section 4 below discusses the risks that arise from these arrangements. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Rating: High Relevance of Design. The design of the project was highly relevant, given the context of low capacity, poor access to basic social services, low levels of participation, and an urgent need to deliver visible results to the population in a fragile context. The CDD approach permitted direct 9 community participation in the decision-making process and implementation of subprojects, thereby permitting rapid results despite low government capacity, particularly at the local level. Design supported Government and Bank strategies (see 1.1 above) by strengthening institutions and providing for urgent social needs. Relevance of Objectives. The PDO was conservatively formulated to focus on outcomes for which the project was directly accountable, and the results targeted were clear, realistic and important. The objectives captured both the need to improve access to basic services through creating infrastructure as well as the desire to do so through a participative, community based approach targeting poor areas of the country. Both the original components and the successive additions to the project responded to the PDO and expected outcomes. One drawback initially was that, although the outcomes set out in the initial Project Paper targeted improved access, the key indicators rather concerned physical outputs (sub-projects completed and functioning etc.). This was corrected in the 2010 Board- approved revisions when the revised outcome indicators measured access across all components. Relevance of Implementation. As the project was implemented, it remained highly relevant. The community-based approach, modular design, and local implementation support structure made the project highly flexible. Over the course of implementation, this flexibility allowed the project to expand to address further needs as they became urgent priorities, particularly those that arose from the global food, fuel and financial crisis (such as school feeding, public works, and agricultural inputs). A testament to the relevance of the project is the extremely high levels of applications for community sub-projects received from communities-in some sectors more than 10 times more applications were received than the project had the ability to finance. Figure 1: Demand from Communities for Socio-economic Infrastructure Sub-Projects 1800 - 1590 1600 - 1400 - 1200 - 1000 - 800 - 690 600 - 400 -250 200 193 16811 200 - 111 3 1 95 1 0 23 3 rmL Education Health Water and Sociocultural Rural roads Socioeconomic Sanitation and other OApplications received from communities M Approved sub-projects 10 The project objectives remained relevant through implementation to the end of the project. While by the end of the project, the situation was no longer one of emergency, Togo still faced acute needs for infrastructure related to basic services delivery, social safety nets, and access to economic opportunities through income generating activities. Indeed, as evidence of the continued relevance, at the end of the project, the Government requested support from the World Bank for a second phase of the project, which was approved in March 2012 in the form of the Community Development and Safety Nets Project (PDCplus). 3.2 Achievement of Project Development Objectives Rating: Substantial The project's development objective and outcome indicators were restructured twice: first in the Additional Financing of October 2008 and again in the Additional Financing of June 2010. According to the August 2006 ICR Guidelines (revised Oct 5, 2011), the project outcome should be assessed against both the original and revised project objectives. Both the original and revised PDO statements can be divided into three dimensions: A. Providing basic socio-economic infrastructures, income generating and labor-intensive activities and access to food; B. Targeting poor communities; C. Empowering communities to identify and directly implement project activities. A. Providing basic socio-economic infrastructures, income generating and labor-intensive activities and access to food Table 3: Outcome indicators relating to community sub-projects and safety nets activities End of Actual Actual as Project outcome indicators Baseline implementation September percentage target 30, 2013 of target Direct project beneficiaries (indicator added in June 2010 restructuring with the sub- 0 199,500 268,119 134% indicators listed below) (of which female) -- 50% 44% 88% * Beneficiaries of public works 25,000 26,623 106% programs * Students receiving at least one meal 57,000 60,934 122% a day 0 * Farmers with access to 0 14,000 13,774 98% fertilizers/seeds * People directly involved in new or improved income-generating 0 1,500 5,981 399% activities * Students enrolled in constructed/rehabilitated schools 0 72,000 97,367 135% (cumulative) 11 * People with access to an improved 0 30,000 63,440 211% water source Area under improved land management (hectares) (refers to public works in the area 0 2,200 ha 2,032 ha 92% of reforestation. Added in June 2010.) Sub-projects functioning one year after completion (%) (outcome indicator from original PAD which was retained in both restructurings) The project exceeded the majority of outcome indicators related to providing basic socio- economic infrastructures, income generating and labor-intensive activities and access to food. The overall number of beneficiaries (268,119) was 134% above the target. The technical audit confirmed that the project had performed very well in terms of the quality of infrastructure built, with 100 percent of sub-project sampled in the technical audit being functional one year after completion. The areas which fell slightly short of targets were i) the provision of agricultural inputs, which was able to reach 98 percent of expected beneficiaries; ii) reforestation activities carried out through public works which covered a land area of 2,032 ha, or 92 percent of the 2,200 ha envisioned; and iii) the proportion of female beneficiaries, which reached 44% as opposed to 50%. The shortfall in regard to female beneficiaries was largely due to girls only constituting only 42% of students in the classrooms built, and 40% of workers on labor-intensive public works (see Section 3.5). The original PDO targeted the completion of at least community 350 sub-projects. This goal was achieved with the completion of 584 community sub-projects, including 290 socio-economic infrastructures and 294 income generating activity sub-projects. The language specifying the number of sub-projects was dropped in the June 2010 restructuring. Component 1.1(a): Socio-economic infrastructure. Construction and rehabilitation of socio- economic infrastructure, such as primary schools, health posts, water points, rural roads, and other socio-economic infrastructure, was the largest aspect of the project in terms of resource allocation (see Annex 2 for a breakdown of sub-projects by type). The project exceeded all outcome targets for numbers of beneficiaries benefitting from improved access to infrastructure. While setting sectoral targets in a CDD project is risky because the choices of communities cannot be known a priori, the project achieved the output targets related to education (190 primary school sub-projects building or rehabilitating 621 classrooms benefitting 97,367 students ) and water (37 water subprojects resulting in 76 water points benefitting 63,440 people2). Due to community preference, fewer health facilities were built than originally estimated (23 v. a targeted 40). In some cases, health authorities were sometimes not able to provide personnel or operating funds and so a requested health sub-project was not approved for financing. The beneficiary survey at the end of the project found that the 23 health centers hosted 144,096 patient visits in 2013 alone. The number of unique beneficiaries of the health centers was not able to be calculated because of the possibility of repeat visits by patients. 1 Targets were for 360 classrooms benefitting 72,000 students. 2 Targets were for 30 water points benefitting 30,000 people. 