Document of The World Bank FOR OFFICIAL USE ONLY Report No: PAD704 INTERNATIONAL DEVELOPMENT ASSOCIATION PROJECT APPRAISAL DOCUMENT ON A PROPOSED CREDIT IN THE AMOUNT OF SDR 51.8 MILLION (US$80 MILLION EQUIVALENT) AND A PROPOSED MULTI-DONOR TRUST FUND GRANT IN THE AMOUNT OF US$20 MILLION TO THE REPUBLIC OF THE UNION OF MYANMAR FOR A DECENTRALIZING FUNDING TO SCHOOLS PROJECT April 16, 2014 Human Development Sector Unit/Education East Asia and Pacific Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY EQUIVALENTS (Exchange Rate Effective March 31, 2014) Currency Unit = MMK MMK 962 US$1 US$1.546 SDR 1 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AY Academic Years MDG Millennium Development Goals BE Budget Estimate MDTF Multi-Donor Trust Fund BP Bank Procedures MICS Multiple Indicator Cluster BOT Board of Trustees Survey CAS Country Assistance Strategy MMK Myanmar Kyat CCT Conditional Cash Transfer MNPED Ministry of National Planning and Economic Development CESR Comprehensive Education Sector MoE Ministry of Education Reform CPP Community Participation Plans MPFMP Modernizing Public Finance Management Project CPPF Community Participation NGO Non-Government Organization Planning Framework CSO Civil Society Organization M&E Monitoring and Evaluation DBE Departments of Basic Education OAG Office of the Auditor General DEPT Department of Education OP Operational Policy Planning and Training DFAT Department of Foreign Affairs ORAF Operational Risk Assessment and Trade Framework DFID Department for International PDO Project Development Objectives Development DLI Disbursement-Linked Indicators PEFA Public Expenditure and Financial Accountability ECOP Environmental Code of Practice PER Public Expenditure Review EFO Externally Funded Output EGRA Early Grade Reading Assessment POM Project Operations Manual EMIS Education Management PTA Parent-Teacher Associations Information System EPIC Education Promotion and QBEP Quality Basic Education Implementation Committee Program ii FM Financial Management RETF Recipient-executed Trust Fund GDP Gross Domestic Product SA Social Assessments GPI Gender Parity Index SDR Special Drawing Rights IDA International Development SPC School Project Committees Association IFR Interim Financial Reports IHLCS Integrated Household Living TEO Township Education Offices Conditions Surveys IMF International Monetary Fund TOR Terms of Reference INGO International Non-governmental TPWG Township Program Working Organizations Group IPF Investment Project Financing UNDP United Nations Development Program IPPF International Professional UNICEF United Nations Children's Fund Practices Framework ISN Interim Strategy Note WBG World Bank Group Regional Vice President: Axel van Trotsenburg Country Director: Ulrich Zachau Sector Director: Xiaoqing Yu Sector Manager: Luis Benveniste Task Team Leader: James Stevens/ Lars Sondergaard  MYANMAR Decentralizing Funding to Schools Project TABLE OF CONTENTS Page STRATEGIC CONTEXT .......................................................................................... 1 A. Country Context .................1...........................1 B. Sectoral and Institutional Context.............1........................ C. Higher Level Objectives to which the Project Contributes ..................10 II. PROJECT DEVELOPMENT OBJECTIVES ......................................................... 11 A. Project Development Objectives (PDO)..............l..................1 III. PROJECT DESCRIPTION ................................................................................. 13 A. Programs to be supported .............................................13 B. Project Financing...............................................15 Project Cost and Financing. .............................. ..........17 C. Lessons Learned and Reflected in the Project Design...........................18 IV. IM PLEM ENTATION.......................................................................................... 19 A. Institutional and Implementation Arrangements ......... ................19 B. Results Monitoring and Evaluation. ............................. .....20 C. Sustainab ility .................................................20 D. Gender ......................................................20 V. KEY RISKS AND MITIGATION MEASURES..................................................21 A. Risk Ratings Summary Table.....................l.... .............21 B. Overall Risk Rating Explanation ............................... .....21 VI. APPRAISAL SUMMARY ................................................................... 22 A. Economic and Financial Analysis (see full analysis in Annex 7)..................22 B. Technical ....................................................24 C. Financia Management............................................25 D. Procurement ........................................... .......25 E. Social (including Safeguards) ................................. .....26 F. Environment (including Safeguards) ...........................28 iv Annex 1: Results Framework and Monitoring................................................................29 Annex 2: Detailed Project Description ............................................................................. 35 Annex 3: Implementation Arrangements ......................................................................... 40 Annex 4. Operational Risk Assessment Framework (ORAF)......................................... 48 Annex 5: Implementation Support Plan.........................................................................53 Annex 6: Community Participation Planning Framework.............................................55 Annex 7: Economic Analysis ............................................................................................... 51 V PAD DATA SHEET Myanmar Decentralizing Funding to Schools Project (P146332) PROJECT APPRAISAL DOCUMENT EAST ASIA AND PACIFIC EASHE Report No.: PAD704 Basic Information Project ID EA Category Team Leader P146332 B - Partial Assessment James A. Stevens Lending Instrument Fragile and/or Capacity Constraints Investment Project Financing Financial Intermediaries Series of Projects [ ] Project Implementation Start Date Project Implementation End Date 02-Jun-2014 30-June-2018 Expected Effectiveness Date: Expected Closing Date: 02-Sept-2014 31-Dec-18 Joint IFC No Sector Manager Sector Director Country Director Regional Vice President Luis Benveniste Xiaoqing Yu Ulrich Zachau Axel van Trotsenburg Borrower: Republic of the Union of Myanmar Responsible Agency: Ministry of Education Director-General, Budget Department, Contact: Maung Maung Win Title: Ministry of Finance Telephone (95-67) 410-450 Email: Maungmaungwin58@gmail.com No.: Project Financing Data(in USD Million) Loan [X] Grant [ ] Guarantee [X] Credit [ ] IDA Grant [ ] Other Total Project Cost: 100.00 Total Bank Financing: 80.00 Financing Gap: 0.00 Vi Financing Source Amount BORROWER/RECIPIENT 0.00 International Development Association (IDA) 80.00 Multi-Donor Trust Fund 20.00 Total 100.00 Expected Disbursements (in USD Million) Fiscal FY15 FY16 FY17 FY18 FY19 0000 0000 0000 0000 Year Annual 19.00 23.00 29.00 29.00 0.00 0.00 0.00 0.00 0.00 Cumulati 19.00 42.00 71.00 100.00 100.00 0.00 0.00 0.00 0.00 ve Proposed Development Objective(s) The objective of the Project is to help improve and expand Myanmar's School Grants Program and Student Stipends Program. Components Component Name Cost (USD Millions) Expansion and Improvement of the School Grants 77.00 Program Expansion and Improvement of the Student Stipends 19.00 Program Capacity Improvement to Strengthen Monitoring and 4.00 Implementation of Programs Institutional Data Sector Board Education Sectors / Climate Change Sector (Maximum 5 and total % must equal 100) Major Sector Sector % Adaptation Mitigation Co-benefits % Co-benefits % Education General education sector 60 Other social services Other social services 40 Total 100 Z I certify that there is no Adaptation and Mitigation Climate Change Co-benefits information applicable to this project. vii Themes Theme (Maximum 5 and total% must equal 100) Major theme Theme % Human development Education for all 60 Social protection and risk management Social safety nets 40 Total 100 Compliance Policy Does the project depart from the CAS in content or in other significant Yes [ ] No [X] respects? Does the project require any waivers of Bank policies? Yes [ ] No [X] Have these been approved by Bank management? Yes [ ] No [ ] Is approval for any policy waiver sought from the Board? Yes [ ] No [X] Does the project meet the Regional criteria for readiness for implementation? Yes [ X] No [ ] Safeguard Policies Triggered by the Project Yes No Environmental Assessment OP/BP 4.01 X Natural Habitats OP/BP 4.04 X Forests OP/BP 4.36 X Pest Management OP 4.09 X Physical Cultural Resources OP/BP 4.11 X Indigenous Peoples OP/BP 4.10 X Involuntary Resettlement OP/BP 4.12 X Safety of Dams OP/BP 4.37 x Projects on International Waterways OP/BP 7.50 X Projects in Disputed Areas OP/BP 7.60 X Legal Covenants Name Recurrent Due Date Frequency Project institutional arrangements X Continuous Description of Covenant Obligation of the Recipient to maintain, at all times during the implementation of the Project, the Project Steering Committee with functions, composition, staffing and resources satisfactory to the Association. Name Recurrent Due Date Frequency Project Operational Manual X Continuous Description of Covenant vm Obligation of the Recipient to carry out the Project in accordance with the Project Operational Manual Name Recurrent Due Date Frequency School Grant and Stipend Program SchoolYear 2014/15 Guidelines Description of Covenant Obligation of the Recipient to adopt the School Grants Guidelines and Student Stipends Guidelines, in form and substance satisfactory to the Association, and ensure that all School Grants and Student Stipends financed out of the proceeds of the Financing are made and administered in accordance with such guidelines. Name Recurrent Due Date Frequency Safeguards X Continuous Description of Covenant Obligation of the Recipient to shall ensure that the Project is carried out in accordance with the provisions of the Community Participation Planning Framework (CPPF), including the preparation of Social Assessments and Community Participation Plans in accordance with the CPPF. Name Recurrent Due Date Frequency DLI Monitoring and Reporting X Annual Description of Covenant Obligation of the Recipient to monitor and furnish reports to the Association on the achievement of the Disbursement Linked Indicators, not later than March 31 of each year during the implementation of the Project (except for the first report which shall be furnished to the Association not later than August 31, 2014), in accordance with the technical standards, arrangements and procedures set out in the Project Operational Manual. Conditions Name Type Cross-effectiveness Effectiveness Condition Description of Condition The Grant Agreement (for the Myanmar Partnership MDTF co-financing grant for the Project) has been executed and delivered and all conditions precedent to its effectiveness or to the right of the Recipient to make withdrawals under it (other than the effectiveness of the IDA Financing Agreement), have been fulfilled. Conditions Name Type Disbursement Withdrawal Conditions conditions and provisions Description of Condition Provisions specifying that: (i) the Recipient may not withdraw the proceeds of the Financing unless and ix until it has furnished evidence satisfactory to the Association that it has achieved the respective DLI(s) and reported a corresponding amount of Eligible Expenditures; (ii) the amount to be withdrawn against each DLI would correspond to the amount allocated to such DLI; and (iii) in the event that any DLI has not been achieved by its target date, the Association may withhold any amount of the Financing allocated to such DLI, disburse it at any later time if and when such DLI is actually completed, and/or reallocate or cancel in whole or in part any withheld amount of the Financing. Team Composition Bank Staff Name Title Specialization Unit James A. Stevens Senior Operations Team Lead EASHE Officer Sirirat Sirijaratwong Procurement Specialist Procurement Specialist EASR2 Frederick Yankey Senior Financial Senior Financial EASFM Management Specialist Management Specialist Lars M. Sondergaard Country Sector Co-Team Lead EASHE Coordinator Pamornrat Senior Social Social Specialist EASTS Tansanguanwong Development Specialist Manush Hristov Senior Counsel Senior Counsel LEGES Satoshi Ishihara Senior Social Senior Social EASTS Development Specialist Development Specialist Miguel-Santiago da Senior Finance Officer Disbursements CTRLN Silva Oliveira Reena Badiani- Economist Economist EASHS Magnusson Puja Vasudeva Dutta Senior Economist Senior Economist EASHS Christopher Robert Senior Financial Senior Financial EASFM Fabling Management Specialist Management Specialist Yuko Okamura Economist Economist EASHS Ei Kalya Moore E T Consultant E T Consultant EASHE Thomas Flochel Young Professional Young Professional EASHS Anna Coronado Program Assistant Administrative Support EASHD Non Bank Staff Name Title Office Phone City Kevin MacDonald S T Consultant Andrew Vickery Financial Management Specialist Emiko Naka Senior Education + 95 1 251 810 Yangon Program Manager, DFAT x Locations Country First Location Planned Actual Comments Administrative Division x1 I. STRATEGIC CONTEXT A. Country Context 1. Myanmar is a culturally and ethnically diverse country, comprising more than 130 ethnic groups across countries and regions. It is one of the poorest countries in the East Asia and Pacific Region, with an estimated gross domestic product (GDP) per capita of between US$500-800. While reliable poverty data are scarce in Myanmar, all indicators point to poverty being concentrated in rural areas: two rounds of household surveys (in 2005 and 2009) supported by the United Nations Development Program (UNDP) highlight significant differences in poverty across geographical areas, and the bulk of the poor are reported to be concentrated in rural areas. Available figures also suggest that social indicators are poor, for example 32 percent of children under five suffering from malnutrition, the highest rate in the region. Agriculture is the mainstay of the national economy, generating approximately 43 percent of gross domestic product, 54 percent of employment and providing livelihoods to more that 70 percent of the population. However, the sector is characterized by a number of constraints, including limited irrigation, little value-added agro-processing capacity, poor policy settings, and widespread indebtedness in rural communities. 2. Decades of underinvestment have taken a toll on public institutions and the civil service, limiting their ability to provide services in rural areas. Health and education services (among others) suffer from having been underfunded for decades. As part of the move toward a "people- centered approach," government has substantially increased spending on health and education, more than tripling their nominal allocations between 2011/12 and 2013/14. In addition, the FY2012/13 budget provided a significant raise to civil servants, whose salaries - in particular for those at the lower levels of the bureaucracy and in rural areas - had often been insufficient to cover living costs. Partly as a legacy of these limitations, households have had relatively high out-of-pocket costs to access public health and education services. B. Sectoral and Institutional Context 3. Education reform is one of the key priorities set out by the government. The 30 Year Education Development Plan (2001/02-2030/31) lays out strategies to promote greater access and to improve the quality of basic education, and some significant progress has been made. Enrollment in primary and secondary schools has increased rapidly and public expenditure in education has tripled over the last 3 years. But significant challenges remain. Myanmar remains far from full primary enrollment; legislation and the policy framework for education remain outdated; and there is a strong demand for the government to address curriculum and language of instruction to reflect Myanmar's diverse language and cultural context. Capacity to implement reform at all levels also remains a concern, particularly as increasing roles and responsibilities of local authorities will be required for decentralization to take place. 4. Education providers in Myanmar include the government, which is by far the largest, the Monastic Education School System and other non-state providers (including Ethnic Education Departments). At present, the government supported education sector includes about 35,500 1 primary and post-primary schools, 3,000 middle or lower secondary schools, and 2,300 high or upper secondary schools.' About 5.2 million students are enrolled in primary school (grades 1- 5), 2.3 million in middle school (grades 6-8) and 0.7 million in high school (grades 10-11). There are about 1,400 monastic schools registered with the government (accounting for less than 1 percent of total enrollment), as well as a much smaller number of non-registered affiliated and community schools and school run by cease fire groups in some States. Enrollment in private schools is virtually non-existent in primary and middle school but accounts for approximately 5 percent of enrollment in high school. Primary net enrollment rates (as measured by the MICS survey in 2009/10) stood at 86.7 percent, dropping to 58 percent and 32 percent for middle and high school, respectively. Net enrollment rates are lowest for the poor and among the bottom 40 percent of the population. There are no clear gender differences in enrollment patterns at a national or state level at a primary school level. At middle and high school level, net enrollment rates for girls are somewhat higher than they are for boys. The higher enrollment of girls at these levels partly reflects higher opportunity costs for boys in the labor market. 5. Within the Ministry of Education (MoE), the general education grades 1-11 are administered by a Department of Education Planning and Training (DEPT) and regionally-based Departments of Basic Education (DBE) 1, 2 and 3 (see Figure 1). The 14 States and Regions have small education offices, but education offices (which are considered part of the ministry structure) in the approximately 330 townships are operationally responsible for basic education delivery. Until recently the overall governance of basic education has been characterized by highly centralized and non-transparent decision making, which is concentrated in the MoE, mainly in the DBEs. A recently launched Comprehensive Education Sector Reform (CESR) process (see below) has provided evidence of a willingness to be more transparent and a desire to pursue decentralization of financial and potentially curriculum decisions to lower governmental levels. ' The status ofa school (e.g. "middle school") refers to thehighest gradebeingofferedin the school. As such, middle schools also haveprimary school students; and high schools haveboth middle and primary students. 2 Figure 1: Structure of basic education system in Myanmar Myanmar National Education Conmnittee Ministry of Education Region and State Minister for Social Affairs Basic Education Council Basic Education Curriculum, Teacher Education Syllabus and Textbook Committee Supervisory Committee Departments of Departments of Departments of Myanmar Education Planning Basic Education Myanmar Education Examination and Training No.1, 2 and 3 Research Bureau Board Region and State Education Offices District Education Officers 20 Education Collages Township education Offices Teachers 273,516 Basic Education High Schools 2,351 Students 8,200,595 Basic Education Middle Schools 3,121 Basic Education Post-Primary Schools 6,553 Basic Education Primary Schools 2 Total 40,992 Private School 12 6. A renewed financial commitment to the sector provides evidence that education has become a high priority on the government's national agenda. Public education spending was estimated to have been about US$354 million equivalent in the fiscal year 2010/2011, representing an estimated 0.7 percent of GDP. The budget for education in 2013/2014 has risen substantially to about US$1 billion equivalent, or about 14 percent of overall government spending. A large share of this new education spending is accounted for by increasing teacher wages, which are low by regional standards. But Myanmar is also focused on the need to increase non-salary recurrent spending, as well as to increase both government and donor- financed investments. This rapid increase in the public funding for education has begun to re- balance public funding with private spending on education. The Bank's recent rapid assessment 3 of education spending estimates private contributions to education to be about two-thirds of overall education spending in Myanmar in 2009/10, which is high by global standards. The recent increase in public spending implies that a more balanced share of the cost of education is born by the public sector. Public education spending is likely to continue to increase in the near term in absolute terms and as a share of overall (public and private) education spending. Table 1: Total Education Expenditure 2009/10- 2013/14 (million kyats) Level ofEdu cation 2009-10 2010-11 2011-12 2012-13 (RE) 2013-14 (BE) Basic Education 183,554 235,376 270,927 521,392 732,590 25,284 30,604 37,848 241,968 366,269 Administration and Others 623 722 947 1,963 2,165 Total Government expenditure on education 209,461 266,702 309,722 765,322 1,101,024 428,399 437,824 473,726 486,990 500,626 637,860 704,526 783,448 1,252,312 1,601,650 Households spending as %of total educationspending 67.2% 62.1% 60.5% 38.9% 31.3% Education expenditure as %of GDP 1.8% 1.8% 1.8% 2.6% 3.0% Total government education expenditure as%ofGDP 0.6% 0.7% 0.7% 1.6% 2.1% Total education expenditure as % oftotal government expenditure 6.6% 6.1% 6.9% 10.1% 14.1% Source: (i) Budget data from PER Database; (ii) GDP figures from IMF Article IV August 2013; (iii) Non-MoE higher education expenditure from relevant ministries; (iv) household expenditure for 2009/10 from IHLCS 2009/10 (and World Bank staff projections for subsequent years). (RE) means "revised budget estimate" and "BE" means "budget estimate". 7. Education governance in Myanmar is gradually moving towards decentralization. Like other government systems, the governance of education is set to make a transition from a high degree of central control to more authority being decentralized to lower administrative levels. In addition to more decentralization of funding - which is the focal point of this project - the government is in the process of reviewing and developing policies to allow curricula and language of instruction to reflect local concerns. The transition is likely to be gradual and uneven, with region, state and district education offices exercising more or less authority depending on the local context. But Myanmar's 330 townships are likely to be the focal point for service delivery in basic education across the country. 8. Decentralization of education will have particular importance in the conflict-affected ethnic minority areas. These regions (affecting approximately 15-20 percent of townships) are 4 among the most underserved areas of the country, largely due to ongoing armed conflicts between ethnically-associated non-state armed groups and the government. A clear and credible path towards decentralized governance of the education sector could be a positive contribution to the ongoing peace process, as it addresses some of the widely-held grievances regarding central control of education. However, government-led efforts to improve schools in conflict-affected areas in the short term may be opposed by influential ethnic leaders and armed groups, especially if it is seen to undermine the ethnic education systems. The peace negotiations are likely to shape governance and service delivery in the ethnic areas, so it will be essential for efforts to promote decentralized education to avoid scenarios that could preempt outcomes of the negotiations, or undermine the government's credibility in the peace process. 9. The education sector reflects Myanmar's evolving political landscape. Education in Myanmar is reported to be highly valued, as demonstrated by the high private contributions to schooling. Underfunding of education has likely exacerbated inequalities, as poorer and more isolated areas comparatively lack financial means and technical capacity. The underfunding of the system is gradually being addressed. But the capacity and willingness of specific sub- national jurisdictions, including townships, and school-level committees (parent-teacher associations and school committees), to represent all groups and to target scarce resources on disadvantaged groups or areas remains a subject of concern. 