Projet Éducation Pour Tous en Haïti : Rapport sur l'achèvement de la mise en œuvre et les résultats

Document of The World Bank Report No: ICR2425 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-H2860 IDA-H5880) ON A GRANT IN THE AMOUNT OF SDR 24.8 MILLION (including additional financing) (US$37.0 MILLION EQUIVALENT) TO THE REPUBLIC OF HAITI FOR A EDUCATION FOR ALL PROJECT IN SUPPORT OF THE FIRST PHASE OF THE EDUCATION FOR ALL PROGRAM November 30, 2012 Human Development Department-Education Sector Caribbean Country Management Unit-Haiti Latin American and the Caribbean Region 1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized CURRENCY EQUIVALENTS (Exchange Rate Effective August 30, 2012) Currency Unit = Haitian Gourdes (HTG) HTGl.00 = US$ 0.0238 US$ 1.00 = 42.05 FISCAL YEAR January 1 -December 31 ABBREVIATIONS AND ACRONYMS APG Adaptable Program Grant BND Bureau de Nutrition et de Developpement (Office for Nutrition and Development) CAS Country Assistance Strategy CCT Conditional Cash Transfer CDB Caribbean Development Bank CIDA Canadian International Development Agency DAA Department of Administrative Affairs DAEPP MENFP's Department for Private Education and Partnership DDEs Regional Education Departments DHS Demographic and Health Survey 2005 DPCE Planning Department DRH Department of Human Resources EFA FTI Education For All Fast Track Initiative EFACAPs Ecole Fondamentale d'Application et de 1'Appui Pedagogique (Practical Application and Teaching Support Basic School) EGRO I/II Economic Governance and Reform Operations (I/II) EPCA Emergency Post-Conflict Assistance Program EU European Union FMA Financial Management Assessment FMRs Financial Management Reports GDP Gross Domestic Product GER Gross Enrollment Rates (total number of students enrolled in a particular level divided by the total population of school-age for that same level). GIS GIS-based school mapping database GOH Government of Haiti HIPC Heavily Indebted Poor Countries ICF Interim Cooperative Framework ICT Information and Communication Technology IDB Inter-American Development Bank ii IFR Interim Financial Report I-PRSP Interim Poverty Reduction Strategy Paper IROR Internal Rate of Return ISN Interim Strategy Note ISR Implementation Status and Results Report LICUS TF Low Income Country Under Stress Trust Fund M&E Monitoring and Evaluation MDG Millennium Development Goals MENFP Ministere de 1'Education Nationale et de la Formation Professionnelle (National Ministry of Education and Vocational Training) MIS Management Information System MTEF Medium Term Expenditure Framework NEPO National Education Partnership Office NER Net Enrollment Rates (total number of students with the appropriate school-age for a particular level, divided by the total population of that age) NGO Non Governmental Organization NSFP National School Feeding Program PCF Post-Conflict Fund PCR Primary Education Completion Rate PCU Project Coordination Unit PNCS Programme National de Cantines Scolaires (National School Canteen Program) PPF Project Preparation Facility SMC School Management Committees TORs Terms of Reference UNESCO United Nations Education, Scientific and Cultural Organization UNICEF United Nations Children's Fund USAID United States Agency for International Development Vice President: Hasan Tuluy Country Director (Special Envoy): Alexandre V. Abrantes Sector Manager: Reema Nayar Project Team Leader: Patrick Ramanantoanina ICR Team Leader: Patrick Ramanantoanina ICR Primary Author: Richard J. Carroll iii HAITI EDUCATION FOR ALL ADAPTABLE PROGRAM GRANT PHASE 1 (APG-1) CONTENTS B. KEY DATES ................................................................................. V C. RATINGS SUMMARY..........................................................................V D. SECTOR AND THEME CODES..................................................................VI E. BANK STAFF ................................................................................ VI F. RESULTS FRAMEWORK ANALYSIS.............................................................ViI G. RATINGS OF PROJECT PERFORMANCE IN ISRS ....................... ............................IX H. RESTRUCTURING (IF ANY) ..................................................................... X I. DISBURSEMENT PROFILE......................................................................X1 .PROJECT CONTEXT, DEVELOPMENT OBJECTES AND DESIGN.......................................... 1. Context at Appraisal........................................................ 1 1.2. Original Project Development Objectives (PDO) and Key Indicators.................... 3 1.3. Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification ............................................................. 3 1.4. Main Beneficiaries....................................................... 3 1.5. Original Components..................................................... 4 1.6. Revised Components...................................................... 5 1.7. Other signiicant changes ................................................................................. 5 I. KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOMES ...................................6 2.1 Project Preparation, Design and Quality at Entry............................ .....6 2.2 Implementation............................................................................................ 8 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization .. ...............10 2.4 Safeguard and Fiduciary Compliance PDOn.........................................11 2.5 Post-completion OperationNext Phase.........................................12 1. ASSESSMENT OF OUTCOMES ................................................................ 12 3.1 Relevance of Objectives, Design and Implementation ............................... 12 3.2 Achievement ofProject Development Objectives .......................... ........ 13 3.3 Efficiency ............................................................. 15 3.4 Justfcation of Overall Outcome Rating .......................................15 3.5 Overarching Themes, Other Outcomes and Impacts ....................... ..... ... 16 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops.. ............. 17 IV. ASSESSMENT OF RISK To DEVELOPMENT OUTCOME.............................................. 17 V. ASSESSMENT OF BANK AND BORROWER PERFORMANCE .............................. .......... 17 5.1 Bank Performance ...................................................... 17 5.2 Borrower Performance ..................................................... 18 VI. LESSONS LEARNED......................................................................... 19 VII. COMMENTS ON ISSUES RAISED BY BORROWER/IMPLEMENTING AGENCIES/PARTNERS ................ 20 Annex 1. Project Costs and Financing ..... ........................................ 22 Annex 2. Outputs by Component ................................................... 23 Annex 3. Economic and Financial Analysis Workshops........................................... 26 Annex 4. Bank Lending and Implementation SupportSupervision Processes ............. ....... 28 Annex 5. Beneficiary Survey Results ................................................ 30 Annex 6 Stakeholder Workshop Report and Results.............................. ....... 31 Annex 7. Summary of Borrower's ICR and/or Comments on Draf ICR ............... ............ 32 Annex 8. Comments of Cofinanciers and Other PartnersStakeholders ........................... 45 Annex 9. List of Supporting Documents IR........ ....................................... 46 1V A. Basic Information - HT Education For All Country: Haiti Project Name: Adaptable Program Grant Phase 1 Project ID: P099918 L/C/TF Number(s): IDA-H2860,IDA-H5880 ICR Date: 11/30/2012 ICR Type: Core ICR GOVERNMENT OF Lending Instrument: APL Borrower: HAITI Original Total XDR 16.80M Disbursed Amount: XDR 24.70M Commitment: Revised Amount: XDR 24.80M Environmental Category: C Implementing Agencies: Ministry of Education Cofmnanciers and Other External Partners: Caribbean Development Bank B. Key Dates Revised / Actual Process Date Process Original Date (s) Concept Review: 10/05/2006 Effectiveness: 09/28/2007 09/28/2007 05/27/2010 Appraisal: 01/29/2007 Restructuring(s): 05/7/2012 Approval: 04/26/2007 Mid-term Review: 05/04/2009 Closing: 01/31/2011 05/31/2012 C.Ratings Summnay C.1 Performanc e R ag b yl CR Outcomes: MS Risk to Development Outcome: Substantial Bank Performance: MS Borrower Performance: MS C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: MS Government: S v Quality of Supervision: MS Implementing MS Agency/Agencies:. Overall Bank Overall BorrowerMS Performance: S Performance: C.3 Quality at Entry and Implementation Performance Indicators Implementation Indicators QAG Assessments Rating Performance (if any) Potential Problem ProjectYes Quality at EntryYes at any time (Yes/No): (QEA): Problem Project at any N Quality of No Yes time (Yes/No): Supervision (QSA): DO rating before Moderately Closing/Inactive status: Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) Central government administration 13 13 Other social services 8 8 Primary education 75 75 Teriay duaton4 4 The me Co de =(as %Oo oft to ta IBa nk finan cing) Administrative and civil service reform 17 17 Child health 17 17 Education for all 33 33 Social safety nets 33 33 E.Bank S taff Positions At ICR At Approval Vice President: Hasan Tuluy Pamela Cox irc Alexandre V. Abrantes (Special LCountry Director: Eno)Caroline =Anstey Envoy) Sector Manager: Reema Nayar Eduardo Velez vi Project Team Leader: Patrick Philippe Ramanantoanina Samuel C. Carlson ICR Team Leader: Patrick Philippe Ramanantoanina ICR Primary Author: Richard J. Carroll F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) To (a) improve access and equity of primary education; (b) operationalize partnerships between public and non public sectors; and (c) build capacity to assess learning outcomes. Revised Project Development Objectives (as approved by original approving authority) No change in the PDO, though the original Development Credit Agreement had erroneously used the program objective instead of the Project objective (see Section 1.2). The May 27, 2010 restructuring paper pointed this out and the Legal Agreement was amended accordingly. (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Revised Achieved at Inictor BaeieIau approval Target Completion or documents) Values Target Years Indicator 1 Children receiving tuition subsidies (person/years) Value quantitative or 0 210,000 390,000 425,000 Qualitative) Date achieved 03/21/2007 01/31/2011 05/31/2012 5/31/2012 General target met and exceeded after upward revision. The revised target included the impact of the Project and of other donor assistance catalyzed by the Comments Project. Specific outcome targets for the Project were not established. Project (incl. % attributed impact was 285,000, with an additional 140,000 children receiving achievement) subsidies through funding from other donors. At the 5/27/2010 restructuring, the indicator was changed from "Number of Poor Children benefiting from Public subsidies to attend Non Public Schools." Indicator 2 : Grade 1 capacity utilization in participating schools Value (quantitative 47% 100% /A /A or Qualitative) Date achieved 03/21/2007 01/31/2011 Comments Originally "Intake Rate for Grade 1 in Participating Schools," this indicator was (incl. % dropped at the 5/27/2010 restructuring because capacity utilization was not an achievement) issue after the January 2010 earthquake. vii Indicator 3 : Number of teachers receiving pre-scrvice teacher education Value (quantitative 0 3,600 /A N/A or Qualitative) Date achieved 03/21/2007 01/31/2011 Comments This indicator was shifted to Meeting Teachers' Needs project, which spun of from component 2 of this (APG-1) Project, and formally dropped at the 5/27/2010 restructuring. See ICR for Meeting Teachers' Needs project for data achievement) for this indicator. Indicator 4: Number of accredited non-public schools participating in per student subsidies. Value quantitative or 0 500 1,100 1,212 Qualitative) Date achieved 03/21/2007 01/31/2011 05/31/2012 05/31/2012 Comments Target met and exceeded. (incl. % The indicator was changed in the 5/27/2010 restructuring from the PAD version achievement) (p.4 of restructuring paper). Share of beneficiary schools submitting use of funds reports to DEAPP and/or Indicator 5 : P~NEPO Value (quantitative 0 80% Not revised 76% or Qualitative) Date achieved 03/21/2007 01/31/2011 05/31/2012 Comments Target largely, but not fully met. Indicator changed from, "Percentage of School (incl. % Management Committees (SMCs) which submit use of funds reports to DEAPP and/or NEPO" at 5/27/2010 restructuring because of capacity constraints in achievement) SMCs. Indicator 6 : Children participating in integrated nutrition/health program (person/years) Value quantitative or 0 70,000 210,000 260,000 Qualitative) Date achieved 03/21/2007 01/31/2011 05/31/2012 5/31/2012 General target met and exceeded after upward revision. Similar to above tuition Comments subsidies indicator, target reflects impact of additional donor funds catalyzed by he Project. Project attributable person/years was 150,000, but there is no Project (incl %i tJ t i(l. specific target against which to assess that achievement. Indicator was changed achievement) at 5/27/2010 restructuring from "Students benefiting from program-financed health and nutrition programs." Number of schools where Grade 2 literacy competency test is applied (to establish baseline) Value quantitative or 0 200 86 84 viii Qualitative) Date achieved 03/21/2007 01/31/2011 05/31/2012 5/31/2012 Comments Revised target largely, but not fully met. (incl. % Indicator was modified at 5/27/2010 restructuring (p. 4 of restructuring paper). In addition, target was reduced to 86 because of higher than anticipated costs and achievemeInt) capacity constraints. (b) Intermediate Outcome Indicator(s) Original Target Actual Value FormallyromAchieved at Indicator Baseline Value Values (from Revised cmplet approval Completion or documents) Target Years Indicaorl Multi-Age Learning Program implemented in at least 100 schools Value (quantitative 0 100 N/A N/A or Qualitative) Date achieved 03/21/2007 01/31/2011 Comments (Cmmd %The sub-component to which this indicator pertains was dropped because of ac. inadequatecapacity in the implementing agency (p. 6 Restructuring paper). achievement) Percentage difference between the unit cost of the NSFP's school feeding Indicator 2: program and that of the average unit cost of 3 largest NGO school feeding roviders Value (quantitative 315% 20% N/A / or Qualitative)KA Date achieved 03/21/2007 01/31/2011 (Cmm.t% This is an efficiency measure that was dropped during the 2010 restructuring a(icivmet %because improving efficiency was not an objective of the Project. G. Ratings of Project Performance in ISRs Date ISR Actual No. DO IP Disbursements (USD millions) 1 12/20/2007 Moderately Satisfactory Moderately Satisfactory 0.20 2 04/11/2008 Satisfactory Satisfactory 2.32 3 11/18/2008 Satisfactory Moderately Satisfactory 5.43 1x 4 11/25/2008 Satisfactory Moderately Satisfactory 5.43 5 04/21/2009 Satisfactory Moderately Satisfactory 7.44 6 07/11/2009 Satisfactory Moderately Satisfactory 8.87 7 07/31/2009 Satisfactory Moderately Satisfactory 8.87 8 08/30/2009 Satisfactory Moderately Satisfactory 8.87 9 09/13/2009 Satisfactory Moderately Satisfactory 9.24 10 10/31/2009 Satisfactory Moderately Satisfactory 7.13 11 05/07/2010 Moderately Satisfactory Moderately Satisfactory 21.29 12 06/21/2010 Satisfactory Moderately Satisfactory 21.29 13 02/09/2011 Satisfactory Moderately Satisfactory 33.19 14 08/08/2011 Satisfactory Moderately Satisfactory 34.99 15 02/21/2012 Satisfactory Moderately Satisfactory 37.85 16 05/28/2012 Moderately Satisfactory Moderately Satisfactory 37.85 H. Restructuring (if any) ISR Ratings at Amount Restructuring Board Restructuring Disbursed at Reason for Restructuring & Date(s) Approved Restructuring Key Changes Made PDO Change DO IP in USD millions 1. Additional financing of SDR8 million (US$12 million) to increase beneficiaries of school feeding and tuition subsidy programs and additional institutional support to MENFP. 