12 The outcome indicator measuring the quality of the socio-economic infrastructure aimed to measure the percentage of infrastructure functional one year after completion (target = 90%). This indicator originated in the original PAD and was maintained through the two restructurings. The technical audit at the end of the project found that 96 percent of socio-economic infrastructure sub-projects were functional one year after completion, thereby exceeding the target. In fact, given the gap between the completion of the socio-economic infrastructure between 2009 and 2011 and the technical audit which was completed in September 2013, the technical audit measured the functionality between two and four years after completion. Four of the 89 infrastructures samples (4 percent) were found to be non-functional. Three of the 89 infrastructures (2 percent) examined by the technical audit were found to be functional but with serious deficiencies. The remaining 83 infrastructures sampled (93 percent) were functional with acceptable technical quality, though 54 (61 percent) presented cosmetic or minor problems not limiting the usefulness of the infrastructure at the age of two to four years. Table 4: Functionality of socio-economic infrastructure (Sub-projects completed from 2009 to 2011) Category "0" Category "1" Category "2" Category "3" Number Not functional Functional but Functional and Functional and Regioni of sub- and does not does not meet meets standards meets technical projects meet technical technical or but with and architectural sampled and architectural architectural cosmetic or standards standards minor problems Maritime 18 1 0 17 0 Plateaux 16 2 0 6 8 Centrale 15 1 0 8 6 Kara 15 0 1 11 3 Savanes 25 0 1 12 12 Total 89 4 2 54 29 Percent of 4% 2% 61% 33% total Source: Technical Audit, September 2013 Component 1.1(b). Income generating sub-projects. The project support 294 income generation sub-projects, nearly doubling the target of 150. The number of people benefitting from income generation sub-projects was almost four times the target (5,981 v. a target 1,500). Income-generating sub-projects were primarily agricultural, including market gardening, grain production, livestock, fish farming, food processing, and milling (see Annex 2). No project supporting entrepreneurial activities can expect universal success among the ventures supported, and in terms of the success rate of sub-project the PDC performed reasonably well. The 2013 evaluation of the income generating sub-projects found that 75 percent were still supporting improved livelihoods for beneficiaries, with 58% 'working well' and 17% 'working moderately well.' 25% were found to be structurally unprofitable or abandoned. A quantitative assessment of the profitability of these sub-projects was extremely difficult due to beneficiaries' tendency to underreport their revenues to the implementing agency for fear of taxation and to 13 make the case for further support. The evaluation of the income generating activities was able to gather complete data on profitability for only 6% among the sample of income generating activity sub-projects, not a large enough sample to draw any conclusions. Qualitatively, interviews with beneficiaries found that 78% thought that they had improved their livelihood through the sub-projects, with 12% reporting their situation being unchanged and 10% reporting being worse off. Beneficiaries also reported that the incomes generated meant they did not have to make high-cost short term borrowing. The income-generating activities sub-component had the most success of any project intervention in benefitting women. According to the evaluation, 67 percent of beneficiaries of this sub-component were female. Component 1.2: School Feeding. The approach of the school feeding program was based on a system of village women who prepared and distributed meals under the supervision of Parent- Teachers Associations (PTA) and experienced local NGOs. The approach, which by the end of the project covered 256 schools in all five regions, promoted local economies by procuring all food locally and employing local women. An evaluation of the school meals program carried out in 2010 showed that the program provided two thirds of the daily nutritional needs of beneficiary children. The evaluation also found that most children (72%) were satisfied with the quantity of food and that while meat or fish was served 3 times per week, fruits were rarely served due to budget constraints. Despite limited access to clean water in some areas, hand washing was systematic in 85% of schools. Convinced of the value of the program, the Government allocated its own resources to continue the school feeding program after PDC in conjunction with IDA resources under the successor to the PDC, the PDCplus. Component 1.3: Labor-intensive public works. A review of the LIPW activity found that 12% more labor days were created than targeted. A total of 2,032 hectares (target 2,200 hectares) were reforested on 128 public and community sites. Another 123 LIPW sub-projects were completed (target 75 sub-projects). These non-forestry LIPW sub-projects included rehabilitation of rural roads, construction or rehabilitation of village ponds and other community water structures. Excluding trees damaged by a fire in the region of Savanes and which were later replanted too late for the technical audit, the success rate for reforestation was 80% against a target of 75%. The technical audit visited a small sample of other type of infrastructure realized through LIPW, which were technically satisfactory but the sample size is too small to draw any overall conclusion. The activity fell short on its (ambitious) gender and youth employment targets (see 3.5 (a) below). The cancellation of the randomized control trial impact evaluation due to the inability to secure trust fund resources prevents the assessment of the project in relation to two output indicators would have required such an evaluation (LIPW beneficiaries who were below the poverty line, and related increases in incomes of targeted LIPW beneficiaries compared to a control group). As no impact evaluation was carried out, data on these two indicators - and on the broader impacts of the project on health, education etc. - are not available. Component 4: Enhanced Cereals Production. This component was designed with a narrow, short-term objective of supporting the purchase and delivery of inputs to increase the domestic 14 production of cereals. The component reached 98 percent of expected beneficiaries covering a land area of 2,032 ha, or 92 percent of the 2,200 ha envisioned. The component met its quantitative targets of fertilizer and seed distribution and recorded yields were approximately 1 t/ha above the average for the region, which was the target. Therefore the component does seem to have been successful as an emergency measure in increasing production and yields in the short-term; the longer term impact was not quantified but is not likely to be significant. Cost recovery was not an explicit objective, and the component recovered 87% of the full cost of inputs. The counterpart funds generated through the cost recovery of inputs were used by the Ministry of Agriculture and the Ministry of Community Development to support young entrepreneurs in agriculture and agri-business, and to strengthen targeted parts of the agricultural value chain. The usage of these funds was agreed between the Government and the Bank task team following inputs from the Bank's agriculture task team covering Togo. The component also supported modest capacity building to help advance work on the design of a sustainable system to deliver and distribute fertilizers and seeds, an area that continues to be supported by the Bank under the Agriculture Sector Support Project. B. Targeting Poor Communities Table 5: Outcome Indicators related to Targeting End of Actual Actual as Project outcome indicator Baseline implementation September percentage target 30, 2013 of target Repartition offinancing of sub-projects consistent with regional targeting criteria less than 10% 3.5% (outcome indicator from original PAD variation variation which was retained in both restructurings) Despite a general lack of data in a fragile state environment, the project sought to ensure to the degree possible that project investments were directed towards the poor. To measure this, the original PAD included an outcome indicator to verify that the allocation of resources for sub- projects closely followed the relative levels of poverty among regions, explicitly excluding the relatively less poor capital of Lom& and its close surroundings. This indicator was maintained through the two restructurings. At the end of the project, the actual pattern of resource allocation only differed by 3.5% from the distribution of poverty among regions as measured by the 2006 QUIBB, thereby achieving the outcome indicator target. 