10. Myanmar has shown commitment to working with development partners to review its education policies, undertake analytical work and develop a new strategy. There is strong donor coordination in the education sector led by Australia as co-chair of the Education Joint Sector Working Group. A rapid assessment phase of Myanmar's CESR process was completed in early 2013, and work on a comprehensive sector plan is underway, expected to be delivered in the second half of 2014. The President has emphasized that the implementation of education reforms must be accelerated and be pragmatic. An Education Promotion and Implementation Committee (EPIC) was set up in October 2013 with a key mandate to develop education policies and national education law as a framework for the sector by drawing upon recommendations and findings from the CESR. While gaps will continue to exist in sector knowledge and in Myanmar's capacity to plan and manage resources for education, the changing political landscape offers opportunities to scale up system-building support. Myanmar identified its school grants and stipends programs as priorities for 'quick wins' in its "Framework for Economic and Social Reforms". The CESR rapid assessment and other sector analytical work have identified the lack of reliable information on learning outcomes as a major constraint to knowledge work and investment programming. Decentralizingfundingfor education 11. The MoE in Myanmar is currently decentralizing funding for education through two national programs: (i) it transfers funding for various operating expenditures through townships to schools based on the number of primary students in the school; and (ii) it transfers funding through townships and schools to pay cash stipends to poor children and scholarships to high achieving students. The school grants program has its origins in the need to provide schools with operating funds following the government's decision to make primary education (grades 1-5) free beginning in school year 2009/10. 5 12. Both of these initiatives were established through ministerial decrees during the 2009- 2010 school year. The decrees established the basic framework of the amounts and the flow of funds, but neither initiative has been established as a formal program with statements of objectives, detailed descriptions of responsibilities, performance indicators, or provisions for monitoring their impact on the education system. Moreover, lacking manuals and training to explain how these programs are to be implemented, township officers and school heads are currently left with little guidance and support to implement the programs. 13. The school grants program is national in coverage and grant levels have grown rapidly, albeit from a very low base. In FY 2012/13, Myanmar transferred approximately US$250 on average to small schools, US$400 to medium size schools and US$500 to larger schools in direct subsidies to support spending on school's operational costs.2 The program supported more than 40,000 schools, and accounted for overall spending of about US$15 million equivalent in 2012/13. These amounts have been more than doubled in the 2013/14 budget. School grants programs in neighboring countries (Cambodia, Lao, Vietnam) provide individual primary schools in the range of US$2,000 to US$3,000 per school annually. 14. The first years of implementing the school grants program suggest several areas that can be improved. Township officers and school heads would benefit from more detailed guidelines and training to explain the basic design elements: how much each school is entitled to receive; when they can expect to receive the funding; what they can spend the funding on; and what requirements need to be fulfilled (e.g. in terms of community involvements in overseeing spending, reporting, etc). Visits to schools and townships suggest that, at present, these basic design features are unclear, especially when it comes to the amount of money they can expect. In addition, it is anticipated the roll out of school improvement plans nationwide would become a key planning tool for schools and guide school grant expenditure. Furthermore, the amounts disbursed are currently tied to specific budget codes which adds complexity and limits schools' ability to use the resources where their needs are the greatest (e.g. currently, US$100 is ear- marked for "office supplies" and another US$100 is ear-marked for "consumables"). This project will support Myanmar in strengthening its school grants program in accordance with regional norms, and of gradually giving school headmasters and communities a greater say in how resources (both the budget for goods and services) get spent. The program will assist in supporting the MoE and Regional and Township education officers through research and analysis of existing systems in preparation for decentralization. 15. The stipends program is also national in coverage, but is small both in terms of financing and the number of beneficiaries. Eligibility is currently limited to approximately 33 students in each of Myanmar's 330 townships, covering 2 students per grade in grades 1-11. The beneficiary population is about 11,000 students nationally. In the 2014/15 school year, the government plans to double the number of students receiving stipends in grades 1-5, raising the number of 2 In 2012/13, a small schoolwas defined as a school(ofany type) having less than 100primary school students. A medium size school was defined as havingbetween 101 and 200 primary school students. And a large school was a school with more than 200 primary school students. The estimates above are basedon observed spendingon goods and services at the school level (totaling 13.9 billion kyat). 6 beneficiaries to 16,000 students, and a similar further expansion is planned for secondary students in the 2015/16 school year. The stipend amounts are US$5 per month over 10 months for primary students, US$6 per month for lower secondary students and US$8 per month for upper secondary students. In total, approximately US$700,000 equivalent is currently allocated annually for stipends. Although the levels of stipends are comparable to the levels of stipends in other countries with stipends programs at a comparable level of development, the coverage of the program is substantially lower and the dispersal across townships is much thinner. 16. The implementation of the stipends program is uneven across and within townships. The uneven implementation reflects the absence of clear and detailed operational guidelines for townships and schools, and the omission of key implementation steps that can help to standardize processes. The stipends program uses a combination of categorical targeting, focusing on orphans, and a broader based poverty criteria for which there is no criteria. All schools are encouraged to nominate students for the stipends program, and selection among the nominated students is made at the township level. Practices for selecting students are not uniform across Myanmar. In some state/regions, school officials consult with Parent-Teacher Associations (PTA) and/or Boards of Trustees in making the stipend awards, but this consultation is not mandated. Townships are encouraged in the existing guidelines to form committees to allocate stipends across schools. In practice, the use of committees is not uniform across townships. There is no structure in place for program monitoring and evaluation (M&E), resulting in limited knowledge of actual program implementation at the DBE and DEPT levels. 17. Central funds for these programs are devolved from the MoE to Township Education Offices (TEOs). Each township is responsible for ensuring that the funds reach the school and are spent on eligible items in accordance with simple guidelines. Townships generally disburse funding to schools in cash. Schools are not currently required to prepare budgets or plans in order to receive funding. Purchases carried out by schools are recorded in account books and monitored by the Township Education Office. 18. Learning outcomes. The CESR rapid assessment identified the lack of reliable information on learning outcomes as a significant obstacle for education reform. Information on learning outcomes is particularly important in a decentralized system where local stakeholders have decision making authority and financial discretion. In this environment, the role of central authorities becomes one of holding localities accountable. Data on how well children are learning is crucial for assessing system performance. Standardized learning assessments are, therefore, a necessary complement to support decentralized systems. 19. There are currently no regular assessments of student learning at the end of each cycle in Myanmar. As such, there is little information about whether students are mastering the key desired competencies at the end of the primary, or middle school cycles. The only regularly administered standardized examination is the Basic Education High School exam at the end of high school. But it is not clear that these exam results are comparable between years. 20. The Quality Basic Education Program (QBEP), financed by Australia, the United Nations Children's Fund (UNICEF), the Department for International Development (DFID), Denmark and Norway and implemented by UNICEF, has helped Myanmar to conduct three rounds of 7 assessments of Myanmar language and mathematics skills. In 2007, 3rd and 5th graders in 20 of Myanmar's approximately 330 townships participated in a first round of assessments. In 2009 and 2011, students in the same townships were tested using the same test. However, the tests were largely driven by the donor project and for the purpose of providing information to the project. As such, none of the results have been made public, and the Ministry has not been using the findings as a measure of the quality of education provided in its schools. 21. Gender status. The education sector is characterized by a high degree of gender equality. There are virtually no differences in net enrollment rates between boys and girls. The primary net enrollment rate was 88 percent for both boys and girls, as measured by the Multiple Indicator Cluster Survey in 2009/10. At lower secondary it was 59 percent for girls and 57 percent for boys, while at upper secondary it was 34 percent for girls and 31 percent for boys. The slightly higher enrolment rate of girls at secondary level is partly attributable to higher male wages in the labor market, resulting in greater financial pressure for boys to drop out earlier. The national averages mask some relatively small variations at the regional level, with some states having higher rates for girls (than boys) and others showing the opposite picture. The small gender differences in net enrollment rates are however dwarfed by differences in enrollments due to socio-economic circumstances: less than 10 percent of high-school age children in the poorest quintile are enrolled in high-school, compared to over 60 percent in the richest quintile. Assessments of learning outcomes supported by QBEP, mentioned above, find no gender differences in math scores at grade 5, but that girls perform better than boys at Myanmar language. Since gender parity has been achieved in primary education enrolment3, the need for gender equality measures may be rejected by education stakeholders. This is of concern to as ostensibly equal national figures mask subnational disparities (available data on literacy rates suggests there are still gender inequalities in ethnic areas including in Northern Rakhine).4 Education programs that do not address gender equality fail to promote women's empowerment which is an issue in Myanmar as the country with the Southeast Asian region's lowest female - 5 political participation rates at just 6 percent. 22. In terms of teachers and administrators, the MoE has the highest proportion of female workers among all ministries. Amongst teachers, around 83 percent are women, with slightly higher rates in primary than in high school. A new female Education Minister H.E. Prof Dr. Khin San Yee has been recently appointed in a male dominated senior leadership team (Vice Ministry and Director Generals). Women are well-represented in leadership positions below the top positions: 62 percent of primary school head masters are women and 55 percent of high school head masters are women. Moreover, looking across the sector - including amongst township education officers and the ministry staff - 70 percent of the positions of deputy directors (and above) are held by women (on average, only 33 percent of women hold deputy director posts (and above) across all ministries). 23. Development partner support. This project has been collaboratively prepared with the Australian Department of Foreign Affairs and Trade (DFAT) which is expected to provide Ministry of National Planning and Economic Development and UNICEF Situation Analysis of Children in Myanmar, Nay Pyi Taw 2012. UNICEF, Quality Basic Education Program Annual Report 2013 5 World Bank, Myanmar Data (accessed 20140321 at http://datatopics.worldbank.org/gender/country/myanmar) 8 support through a Multi-Donor Trust Fund (MDTF) for this operation in the amount of US$20 million. Australia is fully committed to ensuring that the program supported by this investment is implemented, monitored and evaluated thoroughly and professionally as part of a quality-focused effort that ensures that Australia's investments in Myanmar education result in programs that are continually improved and in the progressive strengthening of government's capacity to plan, manage and monitor high quality education programs. More broadly, the education sector is being supported by a group of development partners which provide both financial and technical advice to the ministry. In terms of financial support, the largest ongoing support operation is the second phase (2012-2015) of the QBEP with commitments likely to reach US$80 million for the four year period. The main focus of the program is 28 disadvantaged townships to which it provides additional support toward the goal of "increasing the number and proportion of children accessing and completing a quality basic education." Throughout the first phase (2006-2011) of the QBEP program, a number of interventions have been "piloted" in the targeted townships, including preparing and distributing "essential learning packages", consisting of textbooks and basic learning materials, training teachers on child-centered approaches, administering a student learning assessment, and improving township education management and school planning (including School Improvement Plans). These interventions are being continued under the second phase of the MDEF. In addition, the MDEF has also laid the groundwork for more upstream policy discussions with the ministry, including providing support for the inclusive CESR process which was launched in 2012. 24. Bank support: The project is also closely linked with other efforts supported by the Bank and others, aimed at strengthening the government's ability to deliver core services. For example, this project is being prepared and will implement in parallel to a Bank-supported project aimed at modernizing the public financial management system.6 Among other things, the "Modernizing Public Finance Management" Project will make the budgeting process more strategic; improve transparency and accountability of public resources; strengthen reporting systems; and strengthen the Parliament's Public Accounts Committee. In addition, the World Bank has carried out a Public Expenditure and Fiduciary Assessment7, and is currently carrying out a multi-year public expenditure review (PER), focusing on the education, health and education sectors. 25. The Modernizing Public Finance Management Project (MPFMP) will address reforms in key central ministries and agencies including the Ministry of Finance, Ministry of National Planning and Economic Development (MNPED) and the Office of the Auditor General (OAG). MPFMP will address public finance management reform at the central level as well as broader FM management issues to support improvements to central systems and processes. There are important synergies between the PFM project and this project which will require closer coordination across the two projects as the implementation commences. Key financial risks addressed in both the MPFMP and this document will be mitigated through strengthening of PFM processes at the center (including reforms to the budget planning and execution process for evidence based resource allocations), capacity building across government and monitoring as part of the project activities and regular M&E. It will also be important to monitor the reforms at 6 World Bank (2014): Project Appraisal Document for "Modernizing Public Finance Management Project" (Report No: PAD925) 7World Bank(March 2012): "Public Financial Management Performance Report", Report No. 75897-MM 9 the center as they are likely to have implications for line ministries focused on service delivery. The Overall Risk Assessment Framework Risks (Annex 4) also addresses FM risks identified in the Public Expenditure and Financial Accountability (PEFA) and Education Sector Fiduciary Assessment. Finally the PFMP will also support capacity building and operational effectiveness of the OAG through the development of a medium term capacity development plan to increase the quality and quantity of audit coverage and bring audits in line with international standards. This will further support audit requirements placed on OAG as part of the Education Grants and Stipends program. C. Higher Level Objectives to which the Project Contributes 26. Beyond the project's development objective indicators which will track progress during the next four years (described below), this project will contribute to important longer-term goals in Myanmar's education sector. Most importantly, improving the school grants and stipends programs will strengthen MoE's ability to provide a quality education for all Myanmar's citizens by, for example, helping to ensure that more students complete more years of schooling. These higher level objectives will be monitored through use of the Bank's core indicators for IDA projects that support basic education. In the case of this project, MoE will report on the primary completion rate, on the gender parity index (GPI) and on the establishment of regional baselines for early grade reading performance. 27. The grants program aims to improve educational outcomes through strengthening the supply side. Providing a transparent, reliable and flexible source of resources to schools and encouraging parents to be involved in their allocation would be a powerful signal from the Ministry (and the government, more broadly) that education is a high priority, and that the government is taking concrete steps to empower communities. In addition, increased public spending on schools should reduce households' financial contributions to schools. This, in turn, will ease one of the barriers to attend school for poor households. The additional financial resources provided to schools should gradually result in more learning materials to be available in classrooms. This, in turn, should support students in better learning. 28. The stipends program is a form of demand-side financing that will place resources in the hands of those who demand education, rather than the suppliers of education - the schools. Stipends can be used to help poor families invest in schooling by supporting them with direct costs - such as uniforms - and by helping families to recover some of the opportunity costs, such as alternative income earning opportunities for the child. Therefore, the aim of this intervention is to counter the constraints that prevent children from attending school or from continuing their education. Experience from other countries suggests that, properly implemented and monitored, the stipend program has the potential to lead to higher school attendance rates and lower dropout and repetition rates in the targeted grades. In addition, these programs have been found to reduce poverty and to increase household welfare. 29. These longer term goals - of improving learning and reducing dropouts - are important pathways to help reduce poverty and provide all Myanmar's citizens with opportunities provided by the opening up of the economy. There are three important ways in which a focus on improving education outcomes can help reduce poverty: first, students who complete more years of schooling earn a higher income. In Myanmar, on average, completing one more year of 10 schooling is associated with 6.7 percent higher income. Second, having more years of education helps protect against economic shocks. That is, education makes it more likely for men and women not just to be employed but to hold jobs that are more secure and provide good working conditions and decent pay. In so doing, education can not only lift households out of poverty, but also guard against them falling - or falling back - into poverty. Third, providing poor people with a better education is one way of breaking the strong intergenerational linkages of poverty. 30. Improving the institutions that deliver basic services is also a core element of the Bank's interim strategy note (ISN).8 The strategy focuses on programs that can support the Myanmar government in the country's current triple transition - from an authoritarian military system to democratic governance, from a centrally-directed economy to market-oriented reforms, and from conflict to peace in the border areas. The ISN outlines support around three pillars: the first aimed at supporting government's efforts to transform institutions to allow them to deliver for citizens; the second aimed at building confidence in the ongoing reform process; and the third focused on preparing the way for the resumption of a full country program. 31. To signal its commitment to improving public service delivery, the government has prepared a "Framework for Economic and Social Reform" and, in this Framework, it highlights the government's 'school grants' program and its 'stipends' program as the two highest priorities for 'quick wins' in the education sector. It is on this basis that the Bank has engaged in a rapid response effort to support these programs. The proposed project specifically supports the first two pillars of the ISN: supporting government's efforts to transform institutions to allow them to deliver for citizens and building confidence in the ongoing reform process by helping promote 'quick wins'. II. PROJECT DEVELOPMENT OBJECTIVES A. Project Development Objectives (PDO) 32. The objective of the Project is to help improve and expand Myanmar's School Grants Program and Student Stipends Program. a. Project beneficiaries 33. The project will have four sets of beneficiaries: (i) approximately 8.2 million students in Myanmar's approximately 43,000 government-supported schools who will benefit from attending better-financed schools; (ii) approximately 100,000 poor students in approximately 40 townships who will receive financial support to attend school under the re-designed stipends program; (iii) approximately 1,000 township education officers, assistant education officers and accounting clerks and approximately 43,000 school head masters who will benefit from capacity building training under the project; and (iv) parents and local community leaders throughout Myanmar who will be encouraged to increase their participation in their community's schools. 'The ISN (dated October 30,2012) covers the periodFY2013-14 (World BankReport No. 72458-MM) 11 b. Project development objective results indicators 34. Achievement of the PDO will be monitored using the following results indicators: * Number of students receiving payment in the revised stipends program * Number of townships (out of 330) distributing school grants to 80 percent of their schools according to formula * Number of school heads which have received training in project implementation in the programs 35. These indicators will support the following parts of the PDO: * Expanded coverage of the stipends program: The current stipend program will cover 16,022 students in 330 townships in school year 2014-15. The revised stipends program is expected to cover at least 100,000 students in 40 townships by school year 2017-18. * Improved reliability and transparency of the school grants program: Through the school grants program, schools will receive reliable and transparent funding that will promote school improvement planning. 