1 EDUCATION PUBLIC EXPENDITURE REVIEW GUIDELINES Education Global Practice 2 Education Public Expenditure Review Guidelines © 2017 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e- mail: pubrights@worldbank.org. i TABLE OF CONTENTS ii ACRONYMS iv PREFACE v USER’S GUIDE 1 PART I: CHECKLIST FOR EDUCATION PER STEPS 2 Introduction 3 Public Expenditure Review Steps 16 PART II: CHECKLIST FOR AN EDUCATION PER ANALYSIS 18 Section 1: Background and Overview of the Education System 20 Section 2: Overview of Education Financing and Spending 25 Section 3: Financial Accountability 30 Section 4: Adequacy and Sustainability 34 Section 5: Efficiency and Effectiveness 38 Section 6: Equity 41 TECHNICAL NOTES 85 PART III: EXAMPLES 206 NOTES 210 REFERENCES ii Education Public Expenditure Review Guidelines ACRONYMS ACS Average Class Size LGA Local Government Authority AY Academic Year LGU Local Government Units BIA Benefit Incidence Analysis LLECE Laboratorio Latinoamericano CBA Cost Benefit Analysis de Evaluación de Calidad de la Educación, or Latin American CCT Conditional Cash Transfer Laboratory for Evaluating the CEA Cost-Effectiveness Analysis Quality of Education CGE Computable General LSMS Living Standards Equilibrium Measurement Study COFOG Classification of Functions of MAMS Maquette for MDG Government Simulations CSOs Civil Society Organizations MFM Macroeconomics and Fiscal DD Difference in Differences Management DEA Data Envelopment Analysis MICS Multiple Indicator Cluster Survey DHS Demographic and Health Survey MOF Ministry of Finance ECD Early Childhood Development MTEF Medium-term Expenditure Framework EFA Education for All NEA National Education Accounts EMIS Education Management Information Systems NER Net Enrollment Rate EPSSim Education Policy and Strategy NGOs Non-governmental Simulation Model Organizations FTE Full-time Equivalent NSOs National Statistical Offices GDP Gross Domestic Product OECD Organisation for Economic Co- operation and Development GEM-Education The General Equilibrium Model for Education PASEC Programme d’Analyse des Systèmes Educatifs de la GER Gross Enrollment Rate CONFEMEN (the CONFEMEN GNI Gross National Income Programme for the Analysis of GNP Gross National Product Educational Systems) GP Global Practice PCF Per Capita Financing HNPGP Health, Nutrition, and PEFA Public Expenditure and Population Global Practice Financial Accountability IGR Internally Generated Revenue PER Public Expenditure Review INESM Inter-agency Network on PETS Public Expenditure Tracking Education Simulation Models Survey ISCED International Standard PFM Public Financial Management Classification of Education PIAAC Programme for the IT Information Technology International Assessment of Adult Competencies iii PIRLS Progress in International TVET Technical and Vocational Reading Literacy Study Education and Training PISA Program for International UIS UNESCO Institute of Statistics Student Assessment UNESCO United Nations Educational, PPP Purchasing Power Parity Scientific and Cultural PTA Parent-Teacher Association Organization PTCA Parent-Teacher Community WB-MAMS World Bank’s Maquette for Association MDG Simulations PTR Pupil-Teacher Ratio WDI World Development Indicators QSDS Quantitative Service Delivery Surveys VFM Value for Money RDD Regression Discontinuity Design SA Social Assistance SABER Systems Approach for Better Education Results SACMEQ Southern Africa Consortium for Monitoring of Education Quality SBM School-based Management SDGs Sustainable Development Goals SDI Service Delivery Indicators SERCE Second Regional Comparative and Explanatory Study SES Socioeconomic Status SNA System of National Accounts STEP The STEP Skills Measurement Program STR Student-Teacher Ratio SWAP Sector-wide Approach Program TERCE Third Regional Comparative and Explanatory Study TIMSS Trends in International Mathematics and Science Study iv Education Public Expenditure Review Guidelines PREFACE Public expenditure reviews are one of the World Bank’s core diagnostic tools for informing various stakeholders about the state of education financing in a country. Such reviews assess the efficiency, effectiveness, and equity of expenditures on education and their adequacy and sustainability relative to the country’s educational goals. They review not only public spending, but also private and donor spending. Guidelines for public expenditure reviews in education establish content and quality standards for such reviews in the sector, using technical notes and examples to deepen the user’s understanding of how to meet these standards. The World Bank prepared the last public expenditure review guidance for education in 2004 and revised it in 2009 as part of the Human Development sector-wide initiative. The document included a cross-sectoral Core Guidance and a sector-specific Guidance for each sector. This Guidance covers both cross-sectoral and education sector-specific issues in one volume. The new Guidelines update the contents of the earlier guidelines to reflect recent developments in education financing and respond to demands for more hands-on advice, as follows: (i) they adopt the World Bank’s new initiative—Systems Approach for Better Education Results (SABER) School Finance Framework—as the basis for the analytical dimensions; (ii) they use a decision-tree approach, providing step-by-step guidance for conducting public expenditure reviews under different context, scopes, and types of analytical tools; (iii) they analyze policy recommendations in completed public expenditure reviews, assessing why each is good or weak; (iv) they introduce a new analytical tool, BOOST, which converts government budget data into a useful format, makes analysis relatively easy, and contributes to the development of the comprehensive education-finance data accounts called National Education Accounts; (v) they enhance discussions of governance and public financial management issues that partly determine effective spending; (vi) they strengthen technical notes by including more details on data sources and definitions of concepts and variables; and (vii) they update good analytical examples from completed public expenditure reviews, categorized by topic. Sachiko Kataoka led a team under the guidance of Luis Benveniste (Director, Education Global Practice) to develop the new guidelines. The team included Sue Berryman, Yulia Makarova, and Aliya Bigarinova, and also relied on a number of colleagues who provided insightful comments throughout the process. Specifically, the team is grateful to: Dina Abu-Ghaida, Samer Al-Samarrai, Mohammed Audah, William Dorotinsky, Kebede Feda, Katia Marina Herrera Sosa, Margo Hoftijzer, Jennifer Klein, Shinsaku Nomura, Anna Olefir, Shawn Powers, Holy Tiana Rame, Furqan Saleem, Lars Sondergaard, and Ryoko Tomita Wilcox. The team also would like to thank Husein Abdul-Hamid, Melissa Adelman, Jung- Hwan Choi, Jennifer Klein, Laura Gregory, and Kirsten Majgaard for helping to compile recent public expenditure reviews