15 Table 6: Regional contribution to poverty and allocation of sub-projects (FCFA) Region Regional Regional "Ideal" Variation betwee e alocation for % of contribution regional "ideal" and actual Percent community total to poverty allocation per regioal alocatio varation sub-projects (QUIBB 2006) QUIBB 2006 (d) = (c) * total (e) = absolute value (f) = (e) (a) (b) (C) of(a) of(d)-(a) as % of(d) Maritime 1,336,284,671 23.9% 24.3% 1,357,324,546 21,039,875 1.6% Plateaux 1,291,750,819 23.1% 23.1% 1,290,296,173 1,454,646 0.1% Centrale 896,234,663 16.1% 14.3% 798,754,774 97,479,889 12.2% Kara 947,005,496 17.0% 17.9% 999,839,892 52,834,396 5.3% Savanes 1,108,836,374 19.9% 20.3% 1,133,896,637 25,060,263 2.2% Total 5,580,112,023 100.0% 100.0% 5,580,112,023 197,869,0693 3.5% The Government and Bank team also recognized the desirability of a more fine-grained geographical targeting beyond the region level, but wanted to leave a margin a manoeuver to permit allocation to communities that had recently been affected by severe floods. Therefore, below the region level, AGAIBs were given data at the beginning of the project on poverty level by prefecture4 to guide allocation-but were also instructed to consider flooding and access to socio-economic infrastructure as criteria in resource allocation. Below the prefecture level, there was no poverty data available to guide targeting. This issue was resolved by the development of a new poverty map and selection criteria which were incorporated into the revised PIM. This was only available in 2012, too late to influence selection of the communities under the project, but its use has been key to permitting more granular targeting in the follow-up project. Support to school feeding was targeted geographically and categorically to schools meeting the following criteria: (i) geographically isolated or difficult to access schools; (ii) schools in disaster affected areas (floods, fire); (iii) schools prone to child trafficking due to proximity to borders; (iv) schools with high drop-out rates; (v) informal or community schools without adequate infrastructure or teachers; (vi) schools with low attendance rates; and (vii) schools with high levels of gender disparity. The selection of schools meeting these criteria was made by the regional offices of the Ministry of Education, Ministry of Social Action, and the AGAIBs, and then validated by the steering committee. As with the community sub-projects, resources for LIPW were allocated geographically across regions on the basis of each region's contribution to poverty in the 2006 QUIBB. Below that level, communities were selected based on (i) poverty level; (ii) need for environmental rehabilitation; (iii) lack of economic opportunities; (iv) ability to undertake maintenance; and (v) availability of workforce. At the beneficiary level, community selection of beneficiaries and the wage rate of FCFA 1350 (USD $2.78) per day aimed to discourage less poor people from working. An analysis of the effectiveness of this mechanism was not possible due to the lack of resources for a randomized control trial impact evaluation. 3 Total for this column is the sum of the variation for each region. 4 Each region contains between four and nine prefectures. 16 The Enhanced Agricultural Production component specifically targeted farmers who were members of production groups, who had at least one hectare of agricultural land, and who could pay cash up front for half of the inputs supplied. The aim was to target the most productive farmers in the respective production groups; it is likely therefore that this component did not benefit the poorest. C. Empowering communities to identify and directly implement project activities Table 7: Outcome indicators Relating to Community Empowerment End of Actual Actual as Project outcome indicator Baseline implementation September percentage target 30, 2013 of target Number of communities that following 0 350 584 169% sensitization and training have then presented a proposal which was approved by the A GAIB approval committee. (Original indicator dropped in June 2010 restructuring) Minimum number of subprojects which 0 280 290 104% reflect the priorities of targeted communities and beneficiaries at MTR and at end ofproject (Refers to socio-economic infrastructure sub-projects. Original indicator dropped in June 2010 restructuring) Given the fragile state context and the desire to ensure community participation, the project design called for community associations to play the central role in the choice and implementation of sub-projects. As in all CDD projects, training of communities was of key importance to ensure the success of the approach. Community groups received training on community driven development methodologies, project identification and preparation of a project proposal, procurement, and financial management. Training and coaching of communities was provided by the regional AGAIBs and NGOs recruited to assist communities. As a prerequisite of being approved, all projects had to be identified by communities after receiving training and submitted by communities for approval to a regional approval committee composed of the AGAIBs, local government officials, and sectoral Ministry officials. All 584 socio-economic infrastructure and income generating sub-projects followed this process, exceeding the target of 350. All 290 socio-economic infrastructure projects reflected the priorities of targeted communities and beneficiaries, exceeding the target of 280. This was verified by the end of project beneficiary evaluation (see Annex 5) which found that 99% of a sample of 2,152 community members surveyed felt that the infrastructure sub-project was among the community's top priorities. 17 At the end of the project, an assessment of the long-term retention of trained community members was carried out. Retention of concepts around the role of the community and basic concepts of community driven development was strong, with 95% of community members passing a short quiz on key concepts around this theme. Retention of concepts relating to financial management was high at 97%. As topics became more complex, the number of beneficiaries retaining key concepts decreased but was still significant; 68% retained knowledge on project implementation and monitoring and 59% on procurement (see Annex 2). Overall, 75% (or 1,663) of beneficiaries retained knowledge on at least 4 of the 6 themes measured, short of the unrealistic initial target of 2,100, which would have required over 95% of beneficiaries to meet this combined standard. Communities also played an important role in school feeding and labor intensive public works activities. In school feeding, Parent Teacher Associations (PTAs) managed the delivery of meals in each school, with the support of local NGOs. In public works, the project cycle operated in the same manner as for community sub-projects, with communities taking responsibility for sub- project identification (though wages were paid by a third party to avoid conflicts of interest). Table 8: Weighting of Outcome Ratings Against Original and Revised Project Objectives5 Against Against First Against Overall Comments Original Revised Second PDOs PDOs Revised PDOs 1 Rating Satisfactory Satisfactory Satisfactory - 2 Rating value 5 5 5_- Note: Disbursement 0 12.9 19.3 32.3 during period (US$ m.) 3 Weight (% disbursed in 0% 40.2% 59.8% 100 % period) 4 Weighted value (2 x 3) 2.99 2.01 5 - 5 Final rating (rounded) - Satisfactory 3.3 Efficiency Rating: Substantial Given the emergency nature of the operation and its CDD approach, no evaluation of EIRR was conducted at appraisal or ex post. Still, for most of the project components and sub-components, M&E and cost data provides an understanding of the cost efficiency of project activities. Component 1.1: Community Development Sub-Projects. The Togolese Ministry of Education's forthcoming 2014 Report on the Status of the National Education System (Rapport d'etat du systeme educatifnational-RESEN) finds that the PDC represented the least expensive approach for building classrooms in the country. The similarly community-driven Education and 5 Table follows OPCS Guidelines for Implementation Completion and Results Reports, Appendix B: Attachment to the Guidelines for Outcome Rating of Projects with Formally Revised Project Objectives (p. 44) 18 Institutional Strengthening Project (Projet Education et Renforcement Institutionnel-PERI), also administered by the Bank with financing from the Global Partnership for Education (GPE), had similarly low costs. As shown in Figure 2, the PDC's costs for classroom construction were below those for comparable infrastructure executed by ministries and other public agencies (USD 116/M2 compared to costs of USD 131-230/M2 under other programs). The report finds that delegation of procurement responsibility to communities was a key factor in keeping costs down. Figure 2: Unit costs of school construction in various programs Values standardized to 2012 prices 250 12 200 28 14 a 150 0 E 100 a) 50 0 BID II MEPSA/PERI MEPSA/PIP MEPSA/PIP EPTT UNICEF MEPSA/PERI PDC 2010 2012 2010 2012 2011 2010 2012 2009-12 Centralized approaches with national bidding Decentralized approaches with local bidding U direct costs O indirect costs Source: Togo Rapport d'tat du systime ducatif national; 2014 In regard to Income Generating