'Transparent' means that money is transferred based on a simple but clear formula and 'reliable' means that money arrives on time (and in accordance with that formula). The timeliness of allocations to schools in June and November each year will be reported on by DBEs. * Improved capacity to implement the programs and monitor their progress: MoE will provide training and capacity building related to key implementation issues of the grants, stipends program and early grade reading program, including financial management, monitoring and evaluation and implementation. The number of trained township education officers, clerks and school headmasters will be monitored. 36. These indicators, as well as related intermediate outcome indicators, will also be the basis for the project's Disbursement-Linked Indicators (DLIs). The specific project indicators are summarized in the project description below and elaborated on in Annex 1 (Results Framework and Monitoring). Each indicator will be monitored and reported on by MoE as part of the program and verified in 'spot check' surveys by the Bank. The Bank will engage in monitoring and evaluation activities to provide external confirmation of progress on results. These activities include surveys in selected townships that will be conducted in cooperation with MoE. 37. Core IDA indicators including primary completion rates, GPI, direct project beneficiaries and learning assessment will also be updated as information is available in order to monitor long term outcomes in the sector. In addition, during implementation, the MoE will produce education outcome results baselines in two areas: (i) baselines for early grade reading performance results will be established for one or more DBEs; and (ii) baselines will be established for the drop out/transition rates of stipends students in the 40 stipends townships. 12 III. PROJECT DESCRIPTION A. Programs to be supported 38. The project will contribute to the MoE's existing budget allocation in support of three specific programs: (i) expansion of MoE's existing national schools grants program that benefits all government-supported schools; (ii) expansion of its student stipends program, focusing on 40 townships over the four year project duration; (iii) strengthening of its capacity to monitor and implement these programs and lay the foundation for more focus on learning outcomes in future programs by conducting an early grade reading assessment (EGRA) baseline. The implementation period for the project is 4 years. The Bank will manage a parallel technical assistance program to support program design, monitoring and evaluation. Activities under the proposed operation will be financed at 80 percent by the IDA credit and 20 percent through funds from the MDTF. 39. Component 1: Expansion and Improvement of the School Grants Program (US$77 million): All schools with primary students currently supported by government budget funding are eligible for participation in the school grants program. Expansion of the program, therefore, will mean increasing the size of annual operating grants to schools from approximately US$250, US$400 and US$500 per school for small, medium and large schools, respectively, to targets of US$900, US$1,200 and US$1,800 per school, respectively, over a four year period.9 Improving the program means introducing innovations from global experience, as well as improving the fiduciary management of the program, in particular its financial management. Innovations will be introduced to the program by revising its guidelines and by providing training. Specific innovations include: (i) introduction of well-defined program objectives and performance indicators; (ii) tying of the grant funding to school improvement planning; (iii) introducing increased autonomy on school-level spending; (iv) promoting community participation and oversight through parent teacher organizations; (v) standardizing financial reporting; (vi) provisions for audits; and (vii) linking program progress reporting to MoE's own information systems. 40. Component 2: Expansion and Improvement of the Student Stipends Program (US$19 million): While all government-supported schools in Myanmar are nominally eligible to participate in the existing student stipends program, the small size of the program (11,000 stipends to be awarded nationwide) effectively means that, while most schools apply for stipends funding, few schools are actually selected to participate in the program and those that do participate would have, in most cases, no more than 2 stipend students. Because the new student stipends guidelines will include an increase in coverage for each school and more rigorous targeting and administration, the program will only be expanded to more schools and students in a limited number of townships over time. In school year 2014-15, the stipends program will be expanded to 8 townships and is expected to cover approximately 25 percent of grade 5-11 students in each township. An additional 12 townships will be added in school year 2015-16 and 9 MoEis consideringre-organizing the three categories (small(<100 students), medium(100-200 students) and large (200+ students) used duringthe first years ofthe programinto more categories to allow forhigherper school allowances forlargerschools; the Banksupports this change. 13 an additional 20 townships will be added in school year 2016-17, for a total of 40 townships (out of 330 total) to be supported over 4 years. In school year 2015-16, the stipends program may be expanded to include grades 1-4 as well as grades 5-11 if more recent evidence suggests a need to support earlier grades. The number of total stipends provided by MoE under the revised program is expected to increase from no students to at least 100,000 over 4 years (Myanmar's education system includes about 8.2 million students). Townships will be selected based on drop-out rates, poverty indicators and other criteria to be agreed with the World Bank and Australia as part of the DLI process. 41. Improving the stipends program means introducing innovations from global experience, as well as improving the financial management of program funding. Assessments of the program have suggested a number of ways in which the program design and implementation can be improved. The national program allocates stipends to all primary and secondary grades. Evidence from a number of data sources from 2009-10 suggests that the fraction of students dropping out before grade 4 is small, and that dropout rates are particularly high at key transition moments from primary to middle school, and from middle school to high school. The revised program will therefore focus on children in grades 5-11, at the moments when dropouts appear to be the most prevalent. The stipends program may however be expanded to include grades 1-4 as well as grades 5-11 if more recent evidence suggests a need to support earlier grades. In addition, the stipend value will be progressively raised in lower-secondary and upper-secondary school to reflect the increase in opportunity and direct costs incurred as a child advances throughout the education system. As with the school grants above, innovations will be introduced to the program by revising its guidelines and by the provision of training. Specific innovations include: (i) introduction of well-defined program objectives and performance indicators; (ii) evidence- based selection of initial townships; (iii) evidence-based targeting of schools and students within townships; (iv) better definition and communication of conditionality and of the program in general; (v) community participation and oversight; (vi) introduction of provisions for audits; and (vii) linking program progress reporting to MoE's own information systems. 42. Component 3: Capacity improvement to strengthen monitoring and implementation of programs (US$4 million): This project will focus on training, and on conducting a baseline assessment of early grade reading. On training, MoE will design and begin implementing a national training program during school year 2014-15 aimed at introducing the new school grants and stipends program to township officials and school headmasters. Program content will be prepared as part of the process for the preparation of program guidelines. In the case of the school grants, it will also benefit from the example of similar training programs already introduced by UNICEF in Myanmar. The training program is expected to follow a cascade model used by UNICEF in which training providers are trained at the central level and are then responsible for delivering training sessions at the regional or township levels. Over 4 years, MoE is expected to deliver training to approximately 1,000 township education officers, assistant and deputy township education officers, level three accounting clerks and approximately 43,000 school head masters. 43. Assessment data, capturing student learning achievement and progress, is a critical building block for school planning and effective resource targeting. During project preparation, the Bank has assisted MoE with technical assistance and Australian funded an initial baseline for 14 EGRA in the Department of Basic Education 3 (Yangon area). The survey will be completed in January and a full report and public presentation will be completed by June 2014. Save the Children Myanmar also contributed technical assistance to this initiative and helped to manage the survey. During the 4 year project period, MoE will carry out baseline surveys in DBEs 1 and 2 (lower and upper Myanmar) as part of the project, providing a complete map of the distribution of children's early grade reading skills across the country. The Bank will continue to provide technical support through a parallel technical assistance program (see below). 44. Parallel Bank-Executed Technical Assistance Program: A Bank-executed technical assistance program in the amount of about US$2.5 million, for which financing is being sought from the MDTF, is expected to accompany the project activities. These resources would be used to augment regular Bank supervision activities through a technical assistance program, supporting process evaluations to help inform future revisions of program design, close monitoring of program implementation, analytical activities and surveys. The analytical work would complement MoE's regular monitoring of the programs. The analytical work would be aimed at helping to ensure that the programs supported are continually learning from their previous experience and improving over time. The cycle of learning and improving that is proposed would aim to build the capacity of authorities at all levels to increase the use of monitoring and data in program design and implementation and to show how this could be done during the project cycle. B. Project Financing a. Lending Instrument 45. The proposed project is an Investment Project Financing (IPF) in the amount of SDR 51.8 million (US$80 million equivalent) from IDA and US$20 million equivalent from the MDTF. The project follows a results-based financing approach with disbursements against an eligible expenditures program linked to measurable results, referred to as DLIs. It is expected to be implemented over a period of 4 years. It contributes to the Government's existing funding for school grants, student stipends, capacity building training and survey implementation. 46. It is expected that the grant resources will be secured by effectiveness and be available for joint-co-financing of the Project. A cross-effectiveness provision has been made in the Project Financing Agreement to ensure that the Recipient-Executed Trust Fund (RETF) Grant Agreement is signed and effective at the same time as the Financing Agreement. However, in case the co-financing does not materialize, the Project will be restructured to promptly commence implementation of the activities and either additional resources will be sought through potentially an IDA additional financing mechanism, or the project will be restructured to a three year project. b. Project Financing 47. The results-based approach was chosen as the best design alternative for making sure that MoE's focus will be on the results of the programs and building an information base and not merely on delivering inputs and funding to schools. The initial disbursement - expected to be US$19 million - will be disbursed following achievement and verification of the DLIs that are 15 expected to be met at the beginning of the 2014-15 school year (June 2014, i.e. Year 1). Subsequent annual disbursements, however, will be made following: (i) satisfactory proof of MoE having achieved a particular DLI; and (ii) documentation (in the form of Interim Financial Reports - IFRs) demonstrating that previous advanceslo have been utilized for eligible expenditure items as identified below. Such agreed eligible expenditure program are defined as specific budget headings and budget lines within those headings. A fund flows chart has been documented in the operational guidelines. 48. Eligible expenditures would be limited to the following: a) Spending by the "primary schoor', "middle schoor' and "high schoor' minor heads within the Departments of Basic Education 1, 2 and 3's budgets under the following "primary units" (budget headings under recurrent expenditures): 0201 Internal Travelling Allowance 0301 Labor charges for school level contract workers 0303 Renting vehicles or machinery 0304 Transport charges 0305 Office supplies 0307 Postage, telegram and telephone 0308 Electricity and power 0309 Books, Periodical and Newspaper 0313 Consumable expenditures 0314 Food supplies 0325 Exhibition and convocation charges 0409 Minor maintenance and repairs b) MOE budget headings for stipends (0501 transfers) to primary, middle and high school students, disbursed as direct cash payments to students; and MOE budget headings at the Department of Education Planning and Training and DBE1, DBE2 and DBE3 on 0506 (educative training) and costs incurred under minor head: "refresher training" to cover the costs of professional development to be undertaken at central, regional and township levels. 49. The main characteristics of the disbursement mechanism are: (i) each DLI is associated with a specific US$ value that will be disbursed upon achievement of the indicator; (ii) the DLI indicators can be achieved at any time but are sequential within results areas; (iii) MoE will provide evidence of the achievement of each DLI in accordance with a verification protocol (included in the Project Operations Manual - POM) detailing the technical details/standards of each DLI and the verification process to be followed; (iv) the Bank will administer an independent evaluation of DLI results reporting by contracting a firm or NGO; (v) MoE will provide evidence that spending against eligible expenditures has been incurred and reported in an amount greater than or equal to the value of the achieved DLI; and (vi) the World Bank will 10 That is to say that, in order for Year 2 disbursements to take place, satisfactory evidenceofYear 1 advance utilization has to be presented. The same logic is then applied to subsequent years. 16 disburse the amount triggered by the DLIs based on verification and reporting against the eligible expenditures. 50. The budget codes identified for eligible expenditures under the project are recurrent spending and, in most cases, non-procurable in nature. Moreover, simple local shopping procedures have been defined in the government's school grants guidelines. 51. During the fiscal years 2012/13 and 2013/14, spending for minor maintenance and repairs were not covered under the school grants program. However, as part of this project, the Bank has recommended that MoE consider allowing expenditure for such activities which are frequently allowable under school grants schemes globally. MoE is considering making this change to its programs, but it not ready to do so in the project's first year. In order to accommodate such a change to the government's programs during implementation, the MoE will include a simple environmental code of practice (ECOP) in its school grants guidelines to guide any 'small scale' renovation work such as repairing/cleaning windows, repairing a roof, repairing furniture, etc. No additions, expansion or demolition would be permitted under the school grants or any similar changes to the physical footprint of the school. The average size of a school grant is about US$800 per year and the maximum grant for a large school is US$1,800 equivalent in FY 2014/15. It is also unlikely that the entire amount of a school grant, which is recurrent in nature, would be devoted to one project or activity. These amounts define the 'small-scale' nature of any potential repair work to the school facilities. 52. Schools would be allowed to spend out of their school grants on small amounts of food supplies to facilitate meetings and consultations. Expenditure on food supplies under the project will not include spending on any large or recurrent student feeding or nutrition programs. The small size of the school grants, averaging US$800 annually per school, would not allow the financing of a regular feeding program, and it is unlikely that all grant spending in a school would be used for a single purpose. 53. Schools would also be allowed to spend small amounts on non-consultant services labor. This could entail payments for cleaners, minor services such as painting or small repairs, or incidental teaching assistants. Small services would be covered under the simple shopping procedure included in the school grants guidelines. Assistance with instruction, remedial tutoring or mentoring would be selected based on availability, skills (such as language skills) and experience. No service provided under the grants will be considered a permanent or regular assignment, and any invoice for services will, of necessity, be small. 54. DLIs will comprise improvements in the initial design of MoE's existing programs, improvements to be introduced during implementation (based on implementation experience and M&E findings), and the achievement of intermediate results against defined program objectives. The DLI program is described in Annex 2. Project Cost and Financing 55. The project's US$100 million contribution to Myanmar's programs represents about 51 percent of MoE's estimated US$198 million in projected costs for these programs over 4 years, as shown below. 17 Table 2: Estimated cost of programs to be supported for FY2014/15-FY2017/18 1 4 Year Program Of which IDA/TF Cost Financing % Programs (in million USD) 1. School Grants Program 160 77 48 2. Student Stipends Program 28 19 68 3. Capacity Strengthening Program 10 4 40 Total Costs 198 100 51 Source: World Bank staff estimates based on discussions with MOE counterparts C. Lessons Learned and Reflected in the Project Design 56. Although this project is the first World Bank engagement on education in Myanmar, the project design reflects a host of experience from other countries and from development partners in Myanmar. These lessons were gathered in two ways during project preparation. First, a series of workshops were held to discuss key program design issues; lessons from international experiences were presented and discussed, and a design note was developed by the Bank team to inform key program decisions. Secondly, the MoE and Bank team conducted a study tour to Cambodia to learn more about the Cambodian grants and stipends programs, financed through Australian-provided trust fund resources. 57. International evidence suggests that a key to the success of conditional cash transfer (CCT) programs in reaching their development objectives has been their use of a pilot-phase to improve design and implementation. The initial phase of roll-out is particularly critical to test out the design and monitor program implementation. Evaluations suggest that CCT programs have led to significant reductions in school dropout and repetition rates, and to substantial reductions in poverty. Across the world, these programs have started in pilot form in small, geographically concentrated areas with a limited coverage and have used this pilot phase to test out their design and improve program implementation before expanding further. The government has shown their commitment to a gradual roll-out of the stipends program through increasing the number of townships covered by the program over time. 58. International evidence suggests a number of areas where program design and implementation choices can affect outcomes in CCT programs. During the course of project preparation, evidence was put forward from international experience on how to improve program design. For example, evidence from impact evaluations of different conditionality designs put forward mixed evidence on the impact of conditionality on attendance and enrolments. International evidence suggests that monitoring and enforcing conditions can come up at some administrative cost, and can further complicate program design in low capacity settings. Although the current stipends program includes conditions that need to be met in order to receive the cash transfer, in practice these are not monitored or enforced in practice. Recognizing the higher costs of imposing hard conditions, the revised stipends program adopts a less administratively costly soft-conditionality that reflects local capacity constraints. In addition, a 1Myanmar does not yethave multi-year spending plans; budgets are approved annually by its parliament. As such, the figures in the table are estimates ofthe likely future size oftheseprograms, basedon discussions with counterparts. 18 streamlined monitoring information system suitable for the local capacity is being developed in order to allow for improved monitoring of conditions compliance by project management. Similarly, CCT programs have the greatest impact on dropout rates if they are targeted at the ages where dropouts are most likely to occur. The current stipends program in Myanmar covers all grades in the education system. Evidence from data analysis using household data from 2009-2010, however, suggests that dropouts are most likely to occur during key transition moments from primary to middle school, and from middle school to high school. The revised design of the stipends program therefore focuses on grades 5-11. The stipends program may however be expanded to include grades 1-4 as well as grades 5-11 if more recent evidence suggests a need to support earlier grades. IV. IMPLEMENTATION A. Institutional and Implementation Arrangements 59. The program is expected to be implemented over a 4-year period between June 2014 and June 2018. Disbursements are likely to follow a schedule revolving around the school's calendar (early June through end-March each year). Because the IDA and Trust Fund funding contributes to the ministry's own funding for existing programs, the overall responsibility for implementation rests with MoE departments and township officers. MoE will make funding available to townships and schools in accordance with operational guidelines to be agreed with the Bank and Australia as part of the DLI process. 60. MoE will establish a Project Steering Committee within the MoE, with a dedicated secretariat, which would be responsible for opening and maintaining the IDA and TF designated account for reporting to the Bank, including the arrangements of external audits. The Project Steering Committee is expected to be established by MoE under terms of reference (TOR) satisfactory to IDA prior to project effectiveness. The Project Steering Committee will be coordinated by the DEPT, which will appoint a Secretary to the committee who will be the Bank's primary liaison for program supervision. The Project Steering Committee will meet periodically to monitor program implementation. 61. Reporting against achievement of DLIs is expected to be completed by March 31 of each yearl2. This deadline will be preceded by an annual program review meeting between the Bank, Australia and MoE in which a detailed schedule for DLI evaluation and monitoring will be agreed. The verification of DLI reporting is expected to take place during the April-May period each year. The reporting and verification process will aim to confirm DLI achievement and authorize the annual IDA disbursement by July 31st of each year. 62. In parallel with the DLI reporting and verification process, MoE will provide the Bank with evidence of actual spending on eligible expenditures (covering the period April 1 through March 3113) by June 30 of each year beginning in 2015. The amount of each subsequent advance 12 Except the first yearwhich is expected by 31 August2014. 