2. Adjust indicator targets to reflect changes in priorities, financing and Project 05/27/2010 No MS MS 21.29 environment. 3. Reallocate funds to priority activities including tuition subsidies and school feeding. 4. Physical inspections replace financial reporting to verify tuition subsidy program. 5. Extension of project closing date by 16 months from January 31, 2011 to May 31, 2012. Reallocation of funds to adjust 05/07/2012 S MS 37.85 Grant expenditure categories in preparation for Project closing. x ISR Ratings at Amount Restructurit Board Restructuring Disbursed at Reason for Restructuring & IOBRs) Approved Restructuring PDO Change DO IP in [SD millions Categories I and 2 were slightly overdrawn because of higher than expected participation in tuition subsidy program Categories 3 (school health and ndtrition) and 4 (reinforcement of MENFP's technical capacity) were lower than expected. The latter because expenditures on post-earthquake capacity building were lower than initial estimates. .Disbursement Profile Original Formally Revised -- Actual 40- 30- .20- 0- 11111a1 ** d dd d- d d * XI I. Project Context, Development Objectives and Design 1.1. Context at Appraisal 1. At the time of appraisal, 78 percent of Haitians were living on less than US$2 a day, and more than half living on less than US$1 a day,1 making Haiti the poorest country in the Latin America and Caribbean (LAC) Region. Living conditions were particularly difficult in rural areas, where poverty and extreme poverty rates were estimated to be even higher. With an infant mortality rate of 76 per 1,000 births, illiteracy of 47 percent, and the highest incidence of HIV/AIDS outside Sub-Saharan Africa, Haiti ranked 153 rd of 177 countries worldwide in the United Nations 2005 Human Development Index. 2. The country was emerging from several years of political, economic and social strife. From 1996 until the time of appraisal, the country has suffered from weak or negative GDP growth, accompanied by mounting political and social strife exacerbated by the departure of Jean-Bertrand Aristide in February of 2004. Governance of the sector was also a disadvantage. There was a lack of procedures and transparency in the use of public funds, and poor coordination across technical departments and of externally financed investment projects. Many trained MENFP civil servants left the system for opportunities abroad or as externally-financed consultants. The MENFP's basic capacities to conceptualize, plan, execute, monitor and evaluate educational programs in a coordinated manner were weak. Reliable up-to-date education and demographic statistics for each region and nationwide, essential for planning purposes, were not available to MENFP staff. 3. Until the preparation of the Education for All (EFA) Adaptable Program Grant- Phase 1 (APG-1), 2 the Bank had not been engaged in the Haiti education sector for eight years. In 2006, after many years of political turmoil, Haiti held democratic elections and launched an I-PRSP that focused on expanded access to education and health services. The I-PRSP informed the strategic pillars of the Bank's Interim Strategy Note. In the four years leading up to appraisal, Haiti prepared a National Strategy for achieving Education For All. This strategy was developed in partnership with the United Nations Education, Scientific and Cultural Organization (UNESCO), the Bank, the United Nations Children's Fund (UNICEF) and other donors. This ten-year plan placed Haiti on a path toward achieving Education For All by 2015. 4. The available data for the education sector revealed weak outcomes. The primary net enrollment rate was 71 percent in 2006, with a gross enrollment rate of 127 percent and a primary level completion rate of 66 percent. An estimated 500,000 children aged 6-11 did not attend school of any kind, and only about half of all six year olds enrolled in first grade. 3 More than 80 percent of all primary level students attended non-public schools, financed mainly by parents and religious associations, NGOs and other sources, while the remainder attended virtually free public schools. Non-public schools had become increasingly distant from the GOH and were not as concerned with accreditation 1 IMF Article IV Consultations, August 10, 2010. 2 EFA and APG-1 are used interchangeably and refer to the same project. 3 DHS Survey 2005. 1 standards. The Living Conditions Measurement Survey revealed that 43 percent of all parents (50 percent for parents in the lowest income quintile) did not send their children to school because of the costs of school tuition, books, uniforms and transportation for the private-based education system. 5. Quality of education. Curricula were outdated and not relevant for over-age students who made up the bulk of the primary education population. The APG-1 project appraisal document (PAD) described teaching practices as "chalk and talk", requiring students to recite words and phrases they frequently did not understand. In addition, about 75 percent of all teachers lacked adequate training with many having just a 9th grade education, with no formal teacher training. 6. Affecting both quality and access was the state of children's nutrition as approximately 40 percent of Haitian children suffered from a caloric deficit of an average of 460 kilo-calories per day. 4 Many schoolchildren were commuting long distances from home to school, often without eating any breakfast. At the same time, school feeding programs in Haiti were declining in coverage, due to cutbacks from key international donors, from 800,000 children in 2002 to about 400,000 in 2006. 7. Rationale for Bank involvement. The EFA grant was consistent with the Bank's role at both sectoral and macro-economic levels to assist the Government in implementing its national strategy for Education For All. The Bank's policy dialogue and ongoing support for education governance reform under the Low Income Country Under Stress Trust Fund (LICUS TF) contributed to preparation of the Project's activities toward improved public-private partnerships (PPP) in the education sector. The Bank also used its comparative advantage in development of demand-side education sector financing mechanisms. Through the APG-1, the Bank intended to generate additional funds from other donors, which included US$10 million of parallel financing from the Caribbean Development Bank (CDB). Lastly, the reputational risk for the Bank of not re-engaging in the education sector after eight years was also a rationale for Bank involvement. 8. Complementarity with other projects. The APG-1 complemented other World Bank investments, such as the Rural Water Supply and Sanitation Project, Roads and Territorial Development Project, and the Community-Driven Development Project, which aimed to provide tangible improvements in living conditions. APG-1 also helped the Bank maximize synergy and leverage of its other operations in Haiti to support education. The second Economic Governance Reform Operation (EGRO II) and the Heavily Indebted Poor Countries (HIPC) Initiative included important education policy and programmatic reforms. The Meeting Teachers' Needs project (effective 9/26/2008) and the Emergency School Reconstruction project (effective 6/25/2009) later complemented APG-1, bringing supply-side interventions of more teachers and schools to the demand-side interventions of tuition subsidies and school nutrition programs. Thus, the Bank supported multi-pronged approach to improve access to primary education. 4 Height for age is a strong predictor of school enrollment and is positively associated with poor cognitive function, school attendance, and school performance. 2 1.2. Original Project Development Objectives (PDO) and Key Indicators 9. Program Level PDO: To improve access to primary education for poor children aged 6-12, while improving equity, quality and governance of the education sector. 10. Project Level PDO: To (a) improve access and equity of primary education; (b) operationalize partnerships between public and non-public sectors; and (c) build capacity to assess learning outcomes. 11. The Development Credit Agreement inappropriately quoted the program objective rather than the Project objective. The ICR uses the PDO in the PAD for the purpose of assessment. The key indicators, along with their revisions, are presented in Table 1. 1.3. Revised PDO (as approved by original approving authority) and Key Indicators, and reasons/justification 12. Several of the PDO indicators were revised as shown in Table 1. The Legal Agreement was amended to correct the error in the wording of the PDO. The impact of additional funds from both the Bank and other donors are reflected in the scaled up targets for the tuition subsidy and school feeding programs. The Grade 2 literacy baseline was to be established in fewer schools (86 rather than 200), a target that was reduced because of capacity constraints in schools. The two dropped indicators are explained in the table. TABLE 1: PDO Indicators and Revisions Original from PAD Revised w/ Additional Financing-5/27/10 135,000 children benefit from primary Children receiving tuition subsidies (person education enrollment subsidies. yrs.)--390,000 Grade 1 Capacity Utilization in participating Dropped-capacity utilization is not an issue schools increases to 100% (access). post-earthquake At least 3,600 new teachers receive pre-service Dropped-teacher training is delivered through teacher education. the Meeting Teachers Needs project-P106621 At least 500 accredited non-public schools Number of accredited non-public schools receive per student subsidies allocated by multi- participating in per student subsidies-- 1,100 stakeholder committees. (public/private partnership) At least 80 percent of beneficiary schools submit Share of beneficiary schools submitting use of use of funds reports to DAEPP and/or NEPO. funds reports to DEAPP and/or NEPO-80% (public/private partnership) At least 25,000 schoolchildren participate in Children participating in integrated integrated nutrition/health program. (quality) nutrition/health program (person yrs.) 210,000 Application of Grade 2 literacy competency test Number of schools where Grade 2 literacy in at least 200 schools (capacity to assess competency test is applied (to establish learning outcomes), to establish baseline. baseline)-86 Note: Original indicators from PAD Annex 3, pp. 45-6. 1.4. Main Beneficiaries 13. The main beneficiaries of the Project included: * Students receiving a tuition waiver: Original-135,000 (number of students by year 3, or 210,000 student-years), revised-390,000 (student-years). 3 * Students receiving free school food: Original-70,000 (students-years), revised- 210,000 (student-years). * Teachers receiving training (2,500) and improved curriculum (5,000-10,000). * Agencies that received institutional capacity building including the Ministere de 1'Education Nationale et de la Formation Professionnelle (National Ministry of Education and Vocational Training-MENFP), the MENFP's Department for Private Education and Partnership (DAEPP), Regional Education Departments (DDEs) and Programme National de Cantines Scolaires (PNCS). * School Management Committees (SMCs) at schools receiving tuition waivers benefiting from financial management and reporting training: Original-500, revised- 1,100. 1.5. Original Components 14. APG-1 was the first phase of a planned three-phase program consisting of the following components: Component 1: Improving Access to Primary Education * The provision of tuition waivers for primary school students in private schools. The tuition waivers paid by the Project cover various school costs particularly the cost of teacher salaries, as well as school materials. * Training programs for pre-service teachers. * Strengthening SMCs' capacity for, inter alia: (a) the identification and preparation of tuition waiver subprojects; (b) financial management; (c) supervision of selected school-level projects, including quality improvement projects and school feeding programs; and (d) school management under increased transparency and accountability procedures. Component 2: Improving the Quality of Primary Education * The development of multi-grade learning programs to support learning in rural areas, including all training and all monitoring and evaluation procedures. * The improvement of schoolchildren's performance through: (a) the provision of support for the steady and efficient delivery of school feeding services, including a package of micronutrients supplementation and hygiene education that promotes balanced nutrition, clean water and high sanitary standards, both at school and at home; and (b) the provision of School Feeding Grants to Beneficiary CBOs, including SMCs and NGOs, to finance community-driven School Feeding Subprojects in the poorest areas of the Recipient's territory. Note: This subcomponent is also referred to as the school canteen program and the school health and nutrition program. * The improvement of literacy skills development in the first two grades for public and non-public schools and the improvement in the competencies in primary education curriculum for grade 1 through 6 through, inter alia: (a) the provision of textbooks in local Creole language, and all associated training in relation thereto, including the training of trainers; (b) the strengthening of MENFP's capacity to monitor the acquisition of literacy skills and related quality of teaching and learning at the primary level; and the training of teachers in the utilization of textbooks to be provided under the Project. 4 Component 3: Improving Institutional Governance of MENFP and the Education Sector * The strengthening of the institutional capacity of MENFP to manage public funds and complete public procurement in a transparent and fully accountable manner, through the modernization of: (a) its department of administrative affairs and department of human resources; and (b) the electronic communication between MENFP's central and departmental levels. * The provision of support for the establishment, institutional capacity building and operation of the National Education Partnership Office (NEPO) as a public entity responsible for channeling public funds to non-public SMCs following transparent, rigorous, equitable and monitorable eligibility criteria. * The provision of support for the strengthening of the capacity of MENFP to; (a) increase coordination and effectiveness between its technical departments at central and regional levels; (b) develop and maintain a GIS-based school mapping database; and (c) carry out evaluation and accrediting activities for non-public schools through its department for private education and partnership. Part 4: Project Coordination and Evaluation * The Provision of support for the establishment and operation of the Project Coordination Unit (PCU) within MENFP to be responsible for Project coordination. * The carrying out of studies to evaluate Project impact, including the qualitative and quantitative impact of: (a) the implementation of Student Enrollment Subprojects; and (b) the delivery of school feeding programs and their contribution towards enhancing the quality of education. 