Activities, the end of project evaluation of the income generating sub-projects suggests that for the more profitable enterprises, FRRs can reach 30%, but the data and sampling on which the estimate is based are fragmentary. A complete quantitative assessment of the profitability of these sub-projects was extremely difficult due to beneficiaries' tendency to underreport their revenues to the implementing agency for fear of taxation and to make the case for further support. Overall, 80% of participating groups provided at least some return (in cash or kind) to their members. As mentioned in the previous section, the end of project evaluation of the income generating sub-projects found that 75 percent were still supporting improved livelihoods for beneficiaries, with 58% 'working well' and 17% 'working moderately well.' 25% were found to be structurally unprofitable or abandoned. Qualitatively, interviews with beneficiaries found that 78% thought that they had improved their livelihood through the sub-projects, with 12% reporting their situation being unchanged and 10% reporting 19 being worse off. Beneficiaries also reported that the incomes generated meant they did not have to make high-cost short term borrowing. Component 1.2: School Feeding. Costs per child for school feeding were US$57 per year under the project for a daily hot meal averaging 1,200 Kcal. Adjusted to 700 Kcal per day and a 180- day school year for international comparison purposes, this equaled a standardized cost of $28 per student per year6. For comparison, Bundy et al (2009) found that in 2006, before the global food price crisis, the range of similar on-site feeding programs was estimated to be between US$28 and US$63 per year depending on the type of food provided, or about US$40 on average. Given that global food prices approximately doubled between 2006 and 2010, according to the World Bank Global Food Price Index, the costs of the school feeding program appear to compare well to international standards. Component 1.3: Labor Intensive Public Works. For the LIPW activity, the primary objective was a cash transfer to the beneficiaries, with creation or rehabilitation of physical assets as a secondary objective. In this context, the primary indicator of efficiency was the percentage of resources spent on LIPW sub-projects allocated to wages. This indicator exactly met the target of 70 per cent of the total cost going to wages (Annex 2). The remaining 30 percent primarily financed materials and equipment. Component 3: Management, operating costs and audits. Implementing agency operating costs were reasonable in relation to investment costs, with AGAIB project management expenses on target at about 15% of disbursements for sub-projects (see Annex 2). This appears to be in line with norms from other CDD projects. For example, an IFAD study comparing administrative costs of CDD projects found costs ranging from 10 to 24 percent, with average administrative costs of 16 percent. Component 4: Enhanced Cereals Production. Cost efficiency of the provision of agricultural inputs is difficult to ascertain due to the lack of randomized selection of a treatment and control group which would permit the isolation of the effects of the intervention. Nevertheless, project administrative data puts the average yield of producers of rice and corn receiving the inputs at 2.90 and 2.03 tons per hectare respectively. This compares favorably to the average annual yields over the past ten years which have ranged from 1.57 to 2.83 for rice and 1.15 to 1.29 for corn. 3.4 Justification of Overall Outcome Rating Rating: Satisfactory The relevance of the project to Togo's situation and strategic priorities was high at the time of appraisal and remained high throughout implementation. The project was highly appreciated by Government as the Bank operation that had the strongest impact on poor populations throughout the country. The PDO, as measured by achievement (or more) of a good majority of the key indicators, has been met. Efficiency can be measured by fewer indicators - and the lack of data is a shortcoming - but on the basis of what information can be gleaned and through qualitative 6 Andrews et al (2011), p. 25. 20 assessments, efficiency is rated as satisfactory. An overall 'Highly Satisfactory' rating could have been justified had there been greater clarity on the operation's impact and efficiency. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development Generally, the benefits reached the target group of the poor and vulnerable, because the project was conducted in areas that, according to the QUIBB, were poor. Geographical targeting faced a lack of data at lower levels (see 2.2 above) but there is no evidence that this led to significant skewing or leakage towards the non-poor. However, in the absence of impact evaluation, no clear evidence exists on poverty reduction or distributional impacts other than the a priori reality that activities benefitting the health, educational and nutritional status of poor and vulnerable populations and increasing their incomes will have a poverty reducing impact. An impact evaluation would have allowed useful extension of this generality, for example by quantifying leakage or inequitable distribution of benefits and by quantifying which activities had the most poverty-reducing impact and which were the most cost-effective. The Enhanced Agricultural Production component was carried out in the same poor areas but specifically targeted farmers who were members of production groups, who had at least one hectare of agricultural land, and who could pay cash up front for half of the inputs supplied: it is likely therefore that this component did not benefit the poorest. The proportion of female beneficiaries reached 44% as opposed to the target of 50%. The shortfall in regard to female beneficiaries was largely due to girls constituting only 42% of students in the classrooms built, and 40% of workers on labor-intensive public works. The percentage of female students in the constructed classrooms was below the national average, likely because of the focus on rural areas where girls, particularly poor girls, are less likely to be in school. Given the large proportion of schools among the infrastructure built, the target of 50% female participation was overambitious in the absence of any complementary measures to boost girls' enrollment. The lower than targeted female participation in LIPW was likely due to the types of infrastructure selected, with women tending to participate significantly less than men in road rehabilitation, despite some measures such as on-site child care aimed at boosting the numbers of women workers. CVDs were asked to have at least one woman in a leadership post, and the beneficiary survey found that this was the case in 97% of CVDs. However, overall CVD membership was not sufficiently representative: only 68% of CVDs had women make up at least one quarter of their members against a target of 90%. This calls for a stronger emphasis on quotas to enforce women's representation in the future. (b) Institutional Change/Strengthening The project was initially set in the context of the Government's and Bank's priority of strengthening institutions in support of local economic and social development (see 1.1 above). The important training component (USD 1.7 million) was designed to strengthen capacity at all 21 levels, and the involvement of all stakeholders in the implementation of this CDD project also had an important capacity building impact. Although no evaluation data are available regarding capacity building at community level (other than that 9,400 community members were trained, with a 75% success rate, success being measured by a test of trainee's knowledge and competence following the training), the CDVs and associations touched by the project certainly acquired enhanced capacity for all phases of local development and management and for dealing with public and NGO partners. The capacity of the AGAIB and the PCU was also significantly built up, although some of this capacity - as well as motivation - was reduced as a result of the 2012 staff lay-offs (see 2.2 above). The follow-on PDCplus project is benefitting from this capacity. One question is whether the time is right to transform the PCU into a permanent organization, and a related question is the future of the AGAIB (see Section 6 below on these two questions). (c) Other Unintended Outcomes and Impacts (positive or negative) None. 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops At the end of the project, a beneficiary survey was undertaken to measure beneficiary satisfaction with different aspects of the project, as well as to assess the extent to which community members had retained knowledge around key themes of community training, and to measure the numbers of beneficiaries of infrastructure sub-projects. Beneficiary satisfaction for sub-projects was extremely high. Globally, 98% of beneficiaries were satisfied with the sub-project, of which 55% were very satisfied and 43% satisfied. 