3 Except the first yearwhich covers the periodJune 1 through March31. 19 following the initial advance in 2014 will be subject to documentation of the previous advance against eligible expenditures and achievement of the relevant DLIs. B. Results Monitoring and Evaluation 63. The project will support capacity building for monitoring and evaluation through MoE's training programs, and will serve to make the decentralized funding flows an incentive for better performance and financial reporting. As mentioned above, and described in detail in Annex 1, MoE, the Bank and Australia will work together to monitor a focused results framework that corresponds to the DLIs. Most reporting will be integrated into standard statistical and finance reporting requirements within MoE, and some special reporting on stipends and school grants program actions and deliverables will be required under the project, as defined in the stipends and school grants guidelines. In parallel and subject to financing, the Bank would manage an analytical work program that would complement MoE's regular monitoring of the programs. The analytical work would be aimed at helping ensure that the programs supported are continually learning from their previous experience and improving over time. The cycle of learning and improving that is proposed would aim to build the capacity of authorities at all levels to increase the use of monitoring and data in program design and implementation, and to show how this can be done during the program cycle. The project will also support implementation of an EGRA, which will build capacity for measuring learning outcomes. Information on reporting of data related to DLIs, and the project in general, can be found in Annexes 1 and 2. C. Sustainability 64. Providing direct funding support for existing national programs is expected to promote sustainability and establish lessons to be used going forward both for education and for other sectors. Because the government currently finances the existing stipends and school grants program, fiscal sustainability of the program under the anticipated reforms depends on the government's capacity to absorb any additional costs. The additional costs of the stipends program when reformed (including expansion) suggests that fiscal sustainability presents some risk. This risk is partially mitigated by recent trends in public recurrent and social sector spending and by the willingness of international donors to support well-targeted and well- implemented education program. Further analysis of sustainability is provided in Annex 7. D. Gender 65. Although limited gender differences have been identified with respect to educational outcomes, there is scope to enhance and ensure gender equality within project implementation. During the implementation, the project will encourage the MoE to: (i) promote gender balance in the school committee. Female members will be encouraged along with the other committee members to receive training, provide input, implement and monitor the programs; (ii) include a discussion on gender in the safeguards related training modules; (iii) incorporate sex disaggregated data in all the key documents related to the school stipend program; and (iv) mainstream gender aspects in the design and analysis of project monitoring mechanisms such as in the beneficiary assessment, school and household surveys. 20 V. KEY RISKS AND MITIGATION MEASURES A. Risk Ratings Summary Table Risk Category Rating Stakeholder Risk Substantial Implementing AgencyRisk - Capacity High - Governance Substantial ProjectRisk - Design High - Social and Environmental High - Program and Donor Moderate - Delivery Monitoring and Sustainability High - Other(Optional) - Other(Optional) Overall Implementation Risk High B. Overall Risk Rating Explanation 66. Based on the Operational Risk Assessment Framework (ORAF), the overall preparation risk is high because of the risks of supporting national programs in Myanmar. The fact that these programs are already being implemented in a simplified fashion serves to limit preparation risk somewhat. The overall implementation risk is also high, primarily because of the financial and capacity risks associated with the project. Nevertheless, it is encouraging that MoE is already implementing the basic mechanics of these programs. 67. During discussions with government on the selection of townships where the revised stipends program would be piloted, the team has identified a new political risk, not identified at the project concept stage. There is substantial risk that political expediency - especially in terms of the need to distribute resources equally across states and regions - will drive decision making rather than technical rigor. In this particular case, the need for equal distribution across political regions is driving the government's desire to begin the stipends program in all areas of the country at once - in 15 widely dispersed townships across all state/regions - as opposed to beginning with a manageable pilot - 8 townships in close geographic proximity - and rolling out slowly based on results and experience to 12 and 20 townships in years 2, 3 and 4 respectively. The government has been open about this political imperative. The Bank has made the case to limit the scope of the stipends program in the first years. A potential compromise would be to limit selection risk under the stipends program by reducing the leeway given to townships and schools to undertake selection of stipends students in favor of earmarking stipends funds for certain categories of schools and covering all students in those schools. 21 68. The conflict-affected, ethnic areas will entail significant additional risks for the project, particularly for the stipends component. There is a substantial risk that ethnic political leaders, armed groups, and Civil Society Organizations (CSOs) could oppose efforts to provide stipends and/or school grants in ethnic minority areas. In some areas, the introduction of the project could exacerbate local tensions, and result in additional risk for beneficiary communities and local officials. There is also a risk that the project could become an issue of concern in the peace process, especially if ethnic leaders view the project as a potential threat to ethnic education systems. During the first year of the project, Australia and the World Bank will support new analysis of the conflict-affected areas, to systematically identify key risk areas, and ways to reduce these risks. Also during this time, continued extensive consultations (which were already started as part of the CPPF consultation process) with key ethnic leaders will be conducted to further understand their concerns and propose modest adjustments to program guidelines for these areas to reduce the risk of exacerbating conflict-related tensions. 69. A full description of risk and mitigation measures is presented in the ORAF. VI. APPRAISAL SUMMARY A. Economic and Financial Analysis (see full analysis in Annex 7) 70. The project partially finances two existing government programs (school grants and stipends); consequently, the contribution to economic development of the project's financing depends on the contribution to economic development of these programs. The project anticipates or stipulates various reforms (including expansion) to the programs. To appraise the project's contribution to economic development, the government programs-under the anticipated reforms-are assessed by the economic analysis on: (i) whether their monetary benefits exceed their costs; and (ii) whether their costs are fiscally sustainable. 71. The monetary benefits of both programs accrue through their contribution through human capital to increased productivity. The stipend program aims to increase educational attainment (a well-researched indicator of human capital) while the school grant program overtime will increase the quality of primary and secondary education which will subsequently contribute to increased cognitive ability (another well-researched indicator of human capital). These programs' anticipated impacts on an individual's human capital will generate benefits over his or her working lifetime while the costs of operating the programs are incurred immediately. To assess whether these benefits exceed costs, the economic analysis uses the most recent household survey data and administrative data as well as international experience to estimate an economic rate of return (a measure of the extent to which benefits exceed costs), a social discount rate (as a benchmark rate of return), the relationship between educational attainment and future earnings, and the public and private costs of educational attainment. The results of the cost benefit analyses for the stipend and grant programs are as follows: 72. Cost-benefit analysis of the stipend program: The stipend program aims to increase educational attainment. The contribution of educational attainment to human capital and subsequently to economic development is well researched. In addition, stipend programs as 22 conditional cash transfers have been the subject of numerous evaluations where, by design, their impact on a student's participation in schooling can be assessed. For example, among impact evaluations of CCT programs listed by Fiszbein and Schady (2009), impacts range from students being 1.05 to 1.33 times more likely to be in school, with an average of 1.11 times. 73. Analysis of available household survey data as well as administrative data suggests that the stipend program is an optimal investment even if the impact of the stipend program on participation is low by international standards (i.e., the economic rate of return of the stipend program exceeds the social discount rate). This is true under the current targeting method, but if this method were improved to better target the poorest, the net benefits would be even higher. In other words, even though there is ex ante uncertainty about the impact of the stipend program, the associated risks are mitigated by: (i) the program's net benefit being robust to low levels of impact internationally; and (ii) having a higher net benefit if better targeted. 74. Cost-benefit analysis of the school grant program: The objective of the school grant program represents intermediate steps to improving human capital. During the course of this project, school grants will deliver more reliable and transparent resources to schools. With this in place, the school grant program will, overtime, contribute to raising the quality of primary education by: (i) lowering the private costs associated with schooling; and (ii) by improving the school learning environment. International research suggests limited short-term impacts of school grants on education outcomes but highlight the importance of the anticipated reforms for their long-term impact. There are few randomized evaluations of the impact of school grants or school-based management interventions on education outcomes, and those that do exist suggest limited impact in the short-term. For example, an analysis in the United States found modest impacts on student achievement for the first five years and much larger impacts into the future (Borman et al 2003 in World Bank 2007). The anticipated reforms of the school grant program establish an important foundation that enable the grant program to have a positive impact on education outcomes in the future. 75. Analysis of available household survey data and administrative data as well as international experience suggests that, even if the impact of grants on education outcomes is long-term, the subsequent monetary benefits would still exceed operating costs. It also suggests that if the grant program offsets tuition fees as intended, the poorest would benefit the most from the grant program. 76. Fiscal sustainability analysis: Because the government currently finances the existing stipend and school grant program, fiscal sustainability of the program under the anticipated reforms depends on the government's capacity to absorb any additional costs. The additional costs of the stipend program when reformed (including expansion) suggests that fiscal sustainability presents an important risk. This risk is partially mitigated by recent trends in public recurrent and social sector spending and by the willingness of international donors to support well-targeted and well-implemented education program. Rationale for public intervention 23 77. Available data suggests that the average benefits of pursuing an additional year of schooling outweigh public and private costs for each grade of school. As a result, school leavers may be indicative of a market failure, and a public intervention such as the stipend program is justified. In general, the market failure that typically justifies public expenditure on education is a credit constraint: while a household's investment in a young child's education might yield higher earnings for the child in the future, these future earnings cannot in practice be borrowed against to finance present-day education expenditure. Value-added of World Bank and Australian support 78. A key pillar of the World Bank's ISN for Myanmar is to support government's efforts to transform institutions to allow them to deliver for citizens. This pillar is viewed as an important first step for economic development and poverty reduction, and the project is closely aligned with this pillar. Furthermore, the World Bank has considerable experience and knowledge on conditional cash transfers and school-based management which directly apply to the stipend and school grant programs, respectively. 79. As lead donor in the education sector in Myanmar, the Government of Australia has strong relationships with MoE. This is reinforced by Australia's role as an important bilateral donor partner for Myanmar with engagements across a range of sectors including closely related work with central agencies in governance including public financial management. Australia has considerable experience in school grant programs in the region and will bring contribute expertise to the development and management of this investment. Australia also has expertise and experience to contribute in conflict sensitivity. As mentioned previously, the Government of Australia is expected to support the proposed operation through a US$20 million contribution to the MDTF. B. Technical 80. Technical advice provided to MoE to help re-design and improve its school grants and student stipends program have benefited from the Bank's global experience in these two areas. Findings and research from numerous other countries have been presented to MoE's technical working group in a series of workshops during the project's preparation. Country experience presented included Cambodia, Vietnam, the Philippines, Mexico, Kenya and Indonesia. In addition, a group of 10 MoE leaders and technical experts participated in a study tour to Cambodia during preparation, which served to expose MoE to on-the-ground implementation and experience with school grants and stipend stipends programs. Lessons learned from the Cambodia visit and from global experience have been incorporated into MoE's own guidelines for both programs. An externally funded output (EFO) trust fund provided by Australia during preparation financed the workshops and study tour. 81. In addition, during preparation, MoE undertook a first iteration of an EGRA survey in its Department of Basic Education 3 (covering the Yangon region). This experience has provided an opportunity for MoE to gain first-hand technical knowledge in terms of the science of assessing reading acquisition skills, as well as organizational knowledge in terms of undertaking large scale assessment surveys in schools. Lessons learned from this iteration of EGRA are 24 expected to facilitate roll out of the assessment to MoE's other two DBEs during the project period. The EGRA survey in DBE 3 was also financed by Australia under the EFO. C. Financial Management 82. The overallfinancial management risk is assessed as 'substantial'. The main risks are the inexperience of MoE in managing and handling donor funded projects; inadequate documentation of policies and procedures (although the systems of manual internal control themselves are quite strong); that funds may not reach the correct beneficiary schools or, in the case of the stipends, caregivers and parents; and that some of the schools may use the funds for non-intended purposes. However, these risks are mitigated in several ways. The project supports training during the project period for township and school education mangers (township education officers and school headmasters) and for finance clerks. The capacity and oversight of staff is expected to increase during the project period. The programs being supported also call for regular audits of expenditure at the township and school levels, and there is a results indicator focused specifically on increasing use of school-level bank accounts in order to reduce the systems dependence on cash transfers. Lastly, the programs being supported introduce additional oversight of budgeting and spending by school committees and the transparency of budgets and spending. School budgets are expected to be displayed in the school. 83. The defined eligible expenditure program also reduces the risk, as expenditures will be limited to selected budget heads and the school grants and student stipend guidelines to be adopted will further clarify the eligibility and conditions for use of the funds. The Interim Financial Reports were agreed during negotiations, which will be in format and content similar to their own current financial reports, thus reducing potential reporting and disbursement delays in reporting against IFRs. The definition of the eligible expenditures will also make the annual external audit compliance easier. 84. It has been indicated in the recent PEFA report and International Monetary Fund (IMF) analysis that there is high level of integrity in terms of accuracy of financial records and financial statements in Myanmar despite using a manual system. The MoE's financial system is adequate, they have the required staff, internal control processes, and reporting in place that will be able to support the financial management requirements of the project. In implementation of the grants and stipends program, MoE will continue to use manual book keeping. The use of Excel processes may be used to support their accounting and reporting until the government introduces computerized accounting systems proposed under the Public Financial Management project. D. Procurement 85. Procurement under the project will be carried out in accordance with the World Bank's Guidelines: "Procurement of Goods, Works and Non-consulting Services under IBRD Loans and IDA Credits and Grants by World Bank Borrowers" dated January 2011 and the provision stipulated in the Financing Agreement. 86. Categories of expenditures to be financed under the program are mostly recurrent costs which will be subject to overall financial controls and accounting, and procurement is limited to consumables, supplies, utilities, materials, training costs, travel costs, and subsidies for poor 25 students. School grants funding will be allowed to be used for minor maintenance and repairs of facilitates or furniture (budget heading 0409 "Minor Maintenance and Repair"), which are described as small scale renovations. These would include, for example, cleaning facilities or furniture, repairing or cleaning windows, repairing a roof, repairing furniture, etc. No structural additions, expansion or demolition of facilitates are permitted under the school grants or any similar changes to the physical footprint of the school. The project will not finance consultant services. 87. The average annual school grant is expected to be about US$800 equivalent, with a small share of large schools (those with more than 400 students - currently about 1,700 schools nationwide) receiving a maximum annual grant of US$1,800 equivalent in FY2014/15. Purchases of non-durable goods and consumables, or work orders in the case of small scale renovations, will be made against invoices following the provisions in the operational guidelines and in most cases individual purchases would not exceed US$100 equivalent. No individual purchase or work order can be higher than the grant amount. Schools are not permitted under the grants program to spend against budget codes for capital expenditure, major civil works or durable equipment and goods, such as computers, printers, photocopies or other office or educational equipment. As mentioned above, however the program will support further analysis of procurement systems to support decentralization of school management in Myanmar. 88. Grouping of contracts at the township level is not feasible because it would involve different schools scattered over the township. Additionally, such an approach would be contrary to the concept of the school grants program in terms of the school managing its own funds. However, the Bank team and MoE will discuss options for future strengthening of the procurement system, for future operations. Such options will include a review of standards and procedures for works and equipment procurement. 89. The risk level for each individual school may be considered moderate given that the value of grants to the school is low, and that the subsequent invoices from these grants will be small. However, the overall project risk for procurement is substantial because: (i) huge number of schools will do procurement (purchase); (ii) there is not clear information on the procedures of procurement at school level; and (iii) this will be the first project financed by the World Bank in the sector. E. Social (including Safeguards) 90. Myanmar is reported to have more than 135 ethnic minorities who are present in various parts of the country. The project will primarily support two government programs, namely, the school grants program and stipends program. The school grants program is implemented nationwide, and the project will top up additional funds to the national program. The stipend program provides cash subsidies to needy students and the project will help expand the program in selected townships. While the project townships will be selected during implementation, some townships where ethnic minorities are present or have collective attachment will likely be selected. OP 4.10 is therefore triggered. 91. During the project preparation, a preliminary social assessment including free, prior and informed consultations was conducted in six townships to assess the issues around access to 26 education and the stipend program among vulnerable and ethnic groups, identify potential risks and impacts that may occur as a result of the project, and ascertain their broad community support. The government has agreed to provide support to 8 townships for the first year and in the process of selecting the areas. It is expected that the townships where the project will support the stipend program will be identified during implementation. The Community Participation Planning Framework (CPPF) was developed during preparation in line with OP 4.10 based on the result of the preliminary social assessment and consultations. CPPF sets out policies and procedures to ensure that the Social Assessments (SAs) including vulnerability assessment and free, prior and informed consultations with vulnerable and ethnic groups will be conducted during implementation, that their broad community support will be ascertained, and that the Community Participation Plans (CPPs) will be developed. SAs will be conducted annually in townships where the stipend program rolls out newly, and the CPPs will be developed on an annual basis at the national level based on the results. 