15. The components were designed to provide quick benefits to primary schools students, as well longer term benefits in the form of improved governance and implementation capacity at central and local levels, and an increased supply of trained teachers. 1.6. Revised Components 16. There were several component revisions made in the May 27, 2010 Restructuring/Additional Financing, which was a Level 2 restructuring and therefore did not require Board approval. The pre-service teacher training (component 1) evolved into a new self-standing project, as other IDA funds became available. Targets associated with teacher training (e.g., 2,500 teacher graduates per year) were assigned to the new project (see ICR for Meeting Teacher Needs Project, Report number ICR00002426). The multi-grade learning program (component 2) was never launched as there was little government ownership for the initiative. The pilot school feeding program (component 2) to be run by SMCs was not implemented because of capacity constraints. 1.7. Other significant changes 17. The other main changes that occurred in the May 2010 Project restructuring were: (i) additional financing of US$12 million to scale up the school feeding and tuition subsidy programs; (ii) Reallocation of funds from the pilot school feeding program and multi-grade learning subcomponents to the tuition waiver and school canteen 5 components; (iii) scaled up targets for tuition subsidies and school nutrition; (iv) physical inspections replaced financial reporting to verify tuition subsidy program; and (v) extension of Project closing date by 16 months from January 31, 2011 to May 31, 2012. The second restructuring in May 2012 was a minor reallocation of funds across components to accommodate higher expenditures for tuition subsidies and slightly lower expenditures for health and nutrition and capacity building at MENFP post-earthquake. II. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 18. The APG- 1 was a high risk operation in a high risk country, but with potentially high returns. The challenges to implementation of APG-1 were high not only because Haiti is a fragile state and because the Bank had not been engaged in Haiti for eight years, but also because implementation depended on a careful sequence of steps. The tuition subsidy depended on a number of important factors: (i) the proper functioning of the public financing mechanism that applied school eligibility rules; (ii) proposal evaluation/ranking criteria; (iii) transparent, multi-stakeholder mechanisms to choose participating schools; (iv) use of the commercial banking system to make transfers to school management committees; (v) extensive public information campaigns; and (vi) independent external monitoring of use of financial transfers. The Project was set up to deal with these challenges through the development of the Subsidy Program Operational Manual, which is part of the APG-l's overall operational manual. The subsidy program operational manual included funding criteria, review procedures, and measures to ensure both transparency and accountability in subsidy allocations. 19. The Project was designed with a PCU embedded within MENFP, with a technical unit consisting of a program coordinator and four sub-coordinators, each one responsible for a specific Project component and (ii) an administrative unit with staff responsible for the Project's financial and procurement matters, all established within MENFP's existing structures. This type of PCU was risky because of the low capacity and high turnover at MENFP. 20. The Project's activities depended on the operationalization of public-private partnerships (PPPs), given that 80 percent of schools were non-public. This strategy aimed to create a space for dialogue between public and non public sectors, and to make non public schools accountable to government. Operationally, it meant that the Project would use public funds to support private entities, such as private schools and NGOs, in the provision of educational and nutrition services. The Project was prepared with the expectation that a National Education Partnership Office (NEPO) would be created that would oversee public private partnerships (PPP) in education including the tuition subsidies, as well as the school nutrition program. 21. The APG-1 was designed in close partnership with the CDB. The Bank-financed and CDB-financed projects used the same project management team and project implementation procedures, including harmonized fiduciary arrangements, project reporting procedures, and audits. The Bank handled procurement non objections for the whole program. The Project was also set up in a way that the PCU could manage expanded activities for tuition subsidies with funds from other donors such as IDB. 6 22. The APG-l was implemented in partnership with the Education For All Fast Track Initiative, which, by design, involved all major education donors. This included a single set of monitoring and evaluation indicators by which to measure the Government's inputs and outputs in achieving the national EFA strategy, and a common reporting framework for tracking all education donor activities and financing. This design resulted in one of the Project's major contributions, which was to offer a mechanism to channel a larger amount of funds to Haiti from donors. 23. To ensure relevance of Project activities, a Social Assessment was carried out in September/October 2006, including interviews and focus groups with key stakeholders in Port-au-Prince as well as over 400 stakeholders in four departments: Artibonite, Centre, Ouest, and Nippes. Widespread support and demand for the proposed interventions existed among all consulted stakeholders. Potential benefits raised by local stakeholders of each of the three proposed interventions include, among others, increased enrollment, reduced financial burden on households, improved student performance, reduced dropouts, more stable school budgets, and improved teaching quality. 24. The tuition subsidy component, by far the largest component, was designed to be implemented as a two-tranche disbursement. Tranche 1 for 60 percent of the tuition subsidy amount for each school was released to schools based on a preliminary list of qualifying students provided by each school. The second tranche for the remaining 40 percent was to be disbursed on the basis of submission of a satisfactory financial management (FM) report from the SMC, which had been trained in financial reporting. This was a logical approach that allowed schools access to critical funding to begin the school year, but then to be accountable later in the school year for the number of students it had claimed for the tuition waiver. Actual payments to schools were made by the Department of Administrative Affairs (DAA) to school bank accounts. Withdrawals from accounts required the signature of both the school director and the SMC representative. 25. The accelerated teacher pre-service professional development program also was challenging. It required a new curriculum, new teaching and learning materials, re- training of professors employed at teacher training institutes and a change in organizational culture, all of which had to be done in sequence. It was designed based on Bank experience (i.e. teacher training programs in Burkina Faso, Guinea, and Senegal) that accelerated pre-service teacher training programs are effective when coupled with strong student-teaching programs so as to ensure that student-teachers are applying their training knowledge to teaching practices. The process was endorsed by key officials in the MENFP as well as by the teachers' unions, non public education providers and other education donors. UNESCO also helped with technical assistance for this component. 26. Key lessons learned from previous operations in Haiti were applied in the Project design, including: (i) there was untapped potential of public-private partnerships to improve access to social services; (ii) the GOH had capacity, in the case of the school feeding program, to play a normative, financial and supervisory role, while leaving actual service provision to the non-public sector; and (iii) the need to move beyond supply-side interventions to address critical demand-side constraints and immediate poverty reduction objectives. Finally, the Post-Conflict Fund (PCF) and LICUS TF grants were implemented in collaboration with other multilateral and bilateral agencies, pointing to the importance of effective donor coordination for program success and on-the-ground 7 results. The application of these lessons is consistent with the recommendations from the 2002 Haiti Country Assistance Evaluation (CAE), which argued "social services should be channeled through the private and voluntary sector where possible [private schools in the tuition waiver program and NGOs in the school nutrition program], but with safeguards to promote adequate quality and access by the poor." 5 2.2 Implementation 27. A devastating earthquake occurred on January 10, 2010 that left more than 200,000 people, 5 percent of the population, dead which included staff in the PCU, the MENFP and an unknown number of Project beneficiaries. There was widespread destruction of infrastructure, schools and government buildings, including the MENFP. Some schools financed by the Project were heavily damaged or destroyed and temporarily ceased to operate. From January to June 2010 very little could be done to move the Project forward. Implementation from January 2010 onward should be assessed in the context of this national calamity. 28. The original, pre-earthquake PCU in the MENFP created a bottleneck for Project implementation, mainly because of lack of ownership at MENFP caused by multiple changes of the Minister of Education. The earthquake compounded these problems due to the destruction of the Ministry of Education, and led to the decision to establishing a more autonomous PCU unit. After the earthquake, the PCU staff operated for several weeks from the yard of its rented property and with limited resources, as the building had suffered damage making it structurally unsound. The PCU was subsequently moved to a safer location, which was key to resuming the pre-earthquake pace of Project implementation; this new office also served as a workspace and provided key functions for some MENFP officials who could no longer find office space in the temporary post- earthquake office structures. PCU staff members lost during the earthquake were also replaced, although finding qualified staff in a context in which human resources were further constrained took several months. While the independent PCU kept activities moving forward, efforts to build ownership and capacity in the MENFP waned. However, there was not a viable alternative to reliance on the external PCU in the short- term post-earthquake context. 29. NEPO. The NEPO, which had been planned as a key institution to oversee PPPs, was never operationalized. The reasons were capacity constraints and a lack of political commitment. Although Parliament adopted the law authorizing the establishment of NEPO (Loi Creant et Organisant l'Office National de Partenariat en Education - September 10, 2007), there was infighting over who would have the power to decide which schools should receive tuition subsidies, which proved crippling to efforts to establish the NEPO. Therefore, the per student subsidy program continued to be implemented through the Department for Private Education and Partnerships (DAEPP), in accordance with the Project's contingency plan. The program was implemented under the guidance of a nine-member Steering Committee, which had a similar membership to what would have constituted NEPO (had it been established), but which did not have as 5 Haiti Country Assistance Evaluation, page 21, 2002. 8 much managerial capability. Prior to the 2010 restructuring, the DAAs, DDEs and the PCU had difficulties compensating for the absence of the NEPO. The transmission of use of funds reports from schools to the PCU was time intensive, and the PCU was not able to process use of funds reports adequately, delaying the disbursements of the tuition subsidies, which were at the heart of the Project. 30. Tuition subsidy payments. The tuition subsidy program was valuable post- earthquake because it had pre-established a mechanism to mitigate the impact of the earthquake on families. However, in practice, the original tuition subsidy disbursement process did not function well and needed to be modified during Project implementation: (i) despite Project-financed training sessions, SMCs were still incapable of producing financial reports in a timely manner and needed substantially more time and resources to build the necessary capacity to do so; (ii) there were serious delays in disbursing the tuition subsidies, even for the first tranche, with many schools reporting that they received the grants when the school year was more than half over. The delays in the tuition payments caused hardships for the schools because teachers were paid months late, textbooks could only be purchased on credit and school rehabilitation often had to be stopped for months. A number of schools reported to the ICR team that they never received a second tranche. The explanation given later by the PCU was that some schools had overstated the number of students who were eligible for the tuition subsidy, which led to an overestimate of the amount of the first tranche. In the 2011/12 school year, approximately 120 of the 1,100 participating schools had overstated the number of students. 31. The 2010 restructuring eased the disbursement problem by simplifying the disbursement mechanism whereby the release of the second tuition subsidy tranche was based on a technical audit by independent firms rather than use of funds reports from SMCs. The technical audit approach facilitated the tuition waiver program by compensating for the lack of reports from the SMCs through verification of, inter alia, the number of tuition-subsidy-eligible students in class. This adjustment improved the transfer of tuition subsidy tranches, but they were still delayed until often well after the school year had ended. 32. Early Grade Reading Assessment (EGRA) and Capacity Building. The EGRA provided a basis to assess primary student learning outcomes and was the first assessment of its kind in Haiti. Although the number of schools participating was reduced (from 200 to 86), the number was adequate to provide a meaningful baseline for assessment. The MENFP took advantage of the opportunity to be closely involved in the EGRA exercise and developed capacity to carry out further assessments. Additional capacity was built at the MENFP through the provision of office equipment and training. However, most of the improvements to physical capacity were destroyed in the earthquake. In addition, after the earthquake, there was much turmover in the MENFP with staff leaving the ministry to work outside the GOH, particularly to work in the hundreds of NGOs that were providing humanitarian relief to Haiti. 