94% of beneficiaries felt that the project belonged to the community (4% answered that it belonged to the AGAIB and 2% to the village chief). 97% were satisfied with the financial management of the sub-project and 95% were satisfied with the quality of the infrastructure (37% very satisfied and 58% satisfied). Results from the beneficiary survey regarding community training are discussed in Section 3.2, and detailed numbers given in Annex 2. 4. Assessment of Risk to Development Outcome Rating: Significant The principal risks to sustaining the development impact of the project are: Without continuation of capacity building support, the CDVs, groups, PTAs etc. may have problems in sustainable O&M of assets created. Risks stem from the short term nature of NGO and project support, which prepares communities to maintain the assets before their completion but does not provide continued support in the operations phase after the start-up of operations. The typical input to capacity building of community institutions has been two years of contracted support during conception, implementation and start-up of operations. This may not be adequate 22 for the types of activity considered (O&M of community social infrastructure, income generating activities, school feeding). There is a risk, too, that CVDs will not be able to raise the revenue needed to pay for O&M. Related to the above, there is a risk that the momentum of community-led development planning and management may falter without follow-up support, structured medium and long term community-led development planning and further investment in community-selected sub- projects. The bottom-up development dynamic which has been created may be at risk. There are specific risks of sustainability attached to Income Generating Activities. The failure rate is already 25% or more (see 3.2 above) and global experience shows that access to advisory services into the medium term may be needed to sustain these businesses. There is a risk to sustainability if follow up training and trouble-shooting are not provided. Prospects for the sustainability of the School Feeding program are brighter than before the project due to Government's recent commitment to begin providing funding for school feeding. Combined with Bank support in the follow-on PDCplus project, support for school feeding is assured for the coming years. However, if government financing cannot be sustained, the school feeding program would be at risk. Although PTAs and local structures can play an important part, continuation of the program in these poor communities will certainly require government financing, and also continued follow up from the AGAIB. Specific risks also attach to the assets created by the LIPW activity. Arrangements for management and maintenance are essential to sustainability. Even with a convention (see 2.3 above), the Ministry of Environment and Forest Resources may have difficulty in financing and executing management on so many small, scattered plots. Finally, there is a risk to the overall institutional and operational arrangements for CDD in Togo as long as the apex organization remains an ad hoc PCU. The future of the AGAIB also remains a question (see Section 6 below). 5. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory The Bank team correctly identified this project operation as the best way to help Togo rapidly within the agreed strategies of the I-PRSP and ISP. Design was well founded on local and global experience and on existing implementation capacity. Measures were built in to the project to ensure that implementation capacity would be strengthened, and that safeguards and fiduciary requirements were respected. The operation was appropriately processed as an emergency project. The components were well aligned with the PDO and targeted outcomes, and the flexible design allowed further modules to be added over the two years after appraisal. The risk assessment was well conducted. 23 The Bank faced a difficult dilemma in responding to the urgent needs related to the global economic and financial crisis. On one hand the Government expressed demands for support in regard to multiple discrete activities, all of which were important priorities. On the other hand, incorporating all of these activities in one project created the risk of an overly complex project that could have been difficult to implement and supervise. The Bank judged that the urgent needs outweighed the risks, and sought to make the community-driven approach the common thread linking together the different activities supported by the project (except the Enhanced Cereals Production component). The correctness of this choice was borne out by the project's ultimately satisfactory performance. There is an argument to be made that the Enhanced Cereals Production component could have been designed in a way that could have promoted a more sustainable impact or been more poverty-targeted. However, the nature of the activity - emergency response to the food crisis - precluded a more cautious approach. More attention could have been paid in the design phase to monitoring and evaluation, especially given the pilot nature of some of the activities. The initial Results Framework lacked indicators that would have tracked the key outcome of improved access to services, though at the time of project preparation the institutional emphasis within the Bank on outcome indicators was not as strong as it was by the end of the project, or indeed even by the time of the second Additional Financing. Overall, the Bank team played a key role in ensuring that the operation was designed to achieve the agreed strategic objective and that these objectives were reflected in the PDO. Fiduciary and safeguard requirements were well integrated. Taking account of the relatively minor shortcomings on risk assessment and M&E, Bank performance on assuring quality at entry is rated as satisfactory. (b) Quality of Supervision Rating: Satisfactory The Bank fielded frequent supervision missions well-staffed with all appropriate competences. Aide memoires and ISRs were full and well prepared, and management provided clear guidance throughout. A thoroughgoing mid-term review was carried out. The complexity risk was mitigated by frequent supervision, the presence of a consultant on the ground, and regular communication by phone and email with the Technical Secretariat. The frequent contact and close support was a constant throughout supervision despite turnover of TTLs. Supervision of safeguards and fiduciary aspects was thorough and ensured that extra training was provided and that the minor problems which arise were quickly corrected. Bank missions provided continual hands-on support to Togolese agencies and were appreciated by those agencies as training and technical assistance. 24 Bank team leaders kept a constant eye open for opportunities to strengthen impacts through addition of further activities. On the basis of project success, the Bank team recommended to management a major expansion of the Community Subproject and School Feeding activities. This proactive approach in fact resulted in the appraisal of a follow-on project ('PCDPlus'). One shortcoming was the slow response to getting M&E underway. The Bank team might have acted more resolutely to get the system operational earlier in the project life. Another shortcoming was the limited linkages with other Development Partners. Particular emphasis was placed in the original 2008 Project Paper on donor partnership, particularly with UNDP, the EU and French Cooperation, and it was envisaged that the project would be implemented in close collaboration with these partners, within the overall donor framework of re-engagement with Togo. In the event, there was little evidence of the proposed collaboration, and opportunities may have been missed to associate other Development Partners in financing (for example in the recent negotiation of the EDF financing program). A further issue that arose towards the end of the project period was several changes in TTL responsibility, which may have impaired the continuity and knowledge base of the Bank's inputs. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory The Bank played the key role in designing this project and ensuring quality at entry, and provided proactive and hands-on supervision of high quality and frequency and with continual focus on development impact. With ratings of Satisfactory on both Quality at Entry and Supervision, the overall rating is Satisfactory. Only relatively modest shortcomings kept the rating below the top level of Highly Satisfactory. 