92. This draft CPPF was consulted with a broad stakeholders on March 19, 2014. Both national and international NGOs as well as other agencies involved in education and ethnic minority issues were invited. Myanmar language and English copies of the draft CPPF and the executive summary of the preliminary social assessment were provided to the invited agencies two weeks before the consultation meeting. Overall, participants expressed strong support to the project, as well as measures provided in this CPPF to ensure that the project will benefit vulnerable and ethnic groups. MoE appreciated all comments and offers from stakeholders to help improve the programs going forward and encouraged participants to provide additional written comments. The key inputs from the consultations have been reflected in this CPPF, which include: * School grants program should consider: (i) extending the program support to monastic schools and community-based schools which often are located in the poor and remote areas; (ii) providing longer training and coaching to teachers to better develop and implement school improvement plans; (iii) putting in place a rigorous monitoring mechanisms for the program; (iv) avoiding using school grants as a substitute for parent/community contributions to the schools; and (v) making a full use of existing entities such as PTA and Board of Trustees (BOT) instead of creating a new committee for the program. * Stipends program should consider: (i) students below grade five could also be considered for the program; (ii) NGOs and International Non-governmental Organizations (INGOs) could collaborate with the Ministry to reach out to more disadvantaged students, and in particular disabled children; and (iii) a longer timeframe should be allowed for complaints - seven days may not be sufficient to allow people from remote areas to provide complaints. 93. Regarding the school grants program, MoE will ensure that all schools will receive grants as per national guidelines for school grants program, and that all schools will be treated fairly and transparently in the allocation of school grants. Furthermore, and in order to help ensure that more eligible students and their parents are aware of the stipend and school grants program, a 27 media campaign will be conducted as part of the campaign for the national Education for All policy. 94. The MoE is preparing operational guidelines for both programs which incorporate key principles of CPPF (these are DLIs 1 and 2). MoE has consulted with relevant stakeholders including representatives of vulnerable and ethnic groups on these national guidelines. The draft CPPF was consulted with stakeholders on March 19, 2014. No stakeholders who attended the meetings expressed concerns with the draft CPPF. The CPPF was officially adopted by the MoE and disclosed in-country on March 4, 2014 and at the Infoshop on April 2, 2014. A more detailed description of CPPF is provided in Annex 6. The preliminary Social Assessment carried out for the project was also disclosed in-country on March 4, 2014 and at the Infoshop on April 23, 2014. 95. The project will not support major civil works. The school grants program does not allow capital spending (such as major civil works or equipment procurement). The OP 4.12 is therefore not triggered. F. Environment (including Safeguards) 96. The activities supported by the project (school grants to finance operational costs of schools, a stipend scheme, and costs associated with training government officials) will have no environmental impact. The policy is triggered, however, since eligible expenditures will be used for minor maintenance and repair activities which would involve small scale rehabilitation works for existing schools within the same footprint (e.g., wall painting, window repairs, utility wiring installation, etc.); these could have minor localized impacts that could be managed well by known mitigation measures following site-specific ECOP. The OP 4.01 is also triggered because it covers some social impacts that are likely to occur but which are not covered under OP 4.10 or OP 4.12, such as potential exclusion of non-ethnic minority social groups from the stipend program and other benefits of the project. Measures to ensure that vulnerable groups will actively participate in and benefit from the project are included in the CPPF. 97. Since the school grant can be used for minor repairs and repairs in schools, the MoE has agreed to include a simple ECOP in its school grants guidelines during implementation. Capital expenditure, such as civil works or equipment (to include new construction, extension or major rehabilitation of facilities) will not be funded by the MoE grants program during the life of the project. 28 Annex 1: Results Framework and Monitoring Project Development Objectives PDO Statement The objective of the Project is to help improve and expand Myanmar's School Grants Program and Student Stipends Program. These results are at Project Level Project Development Objective Indicators .Data Responsibility Cumulative Target Values Sore fo Source/ for Unit of End Methodolog Data Collection Indicator Name Core M e Basele YR1 YR2 YR3 YR4 Frequency Measure Target y Number of students .Tonshis Township receiving payment in the Education revised stipends program Number 0 10,000 18,000 40,000 100,000 100,000 Annual reporting to Officer, DBEs, (DLI4, DLI 7, DLI 10) DBEs. DEPT Number of townships (out of 330) distributing Townships schoolgrantsto 80% of Number 0 50 100 150 200 200 Annual reporting to DBEs, DEPT their schools according DBEs. to formula (DLI5, DLI8, DLIl 1) Number of school heads FDE which have received training in project reporting on implementation in the Number 0 10,000 20,000 25,000 30,000 30,000 Annual number of DEPT programs (Not a DLI) people trained 29 Intermediate Results Indicators (for results areal: Expanded coverage of stipends) .Data Responsibility Cumulative Target Values Sote fo Source/ for Unit of End Methodolog Data Collection Indicator Name Core Baseline YR1 YR2 YR3 YR4 Frequency Measure Target y Guidelines All produced/ sections Secretary of revised for No Guidelines Guide- Guide- Guide- revisedProject Number lines lines lines to reflect Annual DEPT terojc stipends Guidelines produced ...Steering program revised revised revised lessons Committee in the field Number of students Township registered in the TwsisEducation revised in Number 0 10,000 18,000 40,000 100,000 100,000 Annual reporting to Offi D revised stipends DE. Officer, DBEs, program DEPT (Not a DLI) Intermediate Results Indicators (for results area 2: improved reliability and transparency of school grants program) .Data Responsibility Cumulative Target Values Sore fo Source/ for Unit of EdMtooo aaCleto Indicator Name Core Measure Baseline YR1 YR2 YR3 YR4 End Frequency Measure Target y Guidelines E] produced/ Guid- Guide- Initial Secretary of the revised for Text No lines lines Annual DEPT Project Steering school grants .lines prol gram ts produce revised res Committee program revised (DLI2) % of schools that disclose the Schools Township Education school grant Percentage 0 10% 30% 40% 60% 60% Annual reporting to Officer, DBEs, expenditure by TEOs. DEPT cateaorv on 30 school notice board (Not a DLI) TEOs % of large rEor schools which numbr o have opened and lar use a banklarge account to Percentage 0 20% 50% 60% 75% 75% Annual schools that DEPT receive their have school grant opened (Not a DLI) ant ac counts Intermediate Results Indicators (for results area 3: improved capacity to implement and monitor programs) . Data Responsibility Cumulative Target ValuesSore fr Source/ for Unit of End Methodolog Data Collection Indicator Name Core Measuoe Baseline YR1 YR2 YR3 YR4 Frequency Measure Target y Annual status report produced Yes/No No Yes Yes Yes Yes Annual DEPT DEPT (DLI6 and DLI9) Number of DEPT township reporting on officers trained Number 0 330 400 450 500 500 Annual number of DEPT in implementing people programs trained (Not a DLI) % of schools that have a Schools Township school Percentage 5% 7% 15% 30% 50% 50% Annual reporting tOfficer, DBEs, improvement TEOs. DEPT plan (DLI12) 1 Within year In 4th year Schools DEPT dronout rates of Droaram reDortina to 31 collected for TEOs beneficiaries of stipends program (Not a DLI) Results Indicators for core IDA results indicators (Longer term development objectives) .Data Responsibility Cumulative Target Values Sore fo Source/ for Unit of End Methodolog Data Collection Indicator Name Core Measuoe Baseline YR1 YR2 YR3 YR4 Et Frequency Measure Target y Number of Statistical ' Year Book beneficiaries 7,861, 9,062, 10,259, 8,684, 11,461,2 aresuot (number) of 333 801 509 829 37 and results fmerNumber 0 Annual indicators DEPT which femaleNubr 033 81 5929 7 on number (%) (Bank Core F: 50% F: 50% F: 50% F: 50% F: 50% of traie indicator)14 staf staff Primary EMIS data, completion rate Percentage 74% 75% 76% 77% 78% 78% Annual DEPT DEPT (MDG2) Gender parity 0.99 0.99 0.99 0.99 0.99 0.99 Annual EMIS data DEPT mndex (MDG3) Yes (all System for Yes Yes Yes three learning (EGRA (EGRA (EGRA DBEs Department of assessment at Yes/No No introduci Yes introducin introduci have Annual DEPT Myanmar the primary ng in g in ng in conducte Research Board level DBE3) DBE2) DBE1) d an I I_ I I I EGRA) 14 The project beneficiaries include students enrolled (growing at the rate ofnew grade 1 entrants), an individuals trained: headteachers and township officials 32 Project Development Objective Indicators Indicator Name Description (indicator definition etc.) Number of students receiving payments in the revised The number of students receiving stipends under the revised stipends program in the 40 stipends program project townships will increase from 0 to about 100,000 students by January 2017. The (DLI4, DLI7 and DLI 10) DLIs are contingent upon reaching 18,000 students in year one, and 40,000 students in year two. Share of townships (out of 330) distributing school grants Townships report that they distributed grants in accordance with the formula spelled out to 80% of their schools according to formula in school grant manual. (DLI5, DLI8 and DLIll) Number of school heads which have received training in MoE has trained 30,000 school head masters in implementing the programs, in project implementation in the programs (Not a DLI) accordance with the programs defined in DLI 3. Intermediate Results Indicators Indicator Name Description (indicator definition etc.) Guidelines produced/ revised for stipends program MoE has adopted stipends guidelines that (1) include clear objectives and indicators of (DLI 1) these objectives that will be monitored over time, (2) use objective criteria and clear procedures for targeting stipends funding to poor and vulnerable students, and (3) define financial management procedures. Number of students registered in the revised stipends This indicator will track how many students were registered according to the registration program processes laid out in the operational guidelines. (Not a DLI) Guidelines produced/revised for school grants program MOE has adopted guidelines that (1) spell out clear objectives and performance (DLI2) indicators to be monitored against these objectives and (2) strengthen financial management. % of schools that disclose the school grant expenditure by Schools report that they are posting details on the school grant on their school notice category on school notice board (Not a DLI) board. % of large schools which have opened and use a bank This indicator will track how many large schools (defined as schools with more than 401 account to receive their school grants (Not a DLI) primary students) have opened a bank accounts and use this account to receive their school grant. TEO will report on this and DEPT will summarize the information in the 33 annual monitoring reports. Annual status report produced MoE's first "status report" on the grants and stipends programs demonstrates progress in (DLI6 and DLI9) accordance with the objectives defined in the guidelines in DLIs 2 and 1, respectively, and focuses attention on lessons learned. Number of township officers trained in implementing DEPT will report annually on the number of township education officers and assistant programs (Not a DLI) officers, level 3 clerks and school headmasters provided with capacity building training for program management and financial management each year. % of schools that have a school improvement plan Schools will provide evidence to TEOs of PTA endorsement of their annualplanning (DLI 12) exercise. Townships will report the number and share of schools with school plans to DBEs and DBEs will report to the MoE. The indicator will be reviewed annually. Within year dropout rates collected for beneficiaries of Within year dropout rates collected for beneficiaries of stipends program. stipends program (Not a DLI) Number of beneficiaries (number) of which female In year 1, this is calculated as: the number of student enrolled plus target number of (%) (Bank Core indicator) people trained: head teachers and township officials. In year 2 and onwards, this is calculated as the previous years' beneficiaries plus the new cohort of first graders and additional people trained that year. Primary completion rate (MDG2) (core IDAresults This indicator will be measured the same way the ongoing multi-donor education project indicator) measures this indicator (and the targets have been set the same way): Percentage of pupils starting Grade 1 who reach the last grade of primary school (boys and girls). The main data source is the Ministry of Education's EMIS data Gender parity index (MDG3) Gender parity index (GPI) at the primary level is the ratio of female enrollment (in primary schools) to male enrollment (in primary schools). Data source: EMIS System for learning assessment at the primary level The indicator will track whether EGRAs are being rolled out in the three DBE (DBE1, 2 (EGRA) and 3). 34 Annex 2: Detailed Project Description Myanmar Decentralizing Funding to Schools Project 98. The project will contribute to the MoE's existing budget allocation in support of three specific programs: (i) expansion of MoE's existing national schools grants program that benefits all government-supported schools; (ii) expansion of its student stipends program, focusing on 40 townships over the four year project duration; (iii) strengthening of its capacity to monitor and implement these programs and lay the foundation for more focus on learning outcomes in future programs by conducting an early grade reading assessment (EGRA) baseline. The implementation period for the project is 4 years. The Bank will manage a parallel technical assistance program to support program design, monitoring and evaluation. Activities under the proposed operation will be financed at 80 percent by the IDA credit and 20 percent through funds from the MDTF. The specific components to be supported by the project are described above (paras. 38-44). 99. The Decentralizing Funding to Schools Project is an IPF which provides support for two existing government programs: school grants and stipend payments to students/parents. It also supports capacity building training in support of these two programs, as well as the roll out of an EGRA initiative. The design of these programs is summarized in the project description above and various significant implementation arrangements are provided in detail below. The detailed design of these programs will be contained in the government's operational guidelines for school grants and stipends and for training of township officers and headmasters. Final agreement on these documents will be conditions of the first annual approval of disbursement linked indicators. The DLI approach is being used in order to focus the government's and the Bank's attention on the results of these programs and on the need to improve these programs over time by reviewing intermediate results (contained in subsequent DLIs) and using findings and experience to introduce periodic improvements. This section will focus on how the DLI approach is expected to lead to improving these programs over time. Results-based Funding Arrangements 100. The IDA and MDTF funding will be made available to Myanmar on the basis of a results-based funding approach. As mentioned previously, the initial disbursement - expected to be US$19 million - will be disbursed following achievement and verification of the DLIs that are expected to be met at the beginning of the 2014-15 school year (June 2014, i.e. Year 1). Subsequent annual disbursements, however, will be made following: (i) satisfactory proof of MoE having achieved a particular DLI; and (ii) documentation (in the form of IFRs) demonstrating that previous advances have been utilized for eligible expenditure items. That is to say that, in order for Year 2 disbursements to take place, satisfactory evidence of Year 1 advance utilization has to be presented. The same logic is then applied to subsequent years. 101. The DLIs have been grouped into three results areas, which correspond to the three PDO indicators: (i) Number of students receiving payment in the revised stipends program; (ii) Number of townships (out of 330) distributing school grants to 80 percent of their schools according to formula; and (iii) Number of school heads which have received training in project 35 implementation in the programs. The DLIs are organized along approximate annual review periods, which are sequential in nature (e.g. within each area, earlier DLIs must be achieved before subsequent DLIs can be achieved). Annual review will be conducted, but the DLIs can be achieved and presented at any time during the project period as long they follow the sequential order indicated above. 102. The main characteristics of the disbursement mechanism are: (i) each DLI is associated with a specific US$ value that will be disbursed upon achievement of the indicator; (ii) the DLI indicators can be achieved at any time but are sequential within results areas; (iii) MoE will provide evidence of the achievement of each DLI in accordance with a verification protocol (included in the Project Operations Manual - POM) detailing the technical details/standards of each DLI and the verification process to be followed; (iv) the Bank will administer an independent evaluation of DLI results reporting by contracting a firm or NGO; (v) MoE will provide evidence that spending against eligible expenditures has been incurred and reported in an amount greater than or equal to the value of the achieved DLI; and (vi) the World Bank will disburse the amount triggered by the DLIs based on verification and reporting against the eligible expenditures. 103. A Bank-administered technical assistance program is proposed to accompany the project. It would focus on improving analytical feedback on various aspects of program design for the government school grants and revised stipends programs. Processes and design parameters that would be expected to evolve over time include: the processes and data used to select stipends program townships (geographic targeting); the processes and data used to allocate stipend quotas and school grants across schools within townships; the processes and data used to select individuals to receive stipends (individual targeting); documentation and information provision (to increase the visibility of both programs); reinforcing monitoring and evaluation procedures to allow the MoE to monitor the status of the programs (including data use for management purposes); feedback mechanisms (grievance mechanisms); and mechanisms to ensure the representation of underrepresented groups (ethnic minorities, poor households) in community oversight of the grants and stipends programs. 104. The technical assistance provided would be closely linked with the findings from regular monitoring of the programs, so that the project design is continually learning from its previous experience and improving over time. The cycle of learning and improving that is proposed would aim to help the government to increase the use of monitoring and data in program design and implementation, and to show how this can be done during the grants and stipends program cycles. 105. The technical assistance program would include the following: (i) Spot-checks to independently verify the status of the programs, as part of due diligence procedures before payments are disbursed against DLIs. These reviews will include verifying compliance with the operational guidelines regarding procurement and financial management. (ii) Annual school and household surveys would capture detailed information on school instructional and teaching equipment, and on the beneficiaries of the programs. The survey questions will be closely linked to key program design and implementation 36 features, and will aim to both evaluate their success and assessing their progress (e.g. in targeting the poorest students). These surveys would aim to be linked to an impact evaluation of the program that would be designed to assess whether the program is reaching its goals. (iii) Process or operational evaluations would assess implementation of the program, to identify procedures that aren't working well, and to get feedback from schools and townships. (iv) Technical advice and feedback for the roll out of the EGRA exercise to Departments of Basic Education 1 and 2. Disbursement-Linked Indicators 106. Three sets of DLIs were negotiated with government, reflecting the three primary government programs to be supported: school grants, stipends and building capacity to better monitor and implement programs. 107. The amounts assigned to each DLI reflect the relative size of each program. In the case of the school grants, the overall amount increases only slightly over time as an incentive to encourage increasing government support for the program over the four year period as the grants program is expected to expand. In the case of the stipends, the smaller but increasing amounts over the 4 year implementation period reflect the relative complexity and risk identified with the program's geographic and individual targeting, and the corresponding need to expand the program gradually. 108. Reporting against achievement of DLIs is expected to be completed by March 31 of each year 15. This deadline will be preceded by an annual program review meeting between the Bank, Australia and MoE prior to January 31 of each year in which a detailed schedule for DLI evaluation and monitoring will be agreed. The verification of DLI reporting is expected to take place during the April-May period each year. The reporting and verification process will aim to confirm DLI achievement and authorize the annual IDA and MDTF disbursement by July 30 of each year. In parallel with the DLI reporting and verification process, MoE will provide the Bank with evidence of actual spending on eligible expenditures by June 30 of each year beginning in 2015, normally covering the period from April 1st to March 31st. 16 The amount of each subsequent advance following the initial advance in 2014 will be subject to documentation of the previous advance against eligible expenditures. 109. The definition of the DLIs, the indicative timing for their achievement and the results indicators against which they will be evaluated are presented in the matrix below. A full description of the DLI protocols for evaluating the DLIs is included in the Project Operational Manual. The protocols provide a detailed process and institutional responsibilities for measuring, reporting on and verifying the DLIs. 15 Except for the first year, which is expected by 31 August 2014. 16 Except the first year which covers the period June 1 through March 31. 37 DISBURSEMENT-LINKED INDICATORS (DLIs) Results Areas: DLIs with DLI Target Achievement Dates and DLI Values'7 School Year 2014/15 School Year 2015/16 School Year 2016/17 School Year 2017/18 I. Expanded DLI 1. Student stipends: DLI 4. Stipends coverage: DLI 7. Stipends coverage: DLI 10. Stipends coverage: Coverage of Poor MoE has adopted the MoE has paid stipends to at MoE has paid stipends to at MoE has paid stipends to at Students in the Student Stipends least 18,000 basic education least 40,000 basic education least 100,000 basic education Student Stipends Guidelines that: (1) students in the Student students in the Student students in the Student Program include clear objectives Stipends Program in Stipends Program in Stipends Program in and performance accordance with the accordance with the Student accordance with the Student indicators to be Student Stipends Stipends Guidelines. Stipends Guidelines. monitored against these Guidelines. objectives; (2) use objective criteria and clear procedures for targeting stipends funding by educational and socio- economic status; and (3) define financial management procedures. DLI Value: US$7,000,000 DLI Value: US$7,000,000 DLI Value: DLI Value: US$3,000,000 US$2,000,000 11. Improved DLI 2. School grants: DLI 5. Improved DLI 8. Improved reliability DLI 1: Improved reliability Reliability and MoE has adopted the reliability and and transparency of school and transparency of school T)ransparency of School Grants Guidelines transparency of school grants: At least 150 grants: At least 200 School Grants that: (1) include clear grants: At least 100 townships have transferred townships have transferred objectives and townships have transferred grants to at least 80% of grants to at least 80% of performance indicators to grants to at least 80% of schools within their township schools within their township be monitored against schools within their in ac ordan e with the in accordance with the formu these objectives; and (2) township in accordance formula set forth in the set forth in the School Grants define financial with the formula set forth in School Grants Guidelines. Guidelines. II.Improved LI2.management procedures. the School Grants 17oD values will be financed as follows: 80% IDA credit and 20% MDTF flnds. 