to develop a database. v USER’S GUIDE This guidance note applies to all countries, regardless of their circumstances. It seeks to remind the analyst of the main features that are normally included in an education public expenditure review (PER). However, analysts should not use it as a checklist to which they must rigidly adhere. Every review must be selective in what it covers, based on such factors as what is needed for the country dialogue; what is already known and available elsewhere; and what is able to be accomplished given constraints of time, data, and funding. Analysts may omit topics, but with a justification in mind. In addition to agreeing in the concept note on the planned topic coverage, it is useful to convey to the reader of the full report the reasons for omitting any major themes. Similarly, the depth of treatment of any included topic will need to be considered, agreed upon, and explained. Conditions such as fragility, conflict, or violence in the country under review could affect whether a PER can be conducted at all. Such factors can also affect the quality of the data needed to conduct a review, as well as the ability of the country to implement any recommendations that the review suggests. CHECKLISTS Part I of the Guidelines provides a checklist for steps, from preparation to dissemination, of an education PER, including cross-sectoral questions, such as the overall budget and potential tradeoffs between social and other sectors. The PER team will first go through a decision tree to determine the scope, objectives, analytical tools, and policy questions for a particular review. The next steps will probably vary for different circumstances. Part II provides a checklist for an analysis to be completed as part of an education PER, organized by six key questions:1 1. Who finances education and how are funds channeled? 2. How much does the government spend and on what? 3. Is the public financial management system set up to enhance financial accountability? 4. Relative to the government’s policies and standards, how much is needed now (adequacy), and what can be afforded in the medium and long term (sustainability)? 5. Are public resources being used efficiently and effectively? 6. Does public spending promote equity? vi Education Public Expenditure Review Guidelines To keep the checklists as skeletal as possible, three different types of links are used to let review teams explore a topic in more depth. Two types of links are internal to this document. They take the analyst to in-depth technical notes and examples of good treatments of a PER topic, with red for Technical Notes and blue for Examples. The third type of link, in purple, signals a hyperlink to an external resource. Users of these Guidelines may also consult Solutions Notes, which provide concrete solutions for specific education financing issues, to help them frame good policy recommendations. These Notes will be made available at http://www.worldbank.org/en/topic/ education. Examples and Solutions Notes will be regularly updated. Technical Note: A Technical Note elaborates on a concept or provides detailed guidance on analytic methods, indicators, and sources relevant to the topic. Example: An example shows an especially good treatment of a topic. It often includes the description of analytic methods used and other guidance. Examples are taken from recent PERs and will be updated as more good case studies become available. If the analyst is working from an electronic version of these Guidelines, clicking on a color-coded oval that appears next to a topic in the checklist will take the reader to the material connected to that link. For hard-copy versions of the Guidelines, all Technical Notes germane to Part I and Part II appear at the end of Part II in numerical sequence. Part III consists of the Examples. PART I: CHECKLIST FOR EDUCATION PER STEPS 2 Education Public Expenditure Review Guidelines Introduction Why is education finance important? All education systems rely on financing to function. Education finance systems pay for the inputs required to implement education policies, such as teachers, school buildings, and learning materials. Availability of financial resources does not guarantee a quality education, but a quality education is impossible to achieve without adequate resources. Some uses of education expenditures can make a marked difference in learning, particularly in the cases of inputs that directly benefit students or resources that compensate for challenges arising from low-income settings. The same money can be wasted if it is allocated to input factors that only marginally affect learning or if policymakers fail to consider the conditions that must be met for factors to translate into learning gains. Governments are under increasing pressure to use education resources efficiently, but often lack guidance on the optimal ways to invest and manage their school finance systems. Research findings have shown that learning outcomes are not strongly related to spending levels (except in cases where education budget is very small), suggesting that the way money is spent—and not simply how much is spent—matters in education finance. Meeting the World Bank’s twin goals of poverty reduction and shared prosperity in the education sector implies the need to use country and donor resources effectively, efficiently, and equitably. A sound PER assesses how resources are used relative to these goals. In Part II, Sections 5 (Effectiveness and Efficiency) and 6 (Equity) are especially germane to these analytic standards. 3 Public Expenditure Review Steps This section provides guidance on steps involved in conducting a PER, from the preparation to dissemination of results, as summarized in Figure 1. It highlights issues common to such reviews, regardless of the sector, and addresses shared considerations and such cross-sectoral questions as the overall budget and potential tradeoffs between social and other sectors.2 Figure 1: Public expenditure review steps 1 STEP 1: Understanding the context and motivation Understand the country context and motivation of the PER. See Figure 2 for seven possible scenarios which may direct the PER team to a particular path. 2 STEP 2: Defining the scope and objectives Review key policy documents such as the education strategy, sector analysis, and latest PER recommendations to define the possible scope, objectives, and policy questions for the review. Agree on them with the client. 3 STEP 3: Securing access to data and information Identify available financing and non-financing datasets as well as policy and legal documents. Depending on data availability, the scope may need to be modified. 4 STEP 4: Analyzing data and information Assess the country’s education finance system, financial accountability, adequacy and sustainability, efficiency and effectiveness, and equity of education financing. PART II will discuss the analysis in depth. 