5.2 Borrower Performance (a) Government Performance Rating: Satisfactory The two Government ministries which, successively, oversaw the project provided adequate support. They convened the steering committee responsible for oversight, approval of work plans and budgets etc. One shortcoming was the lack of engagement with other potential sources of finance. (b) Implementing Agency or Agencies Performance Rating: Satisfactory The PCU's role was primarily to assure the fiduciary functions and to support execution through the AGAIBs. Its notable contributions included: (i) preparation of the PIM, ESMF and RMF; (ii) implementation of the training component; (iii) support to implementation of the Community Sub-Projects; (iv) development and implementation of the procurement plan; (v) developing and implementing sound financial management and accounting procedures; and (vi) direct implementation of the School Feeding activity. The AGAIB's role was primarily to support implementation by community organizations. 25 Management of the PCU and the five AGAIB was generally sound and professional. Staffing was adequate, although some posts were slow to be filled. The 2012 decision to retrench staff (see 2.2 above) reduced the AGAIB's supervision and capacity building performance and slowed up PCU procurement. Overall, although the project became relatively complex in terms of numbers of activities and of agencies involved, implementation was relatively trouble free and most output targets were met or exceeded. Much of this is attributable to the energy and quality of the PCU and of the AGAIB. Again, the slow start-up of M&E was a shortcoming, although the PCU partially compensated through an intensive effort towards the end of the project to ensure collection of all needed results data. (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory This project represented a challenge for Togolese implementation structures, given relative lack of experience with CDD approaches and the increased technical, safeguards and fiduciary risks inherent in the degree of decentralization of decision-making and implementation to community level. The addition of further activities, some of a pilot nature and all demanding in terms of their implementation and supervision, increased the challenge. The Togolese institutions nonetheless managed to implement the project satisfactorily and achieve almost all the targeted outputs. An overall rating of Satisfactory is justified. 6. Lessons Learned The project was highly effective in providing a quick response to Togo's difficult situation after fifteen years of socio-economic decline. The experience showed that rapid impact on poverty in such a situation can be achieved by a CDD approach backed up by an institutional structure that empowers local communities by decentralizing decision taking and strengthens their capacity through sustained support and capacity building. The principal lesson is that effective poverty reduction through community-level initiatives requires well-structured support and implementation mechanisms together with intensive supervision and capacity building. In the particular case of Togo, the resulting recommendation is that the Government and Bank now need to focus on how to institutionalize, strengthen and finance the functions currently carried out by the central (Technical Secretariat) and regional (AGAIB) organizations to provide the continued investment finance, institutional and technical support and capacity building necessary to sustain the CDD momentum in communities already supported and to extend the CDD approach to all poor communities nationwide. 7 The Bank team has already been examining these issues proactively. At the national level, the PDC Technical Secretariat seek over time to transfer its capacity to its parent ministry, for example by twinning arrangements where Ministry staff will learn on the job from TS staff. In regard to the regional level, a recent report for the Bank (Theunynck 2013) examined three options for the future of the AGAIB: (1) sustainability by diversifying financing sources; (2) converting the AGAIB into private entities and giving them a more 'business' orientation; and (3) conversion to a decentralized development agency. These options were explored for other countries, including Benin, Senegal and Madagascar. See: La Delegation de Maitrise d'Ouvrage en Afrique en 2007. Bilan, enjeux et perspectives. Christian Diou, Michel Henry, Babaly Deme, PPIAF, Republique du Senegal. Avril 2007. 26 A more specific recommendation derives from the relatively high level of risk to the development outcomes (see Section 4 above). This is to ensure sustained follow up, capacity building and (ideally) further investment for the communities and activities successfully supported under the present project in order to ensure the sustainability of project impacts and to maintain the undoubted CDD momentum created. Although the rationale for the addition of the enhanced cereals production activity was persuasive, the development impact was limited. The lesson is that the benefits of such a component can be enhanced if it is integrated within an ongoing support program to agriculture, and the recommendation is that as much effort as practicable should be afforded to ensure integration within programs for sustainable development of farmer services and production. The Bank faced a balancing act between keeping the project design simple, particularly in the low administrative capacity environment, and responding to the many pressing needs. As the project progressed, the project became increasingly complex as additional activities were added to respond to urgent needs. Additional financing was preferred to preparation of new projects for speed, in order to piggyback on the ongoing project's satisfactory performance, and in order not to create a proliferation of projects in the Togo portfolio. Fortunately, the added complexity in this case did not affect the project's eventual satisfactory performance, but it did create added risks. The lesson is that selectivity is difficult for the Bank in situations such as this where the Government may not have many alternative avenues to address urgent needs other than the Bank. The recommendation is to ensure in cases where selectivity is difficult that the project teams have the supervision resources necessary to manage such complex projects in low capacity fragile environments where extra implementation support can be neecsary. 7. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies The Borrower prepared a thorough completion report (September 2013). The report's findings, lessons and recommendations (see Annex 7) are fully consistent with those of the Bank and of the present ICRR and do not require further specific comment. (b) Cofinanciers Not applicable. (c) Other partners and stakeholders Not applicable. 27 Annexes Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD Million equivalent) Approved (USD Actual (USD Percentage of millions) millions) Approved Component 1: Community Development 23.2 24.03 104% 1.1 Community Development Sub- 12.5 11.58 93% Projects 1.2 Emergency School Feeding 6.3 7.78 124% 1.3 Labor-Intensive Public Works 4.4 4.67 106% Component 2: Training 1.7 2.05 121% Component 3: Management and Operating 3.0 2.73 91% Costs Component 4: Enhanced Cereals 5.0 3.53 71% Production Total Cost 32.9 32.33 98% (b) Financing Type of Approved Actual Percentage of Financing (USD millions) (USD millions) Approved IDA Grant H4090 Grant 17.2' 16.39 99.5% GFRP Grant TF-93124 Grant 7.0 6.94 99.1% CRW Grant H6020 Grant 8.79 9.0 99.9% Total financing 32.9 32.33 98.3% 8 The grant of SDR 10.6 million was estimated at USD 17.2 million equivalent at negotiations, but was reduced by currency fluctuations to USD 16.4 million. 9 The grant of SDR 5.8 million was estimated at USD 8.7 million equivalent at negotiations, but was increased by currency fluctuations to USD 9.0 million. 