38 Results Areas: DLIs with DLI Target Achievement Dates and DLI Values17 School Year 2014/15 School Year 2015/16 School Year 2016/17 School Year 2017/18 Guidelines. DLI Value: DLI Value: US$21,000,000 DLI Value: US$21,000,000 US$16,000,000 DLI Value: US$19,000,000 III. Improved DLI 3. Training DLI 6. Program DLI 9. Program DLI 12. School improvement Capacity to program development: monitoring: MoE's first monitoring: MoE's second planning: ME has certified Implement and ME has adopted a annual status report on the annual status report on the that at least 50% of all schools Monitor School professional development school grants and stipends school grants and stipends have produced school Grants and training plan for the first programs has provided programs has provided improvement plans. Student Stipends year of Project updates on all progress updates on all progress Programs implementation aimed at indicators defined in the indicators defined in in the building capacity for School Grants Guidelines School Grants Guidelines and township education and the Student Stipends the Student Stipends officers, township Guidelines. Guidelines. financial management staff and school headmasters in the areas of school improvement planning, school grants and stipends program monitoring, and financial management. DLI Value: DLI Value: US$ 1,000,000 DLI Value: US$1,000,000 DLI Value: US$1,000,000 US$1,000,000 39 Annex 3: Implementation Arrangements Myanmar Decentralizing Funding to Schools Project (P146332) Project Institutional and Implementation Arrangements 110. The relevant departments DBE1, DBE2, DBE3 and DEPT of the MoE will be responsible for overall implementation and reporting. A Project Steering Committee will be created within the MoE, which would be responsible, through its dedicated secretary for opening and maintaining the IDA financing account, for supervising the external investment program and for reporting on DLIs and overall program progress to the Bank and other potential donors, including the arrangements of external audits. The Project Steering Committee is expected to be established by MoE under a TOR satisfactory to IDA prior to project effectiveness. The Project Steering Committee will be coordinated by the DEPT, which will appoint a Secretary to the committee who will be the Bank's primary liaison for program supervision. The Bank and MoE have agreed on a Project Operational Manual that describes the functions and composition of the Project Steering Committee, the DLI achievement procedures and arrangements for disbursement, financial management reporting, safeguards management, monitoring and evaluation and auditing. Project administration mechanisms 111. The project supports existing government programs which devolves funding to townships throughout Myanmar. The MoE will be responsible for monitoring and guiding implementation through its program guidelines, its training outreach and its supervision program progress reporting, including reporting for the DLIs. But TEOs and schools will be responsible for the administration of program funding. Townships and schools are, according to the Bank's field assessments to date, effectively implementing the basic mechanics of a simple schools grants program and a small, although countrywide, stipends program. MoE also has significant experience providing training for program administration on its own and as part of its ongoing UNICEF programs. The program guidelines will be used to introduce improvements to the programs in various areas including administration and reporting. The acceptability of the new guidelines will be the centerpiece of the first phase of disbursement-linked indicators. 112. The expansion of these programs at the township and school level over time will require continued upgrading of processes and capacity at these levels, particularly in the area of financial management and reporting. Such improvements will be addressed as part of the new guidelines and the roll of training. Specific attention to the needs of improved financial management as the programs scale up is provided below. 40 Financial Management, Disbursements and Procurement a. Financial Management (FM) 113. The overall FM risk for the project is assessed as substantial due mainly to: (i) inadequate documentation of policies and procedures (although the systems of manual internal control themselves are quite strong); (ii) inexperience with managing and implementing donor funded project; (iii) delays or non-review of financial reports when submitted; (iv) the risk that funds may not reach the correct beneficiaries schools, caregivers, parents and others on time; and (v) the risk that some of the schools may use the funds for non-intended purposes. 114. It has been indicated in the recent PEFA report and IMF analysis that there is high level of integrity in terms of accuracy of financial records and financial statements in Myanmar despite using a manual system. The MoE's financial system is adequate, they have the required staff, internal control processes, and reporting in place that will be able to support the financial management requirements of the project. In the implementation of the grants and stipends program, MoE will continue to use manual book keeping. The use of Excel processes may be used to support their accounting and reporting until the government introduces computerized accounting systems proposed under the Public Financial Management project. To support the project the MoE will prepare simplified instructions to document their flow of funds systems that are in place and would be used under this project including their reporting formats. 115. The financial management assessment concludes that, despite the identified risks and the fact that this is a results-based financing project, the FM arrangements for this project, including funds flow, reporting and auditing arrangements and action plans agreed with Government ensures that the project meets the Bank's FM requirement. 116. The table below identified the FM actions required. Required action Timing 1. Revised Guidelines for the school grants and stipends programs Issuance by 1 June 2014 in accordance The guidelines forthe schoolgrants and stipends program will describe the with DLIs l and 2. roles and responsibilities ofrelevant authorities at each level of MoE and schools. These guidelines will complement existing MoE operational and financing guidelines available to all schools. The specific guidelines for the school grants and stipendsprogramwill complement, where necessary, the government's existing operational and financial management guidelines. 2. IFR template Completed Developmentand agreement of templates of the Interim Financial Reports and Annual Financial reports between the Bank and MoE. 3. Internal Audit Mobilization with in 6 months after Guidance to be provided to MoE by World Bank FM staff as part of signingofthe FinancingAgreement supervision to help improve Internal Audit capacity and to carry out IA activities on an agreed scope in townships and schools. 4. Financial Audit Agreemen ta lready reached with OAG. The FinancialAudit performed annually by the OAG. Audit programs and During implementation, we will plans will outline the coverage of schools having received school grants monitor the program to confirm from Government. coverage and scope. 41 117. The Secretary of the Project Steering Committee, coordinated by the DEPT is responsible for financial management and reporting. 118. The project will be required to prepare half-yearly IFRs. It is expected that these reports will be prepared by the individual DBE finance departments, consolidated by the Committee Secretary and one project IFR sent to the Bank in accordance with the templates agreed with the Bank at the negotiations. The IFR will be expected to be provided within 45 days after each calendar semester to which it relates. 119. The project financial statements will be audited by the OAG of Union of Myanmar, as they have been assessed as capable of carrying out audits of the Bank funded projects. The audited financial statements and auditor management letter will be submitted to the Bank within 6 month after the end of the financial year. 120. Under the project, MoE will be expected to undertake periodic Internal Audit of townships and schools and summarized copies of their half-yearly reports will be shared with the Bank. 121. The financing for each DLI is expected to be allocated to IDA and the MDTF in the share of 80 percent and 20 percent, respectively. b. Disbursements 122. Project funds will be advanced to the MoE in US dollars through a pooled (IDA and TF) account maintained in the Myanmar Economic Bank with a Variable Ceiling. The Myanmar Economic Bank will then convert the funding into Kyat and transfer it to DEPT and DBEs which will further channel the funds to townships and schools using MoE systems and procedures. As previously stated, the advance of funding for the first year will be contingent upon the achievement of the DLIs only. The advance of funding for subsequent years will be contingent upon the achievement of DLIs and documentation of advances previously made which will be based on the Interim Financial Reports. 42 123. The following table is the planned disbursement schedule. IFR reports to be used as Amount DLI supporting documentation (USD verification DLI mil) report verification Reporting (IDA World Review submission discussed by Period to be Disbursement and Bank Period date Bank covered Due Date date MDTF) FY 31 August- 15- 30 1 14 September-14 September-14 19 15 1 June 14- 30-June- 2 31 March-15 31-May-15 31 March 15 15 31 July-15 23 16 1 April 15- 30-June- 3 31 March-16 31-May-16 31 March 16 16 31 July-16 29 17 1 April 16- 30-June- 4 31 March-17 31-May-17 31 March 17 17 31 July-17 29 18 Docume 1 April 17- 30-June- ntation 5 31 March 18 18 31 July-18 only 19 DLI reporting and verification 124. Disbursements against DLIs will follow a clear annual process. The project is expected to become effective and make its first advance in September 2014 based on confirmation by the Bank that DLI 1, 2 and 3 have been met. For subsequent annual review periods, MoE would submit a report on achievement of the DLIs not later than March 31 of each year. Based on this report, the Bank would undertake 'spot check' verification audits of the DLIs achieved during the previous school year (June 1-March 31). The results of the verification audit reports would be discussed with government by May 31 of each year. The Bank would confirm evidence of DLI achievement and expenditure against agreed eligible expenditures by 31 July of each year. On confirmation, the MoE can submit their request for their next advance, which will be supported by the IFR documenting how the initial advance has been spent on the eligible expenditures, reconciliation of the DA and other documents specified in the Disbursement letter. Delays in submission of DLI reports and verification, as well as the documentation of prior advances, would delay the subsequent releases of funds following the initial disbursement. Eligible expenditure reporting and verification 125. The Bank will advance funds based on the achievement of the DLI's in the first year and on the achievement of the DLIs and the reporting of eligible expenditures in subsequent years. The first advance would be expected in September 2014. For subsequent review periods, MoE will provide the Bank with IFRs against eligible expenditures by June 30 (for the previous Myanmar fiscal year). Following the initial advance, replenishment will depend on completion and verification of the DLIs (see above paragraph) and the documentation of the previous advance. The total amount to be advanced, for each Application for Withdrawal, will be equal to 18 Except the iirst year which covers the period June 1 through March 3 1. 43 the sum of value of the individual DLIs achieved during that year less any undocumented advances at that time. Once the undocumented advances are subsequently documented, the value of the DLI's withheld may be advanced. The template of the IFRs has been agreed with MoE during negotiations. Eligible expenditures will be formally audited and the report submitted to government and the Bank by September 30 each year, starting in 2015 (six months after the March 31 close of the Myanmar fiscal year). Any discrepancies between the audited accounts and the interim financial reports as well as any ineligible expenditure identified would need to be refunded to the Bank. 126. Guidelines for the school grants and stipends programs will clarify the roles and responsibilities of the concerned local and central levels in terms of financial management. 127. The project will have a Disbursement Deadline Date (final date on which the Bank will accept applications for withdrawal from the Recipient or documentation on the use of Credit/Grant proceeds already advanced by the World Bank) four months after the closing date. This "Grace Period" is granted in order to permit the orderly project completion and closure of the Credit/Grant account via the submission of applications and supporting documentation for expenditures incurred on or before the Closing Date. Expenditures incurred between the Closing Date and the Disbursement Deadline Date are not eligible for disbursement, except as otherwise agreed with the World Bank. 128. All documentation for expenditure submitted for disbursements will be retained by MoE and be made available to the external auditors for their annual and interim audits, and to the World Bank and its representatives if requested. Should the auditors or the World Bank implementation support mission find that disbursements made were not justified by the supporting documentation, or are ineligible the World Bank may, at its discretion, require the Recipient to: (a) refund an equivalent amount to the World Bank; or (b) exceptionally, provide substitute documentation evidencing other eligible expenditures. The Bank may also withhold disbursement of additional deposits into the DA (if the ineligible expenditures were paid from the DA). If the Recipient does not comply with any of the requirements stated above, within 12 months following notice by the World Bank to the Recipient of the requirement, the World Bank may (a) decline to allow the use of DA under new Credits/Grants made to or guaranteed by the Recipient, (b) establish compliance as a condition for Board presentation of new Credits/Grants, (c) suspend disbursements under the Credit/Grant; or (d) in extreme cases, suspend disbursements under all ongoing Grants/Credits to or guaranteed by the Recipient. c. Procurement 129. Procurement under the project will be carried out in accordance with the World Bank's "Guidelines: Procurement of Goods, Works and Non-consulting Services under IBRD Loans and IDA Credits and Grants by World Bank Borrowers" dated January 2011 and the provision stipulated in the Financing Agreement. Procurement methods allowed for the purposes of procuring goods, works and non-consulting services under the project include: (i) International Competitive Bidding; (ii) shopping, (iii) direct contracting; and (iv) community participation. 44 130. Spending under the project will be limited to categories of expenditures (as reflected in eligible budget codes) and procurement is limited to office supplies, consumables, small scale maintenance (renovation), etc. which do not involve the acquisition by schools of durable goods, civil works or services. Rather, those expenditures will be made in respect of recurrent expenditures and consumables, such as for supplies, utilities, maintenance, subsidies for poor students and training for teachers. Furthermore, there will be no selection of consultants. 131. Due to the nature of the programs being supported, which consists of very small value purchases or work orders (normally less than US$100 equivalent per invoice) carried by a large number of schools at the township level, it is not feasible for MoE to prepare a procurement plan for the program. Decisions to conduct purchases or work orders are made by school officials, led by a headmaster. To date, this has been done with limited transparency through an informal (and undocumented) solicitation for prices. 132. Under the new school grants guidelines, individual schools will prepare their own budgets in a standard format. Budgets will be approved by the school committee which include parent representatives and posted publicly on an information board in each school. Specific procedures of procurement (quotations) will also be specified in the School Grant Guidelines which would be acceptable to the World Bank. Annual expenditure reports will also be signed off by the school committee, which has community representation. The community involvement is expected to contribute a reasonable level of accountability. This will be further enhanced by the development of budgets that will be made public by posting on school notice boards. At the district level, the district management team will monitor and provide guidance to schools under their jurisdiction. The institutional arrangements under the project will be further strengthened through training programs to ensure that the schools are purchasing in line with eligible expenditure categories and that spending is publically transparent. The documentary records provide the amount of funds used for different items as part of financial management procedures. 133. Grouping of contracts at the township level is not feasible because it would involve different schools scattered over the township. Additionally, such an approach would be contrary to the concept of the school grants program in terms of the school managing its own funds. However, the Bank team and MoE will discuss options for future strengthening of the procurement systems for future operations. Such options will include a review of standards and procedures for works and equipment procurement. 134. The risk level for each individual school may be considered Moderate given that the value of grants to the school is low, and that the subsequent procurement contracts each school will award financed from these grants will be small. However, the overall project risk for procurement is substantial because (1) huge number of schools will do procurement (purchase); (2) there is not clear information on the procedures of procurement at school level- and (3) this will be the first project financed by the World Bank in the sector. 45 Environmental and Social (including safeguards) Social 135. The Constitution and the government's education sector policy grant equal rights for all citizens including ethnic minorities in accessing in education. The existing laws and regulations being enforced in Myanmar are considered insufficient to identify and mitigate the project's potential negative environmental and social impacts according to the World Bank's safeguards policies. Myanmar has more than 135 ethnic groups living widespread throughout the country. 136. Consequently, the project triggers the following World Bank safeguards policies: * Environmental Assessment OP 4.01 * Indigenous Peoples OP 4.10 137. The potential environmental impacts are not likely to be significant. Therefore, no environmental assessment is required. However, because the project covers ethnic minority areas, OP. 4.10 is triggered. In addition, OP 4.01 is triggered because it covers some social impacts that are likely to occur but which are not covered under OP 4.10 or OP 4.12, such as potential exclusion of non-ethnic minority social groups from the stipend program and other benefits of the project. Therefore, in accordance with the requirements of Environmental Assessment OP/BP 4.01, the project has been assigned environmental category B. Social issues were appraised in the SA conducted during preparation. The SA focused not only on the indigenous peoples but all beneficiaries of the programs, focusing more attention on the selective stipends program. Findings and recommendations of the SA are included in the Community Participation Planning Framework (CPPF) which is summarized in Annex 6. The CPPF includes measures to ensure that vulnerable groups will actively participate in and benefit from the project. The CPPF was officially adopted by the MoE and disclosed in-country on March 4, 2014 and at the Infoshop on April 2, 2014. The preliminary Social Assessment carried out for the project was also disclosed in-country on March 4, 2014 and at the Infoshop on April 23, 2014. Environmental 138. As mentioned above, project activities are not expected to have any significant environmental impact. The project finances small grants to finance school's operational costs, stipends to poor students, and costs associated with building capacity of government staff. The grants program supports schools in purchasing consumables and covering operational expenses. No significant works or equipment procurement is permitted under the grants. Minor repair and maintenance of facilities and furniture were not part of the government's grants program in FY2012/13 and FY2013/14, but it will be included as an eligible expenditure in this project. To cover this, the government will agree to include a simple ECOP in its school grants guidelines during implementation. Monitoring & Evaluation 46 139. The project will support capacity building for monitoring and evaluation through MoE's training programs, and will serve to make the decentralized funding flows an incentive for better performance and financial reporting. As mentioned above, and described in detail in Annex 1, MoE, the Bank and Australia will work together to monitor a focused results framework that corresponds to the DLIs. Most reporting will be integrated into standard statistical and finance reporting requirements within MoE, and some special reporting on stipends and school grants program actions and deliverables will be required under the project, as defined in the stipends and school grants guidelines. In parallel, the Bank will manage an analytical work program that would complement MoE's regular monitoring of the programs. The analytical work would be aimed at helping ensure that the programs supported are continually learning from their previous experience and improving over time. The cycle of learning and improving that is proposed will aim to build the capacity of authorities at all levels to increase the use of monitoring and data in program design and implementation, and to show how this can be done during the program cycle. The project will also support implementation of an EGRA, which will build capacity for measuring learning outcomes. Information on reporting of data related to DLIs, and the project in general, can be found in Annexes 1 and 2. 140. Myanmar has an EMIS which relies on forms filed out (on paper) monthly by schools, and aggregated by township education officers. The results are summarized annually in a statistical year book. Although schools provide regular information, there is no "database"; the data are collected in hard copies and then computerized and entered into a MS excel file (which is mainly used for the purposes of printing out statistical yearbook). The ongoing multi-donor QBEC project aims at strengthening the national EMIS system by strengthening township level capacity to report on and use "township level EMISs. This project will contribute to strengthening the use of data in the education system- and, hence, the EMIS-in a number of ways: first, because the school grants formula will be based on data generated from the EMIS system. With grant size tied to data provided through the EMIS, there will be increased focus on improving timeliness and quality of the data reported. Second, the project will generate baseline data on first, second and third graders' ability to read (by supporting the ministry in administering EGRA). Third, the stipends program will require more use of EMIS-style (and other) data as part of the selection process (ie which townships to select, and which schools to focus on). 