33. Project risks and mitigation. A significant number of Project risks that were foreseen at appraisal were mitigated during implementation. For example, the Project design was correct in providing a fallback if NEPO was not established even though it had been endorsed by teachers' unions, parent associations and non-public education service providers. The mitigation plan was to apply objective criteria for school subsidy 9 allocation, along with enhanced transparency at ministerial and school level in how schools were selected for the program. Targets for tuition subsidy were met and exceeded even if there were significant disbursement delays. There was the risk that public education spending increases would be insufficient to fund the student subsidy program that was initially financed under Project. 34. One of the Project's major accomplishments was the additional US$60 million in resources that the PCU reported it channeled for the expansion of the tuition waiver and school nutrition programs. The Project, through the PCU, was able to facilitate large amounts of additional resources to the poor during a time of urgent need (the post- earthquake environment). The Project PCU provided a mechanism that allowed alignment of donor procedures, harmonization of objectives and indicators, and joint supervision and assessment. These additional funds also mitigated the risk that the GOH would not establish a medium-term expenditure framework for education, which it did not. The additional funds from CIDA, CDB, IDB, as well as the EFA Fast Track Initiative and APG-2 funded an expanded government program of tuition subsidies covering nearly one million additional tuition waivers. 35. The mitigation of the risk of weak SMC capacity was addressed by the capacity- building program for over 500 SMCs each year, using NGOs with demonstrated experience in this area, for all school years to be covered under Phase One. However, this capacity building proved insufficient, as despite the training sessions, SMCs continued to submit use of funds reports that were late and/or did not meet minimum quality standards. This capacity building is being continued under APG-2, as it is likely a more sustained and intensive training effort is necessary. The prospects for improving capacity in the MENFP within the life of APG-1 could not recover from the physical damage from the earthquake. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization Monitoring and evaluation of outcomes/results 36. Design. Data availability in Haiti was and is extremely poor. The country does not have a valid census, so it is difficult to know with any precision even how many children there are in Haiti. In that challenging context, the design of the M&E framework was still mostly relevant to the Project objectives, but had to rely mainly on output indicators, for example the number of children benefiting from tuition subsidies or the school nutrition program. The indicators partly captured access to education by measuring the number of students receiving the tuition waiver and the school feeding program. Because the tuition waiver is paid only to private schools, and the school nutrition uses NGOs, the number of students participating is also a measure of the operationalization of partnerships between public and non-public sectors. 37. The access indicators did not, however, allow a clear assessment of the extent to which the beneficiaries of the tuition waiver would have gone to school in the absence of the program. However, the design of the program did address this issue by providing for an impact evaluation that would better assess progress toward the PDOs. There was also a weakness in that the indicators did not sufficiently link between the provision of the school canteen program and improved school attendance, which is a benefit claimed in the PAD. The framework was also somewhat weak in measuring the capacity to assess 10 learning outcomes, and additional evidence on capacity needed to be collected at the ICR stage. 38. Implementation. M&E was carried out by the PCU supported by regular supervision missions. Quarterly reports were prepared by the M&E specialist and submitted for review to the Project Advisory Committee, which was responsible for taking necessary steps to improve program implementation. Semester financial management reports (FMRs) were prepared and submitted to both the World Bank and Caribbean Development Bank, providing information regarding both financial and physical execution. These reports helped track the implementation of the tuition waiver and school nutrition programs as well as other Project activities. The impact evaluation was not carried out because of disruptions relating to the earthquake. Thus, the weaknesses in evaluating Project outcomes were not remedied, which left gaps that had to be filled with additional information (for example, stronger links between nutrition and school attendance and additional evidence of the impacts of the EGRA) at the ICR stage. 39. Utilization. The information gained from results monitoring helped define key Project changes. Results indicators were modified at the first Project restructuring in 2010 with the main changes being the scaling up of targets for the tuition waiver and school canteen programs. The new targets reflected impacts not only from the Project, but also from other sources of financing. The M&E indicators had measured the success and high demand for these activities and provided quantitative justification for the additional funding of US$12 million. 2.4 Safeguard and Fiduciary Compliance 40. Information collected through a number of studies and as part of the Interim Cooperation Framework, the Economic Governance Reform Operations (EGRO I and II) supported by the Bank, and the HPIC Assessment carried out by the IMF and the World Bank, pointed out significant weaknesses in the public financial management systems, with respect to budget execution, accounting and reporting systems, and internal and external control processes. 41. Given the fiduciary risks, the implementation arrangements of the APG-1, including procurement and financial management, could not rely fully on the country's procedures. The DAA had overall fiduciary responsibility of the APG-1, while some components were implemented by school directors. SMCs were not able to play a major role in financial management because capacity could not be built up fast enough for implementation of the tuition subsidy program. 42. For the Project as a whole, the GOH complied with FM requirements with submission of regular financial reports. Annual audit reports were often delayed, but no significant breaches were reported, despite the loss of qualified staff during the earthquake and an inefficient computerized financial system. There were no significant issues regarding procurement. Activities were implemented according to the procurement plan. Most of the Project resources were for the payment of tuition subsidies with some contracting for NGOs for the school nutrition program, so the Project did not require a great deal of procurement. 43. The Project was an environmental Category C and no safeguards policies were triggered. 11 2.5 Post-completion Operation/Next Phase 44. The APG-1 was the first of a planned three-phase program to improve access and quality of education for 6-12 year-olds. APG-2 is currently under implementation (closing date of June 30, 2015) and is continuing the main initiatives of APG-1, in particular the school canteen, tuition subsidies, teacher training and institutional capacity building. APG-2 also has a far greater resource envelope than did APG-1 (US$70 million vs. US$37 million with the additional financing). III. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation Ratings: Before Restructuring: Substantial After Restructuring: High Overall: High 45. Objectives. The objectives of the Project continue to be highly relevant and support Haiti's priority to achieve universal primary education as articulated in the Declaration of General Policy of the Prime Minister-May 2012 (Enonce de Politique Generale du Premier Ministre). They also continue to be in line with Haiti's national strategy for rebuilding the education system, the Operational Plan for Education 2010- 2015, which calls for financing across 9 strategic themes including the provision of support to basic education. The Project's core activities contribute to Strategic Objective 3 of the Interim Strategy Note (ISN) FY12-13, building human capital. They also continue to be relevant to the cross-cutting objective of improving governance and capacity building across Haiti's public sector by building institutional capacity in the education sector. 46. Design. The Project design is also highly relevant to the current national priorities before and after restructuring as evidenced by the demand-driven scaling up in APG-1 of the two main activities, the tuition subsidy and school canteen, as well as the teacher training program (which was part of the original design, but became a standalone project), which are continuing in APG-2. The basis for the continued demand for these activities is the still large income barrier for poor students to attend tuition-based schools. In addition, free school lunches continue to be extremely popular in attracting students to school. The EGRA continues to be an important learning assessment tool in Haiti. 47. Implementation. Changes made in the 2010 restructuring improved Project relevance by adapting its design and scope to conditions on the ground. Specifically, institution of the technical audit to replace SMC financial reports facilitated the implementation of the tuition subsidy program. In addition, the restructuring responded to the increased student demand for tuition subsidies and school nutrition. Therefore, rating following restructuring improves from Substantial to High. 48. Institutional strengthening activities continue to be relevant, and another round of effort to build institutional capacity is financed under APG-2. Although the multi-grade learning program was not implemented, it was not because of low relevance. There is a substantial need for this program, particularly in rural areas where there are many over- age students in classrooms. 12 3.2 Achievement of Project Development Objectives Ratings: Before Restructuring: Substantial After Restructuring: Substantial Overall: Substantial PDO: (a) To improve access and equity of primary education Access 49. This objective was pursued by tuition subsidy and school nutrition programs, which accounted for about 85 percent of Project resources. The PDO indicators used to measure achievement of this objective were: children receiving tuition subsidies (person/years), children participating in integrated nutrition/health program (person/years) and the share of beneficiary schools submitting use of funds reports to DEAPP and/or NEPO. The Project improved access by funding 285,000 person-years of tuition for students to attend non-public schools who might otherwise not have been able to attend. This was a critical achievement, especially when considering that an earlier census had indicated that approximately 500,000 children were out of school. 50. At some schools, particularly remote rural schools, virtually all of the students were subsidized under the Project. The ICR mission observed that these schools did not have other sources of funds and would not have been able to pay teachers or buy learning materials without the tuition subsidy program. While a Project-specific target was not established at the 2010 restructuring, the overall target (including other donor financing facilitated by the Project) of 390,000 person/years was exceeded with the support of the Project, as well as additional support catalyzed by the Project, as a level of 425,000 person/years was achieved. 51. There was some concern over the use of the tuition subsidy because the percentage of beneficiary schools providing use of funds reports was below target (60 percent vs. 80 percent). However, the technical audit of individual schools helped ensure that the funds were used for the payment of teachers and the purchase of learning materials, as in addition to verifying the use of funds the audits also verified the presence of teachers and a minimum number of textbooks per student in the classroom. The resources provided through the tuition subsidy program not only paid for recurrent costs such as teachers' salaries, and learning materials, but, as ICR field visits confirmed, capital improvements to most of the recipient schools. Many of the schools were in severe disrepair because of neglect and/or hurricane and earthquake damage. Many were also undersized and/or lacked chairs and desks. So the Project had a double positive effect: it made it possible for additional students to attend primary school, while also repairing and building additional classrooms in which they could learn, which also contributed to access for primary school students. Equity 52. The tuition subsidy program was implemented on the basis of "self-selection," with students voluntarily coming forward to apply to the program. The tuition subsidy was set at US$90, which is at the lowest end of primary tuition costs. Thus, only poorer students would be interested in the program and targeting was carried out through this self-selection. The fact that the tuition subsidy was targeted to poorer students in a country with a 78 percent poverty rate suggests a strong equity component to the activity. 13 It is also evidence that the students would not have otherwise been able to attend school, which supports improved access. Thus, a rating of substantial is justified. 53. With respect to school nutrition, the Project was also scaled up, achieving a total of 150,000 person-years of school lunches and snacks attributable to the Project. A Project-specific target was not established at the 2010 restructuring, but the overall target (including other donor financing catalyzed by the Project) of 210,000 person/years of school nutrition was exceeded as a level of 260,000 person/years was achieved. 54. Beneficiary schools for the school nutrition component were selected at the start of Project implementation by the National School Canteen Program based on existing poverty maps, which indicates students in areas with the highest poverty rates benefited from school canteen services. Indeed, school directors and teachers reported that many students in these schools, prior to the lunch program, were listless and had poor attendance because of lack of nutrition at home. During the ICR mission, attendance was reported to be higher and more regular and sustained, but there is no systematic data on the actual incremental attendance. It was reported in focus groups during the ICR missions that the nutrition program has been a powerful incentive to attend school. PDO: (b) To operationalize partnerships between public and non-public sectors 55. Nearly all services to beneficiaries were supplied by private entities and were overseen by government agencies. The indicator that measured this objective was "the number of accredited non-public schools participating in per student subsidies." The revised target of 1,100 schools was exceeded as 1,212 schools participated. Thus, the Project fulfilled its objective to operationalize these partnerships developing and implementing a mechanism to provide tuition payments to 1,212 private schools to support children attending school. The ICR field work found that in a number of these schools, the sole source of financing was the Project. 