28 Annex 2. Outputs by Component Component 1 Community development projects, school feeding and temporary employment operation Component 1.1 (a): Socio economic infrastructure sub-projects Table A2.1: Intermediate outcome indicators (Component 1.1 (a)) Intermediate outcome indicators End of Actual Comments implementation September target 30, 2013 Intermediate outcome - increased availability of basic socio-economic infrastructure Basic socio-economic infrastructure 200 290 sub-projects completed (number) Classrooms built or rehabilitated 360 621 (number) Health faculties constructed, renovated 40 23 Variance explained by: (i) and/or equipped (number) community requests below expectations, and (ii) the requirement that health personnel be available was not always satisfied. Improved community water points 30 76 constructed or rehabilitated under the project (number) Basic socio-economic infrastructure 90% 96% According to the 'audit assessed as having satisfactory technical technique' quality (%) Table A2.2: Distribution of infrastructure sub-projects Sectors Maritime Plateaux Centrale Kara Savanes Total Education 42 41 29 32 49 193 Health 3 3 4 5 8 23 Water and 8 10 7 6 6 37 Sanitation Socio-cultural and 2 1 0 1 1 5 others Rural roads 8 1 4 4 1 18 Socioeconomic 4 0 5 4 1 14 TOTAL 67 56 49 52 66 290 29 Component 1.1 (b): Income generating sub-projects Table A2.3: Intermediate outcome indicators (Component 1.1 (b)) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - increased availability of income-generating opportunities Income generating activities subprojects 150 294 (number Table A2.4: Distribution of income generating sub-projects Region Type of activity Number Market gardening (maraichage) 6 Buying and selling of grain 9 Plateaux Grain production 15 Livestock 33 Other 7 Sub-total Plateaux 70 Market gardening (maraichage) 6 Grain production 3 MaitmeLivestock 19 Fish 5 Food processing 12 Other 9 Sub-total maritime 54 Market gardening (maraichage) 18 Grain production 20 Buying and selling of grain 15 Milling 1 Livestock 27 Other 8 Sub-total Kara 89 Market gardening (maraichage) 2 Grain production 3 Buying and selling of grain 19 IAAE Milling 11 Livestock 3 Other 8 Sub-total Savanes 46 Market gardening (maraichage) 6 Buying and selling of grain 6 CENTRALE Food processing 10 Livestock 8 Other 5 Sub-total Centrale 35 TOTAL 294 30 Component 1.2: Emergency School Feeding Table A2.5: Intermediate outcome indicators (Component 1.2) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - increased scope and efficiency of implementation ofschoolfeeding program Meals provided to primary school 12.1 million 17.0 million students (number) Schools successfully implementing the 180 256 school feeding program (number) Schools providing school children with 50% 0 TS explains 'the one fruit per day (%) allowance for the meal was insufficient for fruit to be offered' Table A2.6: School feeding: Evolution of schools and number of students Maritime Plateaux Centrale Kara Savanes Total Year 1 : school year 2008-2009 Beneficiary schools 20 25 15 17 15 92 Beneficiary schools 4.200 4.603 3.090 2.927 3.983 18.803 Year 2 : school year 2009-2010 Beneficiary schools 19 24 15 17 15 90 Beneficiary schools 4.317 5.252 3.316 3.353 5.141 21.379 Year 3 : school year 2010-2011 Beneficiary schools 36 42 31 47 26 182 Beneficiary schools 9 091 8 979 5 835 6 780 7 969 38.654 Year 4: school year 2011-2012 Beneficiary schools 37 42 31 47 26 183 Beneficiary schools 9091 8105 6259 6810 8117 38382 Year 5 : school year 2012-2013 Beneficiary schools 58 53 49 61 35 256 Beneficiaryschools 16070 11193 10785 11268 11618 60934 31 Component 1.3 Labor-intensive Public Works Table A2.7: Intermediate outcome indicators (Component 1.3) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - effective implementation ad targeting ofpublic works program Person-days of work provided (number) 1,000,000 1,139,979 Participants below the poverty line (%) 75% Unknown Proposed impact evaluation did not take place Female participation (%) 50% 40% Young people (under 35) employed (%) 75% 58% Total cost allocated to wages (%) 70% 70% Public works schemes completed with 75% 66% Other sub-projects are satisfactory technical quality (%) (reforestation) reported to be operational but 'their technical standards cannot be assessed' (TSreport) Annual net increase in cash income of 30% Unknown Proposed impact targeted participants compared to evaluation did not take control group place Table A2.8: Regional distribution of THIMO operations Maritime Plateaux Centrale Kara Savanes Total Regional contribution to national 24% 23% 15% 18% 20% 100% poverty Forestry component (area 484 462 302 363 403 2015 reforested) Lumber 240 230 150 180 200 1000 Fuel wood 184 176 115 138 153 765 Fruit trees 60 57,5 37,5 45 50 250 Non-forestry component 18 17 11 14 15 75 Retaining ponds, etc. 6 6 4 4 5 25 Ruralroads 12 11 7 10 10 50 Labor 6.000 5.750 3.750 4.500 5.000 25.000 32 Component 2: Training Table A2.9: Intermediate outcome indicators (Component 2) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - CVD members successfully trained in sub-project management (simplified financial and procurement management, reporting, M&E etc.) CVD are inclusive (at least 25% of the 90% 68% Indicator refers to % of members are women), representative CVDs with 25% women and well-trained members CVD members successfully trained 2,100 1,663 Table A2.10: Statistics of communities trained Category of persons Maritime Plateaux Centrale Kara Savanes Total trained Members of the CVD 1371 1001 822 1725 408 5327 trained Members of groups 759 505 40 168 2207 4079 trained Total number of persons 2130 1506 862 1893 2615 9006 trained Proportion numerical 808 793 318 349 1521 3789 of women trained percentage 37.93 52.65 36.89 18.43 58.16 42.07 Component 3: Project management Table A2.11: Intermediate outcome indicators (Component 3) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - Technical Secretariat, Project Steering Committee in place and functioning AGAIB project management expenses 15% 15.1% as a % of disbursements for sub-projects IFR submitted by AGAIB to the TS on a 90 90 monthly basis within 15 days following the end of the month (number) IFR submitted by TS to the Bank within 18 18 45 days from the end of each quarter (number) 33 Component 4: Enhancing production of cereals Table A2.12: Intermediate outcome indicators (Component 4) Intermediate outcome indicators End of Actual Comments implementation September 30, target 2013 Intermediate outcome - Agricultural inputs are purchased and distributed in view ofproducing 14,000 tons ofadditional maize output by the end ofsecond year ofproject implementation Fertilizer procured and distributed to 4,275 4,275 farmers (tons) Seeds procured and distributed to 370 858 farmers (tons) - Of which commercial seed 356 844 - Of which foundation seed 14 14 Table A2.13: Distribution of selected producers Regions Number of villages Number of farmers' Number of covered organizations producers Maritime 165 246 1836 Plateaux 355 456 4570 Centrale 151 203 2188 Kara 219 369 3265 Savanes 146 198 1915 Total 1036 1472 13774 Table A2.14: Fertilizer type and distribution by region Type of Quantity of fertilizer by region (in tons) fertilizer Maritime Plateaux Centrale Kara Savanes Total Fertilizer NPK 575 873 873 644 390 2 850 151515 UREA 46%N 87 434 434 322 198 1 425 Total 862 1307 1307 966 588 4275 34 Table A2.15: Location of production and grain yields Crop Area Production(t) Yield (t/ha) planted (ha) Maritime Rice 215 645 3,00 Corn 1.621 2.983 1,84 Plateaux Rice 432,05 1.512 3,50 Corn 3.828 8.805 2,30 Centrale Rice 52,50 154 2,93 Corn 2.125 4.575 2,15 Kara Rice 389,25 817 2,10 Corn 2.818 4.931 1,75 Savanes Rice 309 927 3,00 Corn 1.506 2.861 1,90 Sorghum 100 200 2,0 Rice 1.397,8 4.055 2,90 Total by crop Corn 11.898 24.155 2,03 Sorghum 100 200 2,0 TOTAL 13.395,80 28.230 35 Annex 3. Economic and Financial Analysis Given the emergency nature of the operation and its CDD approach, no evaluation of EIRR was conducted at appraisal or ex post. A summary estimation of rates of return for Income Generating Activities (Ouedraogo 2013: 28) suggests that for the more profitable enterprises, FRRs can reach 30%, but the data and sampling on which the estimate is based are fragmentary. Three indicators of efficiency were tracked. First, implementing agency operating costs were reasonable in relation to investment costs, with AGAIB project management expenses on target at about 15% of disbursements for sub-projects (see Annex 2). This appears to be in line with norms from other CDD projects. For example, an IFAD study comparing administrative costs of CDD projects found costs ranging from 10 to 24 percent, with average administrative costs of 16 percent. The second indicator concerns unit rate costs, where a 2013 evaluation by the Ministry of Education found that Project investment costs for classroom construction were below those for comparable infrastructure executed by ministries and other public agencies (USD 1 16/m2 compared to costs of USD 13 1-230/m2 under other programs). For the LIPW activity, a third indicator measured efficiency as the proportion of labor costs to total costs, where the result exactly met the target of 70 per cent (Annex 2). 