47 Annex 4. Operational Risk Assessment Framework (ORAF) Myanmar: Myanmar Decentralizing Funding to Schools (P146332) Stage: Appraisal Risks StakeholderRisk Rating Substantial Risk Description: Risk Management: The stakeholders are Government at The project will include a social assessment/community participation plan. Improvements to the central, township and school levels, programs will include modes of consultation with communities, head masters, parents and teachers. community/village leaders, parents, During implementation, periodic surveys and consultations will form a major part of monitoring and teachers and children ages 5-16. evaluation. Government owns and is committed to the Resp: Both Status: Not Stage: Both Recurrent: Due Frequency Year existing school grants and stipends Yet Date: ly programs because it is currently Due implementing them with no external support. The recipient and Bank interests, objectives, and motivation for undertaking this operation are well aligned. Stakeholder risk in Myanmar is, however, still considered substantial One major stakeholder risk is the role of communities, as represented by PTAs or school committees, in helping to oversee the funding and, in the case of stipends, in the selection of stipends students. Local consensus and full participation could pose unique problems in mixed ethnic communities. Capacity Rating H 48 Risk Description: Risk Management: MoE is implementing simple school grants As previously stated, building capacity for financial management by townships will be a central focus and stipend mechanisms. The funding of the project. Training will be provided to township officers annually. flows are established. The bottleneck for scaling up and introducing improvements Funding for these programs will be disbursed through the DLIs based on demonstrated capacity of to these mechanisms will likely be at the townships to manage and report on financial flows, as well as to report on indicators defined in the township level. In particular, townships results matrix. lack qualified accountants do not use modern financial management tools. There The MoE will be expected to update its Operational/Financial Management Manuals to guide program is a risk that financial documentation, implementation at the township and school levels. These manuals will, which currently exist, document reporting and cooperation with auditors policies and procedures for, among other things, routine tasks, administrative activities, records will be weak at the township and school management, and documentation. levels. Resp: Client Status: Not Stage: Imple Recurrent: Due Frequency Year Yet menta 0 Date: : ly Due tion Governance Rating Substantial Risk Description: Risk Management: There is strong ownership by the The objective of the project is to improve the management and governance of existing programs. Government of these programs and the Indicators of improvement will be included in the results matrix and as part of the DLIs. project's goals are fully aligned with the Resp: Client Status: Not Stage: Imple Recurrent: Due Frequency Year newly evolving sector strategy. The Yet menta Date: : ly complexity of Myanmar's governance Due tion system still makes the risk substantial. Design Rating High Risk Description: Risk Management: The basic mechanism of the funding flows Disbursements for the school grants program will be released based on DLIs what demonstrate better for school grants will be maintained. The fiduciary performance and reporting. Limiting this program to only 8 townships in the first year and objective of the project is to introduce rolling out in the second and third years to another 12 and 20 townships, respectively, is expected to improvements and innovations over time make the newly designed program feasible. To limit selection risk under the stipends program it is that will help to improve the impact of the proposed to reduce the leeway given to townships and schools to undertake selection of funding and the transparency and stipends students in favor of earmarking stipends funds for certain categories of schools and accountability of the funding. There is a covering all students in those schools. high risk that adopting changes and innovations will be slower than expected Professional development of township and school officials will be a trigger for expanding both 49 and uneven across various townships. The programs. mechanism for targeting student stipends will be made more rigorous in selected The experience with undertaking EGRA in DBE3 during project preparation will provide a basis for townships and more students will receive expanding EGRA to the other two basic education departments. thebenefit. These new procedures and Resp: Both Status: Not Stage: Both Recurrent: Due Frequency Year responsibilities could be difficult for MoE Yet Date: : ly to manage and monitor. Due There is also a high risk that MoE's capacity to manage a major professional 0 development program for townships and school Head Masters will be a bottleneck to introducing change and innovations. Social and Environmental Rating High Risk Description: Risk Management: There is no environmental risk under this SA and consultations will be introduced to MoE as an innovation for expanding the school grants and project. It is focused on financial flows, stipends programs. small recurrent expenditures, training and surveys. There is, however, high social risk, mostly SA will be carried out during preparation, and a Community Participation Planning Framework (CPPF) with regard to the targeting and school has been prepared prior to appraisal which will set out processes and procedures to conduct free, prior level management of the stipends program and informed consultations with affected ethnic minorities leading to their broad community support in conflict-affected ethnic areas. The for the project. International Professional Practices Framework (IPPF) will also set out capacities of PTAs and school committees mechanisms, including arrangements to address grievances, to ensure participation of affected ethnic to participate in overseeing school planning minorities in project implementation at the school level. and spending is uncertain and will only improve over time. There will be analysis of the conflict-affected areas carried out to help systematically identify key risk areas and propose ways to mitigate/reduce these risks. In addition, the project will be operating in townships nationwide where many of these There will continue to be extensive consultations with key ethnic leaders, to understand their concems townships have multiple ethnic groups and propose modest adjustments to program guidelines for these areas to reduce the risk of residing together. The project would need exacerbating conflict-related tensions. to ensure that ethnic groups receive equal- toc n enic groups reev aeq Resp: Client Status: Not Stage: Imple Recurrent: Due Frequency Year social and economic benefits that are culturally appropriate and gender-inclusive. Yet menta Date: ly 50 Due tion Program and Donor Rating Moderate Risk Description: Risk Management: Promoting decentralization and improving The project will build on the efforts of other donors who are currently involved in providing direct financial decentralization are part of the support to townships and schools in Myanmar. Others donors and NGOs will be expected to include MoE's sector strategy. The government's support for these government programs in their own capacity building work. For example, UNICEF continued commitment to decentralization training of township officers or head masters should take into account their roles in managing these of financing for the sector only presents a programs. Similarly, MoE will use examples from UNICEF programs to help improve its own moderate risk. programs. Resp: Both Status: Not Stage: Both Recurrent: Due Frequency Year Yet 0 Date: : ly Due Delivery Monitoring and Rating High Sustainability Risk Description: Risk Management: There is a high risk that government The project will support capacity building capacity in M&E and provide incentives to better information systems will not produce performance and financial reporting. But, given the low capacity of MoE to produce timing data flows, timely information with which to monitor the project will provide independent verification of the disbursement linked indicators and of the results and evaluate programs and make timely framework in general. Subject to funding availability, the Bank will undertake a parallel program of decisions on disbursement through DLIs technical assistance to support monitoring and evaluation. . . Resp: Both Status: Not Stage: Both Recurrent: Due Frequency Year There is only a low risk that Myanmar will Yet Date: l not have the financial capacity to continue Due E D ly to fund these programs going forward. Other (Optional) Rating Risk Description: Risk Management: Resp: Status: Stage: Recurrent: Due Frequency Date:: Other (Optional) Rating Risk Description: Risk Management: 51 Resp: Status: Stage: Recurrent: Due Frequency Date: Overall Implementation Risk: High Risk Description: The overall implementation risk remains high because of the uncertainty of government capacity to improve their systems in a timely manner and to report on and evaluate these improvements. 52 Annex 5: Implementation Support Plan Myanmar Decentralizing Funding to Schools A. Strategy and Approach for Implementation Support 141. The implementation support strategy for this intervention will comprise regular dialogue with the MoE, semi-annual joint review of program implementation, and regular oversight of project fiduciary activities, which includes detailed analysis of protocols to verify compliance with DLIs. Regular dialogue and implementation support will facilitate early identification of problems and obstacles and will enable timely provision of technical advice and support to address any obstacles. Joint reviews will take place twice a year, aimed at reviewing the progress and achievement of agreed results. Linking disbursements to the achievement of specific development indicators will help to focus Bank supervision on education development issues. 142. As part of these reviews, lessons learnt from implementing the guidelines will be drawn out and sought incorporated into future revisions of the guidelines and further training of township officers and school heads. 143. With regard to the supervision of fiduciary aspects of the operation, the World Bank will conduct financial management supervisions at appropriate intervals. As a member of the task team, an accredited Financial Management Specialist will participate in regular project supervision. Additionally, the Bank's team will continue to work with MoE to further improve its financial and budget management capacity and performance, including its ability to monitor and oversee recurrent education expenditures which are delegated to townships and schools. During implementation, the Bank will supervise the financial management arrangements for school grants and stipends through: (i) reviewing the program's quarterly interim financial reports as well as the annual audited financial statements and auditor's report and remedial actions recommended in the auditor's management letters; and (ii) reviewing budgeting and financial planning arrangements, disbursement management and financial flows, as applicable. 144. The table below presents the implementation support plan for the project. B. Implementation Support Plan 145. Implementation timeline and support Time Focus Skills needed Resource estimate Partner role 4 year Implementation ofthe Education Specialist Australia funding program revised school grants Economist of Trust Fund program Operations Support Implementation ofthe Social Protection or stip ends program Education Specialist Implementations Systems Specialist/Consultant Monitoring of M&E Education compliance under the Specialist/Economist/ stipends program Implementation Systems I Specialist/Consultant 53 Time Focus Skills needed Resource estimate Partner role Supervision of the Education training activities and Specialist/Professional training content development consultant Implementation of a Education Specialist/PR public awareness Consultant campaign Fiduciary oversight: Education Specialist Protocols to verify compliance with DLIs Financial Management, Financial Management including support for Specialist MoEcapacity development Survey res earchunder Education WB-executed TF specialists/Economist Survey research consultants 146. Skills mix required Skills needed Number of staff weeks Number of trips Comments TaskTeamLeader 20 6 Washington based Human Resources Operations 12 0 Local World Bank staff Education Economist 20 8 Bangkokbased SP Economist 8 4 Bangkokbased EGRA Specialist 12 6 Washington based Implementation Systems 4 2 Specialist/Consultant Operations Analyst 18 0 Local World Bank staff Program Assistant 18 0 Local World Bank staff PR Consultant 4 2 Financial Management Specialist 10 4 Trips to be combined with otherproject support 54 Annex 6: Community Participation Planning Framework Myanmar Decentralizing Funding to Schools A. Summary of the Environmental and Social Safeguards Approach 147. The CPPF has been prepared and adopted by the MOE. The CPPF provides general guidelines and procedures to ensure that the project is implemented in a socially sustainable manner and in line with OP 4.01 and OP 4.10. Specifically, the CPPF aims to: (i) ensure that the poor, vulnerable and ethnic groups living in the participating townships are meaningfully consulted and that they receive project benefits in a culturally appropriate manner; and (ii) provide those excluded from the stipend program avenues to raise concerns and means for redress. B. Principles of CPPF 148. The national guidelines will be developed for the stipend and school grants programs, separately, in line with this CPPF. The guidelines, when developed, will be widely consulted with key stakeholders in the country including but not limited to ethnic minority communities. 149. The CPP will be developed, and then updated annually as the stipend program rolls out to new townships, in line with the provisions of this CPPF. 150. As part of the preparation of the CPP, a Social Assessment (SA) will be conducted which includes vulnerability assessment to identify groups of people whose socioeconomic standings in local communities may subject them to risk of exclusion from stipend program. Free, prior, and informed consultation will also be conducted as part of SA with potential beneficiary groups of the stipend program leading to their broad community support to the stipend program. 151. The public in general, and students eligible to the stipend program and their parents in particular, will be adequately informed of the objective, eligibility criteria and selection procedure of the stipend program. 152. Selection criteria will be refined so that a broad range of poor and vulnerable social groups become eligible to the stipend program. 153. Mechanisms will be set up to address grievances accessible to project affected people and monitor implementation of this CPPF. C. Institutional Arrangements for CPPF Implementation 154. The DEPT and the relevant DBE under the MoE will assume overall responsibility for the implementation of this CPPF. Regarding the stipend program, DBEs will carry out many activities provided under this CPPF in collaboration with TEO and Township Program Working 55 Group (TPWG)'9 at the township level, as well as beneficiary schools. DEPT will pull together data and inputs collected by DBE at the township level and develop, and annually update, the Community Participation Plan (CPP), Detailed implementation arrangement will be spelled out in the national stipend program guidelines. Regarding the school grants program, DEPT will ensure that all schools will receive grants as per national guidelines for school grants program, and that all schools will be treated fairly and transparently in the allocation of school grants. D. Implementation procedures of CPPF Step 1 - National Guidelines for the Stipend Program and School Grants Program will be developed Step 2 - The TPWG, and the School Project Committees (SPC) that will include representatives of vulnerable people, ethnic minorities and female parents, will be established. Step 3 - The MOE will provide training for DEPT and DBEs to implement the CPPF. The responsible DBE officers will further provide appropriate training to TEOs, TPWGs and school headmasters on core CPPF principles. Step 4 - National awareness campaign will be conducted and programs brochures will be distributed in Myanmar and major ethnic languages, in order to help ensure that eligible students and their parents are aware of the stipend and school grants program. Step 5 - Select Townships for the Stipend Program: It is expected that the stipend program will be provided in 40 townships during the four Academic Years (AY) 2014-2018. The program is expected to cover about 25 percent of students in each township. The criteria for the selection of townships and schools will be provided in the national guidelines for the stipend program, but they will include levels of poverty, remoteness and drop-out rates. Step 6 - Relevant Township Education Office officers, in collaborations with T relevant DBE officers, the TPWG and SPC, will carry out Social Assessment (SA) which consists of vulnerable assessment and free, prior, informed consultation with vulnerable and ethnic groups on an annual basis in those new townships that are included into the stipend program in the respective year. Prior to the annual SA, ethnic and vulnerability screening will be carried out based on the literature review and interviews with key informants to select townships where SA will be conducted. The townships where SA will be conducted will be selected to ensure that all ethnic and vulnerable groups that will be affected under the program are represented. Step 7 - DEPT will develop, or annually update, the CPP as the project rolls out to new townships, based on the result of SA. The CPP will address all requirements under OP 4.10 for the Indigenous Peoples Plan. Copies of CPP will be made available to beneficiary schools. CPP will also be translated into relevant ethnic languages of the project townships. 19 Township Program Working Group comprises representatives from township education office, schools, ethnic minorities and civil society organizations working on providing education. 56 Step 8 - Implementation of CPP: DEPT and DBEs will assume an overall responsibility for the implementation of the CPP. The Township Education Officers (TEO) will be in charge of day- to-day implementation of CPP together with TPWGs, and regularly review the progress of the stipend program. The implementation of CPP will be reported to MoE and the World Bank annually as part of MoE's annually reporting on program implementation. 155. The project will empower local communities in school management in general and implementation of stipend and school grants programs in particular through establishing the School Project Committee (SPC) which will be play an oversight role in the monitoring of the stipend and the school grants programs. For the stipend program, the SPC would meet beneficiary students/families and receive feedbacks especially on stipend payments. The SPC will also serve as the focal point for the beneficiary assessment, and school and household surveys. Summaries of the beneficiary assessment will assess implementation of CPPF including access of vulnerable and ethnic groups to project benefits, and their copies will be translated into relevant ethnic and Myanmar languages and made available to people in project townships. As for the school grants program, support will be provided so SPCs will be able to provide inputs to develop the school plan and endorse the final school improvement plan. SPCs will regularly monitor the implementation of the school grants according to the endorsed school improvement plan. 156. The DEPT and DBEs will develop grievance redress mechanisms that are accessible to social groups who are vulnerable to exclusion from project benefits. DEPT, DBEs and the TEOs will make sure that local population is informed of these processes and mechanisms. Detailed processes and procedures of feedback/ complaint handlings procedures will be developed and included in the MoE's program guidelines. 57 Annex 7: Economic Analysis Myanmar Decentralizing Funding to Schools A. Introduction and main findings 157. Purpose of the economic analysis: Per OP/BP 10.00, the economic analysis appraises the project's (1) expected contribution to the country's economic development, (2) rationale for public sector provision, and (3) value-added of World Bank support. The key method for appraising a project's contribution to economic development includes a cost-benefit analysis and fiscal sustainability analysis (OPSQ 2013). The findings of the analysis are as follows: 158. Expected contribution to Myanmar's economic development: The project partially finances existing government programs (school grants and stipends); consequently, the contribution to economic development of the project's financing depends on the contribution to economic development of these programs. The project anticipates or stipulates various reforms (including expansion) to the programs. To appraise the project's contribution to economic development, the government programs-under the anticipated reforms-are assessed by the economic analysis on (1) whether their monetary benefits exceed their costs and (2) whether their costs are fiscally sustainable (OPSQ 2013). The conclusions of the analysis are: * Stipends program: Available data and international evidence suggest that the monetary benefits (through increased human capital) exceed the costs of the program and more so when the program is reformed to better target the poorest. It is also robust to uncertainty over the impact of stipends on enrolment rates. * School grants program: The anticipated reforms to the school grant program establish fundamental processes that lay the foundation for impacting education outcomes in the future. Available data and international evidence suggest that even if impact on education outcomes is long-term as expected, the subsequent monetary benefits would still exceed operating costs. Available data also suggests that the program benefits the poorest the most. * Fiscal sustainability: Because the government currently finances the existing stipends and school grants program, fiscal sustainability of the program under the anticipated reforms depends on the government's capacity to absorb any additional costs. The additional costs of the stipend program when reformed (including expansion) suggests that fiscal sustainability presents an some risk. This risk is partially mitigated by recent trends in public recurrent and social sector spending and by the willingness of international donors to support well-targeted and well-implemented education program. 159. Rationale for public sector provision: Available data suggests that the average benefits of pursuing an additional year of schooling outweigh the public and private costs for each grade of school. As a result, school leavers may be indicative of a market failure, and a public intervention such as the stipend program is justified. In general, the market failure that typically justifies public expenditure on education is a credit constraint: while a household's investment in 58 a young child's education might yield higher earnings for the child in the future, these future earnings cannot in practice be borrowed against to finance present-day education expenditure. 160. Value-added of World Bank support: The World Bank has considerable experience and knowledge on conditional cash transfers and school-based management which directly apply to the stipend and school grant programs, respectively. 161. This annex is structured as follows. Section 2 presents the cost-benefit analysis of the stipends and grants schemes. Section 3 presents the analysis of fiscal sustainability. B. Cost-benefit analysis Methodology 162. Cost-benefit analysis of Myanmar's stipend and school grant programs under anticipated reforms provide an appraisal of the project's contribution to economic development. A bulk of the financing disbursed by the project supports these two programs in anticipation of or conditional on various reforms; consequently, whether this financing contributes positively to economic development depends on whether the programs contribute positively to economic development. Cost-benefit analysis comparing the monetary benefits of the programs to their costs is a key method for appraising an intervention's contribution to economic development (see OPSQ 2013). 