5 STEP 5: Validating key findings and policy recommendations Discuss and validate key findings and policy recommendations with the client, possibly holding separate sessions for technical Ministry of Finance (MOF) and Ministry of Education (MOE) staff and for MOF and MOE policymakers. Disseminate the final report to various stakeholders. 4 Education Public Expenditure Review Guidelines STEP 1: Understanding the context and motivation The very first step is to understand the country context and motivation of the public expenditure review. Figure 2 suggests seven possible scenarios under which the review could be conducted. The scope, objectives, and key questions may vary widely, depending on the scenario. Figure 2: Seven scenarios for country context and motivation What are the country context and motivation? 1. (Re)new 7. Capacity engagement building 2. Emergency 6. Governance support challenges 3. Project 5. Fiscal preparation sustainability 4. Efficiency gains Scenario (1): (Re)new engagement. We have not been engaged in the education sector of the country for a long time or have never been engaged, and need to (re)gain our knowledge about its education financing to renew or initiate our engagement. • When was the last education public expenditure review conducted for the country? • If a review has not been conducted in the last five years, does other sector analytical work exist that can inform you about key issues to be addressed? • Can those available documents give you a good understanding of key challenges to be addressed? ӺӺ These questions affect the decisions on scope and objectives. To explore key issues, focus on all six sections of the guidelines. Scenario (2): Emergency support. The country has recently experienced a crisis (natural, political, or economic), and the Bank needs to assess quickly what kind of financial support the government needs. • What are the major funding items that need urgent support? Possibilities might include repair of damaged schools to restore access, payment of teacher salaries to avoid teacher absenteeism, and financing of cash transfers or school meals to avoid dropouts. • Can the Bank, in collaboration with other development partners, establish appropriate funding mechanisms? ӺӺ Under a crisis situation, the Reference Guide on External Education Financing, published by Inter-Agency Network for Education in Emergencies (INEE), may be more suitable than these Guidelines. 5 Scenario (3): Project preparation. We want to conduct a public expenditure review to inform the preparation of a new education project. • Has the subsector for the new project already been decided? ӺӺ If yes, the review should focus on the chosen subsector, with a thorough examination of six key questions for the subsector. ӺӺ If no, the review should analyze existing key documents and overall education spending, examine subsectoral allocations, and identify education finance policy issues to be addressed, by going through all six sections. Scenario (4): Efficiency gains. The government is facing budgetary constraints and has asked the Bank to analyze public expenditures on education to find fiscal space.3 • Does the government (Ministry of Finance) have a budget-reduction target for the education sector? How imminent is such a cut? ӺӺ In the short term, there is probably little that the Ministry of Education can do besides cutting capital and non-salary, recurrent spending because any reductions in personnel costs usually must be phased in. ӺӺ In the medium term, the public expenditure review should explore potential savings across the entire sector, including rationalizing human resources. At the same time, the review may play a role in defending the education budget if an analysis finds little room for efficiency gains without harming access and quality. Focus on Section 5: Efficiency and Effectiveness. Scenario (5): Fiscal sustainability. The government plans costly reforms, such as an extension of general education, and needs to know the capital and recurrent cost implications of the plans, as well as their fiscal sustainability and equity implications. • A simulation analysis would be essential to inform the government about the feasibility of such a plan ӺӺ Focus on Section 4: Adequacy and Sustainability and Section 6: Equity. Scenario (6): Governance challenges. The government is concerned that public funding is not reaching schools and wants to identify possible issues with public-finance management. • What are major spending problems? ӺӺ It is likely that certain public financial management (PFM) weaknesses are underlying causes of the spending problems. Focus on Section 3: Fiscal Accountability. ӺӺ Also consider other analytical instruments such as Public Expenditure Tracking Surveys (PETS) and Quantitative Service Delivery Surveys (QSDS), provided that there is sufficient budget to conduct them. Go to Technical Note 12: Public Expenditure Tracking Surveys (PETS), Quantitative Service Delivery Surveys (QSDS), and Service Delivery Indicators (SDI). Scenario (7): Capacity building. The government would like to strengthen its analytical capacity and wants the Bank to help it conduct a PER. • In this case, the main objective is to support the government. The client’s engagement and ownership are as important as the final output. This will require more time and budget for close communication with the client. 6 Education Public Expenditure Review Guidelines STEP 2: Defining the scope and objectives There are many factors to consider when defining the scope and objectives of an education PER. Such a review is primarily concerned with public revenues and expenditures as expressions of public policy and public involvement in the economy. A public expenditure review also examines private spending. Rarely can public resources meet all financial needs, and private spending may play a major role in providing educational services. Many kinds of analyses can be conducted in a public expenditure review, but obviously, such analyses must be aligned with the project’s budget, time frame, and data availability, as well as its overall focus and goals. The team needs to: • Review key relevant documents such as the education strategy, sector analyses, and recommendations of the last public expenditure review, to identify education finance policy issues to be addressed. • Decide the subsector focus. The review can have broad coverage (all subsectors, possibly even including adult education) or narrow coverage (a single subsector). State your focus clearly, why you decided on it, what has been excluded, why certain issues or subsectors have been excluded, and whether or how these exclusions might affect the due diligence purpose of the review. Broad, unfocused public expenditure reviews will be less effective than sharply focused ones. • Place education spending within a macro context. Confer with the Bank’s country economist— and, as needed, with the Ministry of Finance (MOF)—about the country’s macroeconomic situation and its implications for education budgets