56. The Project also outsourced the implementation of the school canteen program to private NGOs. A total of four NGOs were able to carry out and operationalize the school canteen program. Of these four, one NGO was new to the provision of school canteen services, and served more than 12,000 students. Thus, the Project expanded PPPs by bringing an additional NGO to provide nutrition services to schools. The Project also initiated the teacher training program in which private institutions would provide, for the first time, accelerated pedagogical training for new primary school teachers. 57. PPPs were also deepened because of the Project. Private schools needed to register and become accredited in order to benefit from the government program. Therefore, these schools had to be more responsive to national priorities and standards. One shortcoming in achieving this PDO is that NEPO has not yet been established to oversee PPPs. However, the law authorizing the establishment of NEPO was approved by Parliament and the Steering Committee assured the role that NEPO would have played. PDO: (c) To build capacity to assess learning outcomes 58. The indicator that was used to measure this PDO was "the number of schools where Grade 2 literacy test is applied". The indication of increased capacity to assess learning outcomes was that 84 schools implemented the Grade 2 literacy test, which largely met the target that was revised during the first restructuring. This was an important activity in measuring progress in literacy outcomes, because a baseline was 14 established against which to measure improvement in learning outcomes. There is additional evidence that the capacity to assess learning outcomes was developed. For example, ICR interviews underscore the strong participation of the MENFP. The Grade 2 literacy test was the first time that the ministry was involved in such an exercise. The MENFP worked closely with the Bank and participating schools in the design and the supervision of the test. This evidence of improved capacity is highlighted in the study, "Haiti: Early Grade Reading Assessment (EGRA), Rapport pour le Ministere d'Education et la Banque Mondiale-Resultats en francais et en creole," April, 2010, which covered a range of issues in primary education including relative performance of boys and girls, availability of teaching materials and differences in comprehension in Creole and French. Finally, the EGRA has become a standard assessment tool financed by a number of agencies working in Haiti, including Catholic Relief Services, Save the Children, USAID and IDB. 3.3 Efficiency Ratings: Before Restructuring: Modest After Restructuring: Substantial Overall: Substantial 59. The economic analysis at appraisal considered three components to assess economic benefits: 1) primary education per student subsidies, 2) multi-age learning programs, and 3) student nutrition and health programs. Because the multi-age learning program was restructured out of the Project, only items 1 and 3 are addressed here. 60. The benefits from these components are based on additional completed years-of- schooling for participating children. These years-of-schooling are summed up to calculate a total increase in years of schooling for all program participants. A factor is applied to capture the relationship between a year of student participation in each program and the actual incremental years of schooling (0.55 years for the tuition subsidy and 0.25 years for the school canteen). A stream of lifetime income is simulated for an individual with one additional year of schooling. The marginal income gains from one additional year of schooling (US$87.21) are multiplied by the total increase in years of schooling across all participants to create the benefit stream. The actual costs of the tuition subsidy and school canteen components are subtracted to calculate net benefits and the internal rate- of-return (IRR). The assumptions used for the analysis are conservative, in particular, no real wage growth. Assumptions and net benefits streams are in Annex 3. 61. Under these assumptions, the estimated internal rate-of-return for these two main activities of the Project is estimated to be 13.0 percent (compared with an appraisal IRR estimate of 21.4 percent for all three components), and the net present value at a discount rate of 12 percent is estimated at US$3.7 million, supporting a rating of substantial for efficiency after restructuring. 3.4 Justification of Overall Outcome Rating Ratings: Before Restructuring: Moderately Satisfactory After Restructuring: Moderately Satisfactory Overall: Moderately Satisfactory 15 62. The relevance of the Project objectives, design and implementation is substantial before restructuring, but high afterwards and high overall. Efficiency is modest prior to restructuring and substantial thereafter and overall. The dropped components were minor compared with the total size of the Project. Outcomes were strong as shown by five of the six PDO indicator targets being met or exceeded. The moderate shortcomings were that proxies for equity and access had to be used for assessment. Data were not available on the economic status of the beneficiaries. However, even without this data, the high incidence of poverty in Haiti provides reasonable assurance that equity objectives were met. Table 2 illustrates how the ratings were weighted to arrive at an overall outcome rating of moderately satisfactory. TABLE 2: Achievement of PDO: Rating Summary Assessment 1 Assessment 2 Overall Sept. 2007-May 2010 June 2010-Closing Rating MS MS Rating Value 4 4 Weight 56% 44% Weighted Value 2.25 1.75 4.00 Final Rating MS 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 63. In a country with a 78 percent nationwide poverty rate and with tuition subsidies and school nutrition targeted to the poorer segments of society, especially in rural areas, nearly all of the beneficiaries were poor. 64. One of the most important impacts of the Project was that it provided a vehicle for expanded and coordinated donor assistance. This feature of the Project is important in a fragile state and proved critical in channeling resources to Haiti after the earthquake. (b) Institutional Change/Strengthening 65. There were both significant strong points and weaknesses in institutional development. The strong points included a re-establishment of government stewardship of schools through the tuition subsidy and school canteen programs, and the schools themselves because they were able to operate with Project funding and improve the physical infrastructure as well. The weaknesses in institutional strengthening include the SMCs inability within the life of the Project to develop capacity sufficiently to manage the tuition subsidy at the school level; the heavy, but necessary, reliance on a PCU that took on many ministerial tasks to keep the Project functioning after the earthquake; and the weak improvements in the DAA, when the NEPO could not be established. On the SMCs, however, the Bank and the GOH consider them important and are implementing another round of capacity building to enhance their role in supporting their associated schools. The SMCs did participate in decisions concerning the use of tuition subsidy 16 funds, often electing to improve the schools' physical structure. The attempts at institutional strengthening at central and local levels were generally below expectations and are also being followed up in APG-2. (c) Other Unintended Outcomes and Impacts (positive or negative) NA 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops NA IV. Assessment of Risk to Development Outcome Rating: Substantial 66. The risk to sustained tuition waiver and school nutrition support is substantial because it continues to rely on funding from other donors. APG-2 is ensuring that there is continued support for tuition waivers and school nutrition through 2015 and has marshaled other donor support. Moreover, there does not appear to be a serious risk that donors will abandon the program in the medium term. The President has also taken the program nationwide and the GOH plans to support the programs through a newly legislated tax (authorizing law Projet de Loi Portant Creation, Organisation et Fonctionnement du Fonds National Pour L'Education-FNE). An advance from the treasury funded the GOH-financed tuition subsidies for the 2011-2012 school year. The national program is estimated to have benefited approximately 900,000 students. The substantial risk however, is that the source of funding for this national program, both for school nutrition and tuition waivers, has not been officially established in the education budget, and its funding is not transparent. 67. At the regional level, there were some capacity improvements in the MENFP's ability to assess learning outcomes and sustainability is enhanced by the fact that other NGOs, financed by other donors, are working with MENFP in using the EGRA assessment tool. V. Assessment of Bank and Borrower Performance 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Moderately Satisfactory 68. The Bank supported a project that was highly relevant to national priorities, but that was also challenging to implement. The Bank understood that the success of the Project depended on the principle that donor activities be coordinated, a major lesson from Bank experience in fragile states. The Project design facilitated attracting and coordinating additional funds when changing circumstances substantially increased the need for Project outcomes (tuition subsidies and school nutrition). The fact that the tuition subsidy program was already designed and functioning allowed for a faster response to the country's needs post-earthquake. This achievement is particularly strong when considering that the Bank had not been engaged in Haiti's education sector for eight years. 17 69. The moderate shortcomings were related mainly to the Bank's overestimate of the potential for capacity building, in particular with respect to the SMCs, which prior to the 2010 restructuring constituted a risk to achieving access outcomes linked to the tuition subsidy component. Another moderate shortcoming was that the Bank could have been more effective in selecting objectives with hard data with which to monitor them rather than the objectives of access and equity, which relied on proxies for measurement. In any case, an impact evaluation was supposed to more fully measure this impact. Finally, the Bank could have assessed that the multi-grade learning program, while relevant to country needs, did not have sufficient government ownership to go forward. (b) Quality of Supervision Rating: Moderately Satisfactory 70. The changes in the Project that the Bank supported at restructuring were key to improving development outcomes, as they proactively adjusted indicators and targets and allowed for adjustment to shifted priorities. These changes included scaling up successful components with additional financing and modifying other components and implementation mechanisms as needed. These changes were an application of the fragile state lesson that project restructurings should refocus project resources on what is working well. In the case of the tuition subsidy, the Bank made a good adjustment in shifting the documentation responsibilities for second tranche release for the tuition subsidy from SMCs to schools and verified through technical audits. The additional financing could be used effectively, in large part, because the Bank had supported an upgraded PCU. This PCU was particularly important when the capacity of the MENFP was destroyed in the earthquake. The PCU was able to channel additional donor funds that were not part of the Project, but which also went to tuition waivers and school nutrition. 71. The Bank also worked effectively with the GOH and the PCU and implementing agencies to ensure that Project activities were back on track within a minimum of time after the devastating January 2010 earthquake, aided by the strong design that was already in place. A moderate shortcoming in supervision was that the 2010 restructuring did not establish a Project-specific target for either the tuition waiver or the school canteen programs, thus complicating the assessment of achievement of objectives. Another moderate shortcoming was that the Bank did not revisit the access and equity objectives when it was clear that the impact evaluation would not be carried out. (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 72. The overall rating of Bank performance is moderately satisfactory, consistent with the quality at entry and supervision ratings, as well as the overall outcome rating. 5.2 Borrower Performance (a) Government Performance Rating: Satisfactory 73. The tuition subsidy and school canteen programs had support at the highest level of government with the President expanding the tuition subsidy program to a nationwide 18 level and initiating legislation to fund both programs. The Ministry of Finance provided satisfactory support to the Project and was flexible in processing the additional funding from the Project. (b) Implementing Agency or Agencies Performance Rating: Moderately Satisfactory 74. The efforts to build capacity for assessing learning in primary schools were partially successful. Capacity building at MENFP for project management was less successful, which can be largely attributed to the earthquake that occurred mid-way through the Project and which destroyed much of the MENFP office infrastructure. One moderate shortcoming was the delays in signing agreements with schools and NGOs for providing school nutrition services. Another moderate shortcoming was the inability of the GOH to overcome political obstacles in appointing the Director General and managing staff of the NEPO. 75. The MENFP supervised the National School Canteen Program on a regular basis. Supervision of executing NGOs did not reveal any significant financial management or technical audit issues. A moderate shortcoming in supervision was that food quality could have been monitored more closely. In addition, some schools had significant difficulties in meeting their obligations to provide some cooking ingredients, eating utensils and fuel wood for cooking. 76. The original PCU in the MENFP created a bottleneck for Project implementation, mainly because of lack of ownership at MENFP, which was a moderate shortcoming. However, the revamping of, and relocating, the unit during the 2010 restructuring (and after the earthquake) greatly improved implementation, in particular, the processing of tuition subsidy transfers to schools. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 77. Government performance is rated satisfactory, while Implementing Agency performance is rated moderately satisfactory, and the outcome rating is moderately satisfactory, so the overall rating for the Borrower is moderately satisfactory. VI. Lessons Learned 78. Demand driven interventions can work well in achieving greater access to education and offer good opportunities for upscaling project activities in countries where provision of public services is limited. With the dominance of private schools in Haiti, the payment of primary school tuition and providing school meals removed two of the most critical obstacles to attending class. Allowing schools to nominate children for the tuition subsidy proved to be an effective way to tap the excess demand for primary education and to target poorer students. 79. Donor coordination is critical in a fragile state whose institutions may not be able to program efficiently all of the aid available to them. In APG-1, one of the most important contributions was to coordinate the large amount of resources available to education in Haiti. APG-1 was able to take on additional donor funding that was necessary to upscale the successful tuition subsidy and nutrition programs. Haiti's 19 education sector might not have been able to effectively absorb the large influx of resources were it not for the design of APG-1. 80. Achieving quick wins in a fragile state may necessitate PCU-oriented project management. The GOH wanted quick wins in the service sector in part to demonstrate that the government could deliver important social services, yet its institutions were weak. With Haiti also highly susceptible to natural disasters, it made sense to concentrate Project management in a small unit of experienced, well-qualified professionals that was well-positioned to continue Project activities. This PCU was also key in effectively channeling the additional resources that were available to the Project and from other donors after the earthquake. However, as was the case in APG-1, it is essential that the PCU remain under the authority of the ministry and follow the ministry's strategy and objectives. 81. Outsourcing provision of basic services can be an effective form of implementation in a fragile state. APG-1 expanded the use of NGOs for school nutrition and SMC capacity building and partnered with non-public schools to improve school access. The GOH itself could not have as effectively delivered these basic education and nutrition services. 82. Provision of school nutrition is important to encouraging attendance. Focus groups of parents reported that the provision of snacks and lunches at school was one of the main reasons that they sent their children to school. Some parents said that on some days of the week the school lunch is the only balanced meal their children eat. 83. If capacity cannot be built as quickly as planned, other capacities in-country may be available to implement project activities. When the SMCs could not provide the use of funds reports on which the tuition subsidies were based, the Project was able to rely on local technical audit capacity to verify compliance with the subsidy program. The establishment of a PCU staffed with capable consultants (as part of the 2009 restructuring) greatly boosted Project implementation. More importantly, after the earthquake that devastated the MENFP, the PCU served as a critical source of capacity and performed important ministerial tasks that kept Project activities moving forward. VII. Comments on Issues Raised by Borrower/Implementing Agencies/Partners (a) Borrower/implementing agencies The corrections below have been entered into the ICR: 1 -A first remark concerns the measurement of Indicator 5: Share of beneficiary schools submitting use of funds reports to DEAPP and/or NEPO. The value determined by the follow-up report in November 2011 is 76.25% instead of 60%. Then I think it would be necessary to add to the comment that this rate is also due to the fact that the subsidy for this year was paid late. The comment could also recall the decision of 2010 that the reports no longer be a condition for the disbursement of the second installment of the grant but stay in the schools for consultation as needed. 2 -Small precision on the measurement of the indicator 7 ((Number of schools where Grade 2 literacy competency test is applied (to establish baseline): the test for assessing reading skills was conducted on 84 schools instead of 86. In addition, the study was 20 conducted between 2008 and 2009, and data collection in March 2009. (see page VII report EGRA). On the other hand, the restructuring of October 2010 changed the indicator from 200 to 86, which is a reduction of 57%. This is a significant reduction regardless of the reasons for it! For this reason, the qualifier "slightly" should be removed from the expression "slightly modified" published in the comment. 3 -In Annex 4: page 27: Task team members, the Senior Education Specialist: Patrick Ramantoanina is not included. (b) Cofinanciers (c) Other partners and stakeholders (e.g. NGOs/private sector/civil society) 21 Annex 1. Project Costs and Financing Project Cost by Component-IDA and Borrower including contingencies (in USD Million equivalent) L ompou ActualLatest P e Appraisal Estimate Atulate Percentage of mponents (USD millions) Esti Appraisal millions) Improved access to basic 18.54 26.81 145% education Improved quality of basic 3.04 7.42 244 Institutional strengthening 0.89 1.80 202% Project coordination 1.54 1.69 110% PPF 1.00 1.00 100% Total Project Costs 25.00 38.19 153% Front-end fee PPF 0.00 0.00 .00 Front-end fee IBRD 0.00 0.00 .00 Total Financing Required 25.00 38.19 153% (b) Financing Appraisal Actual/Latest Type of Estimate Estimate Percentage o Cofinancing (USD (USD Appraisal millions) millions) Borrower 0.00 0.00 IDA Grant Grant 25.00 38.19153% Total 25.00 97.93 22 Annex 2. Outputs by Component Objective Target Outputs Actual Outputs Improve access 1. Access: 135,000 children benefit from 230,000 students and equity of primary education enrollment subsidies. benefited. primary education, operationalize 2. Grade 1 Capacity Utilization in Indicator dropped as partnerships betweenhipubi participating schools increases to 100 irrelevant post-earthquake. between public percent (access). and non public sectors, and build capacity to assess 3. At least 3,600 new teachers receive pre- Shifted to new Meeting learning outcomes. service teacher education. Teachers' Needs project. 4. At least 500 accredited non-public 1,212 schools benefited. schools receive per student subsidies allocated by multi-stakeholder committees. (public/private partnership) 5. At least 80 percent of beneficiary schools 60% submit use of funds reports to DAEPP and/or NEPO. (public/private partnership) 6. At least 25,000 schoolchildren participate 73,000 received nutrition. in integrated nutrition/health program. (quality) 7. Application of Grade 2 literacy 84 schools as revised. competency test in at least 200 schools (capacity to assess learning outcomes), to establish baseline. Demonstrate Execution of medium-term public education No MTF for education was Government expenditure framework, such that public prepared. commitment to education recurrent spending (FY 2007- achieving and 2009) continues to be not less than 21 sustaining EFA percent of total public recurrent spending Outputs and (minus debt service); and the share of outcomes. primary education spending is not less than 50 percent of total public recurrent spending for education. Component One: Increase At least 100,000 previously out-of-school Could not be monitored percentage of 6- children aged 6-12 are enrolled in non because of lack of data. year Olds enrolled public schools because of per student However a total of 230,000 in primary subsidy program. (Access) primary students received education in grade A tuition waivers. 23 1 in districts supported by the Project 6 Component One: Increase supply of At least 3,600 teacher-trainees receive at Activity shifted to new new primary least one year of formal pre-service teacher training project. teachers education. Components One and Two: No subsidized children over age 8 in grade Component was cancelled. Reduce numbers 1 in participating schools; of over-age students, improve Multi-grade learning programs validated by internal efficiency MENFP, printed and distributed to multi- Component was cancelled. (i.e. reduce grade students in at least 100 schools; dropout and repetition) and Pilot multi-grade learning in at least 100 improve quality of schools. Component was cancelled. learning. Component Two: Improved Literacy Baseline established regarding the Completed. Skills in Creole percentage of children in program intervention schools who can satisfactorily, at the end of Grade 2, read out loud and understand 60 words in 60 seconds Component Two: Improved primary Percentage of school under a school feeding 100% education student program covered by de- worming program; health and nutrition Unit costs of the NFSP no more than 25 conditions percent higher than average unit costs of Data not available. three largest NGOs providing school feeding. Component Three: Improved Establishment of a student subsidy Completed. partnership mechanism which includes representation of between public key education stakeholders and application and non public of objective criteria for subsidy allocation, education sectors, including respect for standards for and increased accreditation and quality; regulatory capacity Partially completed with of MENFP of non Establishment of an adequate MIS system the creation of a database at Defined as: Number of six-year old pupils enrolled in grade 1 in September of year t, divided by the total number of six-year olds in the country. 24 public education for tracking all per student subsidies and the PCU level. providers. school grants, in either DAEPP or NEPO. Component Four: Completion of quantitative and qualitative Not done because of evaluations of impact of subsidy on capacity issues and Conduct beneficiary households and schools complications due to independent earthquake. impact assessment of program interventions, including both baseline and intermediate outcomes. Overall Poverty Reduction Impact: Value of the per student subsidy as a share Insufficient data available Reduce of average beneficiary household income, as there was no household educational with at least 80 percent of subsidies going to or beneficiary survey. expenditures of households in lowest three income quintiles. poorest households in Haiti 25 Annex 3. Economic and Financial Analysis The cost-benefit streams of the tuition waiver and school canteen programs are provided below as well as the main assumptions on which they are based. Benefits from the multi- age learning program were estimated in the PAD, however the component was not implemented. Annex Table 3.1: Cost Benefit Streams for Tuition Waiver and School Canteen Total Tuition School Project Waiver Canteen Net Year Costs Benefits Benefits Benefits ($000) ($000) ($000) ($000) 1 200 0 -200 2 5,200 0 -5,200 3 4,700 0 -4,700 4 12,250 0 -12,250 5 12,700 0 -12,700 6 2,800 0 -2,800 7 0 0 0 8 0 0 0 9 0 0 0 10 0 0 0 11 0 0 0 12 0 1,439 545 1,984 13 0 4,797 1,090 5,887 14 0 8,394 1,962 10,357 15 0 11,992 2,834 14,826 16 0 13,671 3,271 16,941 17 0 13,671 3,271 16,941 18 0 13,671 3,271 16,941 19 0 13,671 3,271 16,941 20 0 13,671 3,271 16,941 21 0 13,671 3,271 16,941 22 0 13,671 3,271 16,941 23 0 13,671 3,271 16,941 24 0 13,671 3,271 16,941 25 0 13,671 3,271 16,941 26 0 13,671 3,271 16,941 27 0 13,671 3,271 16,941 28 0 13,671 3,271 16,941 29 0 13,671 3,271 16,941 30 0 13,671 3,271 16,941 IRR 13% NPV@12% $3,687.50 26 Annex Table 3.2: Additional Student-Years of Tuition and Nutrition Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Total Tuition Student 0 30 70 75 75 35 285 years (000) Nutrition Student 0 25 25 40 40 20 150 years (000) Project Costs 200 5,200 4,700 12,250 12,700 2,800 37,850 ($000) Annex Table 3.3: Education Attainment in Years (per year of participation in component) Component Additional Years Primary School per Student Subsidies 0.55 School Feeding and De-worming 0.25 Annex Table 3.4 Assumptions Cumulative inflation 2008-2011 34.6% Avg. Worker's earnings 538 percent owing to 1 year of school 0.162 $ owing to 1 year of school 87.21432 Additional years/year tuition factor 0.55 Additional years/year nutrition factor 0.25 Assume students graduate from school in 6 years w/avg. starting point in 3rd grade. Thus benefits begin in year 12. 27 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members pNames Title Unit Responsibility! Lendng; itleSpecialty Lending Raja Bentaouet Kattan Senior Education Specialist LCSHE Education Katherine M. Shafer Coleman Consultant LCSHE Education Gabriel Demombynes Senior Economist AFTP2 Economist Peter Anthony Holland Senior Education Specialist LCSHE Education Maria Lourdes Noel Senior Program Assistant LCSEN Education Luc Razafimandimby Senior Economist LCSPE Fily Sissoko Lead Financial Management Spec AFTFM Finance Co sult t AFTH3- - . Georges Solaux Consultant HIS Education Jacqueline Beatriz Veloz Lockard Program Assistant AFTFW ILockward Yao Wottor Senior Procurement Specialist LCSPT Procurement Supervision/ICR Patrick Ramanantoinina Task Team Leader-Education Spec. LCSHE Lucy Katherine Bassett Social Protection Specialist LCSHS- Education DPT Raja Bentaouet Kattan Senior Education Specialist LCSHE Education Anne-Marie Boua Consultant LCSHE Education Gabriel Demombynes Senior Economist AFTP2 Economist Valandie Kimberly Dorceus Program Assistant LCCHT Michael Drabble Senior Education Specialist LCSHE Education Peter Anthony Holland Senior Education Specialist LCSHE Education Axelle Latortue Consultant LCSHE Education Patricia E. Macgowan Consultant LCSPT Education Marsha Michel E T Consultant LCSHD Joseph Kizito Mubiru Sr Financial Management Specialist LCSFM Finance Dominique Puthod Operations Officer LCSHE Education Fily Sissoko Lead Financial Management Spec AFTFM Finance David Seth Warren Sector Leader LCSHD Education Yao Wottor Senior Procurement Specialist LCSPT Procurement 28 (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks c tsNo.dingg otravel and consultant costs) Lending FY06 0 57.86 FY07 22.8 339.81 Total: 22.8 397.67 Supervision/ICR FY07 0 10.18 FY08 16.3 217.67 FY09 22.6 179.42 FY10 22.8 407.85 FY1 1 30.3 290.7 FY12 20.5 127.59 FY13 0.6 Total: 113.0 227.85 29 Annex 5. Beneficiary Survey Results NA 30 Annex 6. Stakeholder Workshop Report and Results NA 31 Annex 7. Summary of Borrower's ICR and/or Comments on Draft ICR REPUBLIC OF HAITI EDUCATION FOR ALL (EFA) PROJECT END-OF-PROJECT REPORT Education for All Project (EPT) Port-au-Prince July 2012 32 SUMMARY ABBREVIATIONS AND ACRONYMS ..................................... ..... 34 INTRODUCTION ......................................................... 35 EXECUTIVE SUMMARY ................................................... 36 I. SITUATION AT THE START OF THE PROJECT ................................. 37 II. IMPLEMENTATION..................................................... 38 III. RESULTS AND ASSESSMENTS ........................................... 39 IV. PERFORMANCE OF THE WORLD BANK AND THE HAITIAN GOVERNMENT............ 42 V. SUSTAINABILITY ...................................................... 43 VI. PROJECT LESSONS .................................................... 