36 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility! Names Title Unit it / Specialty [Lending [Hugues Agossou Sr Financial Management Specia AFTME I Itchi Gnon Ayindo Senior Procurement Specialist AFTPW I Joseph Baah-Dwomoh Consultant AFTA1 Paola Ciardi Consultant AFTH2 - HIS Yvette Dan-Houngbo Economist AFTP4 Homa-Zahra Fotouhi Representative MNCDJ [Paul Geli IConsultant MNSSP Daniele A-G. P. Jaekel Operations Analyst AFTHW [Nathalie S. Munzberg - Senior Counsel LEGEN Sylvie Nenonene Communications Officer AFRSC Rodrigo Serrano-Berthet Senior Social Development Spec LCSSO Chantal Leontine Tiko Program Assistant AFMTG Marietou Toure Diack Program Assistant HRDTA Maurizia Tovo jLead Social Development Specia ECSSO Christina A. Wood Senior Economist MNACE Giuseppe Zampaglione ICountry Manager LCCHN Thijs de Jong IConsultant AFTH2 - HIS Supervision/ICR Giuseppe Zampaglione Country Manager LCCHN TTL Maurizia Tovo Lead Social Development Specia ECSSO TTL Anne Mossige Sr Social Protection Specialist AFTSP - HIS TTL [John Van Dyck Sr Social Protection Specialist AFTSW TTL Gbetoho Joachim Boko Social Protection Specialist AFTSW Hugues Agossou Sr Financial Management Specia AFTME Itchi Gnon Ayindo Senior Procurement Specialist AFTPW Ousmane Badiane Lead Rural Development Special AFTA1 Elena Celada Consultant AFTSW Carlo Del Ninno Senior Economist AFTSW Paul Geli Consultant MNSSP Laurent Guerin Consultant AFTSE Serge Herve Houssou Consultant AFTEE Alain Hinkati Sr Financial Management Specia AFTMW Esinam Hlomador-Lawson Program Assistant AFMTG Guy Kossi Doumegnon Lakpo Information Analyst AFRIT Janine E. Mans Junior Professional Associate AFTP3 Sylvie Nenonene Communications Officer AFRSC F. Alain Onibon Agric. Economist AFTA1 Andy Chi Tembon Senior Health Specialist AFTHE Chantal Leontine Tiko Program Assistant AFMTG Carolyn Turk Country Manager AFMRW 37 (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle No. of staff weeks USD Thousands (including travel and consultant costs) Lending FY08 4.3 75.7 FY10 4.8 44.0 Total: 9.1 119.7 Supervision/ICR FY09 15.6 125.5 FY10 18.5 130.1 FY11 11.5 135.2 FY12 14.8 79.4 FY13 17.5 65.4 FY14 8.2 34.2 Total: 86.1 569.8 38 Annex 5. Beneficiary Survey Results At the end of the project, a beneficiary survey was undertaken to measure beneficiary satisfaction with different aspects of the project, as well as to assess the extent to which community members had retained knowledge around key themes of community training, and to measure the numbers of beneficiaries of infrastructure sub-projects. Beneficiary satisfaction for sub-projects was extremely high. Globally, 98% of beneficiaries were satisfied with the sub-project, of which 55% were very satisfied and 43% satisfied. 94% of beneficiaries felt that the project belonged to the community (4% answered that it belonged to the AGAIB and 2% to the village chief). 97% were satisfied with the financial management of the sub-project and 95% were satisfied with the quality of the infrastructure (37% very satisfied and 58% satisfied). The beneficiary survey also measured the participation of women in the community associations (CVDs) supported under the project. Unfortunately, while the project performed well in ensuring that women were represented in leadership posts in CVDs, it did not meet its goals for overall women's' participation. CVDs were asked to have at least one woman in a leadership post, and the beneficiary survey found that this was the case in 97% of CVDs. However, overall CVD membership was not sufficiently representative: only 68% of CVDs had women make up at least one quarter of their members against a target of 90%. This calls for a stronger emphasis on quotas to enforce women's representation in the future. Overall, as a member of the community association, were you satisfied with your role in the implementation of the sub-project? Very Unsatisfied Satisfied Very satisfied Sample size unsatisfied __%_ %_ _ 1% %_ _ __ _ _ Gender Male 0.43 0.93 39.77 58.87 1525 Female 1.25 1.08 50.34 47.32 676 Region Savanes 0.00 0.82 27.05 72.13 471 Kara 0.70 1.41 45.07 52.82 492 Centrale 0.00 0.00 52.50 47.50 336 Plateaux 1.42 0.00 44.68 53.90 439 Maritime 1.15 2.30 48.28 48.28 467 National 0.68 0.98 42.95 55.39 2205 39 What is your impression of the quality of the infrastructure? Unsatisfied Satisfied Very satisfied Sample size Gender Male 5.15 57.62 37.23 1521 Female 4.69 58.38 36.93 676 Region Savanes 5.74 36.07 58.20 471 Kara 2.13 31.21 66.67 489 Centrale 0.00 45.00 55.00 336 Plateaux 9.22 86.52 4.26 439 Maritime 6.90 89.66 3.45 467 National 5.00 57.77 37.23 2201 Distribution of women in CVDs according to position of leadership and region Deputy Vice General Deputy Number of President General Treasurer Total President Secretary Treasurer Women in Secretary the CVD Number Number Number Number Number Number Number % Region Savanes 2 17 4 11 49 29 112 64.00 175 Kara 0 5 10 11 50 27 103 62.05 166 Centrale 4 4 0 4 29 13 54 58.70 92 Plateaux 3 2 3 6 41 28 83 58.04 143 Maritime 4 8 3 12 24 37 88 58.28 151 National 14 36 20 44 193 133 440 60.52 727 Annex 6. Stakeholder Workshop Report and Results Not applicable. 40 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR The Borrower's ICR (September 2013) does an excellent job of describing and evaluating project design, implementation and results. The narrative is consistent with the Bank's ICR. The report highlights several issues: * The problems encountered in 'picking winners' for the Income Generating Activities and in providing the necessary flexible and sustained support. A particular issue raised is the need for longer term, paid engagement by the NGOs supporting these start-ups. * The implementation of the School Feeding program. This required constant follow-up, which proved onerous. * For the LIPW activity, the report highlights the payment difficulties and their consequences, and also the difficulty of maintaining the high labor content for certain subprojects which required mechanized civil works or high raw materials inputs. * The training component was hindered by shortcomings in the quality of some NGO services, and also by the short term nature of the input. Continued hand-holding of the community organizations over a longer period would have been preferable. * The Enhanced Cereals Production component was negatively affected by the low recovery rates and by the poor handling of the seed supply activity. * The effectiveness of the PCU and the AGAIB was impaired by the staff reductions in 2012. Lessons learned were: + The project was an effective instrument for reducing poverty. + The CDD approach empowered communities as agents for their own development. + The project was efficient in delivering infrastructure at low capital cost. + In order to judge project impacts, a working baseline and M&E system are essential. + Successful implementation was enhanced by the efficient inter-sectoral cooperation amongst agencies. + The Bank provided not only needed guidance but also hands-on capacity building, and this contributed to the successful project outcome. Recommendations were: > Standardize design for Subproject infrastructure as far as possible. > Revise the models used for latrines, as the model used during the project was not readily accepted by communities. 41 Adopt a flexible approach to the NGO contracts for Income Generating Activities in order to allow, where needed, for more sustained and longer term support. > Integrate literacy programs into the CDD approach in order to raise the level of understanding of communities. > Revise the approach for School Feeding to make it more efficient and to involve the health and education services. > Develop a long term resource mobilization strategy for the continuation of project activities. Annex 8. Comments of Financiers and Other Partners/Stakeholders Not applicable. 42 Annex 9. List of Supporting Documents Agbo Egbenovi, March 2010. Evaluation de l'Operation de Fourniture de Repas aux Ecoliers des Zones Vulnerables du Togo Andrews, Colin et al. Social Safety Nets in Fragile States: A Community-Based School Feeding Program in Togo. Social Protection Working Paper No. 1117. World Bank, August 2011. Bundy, Donald, et al. Rethinking School Feeding: Social Safety Nets, Child Development, and the Education Sector. World Bank, 2009. Dzogbedo Agbenyo, June 2013. Audit Environmental et Social des Microprojets du Projet de Developpement Communautaire (2008-2013) Elessessi Folly Azianto, September 2013. Audit Technique des Microprojets d'Infrastructures du PDC. Gamo Yawo Sitsofe, September 2013. Rapport final d'Evaluation (Beneficiary Evaluation). Minist&re des Enseignements Primaire et secondaire. Rapport d'tat du systeme educatif national; 2014 (forthcoming) Ministere du Developpement a la Base, de l'Artisanat, de la Jeunesse et de l'Emploi des Jeunes, September 2013. Rapport d'Achevement du Projet de Developpement Communautaire. Ouedraogo Hermann, August 2013. Audit Technique et Financier des Microprojets d'Activites Generatrices de Revenus Realises par les Communautes a la Base. Projet de Developpement Communautaire, January 2011. Rapport d'Audit Technique de la Composante 'Augmentation de la Production Cerealiere'. Theunynck Serge, September 2013. Note Technique. 43