downstream. • Make sure that your education public expenditure review—if part of a multisectoral (comprehensive) PER—fits the purposes of the comprehensive PER, but still respects your main objectives. In the multisectoral case, where several sector-specific reviews feed into a comprehensive PER, the concept note for the overall review should clarify the objectives of the overall PER and the expectations for each of the sector-specific reviews. Although the expectations for the sector-specific reviews are often poorly defined, be sure to define the focus of the education PER and then clear it with the team leader for the comprehensive exercise. • Work to ensure that a comprehensive public expenditure review is more than the sum of its parts. Typically, each sector does its analysis in isolation, and the team leader then combines all sector- specific reviews into one document, including three to five recommendations per sector. By itself, the education review team cannot create incentives for members of comprehensive PER teams to realize the potential synergies within the context of a comprehensive public expenditure review. However, the leader of the education review can at least ask the leaders of the comprehensive review and of the other sector-specific reviews how the overall team might benefit from this specific, comprehensive public expenditure review. 7 • Check that the time and budget4 allocated to you for the education review are in line with your intended scope and its data requirements. Ultimately, the budget and time frame available to prepare and undertake a PER will be decisive in specifying its scope and in setting priorities. Be aware that resources may be needed for the following: ▷ Missions to assemble data and information (travel costs and staff days). ▷ Extensive data recoding and cleaning or new data collection. The early childhood development (ECD), technical and vocational education and training (TVET) and higher education subsectors often lack data required to conduct a review. If the scope of the review includes any of these subsectors, the team may need to collect new data. ▷ Days required to analyze data, write the draft public expenditure review, and revise it after peer reviews. ▷ Missions to discuss results and disseminate them more broadly, such as running a workshop for stakeholders (travel costs and limited staff days). ▷ Follow-up policy discussions. • Once you predefine the scope, discuss it with the key counterparts (typically, the Ministries of Education and Finance) and agree on the scope, objectives, and policy questions for the review during the first mission. Based on the discussion with counterparts, you may need to modify the scope or objectives. This process will help enhance the client’s ownership of the report and bolster the chances that findings and policy recommendations will be accepted and implemented. The following meetings may prove helpful: ▷ Connect with high-level policymakers (Ministers of Education and of Finance, their Deputy Ministers, or equivalent) to explain the objectives of this work and potential benefits to them, and to obtain their approval for this work. Note that since the Ministry of Finance is the primary client for a PER, the Ministry of Education may know little or nothing about this work. Thus, it is essential to obtain education policymakers’ approval before asking their staff to compile and provide data for the project. ▷ After you get the decision maker’s go-ahead, set up meetings with (i) the Education Management Information System (EMIS) team, finance or accounting team, assessments team, and any other unit that has data that you need; and (ii) heads of departments over seeing the levels or topics of education that you plan to analyze to ask them what they see as the main challenges in their respective areas (e.g., early childhood development, secondary education, or teacher policy). ▷ Set up meetings and reach out to key players outside the Ministry to get a broader perspective. Try to visit at least a few schools (ideally, ones different enough to give you a broad picture of the system, e.g., urban and rural, primary and secondary, public and private, wealthy and poor). These school visits can be combined with conversations with local officials who oversee schools at the subnational level, as well as with principals, teachers, and even parents. If you think teacher unions or civil society organizations (CSOs) might offer a useful perspective, ask to meet with them as well. 8 Education Public Expenditure Review Guidelines STEP 3: Securing access to data and information The nature and number of subsectors to be analyzed within the scope of the review helps to determine the datasets and documents that the PER team will need; data availability, in turn, affects decisions on the final scope of the project. Once there is an overall agreement on the scope, the team needs to secure access to reliable data. Good-quality data are accurate, timely, and relevant for decision-making.5 Be clear ahead of time about data sources, their availability, and their consistency, because they can constrain the scope of the report and its quality. When the education review is part of a comprehensive public expenditure review, discuss with the overall team leader how to maximize efficiencies, such as hiring a shared consultant or research assistant for data collection and analysis. Also determine if the team conducting the macro analysis for a comprehensive review will create a government-budget database that the education review team can use (e.g., BOOST). A public expenditure and financial accountability (PEFA) country report might be available, which could prove useful for an assessment of the Public Financial Management (PFM) system relevant to education. In addition, examine the websites of the country’s Ministries of Finance and Education and national statistics agency. These websites often post significant amounts of data, or, at the least, alert you to the existence of databases to which you can request access. Public expenditure reviews use various types of data. Financial data usually come from the Ministry of Finance and sometimes from the Ministry of Education’s budget department. If financing is decentralized for any subsectors within the scope of the review, financing data may be available by subnational unit at the national-level Ministry of Finance. In other cases, you may need to sample subnational units to assemble the needed data. If your scope includes autonomous institutions, such as universities or technical and vocational schools, you will probably have to obtain financing and expenditure data from the institutions themselves. Private and household contributions may be recorded separately, or estimated based on household survey data. Non-financial data, such as the numbers of students, teachers, schools, or classes, usually come from the education management information systems (EMIS) or equivalent, managed by the Ministry of