163. In cost-benefit analysis, the value ofbenefits is compared with costs by estimating an economic rate of return. The economic rate of return is essentially the rate of return of an 20 investment from a social perspective . Typically, an investment project (whether in education, health, infrastructure, etc.) involves an immediate cost followed by a future stream of benefits. For example, in education, going to school one additional year generates a stream of additional earnings in the future depending on the impact of schooling on earnings (often known as the "return to schooling"); however, it also requires immediate costs including foregone earnings as well as private costs (tuition, books, uniforms, etc.) and public costs (teacher salaries, etc.). Like a rate of return in the general sense, an economic rate of return can be compared to alternative investments to decide whether an investment is worthwhile from an economic point of view. In cost-benefit analysis, the economic rate of return is compared to a "social discount rate". 164. The social discount rate is a benchmark for the economic rate of return and has been estimated to be 5.43 percent in Myanmar. The social discount rate is a measure of how much society values future benefits less than present benefits. If an investment yields a rate of return higher than the social discount rate, then the investment is optimal from a social point of view. Analysis of the 2009 household survey Integrated Household Living Conditions Surveys (IHLCS) suggests a social discount rate of 5.43 percent (estimates available on request)21. The 20 Technically it is the discount rate where the present value ofeconomic costs and economic benefits are equal. 21 The method is based on the socialrate oftime preference method (Ramsey 1928) commonly used for long-term projects (see Zhuang et al.2007; Stern 2006) following Kula (2005) and Percoco(2007). Assumed future growth rate is forecasted GDPper capita growth (IMF 2013). As sumed utility discount rate is the ten year average ofthe 59 rate suggests that an investment would need to yield an economic rate of return of over 5.43 percent in order to be socially optimal. In order to assess whether the stipend and grant programs are beneficial, their economic rates of return are estimated and compared Cost benefit analysis of the reformed stipendprogram 165. The objective of the government's stipend program is to increase schooling attainment, and the impact of cash transfer programs on attainment is well researched. CCT programs have been the subject of numerous evaluations where by design their impact on a student's participation in schooling can be identified (Table 1). Table 1. Evaluation of CCT impact on enrolment or attendance Percentage Ratio of point likelihoods Baseline increase in of being Study Country Target enrolment enrolment enrolled Filmer & Schady 2009 Cambodia Grades 7-9 65.0 21.4*** 1.33 Macours & Vakis 2008 Nicaragua Ages 7-15 90.5 6.6*** 1.07 Schady & Araujo 2008 Ecuador Ages 6-17 75.2 10.3** 1.14 Galasso 2006 Chile Ages 6-15 60.7 7.5*** 1.12 Ahmed & al. 2007 Turkey Primary students 87.9 -3.0* 0.97 Secondary students 39.2 5.2 1.13 Maluccio & Florest 2005 Nicaragua Ages 7-13 72.0 12.8*** 1.18 Schultz 2004 Mexico Grades 0-5 94.0 1.9 1.02 Grade 6 45.0 8.7*** 1.19 Grades 7-9 42.5 0.6 1.01 Glewwe and Olinto 2004 Honduras Ages 6-13 66.4 3.3*** 1.05 Average 1.11 Source: Fiszbein and Schady 2009: 128. Statistical significance at the 10, 5, and 1 percent levels denoted * *, respectively. Ratio of likelihood of being enrolled is the ratio of the baseline enrolment plus the increase in enrolment divided by the baseline enrolment. The studies shown here are either randomized control trials or regression discontinuity design evaluations. 166. The contribution of educational attainment to human capital and subsequently economic development is also well researched. Hundreds of empirical studies have estimated the relationship between educational attainment and earnings (for a recent review see Patrinos and Psachoropolous 2010). This research establishes that human capital, as defined by educational attainment, is an investment with both a private return consisting of after-tax earnings for individuals and a social return consisting of pre-tax earnings and other benefits; crude deathrate following (Lopez2008). All otherparameters estimated fromthe 2009 IHLCS following Kula (2005) and Percoco (2007). More details are available uponrequest. 60 consequently, investing in human capital contributes to poverty reduction and economic development by increasing an individual's productivity and future contribution to society. 167. In Myanmar, an additional year of schooling is associated with 6.7 percent higher earnings among households engaged in agricultural production. While the 2009 IHCLS does not contain data on individual earnings, it does contain information on household consumption and savings. Because of this data limitation, the association between educational attainment and earnings is estimated for households engaged in their own agricultural production. This allows aggregate household earnings to be associated with aggregate years of schooling among working members and other farm characteristics including farm size and productive assets (estimates available upon request). Among farming households, an additional year of school per worker is associated with 6.7 percent higher earnings per worker when accounting for differences in household farm characteristics. International estimates based on individual earnings suggest a higher association between years of schooling and earnings; consequently, this 6.7 percent association serves as a conservative estimate for valuing years of schooling in this analysis. 168. Given this association between schooling and earnings, the benefits of an additional year of schooling exceed the private and public costs: school leavers suggest a market failure. As discussed above, the decision to complete one more year of schooling generates a stream of benefits (estimated conservatively as 6.7 percent higher earnings) but also involves an immediate cost including foregone earnings as well as private education expenses and public expenses. An economic rate of return for completing an additional year of schooling for each grade level can be estimated using (1) a 6.7 percent increase in future earnings, (2) one year of foregone earnings, (3) estimates of private expenditure for each level of education using the 2009 IHCLS and (4) estimates of public expenditure for each level of education using administrative data (Figure 1). The resulting economic rates of return for completing an additional year of schooling for each grade always exceeds the social discount rate implying that completing an additional year of schooling is optimal from social point of view. Consequently, school leavers may represent a market failure22 22 in absence ofheterogeneous returns to schooling (seeHeckman and Carneiro 2003; Card 1999) 61 Figure 1. Economic rate of return for completing 80% -an additional year of schooling * 60% ---- Social discount rate 0 40% E 0 0 20% 0% KG 1 2 3 4 5 6 7 8 9 None Primary Middle High grade level completed Blue columns denote the internal rate of return of staying in school an additional year (y-axis) for each completed grade (x-axis). This is based on the estimated relationship between schooling and earnings, the private and public costs of schooling (Table 5), and forecast of future GDP per capita growth (IMF 2013). 169. Available data and international evidence suggest that the monetary benefits of the stipend program exceed its costs. Providing a student with a stipend for the duration of his or her schooling represents an immediate cost followed by a stream of increased earnings depending on the impact of the stipend on his or her educational attainment. Given the estimated value of an additional year of schooling as well as the public and private costs of schooling and cost of the stipend (Table 5), the economic rate of return of receiving a stipend exceeds the social discount rate even when the impact of the stipend on enrollment is relatively low by international standards (Table 2). The economic rate of return is higher when poorer students are targeted rather than average students or orphaned students as currently targeted. 170. Risks associated with the stipend program are mitigated by (1) being robust to lower impacts on enrolment and (2) better targeting of the poorest. The impact of the stipend program is unknown ex-ante. However, even if the impact is relatively low by international standards, the economic rate of return exceeds the social discount rate (Table 2). Additionally, better targeting poorer children as anticipated by the reforms corresponds to a higher economic rate of return further mitigating the risk of stipends having a low impact. 62 Table 2: Economic rate of return of stipend program by impact and targeting scenarios Impact of stipends on likelihood of being enrolled Low: Average: High: 1.05 times 1.12 times 1.33 times more likely more likely more likely e.g.: Honduras e.g.: Chile e.g.: Cambodia Stipends for secondary students targeting: any secondary student 6.4% 7.7% 8.4% those that are orphans 6.4% 7.8% 8.6% those among the poorest 25% 6.5% 8.1% 9.0% those among the poorest 10% 6.5% 8.2% 9.1% Stipends for primary and secondary students targeting: any student 6.0% 7.3% 8.1% those that are orphans 5.9% 7.4% 8.3% those among the poorest 25% 6.6% 8.0% 8.9% those among the poorest 10% 7.0% 8.9% 9.5% Internal rate of return is based on the additional stream of future benefits and costs from receiving a stipend for the specified impact and targeting scenarios. The stream of future benefits are the additional earnings from increased schooling estimated using IHLCS 2009. The steam of future costs are the public and private costs of school presented in part in Table 5 estimated from the 2009 IHLCS. Cost-benefit analysis of the reformed school grant program 171. The objective of the school grant program represents intermediate steps to improving human capital. During the course of this project, school grants will deliver more reliable and transparent resources to schools. With this in place, the school grant program will, overtime, contribute to raising the quality of primary education by (1) lowering the private costs associated with schooling and (2) by improving the school learning environment. This links to human capital through higher educational attainment and learning outcomes. In addition to the link between human capital and educational attainment, there is an emerging body of literature on the link with learning outcomes. Depending on the country context, a standard deviation difference in cognitive ability typically measured by test scores corresponds to a difference in income ranging from 5 percent in Ghana to 48 percent in South Africa (Patrinos and 63 Psachoropolous 2007). Data is not currently available to estimate the return to cognitive ability in Myanmar. 172. International research suggests limited short-term impact on education outcomes but highlight the importance of the anticipated reforms for long-term impact. There are few randomized evaluations of the impact of school grants or school-based management interventions on education outcomes, and those that do exist suggest limited impact in the short- term (Table 3). For example, an analysis in the United States found modest impacts on student achievement for the first five years and much larger impacts into the future (Borman et al 2003 in World Bank 2007). The anticipated reforms of the school grant program establish an important foundation that enable the grant program to have a positive impact on education outcomes in the future. For example, the project aims to strengthen parent participation in school management. Parent and local stakeholder participation increases the information available for decision making and increases accountability of decision making (Bruns et al. 2011; Arcia et al. 2011; Barrera-Osorio et al. 2009:33), and decentralization can fail and even have negative effects without strong participation by parents and communities, especially in disadvantaged areas (Galiani et al 2008 in education; Bardhan and Mookherjee 2005, 2006 more generally). Even when communities are legally involved in school decision making, they often lack de facto influence or capacity to be effective participants (e.g.: see Blimpo and Evans 2011; Bruns et al. 2011:71; Duflo et al. 2007; Parker 2005). Table 3. Summary of randomized evaluations of early impacts of school grant programs Effect on Effect on Country Cognitive ability enrolment Intervention School grants Indonesia 0.17 - 0.22 SD no effect US$870 school grant Gambia no effect no effect US$500 school grant Niger n.a. 10% increase US$209 school grant Sources: Pradhan et al. 2010 (Indonesia); Blimpo & Evans 2011 (Gambia); Beasely & Huillery 2013 (Niger) 173. The economic cost ofthe school grant program per student is small and benefits the poorest more if the school grant offsets tuition as intended. The current working estimate of the total cost of the school grants is 16.2 billion Myanmar Kyat (MMK) which would be 3,153 MMK per student with a forecasted enrolment of 5.1 million primary students. A key question is how private expenditure would respond to this school grant compared to an absence of the grant. Total private expenditure per primary student is estimated to be 11.0 thousand MMK from the 2009 IHLCS adjusted for inflation. If private expenditure remains unchanged, then the grant would represent 29 percent of per student private expenditure and an 80 percent increase in school cash receipts (Table 4). However, it is possible that household expenditure could decline in response to the grants. For example, Das et al. (2011) find household education expenditure reducing in response to anticipated school grants. Additionally, under the free and compulsory 64 basic education program, tuition fees would be eliminated. Tuition fees at the primary level are estimated to be 2.8 thousand MMK based on the 2009 IHLCS. In absence of tuition fees, the school grant would represent an increase in expenditure of 300 MMK or 3 percent of private education expenditure and an 8 percent increase in school cash receipts. Because the poorest paid the lowest levels of tuition, the net increase is the highest among them. 174. If the impact of the school grant program is comparable to international programs, a positive net benefit is possible even if it takes many years to achieve this impact. The impact on educational outcomes is generally not short term and depends on other factors. However, because the cost of the grant program per child is so small, a moderate impact on educational outcomes can occur many years in the future and still yield a positive net benefit. For example, if the school grant program increased enrolment by 10 percent as in Niger (Table 3) and if the grant program does not offset any private expenditure for education, then the program could operate for 64 years without any impact on enrolment and still yield an economic rate of return higher than the social discount rate. Similarly, if the school grant program had a low impact on cognitive ability (e.g.: 0.06 standard deviation in Duflo and Kremer 2007) and if cognitive ability had a low association with earnings (e.g.: 5 percent as in Ghana in Patrinos and Psachoropolous 2010), then the program could operate without any impact for 21 years. Table 4. Grants and private expenditure per primary student (thousands MMK) Student from average poorest household 25 percent Private spending (based on IHCLS 2009 adjusted for inflation) Total spending 11.0 5.4 Tuition fees (eliminated with grants) 2.8 1.0 School contributions 1.1 0.6 Average school grant per child 3.2 3.2 Net increase if grants offset tuition only 0.3 2.1 Source: 2009 IHCLS for private expenditure and the School Grant Discussion Note for school grant per capita. Private per student expenditure adjusted for inflation based on IMF 2013 inflation rates. I. Fiscal sustainability analysis 175. The fiscal sustainability analysis assesses whether the additional costs of the government's reforms to the stipend and school grant program are sustainable by the government. The current stipend and school grant programs are financed by the government's recurrent budget. While the monetary value of the benefits of the stipend and school grant 65 programs likely outweigh their costs, the contribution of the Bank's financing of these programs also hinge on the government being able to sustain the programs. 176. Additional costs of reforms to the grant program are not yet available. This section will be updated once these costs become available. 177. Expanding the stipend program would cost approximately 11.8 billion MMK annually representing 6 percent of the education ministry's non-salary budget. If non- salary expenditures increase by 31 percent, equivalent to forecasted real GDP growth over that period, then the program would represent 4.6 percent of the non-salary budget (Table 5). 178. Under a reasonable range of impacts on enrolment by the stipend program, expanding the stipend program would increase recurrent costs by 1.75 to 1.92 percent. Under the current planned reform, the stipend program would be expanded to 170 thousand middle and high school students. If the stipend program were targeted perfectly, then this would imply an additional 170 thousand children increasing recurrent public expenditure by 4 percent (Table 5). However, from international experience, the impact of cash transfer schemes on enrolment is much lower (Table 1). A reasonable range of impacts would translate into a range of recurrent cost increases (assuming the same public expenditure per student) between 1.75 to 1.92 percent (Table 5). Table 5. Public cost of grant and stipend programs and their anticipated impacts (MMK millions) 2013/14 budget est. 2017/18 (31% growth) Total Non-salary Total Non-salary Recurrent expenditure 685,954 196,883 898,600 257,916 Additional grant program costs 16,201 16,201 16,201 16,201 as percent of recurrent expenditure 2.36% 8.23% 1.80% 6.28% Additional stipend program costs 11,805 11,805 11,805 11,805 as percent of 2013/14 estimate 1.72% 6.00% 1.31% 4.58% Additional stipend costs from increased enrolment (by impact scenario) as percent of recurrent exp. 1.05x (low impact e.g.: Honduras) 1.75% n.a. 1.34% n.a. 1. 12x (avg impac t e.g.: Chile) 1.79% n. a. 1.37% n. a. 1.33x (high impact e.g.: Cambodia) 1.92% n.a. 1.46% n.a. if all recipients would have dropped out 4.00% n.a. 3.05% n.a. Additional grant and stipend program costs are the increases in recurrent costs resulting from changes in these programs described in the PAD. Presented figures are based on the Cash Transfer Discussion Note and the School Grants Discussion Note. Non-salary expenditure is all categories except "Staff (Pay & Allowance Honoraria, etc.)". A 31 percent increase in real GDP is forecasted from 2013 to 2017 (IMF 2013) and is used to forecast recurrent expenditure. Costs from increases in enrolment anticipated from the stipend program depend on the stipend program's impact Impact scenarios are the same as in Table 4; the final row presents the case if all stipend recipients would have dropped out without the stipend. 66 179. Uncertainty in the fiscal sustainability of the project's investments in the stipend and grant programs is partially mitigated by recent increases in recurrent and social sector budgets. The recent PER (World Bank forthcoming) reports a significant realignment of Union government expenditure beginning in the 2012/13 fiscal year. There has been a significant increase in public expenditure and with a shift in expenditure from capital to recurrent. Recurrent expenditure has increased for social sectors including education. Social sectors including education have seen an increase in capital expenditure while defense and construction ministries have seen declines. However, despite these recent trends, there does not exist a long term budget plan. Significant increases in education expenditure (or significant reallocation of the education budget) would be needed to sustain these programs. 67 References Arcia, G., K. Macdonald, H. A. Patrinos, E. Porta (2011). SchoolAutonomy and Accountability. Washington, D.C.: The World Bank. Barrera-Osorio, F., T. Fasih, and H. A. Patrinos with L. Santibafiez (2009). Decentralized Decision-Making in Schools: The Theory and Evidence on School-Based Management. Washington, D.C.: The World Bank. Bardhan, P. and D. Mookherjee. 2005. "Decentralizing Antipoverty Program Delivery in Developing Countries" Journal of Public Economics 89(4): 675-704. Bardhan, P. and D. Mookherjee. 2006. "Decentralization and Accountability in Infrastructure Delivery in Developing Countries." Economic Journal 116(508): 101-127. Beasely, E. and E. Huillery 2013. School resources, behavioral responses and school quality: short-term experimental evidence from Niger. Sciences Po Publications Blimpo, M. P. and D. K. Evans 2011. School-based management and educational outcomes: lessons from a randomized field experiment. Washington, D.C.: The World Bank Bruns, B., D. Filmer, and H. A. Patrinos (2011). Making Schools Work. Washington, D.C.: The World Bank. Card, D. 1999. The causal effect of education on earnings. In Handbook ofLabor Economics, 0. C. Ashenfelter and D. Card (Eds). Amsterdam: North Holland Chaudhury, N. and D. Parajuli 2010. Pilot evaluation: Nepal Community School Project. Unpublished manuscript. World Bank Das, J. 2008. Unpublished manuscript cited in Bruns et al. 2011 Das, J., S Dercon, J. Habyarimana, P. Krishnan, K. Muralidharan, V. Sundararaman (2011). School inputs, household substitution, and test scores. Policy Research Working Paper 5629. Washington, D.C.: The World Bank. Duflo E., P. Dupas, and M. Kremer (2007). Peer Effects, Pupil-Teacher Ratios, and Teacher Incentives: Evidence from a Randomization Evalation in Kenya. Cambridge, M.A.: Poverty Action Lab, MIT Fiszbein, A. and N. Schady 2009. Conditional Cash Transfers: Reducing Present and Future Poverty. Washington, D.C.: The World Bank. Galiani, S., P. Gertler and E. Schargrodsky. 2008. "School Decentralization: Helping the Good Get Better, but Leaving the Poor Behind." Forthcoming in Journal of Public Economics. 68 Heckman, J. J. and P. Cameiro 2003. Human Capital Policy. Working Paper W9495. Cambridge, M.A.: NBER International Monetary Fund (2013). World Economic Outlook: Transitions and Tensions. Washington, D.C.: International Monetary Fund Kula, E. (2004). Estimation of a social rate of interest for India. Journal ofAgricultural Economics. Vol. 55, No. 1 Lopez, H. (2008). The social discount rate: Estimates for nine Latin American countries. World Bank Policy Research Working Paper 4639. Washington, D.C.: The World Bank Parker, C. E. (2005). Teacher Incentives and Student Achievement in Nicaraguan Autonomous Schools. In: Vegas, E. (Ed.) Incentives to Improve Teaching: Lessonsfrom Latin America. Washington, D.C.: The World Bank. Patrinos H A and Psacharopoulos G (2010), Returns to Education in Developing Countries. In: Penelope Peterson, Eva Baker, Barry McGaw,(Editors), International Encyclopedia of Education. volume 2, pp. 305-312. Oxford: Elsevier. Percoco, M. 2008. A social discount rate for Italy. Applied Economic Letters. 15. Pradhan, M, D. Suryadarma, A. Beatty, M. Wong, A. Alishjabana, A. Gaduh, R. P. Artha 2010. Improving educational quality through enhancing community participation. Results from a randomized field experiment in Indonesia. World Bank Policy Research Working Paper 5795. Washington, D.C.: The World Bank Ramsey, F. 1928. A mathematical theory of saving. The Economic Journal. 38 (152) Stern, N., S. Peters, V. Bakhshi, A. Bowen, C. Cameron, S. Catovsky, D. Crane, S. Cruickshank, S. Dietz, N. Edmonson, S.-L. Garbett, L. Hamid, G. Hoffman, D. Ingram, B. Jones, N. Patmore, H. Radcliffe, R. Sathiyarajah, M. Stock, C. Taylor, T. Vernon, H. Wanjie, and D. Zenghelis. (2006). Stern Review: The Economics of Climate Change, HM Treasury, London. World Bank 2007. What do we know about School-Based Management? Washington, D.C.: The World Bank Zhuang, J., Z. Liang, T. Lin, F. De Guzman et al. 2007. Theory and practice in the choice of the social discount rate for cost-benefit analysis: a survey. Economics and Research Department Working Paper Series. No. 94. Manila: Asian Development Bank. 69