43 33 ABBREVIATIONS AND ACRONYMS CIDA: Canadian International Development Agency AFD: Agence Frangaise de D6veloppement APE: Parents' Association APG: Adaptable Program Grant CDB: Caribbean Development Bank BDS: School District Office IDB: Inter-American Development Bank BUNEXE: National Office of Government Examinations CEFEF: Basic Education Teacher Training Center COSPE: Consortium of Private Sector Educational Organizations DAA: Administrative Affairs Department DAEPP: Department for Support and Partnership for Private Schools DDE: National Education Department DPCE: Directorate of Planning and External Cooperation DSNCRP: Growth and Poverty Reduction Strategy Paper ENI: Teacher Training School EFA: Education For All AIT: Accelerated Initial Training IDA: International Development Association MENFP: Ministry of Education and Vocational Training WFP: World Food Programme PNCS: National School Meals Program SNA/EPT: National Strategy-Education For All SIP: Professional placement SPR: Practical placement (classroom) UCFE: External Financing Coordination Unit EU: European Union UTP: Project Technical Unit UNESCO: United Nations Educational, Scientific and Cultural Organization USAID: US Agency for International Development 34 Introduction This document represents the end-of-project report for the EFA (Phase I). It responds a commitment by the Government of Haiti to submit a report to the World Bank on completion of the project activities. It describes the project context in order to shed light on its evolution, successes and failures, assess the involvement of stakeholders, and draw lessons that could help to improve the preparation and management of subsequent phases or of other projects in Haiti. The Education For All project (EFA) was launched in 2007, and was to have been implemented in three phases. Its goal is to help Haiti viably achieve the goal of EFA, to improve the quality of basic teaching, and to strengthen partnerships and performance in the education sector. It is consistent with the commitments made by the Government of Haiti at the World Education Forum in Dakar to achieve the objective of EFA before 2015. The first phase of the project was to extend from 2007 to 2010, according to the initial donation agreement, but was extended to May 31, 2012, following the influx of new funding, particularly from the IDA (IDA II), the EFA Fast-Track Initiative Catalytic Fund, CIDA, and the IDB (school funding). These funds supported additional activities (such as recruiting new students), and the implementation of the Kout Pous pou Tounen Lekol program, as well as extension of the project deadline to May 2012. The EFA project included the following four components: -Component 1: Improved access to basic education -Component 2: Improved quality of basic education -Component 3: Strengthening of institutions (governance) -Component 4: Project coordination and assessment 35 Executive summary The EFA project was introduced during a period characterized by increasing political stability and the beginnings of an economic upturn, despite the ongoing presence of worrisome socioeconomic issues. The earthquake that struck Haiti on January 12, 2010, crippling the country's economy and triggering a breakdown in social conditions, had major repercussions on the EFA project. The education sector faces sizeable challenges. The MENFP is in no position to regulate this sector given that approximately 90% of country's schools are privately run. Schools are either clearly lacking or run down, and there is a severe shortage of qualified teachers. The EFA project was designed and launched to meet the pressing need for improved access to quality education in Haiti, as well as a need to reinforce governance structures. It was a logical response to commitments made by the Government of Haiti at the Jomtien (1990) and Dakar (2000) conferences to achieve the goal of education for all by 2015, on the one hand, and the sector objectives defined in the DNSCRP and the SNA/EPT, on the other hand. In fact, its objective was to help achieve EFA in a sustainable manner, improve the quality of basic education, and strengthen partnerships and governance in the education sector by improving education performance indicators in Haiti. Overall, the project achieved its objectives. Despite implementation problems and the need for restructuring, the results can be rated moderately satisfactory. Certain target objectives were even exceeded. The Bank's performance is rated as satisfactory. It provided substantial support and supervision in improving the project management process, although its actions may have occasionally been perceived as too interventionist. The government's performance is rated as moderately satisfactory. It expressed renewed interest in and demonstrated concrete political support for this project, which is consistent with its goals for the education sector. This was the first time that the MENFP managed such an extensive project. However, the monitoring and coordination structures were not always effective, which had a negative impact on project follow-up and implementation. The efforts planned to strengthen institutions, particularly the transfer of knowledge through the addition of experts to the MENFP teams, have not yet yielded the hoped-for results. Several lessons were learned during implementation of the EFA project, namely with respect to strengthening institutional capacities and overcoming obstacles, the importance of communication and the involvement of decentralized stakeholders in education-related projects, the essential role of steering committees, and the need to avoid offsetting school funding. 36 I. Situation at the start of the project a) Economic and social conditions 1. At the time the EFA project was first introduced, the country seemed to have turned the page on its long history of political crises. The sense of insecurity had lessened and the economic reforms implemented were beginning to yield results. Inflation had dropped to 7.9% in 2007, compared to almost 40% in 2003. 2. However, the economic and social situation was still alarming. With an estimated GDP of US$6.2 billion, Haiti is the last remaining least-developed country (LDC) in the Western hemisphere. According to the 2004 poverty map, 55% of households were subsisting on less than $1 per day, and 76% of the population were living below the poverty line with less than $2 per person per day. The education sector was already feeling the impacts of the perilous socioeconomic situation. Families cannot always afford the high tuition costs and the State is struggling to meet the need for education. It is estimated that 80% of schools in Haiti are private. b) Priorities of the education sector 3. The priorities of the education sector were outlined in the following two documents: the Growth and Poverty Reduction Strategy Paper (DSNCRP 2008-2011) and the National Strategy-Education For All (SNA/EPT). 4. The planned changes involved redirecting educational services to poor students, improving teacher qualifications, emphasizing preschool education programs, improving the basic education system by focusing specifically on the problem of overage students and by providing families with assistance to send their children to school, adapting curricula to the socioeconomic conditions, and identifying ways to effectively oversee the education sector. c) Activities of the World Bank and other sector financial backers 5. The education sector is supported by a number of financial backers at virtually all levels, i.e. governance, improvement of programs and curricula, improvement of training for education stakeholders, support for basic education and early childhood education, vocational training, and the construction of new schools. 6. The World Bank strongly supports implementation of the SNA/EPT operational plan, and provides funding for the EFA and AIT projects, and the Emergency School Reconstruction Project (ESRP). It emphasizes basic education and early childhood education. The Inter-American Development Bank (IDB) is another important financial backer of the SNA/EPT, contributing to financing for both the EFA and the ESRP. The Canadian International Development Agency (CIDA) provides funding mainly for primary education, and the European Union (EU) finances teacher training, classroom renovations, and a regional pilot project for overage students. Other financial backers include the Agence Frangaise de D6veloppement (AFD), the United States Agency for International Development (USAID), the Spanish Agency for International Development Cooperation (AECID), the Swiss Agency for Development and Cooperation, the World Food Programme (WFP), and specialized UN organizations such as UNESCO and UNICEF. 37 II. Implementation a) Implementation measures/division of responsibilities 7. Project implementation was overseen by the MENFP, which received the necessary support from external consultants and a project technical unit, set up within the Ministry to reinforce its structures. The MENFP was responsible for planning, implementation, coordination with private stakeholders, and for managing financial resources. A monitoring and follow-up system was designed with the External Financing Coordination Unit (UCFE), the UTP/EPT and the MENFP technical departments. 8. Practically speaking, the steering committee was unable to meet on a regular basis to provide the necessary guidance, track project progress, or identify solutions to problems. For its part, the UTP not only provided support to the Ministry's technical departments, it played a central role in the project, particularly following the earthquake on January 12, 2010. b) Project preparation work 9. The project preparation work was satisfactory, and was discussed by the World Bank and the Government of Haiti to determine the best options for its implementation. The Government of Haiti wanted to take advantage of the opportunity to strengthen the MENFP, and the World Bank expressed its interest in this initiative, as opposed to creating a new project unit or strengthening an existing one. However, support measures for various MENFP departments were anticipated, either by recruiting consultants or by creating a technical unit within the Ministry-the UTP/EPT, tasked with capacity and financial management. 10. Various alternatives and approaches were studied and subsequently abandoned for valid reasons, and focus was placed on projects led by other financial backers to identify areas of complementarity rather than redundancy. The project preparation team considered the lessons learned from similar education projects in Haiti in the 1970s and 1980s, and other projects funded by the Bank in the social sector. This exercise revealed the untapped potential of public-private partnerships for improving access to social services, the government's capacity to play a prescriptive, financial and supervisory role in the National School Meals Program, while leaving service delivery to the private sector, and the need to address constraints related to the demand for services and the immediate objective of reducing poverty. 11. Moreover, the various political, technical, financial and institutional risks were emphasized, and mitigation measures identified. 12. However, the MENFP's capacities were likely overestimated; in addition, not enough advance consideration was given to the problem of consultants and public servants working together, specifically because of the differences in their fees and salaries, but also because of the different mentality of the civil service culture. c) Monitoring and assessment framework 13. The monitoring and assessment framework was designed based on general project advancement indicators. Certain tools were used to track the project: quarterly component progress reports sent to the IDA and the CDB, prepared by a UTP/EPT follow-up and assessment expert. As part of his mandate, he drafted a quarterly follow-up 38 report on the progress of the various components and subcomponents of the project, as well as progress made in implementing the annual operating report and the annual procurement plan. 14. However, the project steering committee, which was to have been responsible for project follow-up, did not actually play this role. Moreover, the initial assessment for establishing baselines planned for mid-2007, and that should have been the foundation for the final project evaluation, was not done. 15. The absence of baselines and the difficulty quantifying certain indicators made assessing the impact of actions problematic. The absence of a final evaluation means there is a lack of accurate data on results, and especially on quality. d) Restructuring and changes to targets 16. An initial budget restructuring done in 2009 led to the recuperation of a portion of the funds for components experiencing absorption problems (component 2, subcomponent A and subcomponent C), which were reallocated to the school meals program and coordination tasks. Accordingly, certain indicators needed to be revised to reflect the lower targets for the multi-grade component and application of the reading test. The project was restructured again in May 2010, and the results indicators revised. Project targets defined at the evaluation and after restructuring: Results indicators Targets at Targets after May evaluation 2010 restructuring Number of poor children receiving public funding to 135,000 250,000 attend private schools Rate for Grade 1 in participating schools 100% Number of private schools participating in the student 500 1,100 grant program, located in underserved poor areas Application of the reading test in at least 200 schools 200 86 (capacity to assess learning results) to establish a baseline Percentage of school boards that account for MENFP 80 funding Application of reading tests at the end of Grade 2, to 200 86 establish PDA baseline Students served by the nutrition/health program 25,000 115,000 III. Results and assessments a) Adequacy of the project design and objectives (based on priorities) 17. The Government of Haiti committed to the goal of EFA at the World Conference on Education for All in Jomtien, Thailand in March 1990, and then reiterated this pledge at the World Education Forum in Dakar in 2000. 18. The project objectives were relevant in that the EFA project aimed to improve access to basic education for poor children aged 6-12 while promoting equality, quality and governance in the education sector in Haiti. Therefore, it meets national priorities. 39 The project components and subcomponents were defined based on the SNA/EPT and were all focused on achieving the goals set forth in the Dakar Declaration. 19. The implementation strategy consisted in building on MENFP structures, strengthened by the support of external experts, to reinforce institutional capacities of MENFP departments while ensuring efficient project implementation. However, applying this strategy was not without its problems, and the damages suffered by the Ministry during the January 12, 2010, earthquake, as well as the need to relaunch the project in extremely difficult circumstances, meant that the initial institutional strengthening strategy to be adapted. The UTP/EPT took a more active role in the project, and procurement and financial management were entirely overseen by consultants. b) Results indicators in light of project development objectives 20. Despite a very difficult implementation context and delays encountered, overall the project achieved its development objectives since all activities scheduled for the various components and subcomponents were carried out, sometimes even exceeding target objectives. The institutional strengthening component was affected more than the others by the earthquake and therefore only achieved average results. Moreover, the absence of baselines and a final project evaluation limits the possibility of an in-depth analysis of the results obtained. Component 1: Improved, equal access to basic education Subcomponent A: tuition grants 21. 