Education. Detailed student assessment data are useful for measuring spending against performance. Whether for financial or non-financial data, you should get both aggregate data at the national level and data as disaggregated as possible at the school or provider level, or at least at the locality or municipality level, for the level(s) of education on which you plan to focus your analysis. In sum, public expenditure reviews use the following six types of data (see Figure 3). The Technical Notes for these six types of data describe common sources for each type and assess the usual quality of each: 9 Figure 3: Six types of data that can be used for PER analysis Government’s official Cross-national data documents (policies, laws, (international learning and regulations) assessment, cross-national statistical databases) Country system-level data Research reports PER (budget and sectoral data) (donor reports, academic analysis research papers) National surveys (census, New data collection household surveys) ӹӹ Government’s official documents, such as policies, laws, and regulations governing inputs to, and financial arrangements for, the learning process (see Technical Note 1). ӹ Country system-level data (e.g., BOOST, EMIS, national assessments; see Technical Note 2). ӹ National surveys such as census and household surveys (see Technical Note 3). ӹ Cross-national data such as international learning assessment and cross-national statistical databases (see Technical Note 4). ӹӹ Research reports (see Technical Note 5). ӹӹ New data collection (see Technical Note 6). The team also needs data for comparisons. Many indicators, especially various measures of spending on education, do not carry an intrinsic absolute value. We do not know, for example, what “adequate” levels of spending on education would be without situating the PER country within the range of values associated with other countries. Hence, it is essential that a public expenditure review also analyzes the following: • Statistical comparisons with other countries in the same region or at similar income levels, and with other countries that serve as aspiration goals for the country, such as those in the Organisation or Economic Co-operation and Development (OECD). Comparisons can be affected by demography (cohort sizes for school-age individuals), participation rates by subsector, unit costs by subsector, and cross-country differences in the structures of education systems.6 Obviously, financing requirements for small school-age cohorts or cohorts with low enrollment rates are less than those for large cohorts or cohorts with high enrollment rates (especially in the more costly subsectors). These comparisons are descriptive by nature and provide neither a diagnosis nor an explanation. Nevertheless, comparisons can show that a country performs significantly worse (or better) than “expected.” Such comparisons can help set some “reasonable” quantitative targets for policymakers and even provide an incentive to reach them. Technical Note 4: Cross-national data • Trend data that allow comparisons over time. These are essential for gauging how quickly the country may be reaching its target. If improvements are slow in coming, trend data can provoke a deeper exploration into the reasons for the lagging performance. 10 Education Public Expenditure Review Guidelines STEP 4: Analyzing data and information These Guidelines provide guidance on how to conduct a public expenditure review to answer the following six key questions: Key Questions 1. Who finances education and how are funds channeled? 2. How much does the government spend and on what? 3. Is the public financial management system set up to enhance financial accountability? 4. Relative to the government’s policies and standards, how much is needed now (adequacy), and what can be afforded in the medium and long term (sustainability)? 5. Are public resources being used efficiently and effectively? 6. Does public spending promote equity? The key concepts of each dimension of education finance are briefly described below and discussed in more detail in Part II. ▶▶ Questions 1–2: The financing system is the foundation for a country’s public investment in education. These questions ask how the education budget is financed, who spends it, and how it is spent. ▶▶ Question 3: Financial accountability in terms of education finance sheds lights on public financial management (PFM) systems, and assesses whether such systems are set up in such a way that government policies get implemented as intended and achieve their objectives. ▶▶ Question 4: Adequacy of education finance responds to the question of “How much is enough now?”, given the government’s policies and standards. Sustainability of education finance assesses whether planned spending levels on education are affordable in the longer term, given the country’s macroeconomic prospects, sector policies, and demography. ▶▶ Question 5: Efficiency and effectiveness in education finance require that systems invest in those inputs with the largest marginal returns, as measured through outputs and outcomes relative to costs, given a country’s particular stage of development. ▶▶ Question 6: Equity in educational opportunity is a key goal of public education systems. It measures the availability of a quality education to all students, regardless of background. 11 STEP 5: Validating key findings and policy recommendations You want the public expenditure review to impact the client’s policies. Impact usually increases if the review team respects certain principles. • Make evidence-based policy recommendations with an explicit link between the analysis and findings and recommendations. Avoid: ▷ Findings and recommendations that do not follow from the analysis ▷ Analyses that do not lead to any findings or recommendations • Prioritize your recommendations. Policymakers will not act on a laundry list of recommendations. Identify a limited number of key issues—three to five per sector—where getting some traction is most critical. “Nesting” your recommendations in hierarchies can be helpful. The main recommendation (e.g., improve equity of access) can be aimed at policymakers, with the technical specifics aimed at technocrats (e.g., change the balance between programs A and B; improve program C by taking actions 1, 2, and 3). Phasing—short-, medium- and long-term—is another way to “chunk” recommendations into digestible form. • Make the conclusions and recommendations appropriately specific, not bland and general. Place recommendations in a feasible social, political, and administrative context. The World Bank’s public expenditure reviews have a tendency to preach the good and the moral without an appreciation of the realistic and feasible. Can each recommendation be made operational and actionable for the government? If possible, estimate the fiscal costs of any major reforms proposed. What political costs and implementation barriers are involved in each? • Keep report recommendations short and accessible. It may be difficult to do this when laying out the analytical grounds for recommendations in a report that covers multiple sectors. One potential way to tackle this issue is to include a main report and a second volume with the technical background papers. • Discuss the draft of the main findings and policy recommendations with the client to validate the findings and assess the feasibility of policy recommendations. While the review should present critical observations supported by analytical evidence, it also needs to be acceptable to the government to boost the likelihood of policymakers implementing the recommendations. • Disseminate the final report to various stakeholders in the education community of the country, ideally through an in-country workshop. It is not easy to meet these criteria for good policy recommendations. Good examples of policy recommendations that meet most of the criteria (being tied to findings, prioritized, concrete, realistic, feasible, and sometimes costed) can be found in Example 1: Policy recommendations. Tables 1 and 2 show brief examples and assessments of recommendations from the universe of completed PERs, with Table 1 focused on relatively useful ones and Table 2 on less useful ones. 12 Education Public Expenditure Review Guidelines Table 1: Are these recommendations useful? Probably. Recommendation Why is this useful? With respect to developing an equalization The recommendation clearly states the problem mechanism, creative solutions are needed to and describes three concrete options for solving ensure adequate financing for schools with low it. These options vary in their political economy capacity to raise private funds. Although public and budget implications, giving the Government funds are distributed in a generally progressive some flexibility in trading off between fiscal and manner, private fees raised by schools flow political costs. overwhelmingly toward the more well-resourced institutions. Given the overwhelming share of private financing for primary and secondary education, an equalization mechanism is clearly needed to counteract the regressive nature of current funding flows. A wealth of international evidence exists on such mechanisms. One option could be to examine the current BSP program and evaluate whether it can be repurposed to allow a small share of school-raised funds to flow from more affluent schools to less affluent ones (within a district or across neighboring districts). A second option could be to revive the defunct Equalization Grant that was previously used to provide additional resources to underfunded schools. A third option could be to increase the relative funding coefficients for P3 and S3 schools under the Per Capita Grant. One approach currently being considered by the MoE is to introduce a new Building Levy, which would collect a small amount from each SDC budget into a central pool of funds, which would then be used to equalize resources among schools without increasing the fees charged to households. Regardless of the sources of funds—whether public or private—the MoE needs to ensure that all of the country’s schools have at least a basic minimum level of resources available to them to provide a quality education. 13 Recommendation Why is this useful? Continue effort to right-size the school network This review was done in response to a and teaching force. A simple estimation suggests request by the country’s Ministry of Finance to that if the sector could attain student-teacher the Bank for sequenced and targeted advice, ratios comparable to the OECD countries over during a period of austerity, on fiscal reform the medium term, it could save approximately options across priority budget areas, including 11 percent of the education budget, or around education. The recommendation shows the cost 0.6 percent of GDP per annum, in the wage bill to the country of its low student-teacher ratios. It alone. Additional savings could be achieved by suggests a practical way to start the process of consolidating the school network (including increasing student-teacher ratios. It shows the within-schools consolidation of classes). As an savings from such a step, and flags the politically initial step towards school size optimization, the attractive possibility that some of these savings sector could increase the student-teacher ratio in could be used to raise the salaries of the urban schools to the average OECD level, which, remaining teachers. It does not discuss options for according to Ministry of Education estimates, could shrinking the teaching force (e.g., hiring freeze vs. result in savings in amount of 0.03 percent of GDP. early retirement options) or lay out how long it These savings could be used to upgrade facilities, would take each alternative to achieve the raise teacher salaries, invest in teacher qualification cost savings sought. and procure other learning equipment, such as computers. Increase the allocation to the education sector This recommendation goes to one of the root budget. Many of the key issues facing the education causes behind the multiple problems besieging sector stem directly and indirectly from underfunding the sector. Companion recommendations of the education budget. The education budget as a discuss concrete means by which the sector share of GDP stands at about 1.8 percent executed could efficiently ramp up inputs should the sector (or 2.3 percent allocated), which is below the be better funded. This recommendation does not recommended GPE levels as well as the SSA average discuss how the sector budget could be (4.7 percent). Our estimations show that an increase increased because the country has the required to 4.7 percent (in line with SSA average) would be resources. The problem in this case is political will. sufficient to help the sector address three key issues: (i) it would cover the estimated cost of absorbing the out-of-school children into the education sector, (ii) it would allow the full onboarding of all teachers who are currently not in the system, and (iii) it would allow an additional reduction in fees and other costs passed on to households that now operate as barriers to schooling for children from poor households. It is imperative that the government effectively prioritizes the education sector in its budget allocation process as outlined in the MTEF. In order to do so, the spending on the education sector as a share of the total spending should also be revised upwards, closer to the recommended 20 percent, almost doubling the current allocation share. Sources: Cited from unspecified, but actual, public expenditure reviews. 