1,212 accredited private schools received a grant in 2011- 2012. The target of at least 1,100 schools was largely exceeded. Similarly, 630,368 students (2007-2012) benefited from the grant. Subcomponent B: Accelerated initial training for teachers 22. After being a subcomponent of the EFA project, Accelerated Initial Training became a project on its own in 2008. A separate report is dedicated to the AIT. Subcomponent C: Strengthening of school boards 23. Under this subcomponent, 1,236 schools were affected and 3,624 school board members were trained. Component 2: Improved quality of basic education Subcomponent A: Improved reading skills 24. The reading skills test at the end of Grade 2 was to have been applied in 86 schools to establish a baseline (capacity to assess learning results). It was applied in 84 schools between 2008 and 2009, and data were collected in March 2009 from 84 schools in Nippes and Artibonite. Subcomponent B: Nutrition/health program 25. The nutrition/health program was to have encompassed at least 115,000 children. This figure was more than doubled, specifically due to additional funding for this subcomponent. An estimated 282,720 students benefited from this program between 2007 and 2012. Subcomponent C: Improved Creole reading skills 40 26. Planned for 2007, the activities of this subcomponent effectively took place between 2009 and 2010 (evaluation test). The skills test was scheduled to be applied at 200 schools (minimum), but this was lowered to 86 given the project's progress; in the end, the test was applied at 84 schools. Component 3: Institutional strengthening 27. Institutional strengthening at the MENFP was a central objective of the project, and was the main reason why the Ministry departments were chosen to oversee the project; the goal was to improve governance in general at the MENFP. In order to carry out the planned activities under the various project components and subcomponents, the technical departments involved in the project needed support to improve their planning abilities and work environment, and to acquire equipment and upgrade the education management information system. The DAA required support to oversee the project's financial management and procurement processes. In the end, the MENFP was to have achieved a level of transparency and competence to enable it to manage funds from the IDA and CDB, and the EFA Fast-Track Initiative Catalytic Fund, and to comply with National Commission on Government Procurement (CNMP) procedures. 28. The earthquake of January 12, 2010, had highly negative repercussions on the institutional strengthening component. Other than the fact that the catastrophe led to tremendous losses of human life and equipment, it also caused a physical separation from the project support structure due to the collapse of the MENFP building. 29. However, training sessions for MENFP executives on the various actions taken led to a greater understanding of the project by the Ministry. Also, the training given to service providers better equipped the latter to present their service proposals. Subcomponent A: Strengthening of the DAA and DRH 30. Strengthening of these entities is difficult to gauge. The project's financial management and procurement processes were overseen by external consultants recruited to support the MENFP agents. The reports produced, namely follow-up and financial reports, were largely drafted by the consultants. As such, this does not reveal much information about progress made by the Ministry. Subcomponent B: Implementing the ONAPE 31. The groundwork for implementing the ONAPE was done by the UTP/EPT. However, it had no sway over political decisions in advancing this file. The law that created the ONAPE was passed by both houses of parliament and published in Le Moniteur. However, the process could not be finalized during the first phase of EFA. Subcomponent C: centralized technical departments 32. Under the institutional strengthening component, central and decentralized MENFP departments received the support needed to strengthen their logistic and operational capacities. They received ad hoc human resources support to carry out their mission. 33. The MENFP has a new Education information management system, which can address the problem of the lack of data on education in Haiti. Vehicles and motorcycles were provided to the DDE and the MENFP's central departments to help executives get around and supervise the schools more closely. The 2010-2011 school survey was done 41 using equipment provided to the DPCE. 48 BDSs received computer equipment and generators to help them operate normally, and thus monitor the schools and produce data on education. IV. Performance of the World Bank and the Haitian government a) Bank 34. The Bank team was actively involved in the project preparation work, Supervisory missions took place regularly and led to relevant recommendations on ways to improve the project. When it came time to restructure the project, the Bank team played a central role in redefining actions and targets. The Bank's support was key, and recommendations made on financial management by the successive supervisory teams helped to improve the project's procurement and financial management processes. 35. However, response times to requests for the Bank's "no objection" were occasionally too long. Moreover, the Bank's heavy involvement in the project management process may have been perceived as an attempt to limit the Haitian stakeholders' decision-making capacities, which led to misunderstandings between the principal contractor and the Bank as to what the MENF wanted and what the Bank thought it was necessary to do. 36. The World Bank's performance is rated as satisfactory in terms of preparation and supervision. b) Government 37. Managing such a large project was a first for the MENFP and was a major challenge given its weaknesses in terms of both material and human resources. Under these conditions, the technical departments did their best to complete the various project activities, with variable results. 38. Constant progress was made with the procurement process throughout the project, to achieve an overall satisfactory level. The Bank's standards and requirements were generally respected. A procurement plan was drafted each year for the project activities. However, there were persistent shortcomings in terms of setting up an adequate monthly follow-up system. 39. Financial management was satisfactory despite the shortcomings. The financial management department experience problems at the outset and was especially hard hit by the January 2010 earthquake, in which its specialist perished and a large amount of documents were destroyed. However, the financial information needed for the Operating Manual or for monitoring purposes was sent regularly, although not always on time. Conversely, the internal control unit was the weak link in the chain and was unable to provide sufficient support for financial management and follow-up on recommendations. Moreover, malfunctions by the GESCOMPTE software prevented the adequate tracking of disbursements and resulted in delays in the presentation of financial information. The acquisition and installation of the new TOMPRO financial management software will resolve this problem and ensure the better flow of financial information. 40. The objective to strengthen the DAA's financial management capacities was not achieve, despite the fact that the day after the financial manager was lost in the earthquake, a MENF contractor stepped in to ensure the project stayed on track. 42 41. In most cases, the principal contractor reacted to recommendations and proposals by the Bank. Most of the activities scheduled for the various project components were carried out, even though the project had to be restructured and ended up being a year behind schedule. 42. The project follow-up and coordination system did not work very well, which hindered the MENFP's response time in advancing activities or anticipating obstacles. V. Sustainability 43. The project aimed to achieve the objective of Education For All. The scheduled activities and their spin-offs were highly relevant. The grant to alleviate the cost of basic education for families and the nutrition/health program had an impact on school attendance. However, these interventions are costly and the national budget cannot support them in the short term. These actions will require support from financial backers over several years until the Government of Haiti can take over the reins; the same applies to the training of student teachers. Moreover, the importance of paying allowances was revealed as being key to the project's success; payment delays can quickly jeopardize the training program. 44. Institutional strengthening requires long-term support. As of the end of phase 1 of EFA, contrary to forecasts, the DAA is unable to single-handedly take over the financial management and procurement processes. Consultants need to be assigned to this department to train agents as part of a concrete plan involving assessments and preparing for a gradual takeover. However, this coexistence will not be easy if senior management does not get involved in monitoring the strengthening process. 45. The renovation work and supply of equipment can only create a lasting impact if accompanied by support for management and maintenance, and if the necessary financial resources are made available; otherwise, the equipment risks being damaged through misuse or poor maintenance. VI. Project lessons 46. Institutional strengthening implies the physical presence of consultants in the Ministry departments, and a willingness to do the work rather than have the work done. This is a long-term exercise that requires greater sensitivity by the agents and their senior managers. It also requires the implementation of a skills acquisition plan with assessments, early identification of shortcomings, and the necessary theoretical knowledge. It is also important to plan for the gradual withdrawal of consultants. 47. In the Haitian education sector, any innovation quickly arouses suspicions and is met with significant resistance. An understanding of the causes of this resistance and the issues faced by the different stakeholders is needed to improve project outcomes. 48. Time and effort spent on seeking a consensus between the stakeholders is not wasted; it improves the project's chances of success. 49. The Ministry's involvement and accountability in department-specific projects considerably increase the chances of success. 43 50. Communication, transparency and the efficient circulation of information between the various stakeholders is essential, as it prevents misunderstandings, controls rumours, and overcomes a number of other problems. 51. In projects involving the payment of allowances, every effort must be made to pay them on time, as delays will cause recipients to become discouraged, in turn holding up the project. 52. Having steering and monitoring committees that perform up to standard helps with the coordination of stakeholders and the exchange of information; it also allows for prompt action in the case of project delays. 53. There is a high level of incompetence in the public sector. However, this should not systematically lead to the creation of provisional or parallel project management structures. Projects need to be spearheaded by the public sector, in the aim of strengthening the country's institutions in the long term. 54. The relationship between consultants and civil servants is still very troubled, largely due to discrepancies between the fees charged by the former and the salaries earned by the latter. The possibility of paying a bonus to civil servants involved in this type of project is being evaluated. 55. The distribution of grants may lead to the opening of schools to benefit from this funding. As such, we need to make sure this does not lead to a deterioration in the system due to opportunistic schools with no capacity for quality teaching. 56. Schools should be focused on achieving MENFP accreditation 44 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders [Forthcoming] 45 Annex 9. List of Supporting Documents Declaration of General Policy of the Prime Minister-May 2012 (Enonce de Politique Generale du Premier Ministre) Financing Agreement (Education for All Project) between Republic of Haiti and IDA, April 27, 2007. Government of Haiti -The Operational Plan for Education 2010-2015 "Haiti: Early Grade Reading Assessment (EGRA), Rapport pour le Ministere de"Education et la Bankque Mondiale-Resultats en francais et en creole," April, 2010. Haiti Education for All (FA) Adaptable Program Grant Phase I (APG-1) Summary of Programmatic Reports for 2007-10, March 9, 2011. Project Appraisal Document of a Proposed Grant in the Amount of SDR16.8 million (US$25 million equivalent) to the Republic of Haiti for an Education for All Project in support of the First Phase of the Education for All Program, March 21, 2007. Project Paper on a Proposed Restructuring and a Proposed Additional Financing Grant in the Amount of SDR8.0 million (US$12.0 million Equivalent) to the Republic of Haiti for a education for All Adaptable Program Grant Phase 1, Grant H286-0-HA, May 10, 2010. Restructuring Paper on a Project Restructuring of Additional Financing for the Education for All Project in support of the First Phase of the Education for All Program Grant H588-0-HT Approved on June 10, 2010 to the Republic of Haiti, May 7, 2012. World Bank -Interim Strategy Note for FY12-13 List of Supporting Documents from Borrower's ICR EFA-SP Project Assessment Document -March 28, 2008 EFA-SP Operating Manual Accelerated Initial Training Project, Implementation Framework EFA-SP Activity Report, August 2008 -September 2010 Semi-Annual Report, October 2010 -March 2011 46 Semi-Annual Report, April -September 2011 Reminder on the World Bank's Mission from February 2011 National Strategy -Education For All, September 2007 47 HA T ATLA NTIC OC E A N 72°This map a produced by the Map Design Unit f The word Bak. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank 0 SELECTED CITIES AND TOWNS ile de la Tortve Group, anyu dgme on the legalstaus ofany territory, orony 0 DEPARTMENT CAPITALS 20°N Palmiste a 20'N ® NATIONAL CAPITAL Port-de-Paix RIVERS MAIN ROADS 0 Mle St.-Niol NORD -OUEST MAIN ROADS~ Cap-Haitien 4Ciit RAILROADS DEPARTMENT BOUNDARIES FosMrnLibérr- e NO RD Sat°g - -INTERNATIONAL BOUNDARIES ee Santiago Grande RiU Be du-Nor 7l Gonaivs Ennery deSaint- NORD -E SnT .GonivesRaphaél Saint Michel de l'Attalaye -l- Golfe de -< cv la Gonäve X ARTIBO ITE _,MaTssade -HAITI o Hinche Verrees DOMINICAN CENTRE 19°N\ Locdce REPUBLIC ... a ÎedoLCayenne Péfigre," la Gode 0 Anse-à-GaletsMreais7 MirebalaisSanJuan Pointe-à-Raquetteoa Grande Jeremnie Cayemite Roseaux Groix de PORT-AU-PRINCE Bioqets GRANDE- Anse d'Hainault, A NS ELéogåne ttone Lago To M c, d e \a H o t t N IP PE S .iaoäe O UE S ameEnriquillo, Barahona Goett Chaine de la Selle C m -einS U D ...."... les AnglaisCapPriA Vieux Bourg S UD -E ST Be||,e- d'Aquin .amlAnse Thiotte Côes-de-fer Marigot L.es Cayes Port-Saluf P3e à Vache 18°N 18°N 0 10 20 30 40 Kilometers TO z l Caribbean Sea Oweo 0 10 20 30 Miles 8 74° 73°W 72°W