14 Education Public Expenditure Review Guidelines Table 2: Are these recommendations useful? Probably not. Recommendation Why is this not useful? Expand pre-school education, while carefully This recommendation is not connected to the considering the efficiency and effectiveness of problem that it addresses. It gives no sense of the different models. rate at which the subsector could feasibly expand. It fails to provide any specifics on what constitutes an attractive preschool model, or any data to support its recommendation to expand preschool education, such as the relative cost-effectiveness of such programs. Unlike comparators, the sector allocates a relatively This recommendation is vague. What does higher share of its total spending on education to “analysis of impact of spending” mean? Given post-secondary education. After a more detailed the political costs, what are realistic options analysis of the impact of spending at post- for expanding cost-sharing arrangements? Is secondary level, consider introducing cost-sharing there any reason to think that the sector can arrangements at this level so that public spending retain savings at the post-secondary level in can be reallocated to lower levels. order to reallocate them to lower levels? Reform the curriculum to strengthen foundational This recommendation is not tied to the problems skills at primary level and increase actual hours that it addresses. It remains unclear whether of instruction. the instructional time issue is due to low hours as specified by policy or a gap between policy and provision. The recommendation gives no guidance on fruitful curricular-reform options. If policy changes are needed to address the instructional-time problem, it gives no guidance on the potential costs of increasing instructional time, or advice on how to allocate the additional instructional time among subjects. If the problem is due to a gap between policy and provision, the recommendation fails to propose any incentives or penalties that might be used to close the gap. 15 Recommendation Why is this not useful? Better targeting of vulnerable schools and pupils This recommendation, although tied to a problem, would contribute to reducing inequity in education is also vague. Are some “additional resources” more outcomes. In a context of high inequality in learning important than others? Does “better targeting” achievements, as evidenced by the PISA results, indicate that there are now programs in place a revised targeting approach seems necessary focused on vulnerable schools and pupils, but that to ensure that vulnerable pupils and schools these resources have been misdirected in some are provided with additional resources that can way? If so, where are the leakages, and how should contribute to improving their learning outcomes. they be reduced? If there are no such programs, Such targeting would need to be well coordinated what options might be pursued? with other measures to support vulnerable households, which are often managed outside of the Ministry. Address early dropouts and gender disparity This recommendation flags that early dropouts and through the development of supply and demand- gender gaps are problems whose solution requires side interventions. a two-pronged supply-and-demand approach.  It gives no guidance on the types of approaches or tradeoffs among them.   PART II: CHECKLIST FOR AN EDUCATION PER ANALYSIS 17 Part II takes users through what a comprehensive public expenditure review of the education sector should examine. There is no standard outline for such reviews. A review may focus on selected topics or specific subsections, depending on the client’s priorities, the World Bank’s due diligence concerns, the deadline for completing the PER, and budget. In other words, these Guidelines address the six key questions discussed in Part I in a comprehensive manner, but most public expenditure reviews may cover only some of these topics and subsectors. Part II is structured to respond to the six questions in order. Section 1 starts with background and an overview of the education system. Section 2 discusses an overview of education financing and spending (Questions 1–2). Sections 1 and 2 are descriptive and provide a broad picture of the education finance system. Section 3 reviews financial accountability mechanisms qualitatively (Question 3). Sections 4–6 are heavily analytical and examine three principles of education finance: adequacy and sustainability, efficiency and effectiveness, and equity (Questions 4–6). Key Questions 1. Who finances education and how are funds channeled? 2. How much does the government spend and on what? 3. Is the public financial management system set up to enhance financial accountability? 4. Relative to the government’s policies and standards, how much is needed now (adequacy), and what can be afforded in the medium and long term (sustainability)? 5. Are public resources being used efficiently and effectively? 6. Does public spending promote equity? 18 Education Public Expenditure Review Guidelines Section Background and Overview of the 1 Education System Background A background section should provide the sectoral background information as well as the objectives of the public expenditure review. • Sectoral background. Provide a brief summary of sector characteristics relevant to the review. This section should be concise and focused on information relevant to the review. • Major progress and remaining challenges since the last public expenditure review. Acknowledge the client’s reform efforts and achievements, identify why some recommendations have not been implemented, and flag persistent challenges. • Objectives and scope of this review. Define the review’s objectives and scope, including the key policy questions that it addresses. Give the reasons for excluding any important subsectors or topics, and assess how these exclusions might affect a comprehensive analysis of the sector’s financing. • Data sources and analytic methods. Describe data sources (see Technical Notes 1–6) and analytic methods. This section should also address any challenges related to data availability and quality. Example 2: Data sources Overview of the education system An overview of the education system should be presented at the outset to provide basic information about the system and to set the context. • A diagram of the structure of the country’s education system helps the review team and readers of the completed review understand how it is organized. ▷ Structure. Levels of education by grade and official ages for each grade and level. What are the government’s policies on multiple shifts and multi-grade? ▷ Flows (pathways) among levels and types of education. Allowable pathways usually have implications for equity, efficiency, and learning outcomes. For example, students may not be allowed to enter university from an upper-secondary vocational education program. ▷ Type of schools. What types of education providers manage schools? Schools include public, private (financially dependent on the government), private (privately managed without government financing), and religious. Private providers can include nonprofit entities such as non-governmental organizations, or for-profit groups. 19 • Does the education system require private school students to take the same examinations as public school students? How bureaucratically easy or difficult is it for private providers to go into business? Does the state have policies to let private providers access capital more easily? • The overview should provide basic educational statistics (see Technical Note 7: Definitions and notes on indicators). Internationally comparable indicators are available in World Bank EdStats (see Technical Note 4: Cross-national data). Relevant statistics include: ▷ Number of schools by level, location, and type ▷ Number of students in school by level, gender, school location, and school type ▷ Net enrollment rate (NER) by level; if this rate is not available, gross enrollment rate (GER) by level ▷ Number of out-of-school children ▷ Dropout rate by grade ▷ Repetition rate by grade 20 Education Public Expenditure Review Guidelines Section Overview of Education Financing 2 and Spending Key Question 1: Who finances education and how are funds channeled? Education funds can come from government or non-governmental and external sources, and can be channeled through a variety of agents to a broad range of education providers. It is essential to review overall education spending by financing sources, which may include not only public funding, but also private and donor funding. Public expenditure reviews traditionally have not examined the revenue side of the education budget, but it is important to explore how countries raise revenues to finance education. Public funding may be raised at central and local government levels, as well as at the school level. Consider creating a flow chart of financing sources and channels. Policies on fiscal decentralization and school autonomy, together with intergovernmental financing arrangements, define how education funding is channeled down to schools or generated at the school level. These policies can also result in horizontal and vertical imbalances,7 and define how much of education spending is discretionary at each level of government and at the school level. The central government may allocate the education budget to local governments conditionally or unconditionally, based on a transparent methodology or based on negotiation. Schools may or may not have autonomy over budget. Private funding can account for a significant portion of education funding, typically at preschool and tertiary education levels, often at the secondary level, and sometimes for technical and vocational programs. Household surveys can be used to estimate families’ education expenditures. Donor funding may play an important role in financing education in some countries. It could be integrated into the government budget or be off-budget. • How much is spent on education in total, and who finances it, i.e., public, private, or international sources? Technical Note 8: Definition of source of funds Example 3: Analysis of source of funds • How does the government raise revenues to finance education? ▷ At the national level, which sources generate revenues for the national education budget by law (e.g., general revenue or taxes, profits from natural resources, profits from nationalized industries, revenue from lottery, dedicated sources, stabilization funds)? ▷ Do local governments raise revenues, or are all taxes collected by the central government? Example 4: Analysis of revenue sources • Which levels of government finance education? ▷ If financing is split between the central and subnational levels, which levels of government pay for what? Local governments may fund certain levels of education, such as preschool and basic education, but not upper secondary education. They may fund certain inputs, such as school maintenance, but not capital expenditures. Example 5: Analysis of decentralized financing 21 ▷ Are there fiscal transfers from the central government to local governments for financing education? ■ Are these grants earmarked for education or unconditional—i.e., local government can spend the money as it wishes and, in theory, could decide to use none of it on education? ■ How are the grants calculated? ■ Is there evidence of vertical imbalances? ■ Is there evidence of horizontal fiscal imbalances? If so, do they emerge from variations in Example 6: Analysis of education financing by level of government and intergovernmental fiscal transfers • Do schools receive grants? If yes, how are they allocated? How are allocation formulae defined? • Do schools charge parents fees for school? What types of school fees are charged at the primary and secondary levels? Technical Note 9: School fees and other informal payments • Are schools allowed to keep school-generated revenues, such as fees or revenues from entrepreneurial activities, the sale of products in vocational schools, or the rental of school space? For what purposes can public schools use private or international revenues? Example 7: Analysis of school budget by financing source • How much do households with children in school spend on their schooling and for what goods and services? What share of their expenditures consists of mandatory payments (e.g., tuition fees) versus voluntary spending? Example 8: Analysis of household surveys on private spending • Do households receive cash transfers, including vouchers to attend private schools? • Does the private sector, or do other actors, contribute to education spending, e.g., co-financing technical and vocational education and training, or cash transfers to households? • How much do international actors finance education? What is the nature of this spending? How dependent is the government on donors? Example 9: Analysis of donor funding 22 Education Public Expenditure Review Guidelines Key Question 2: How much does the government spend and on what? Rationale for public intervention in the education sector The public expenditure review must scrutinize and justify public intervention in the education sector. Assessments of the following issues should be a subtext of your report: • Proponents of government intervention in education cite a need to ensure equity and efficiency in the sector. The public sector has an equity role because markets generally fail to ensure equal opportunities for all citizens—and, in fact, often create inequities. The efficiency concerns emerge from market failures and imperfections that are commonly associated with information asymmetries,8 externalities,9 and economies of scale. • Government involvement can include regulation, the provision of information, financing, and service provision. Although the government always has regulatory and information-dissemination roles, and is almost always involved in financing education, it does not have to provide such services themselves to ensure equitable access and quality programs. In fact, if the private or nonprofit sector provides education to the country’s standards at lower cost, the government may better serve consumers of education services by subsidizing the consumers, the non-public providers, or both. • Opponents of public intervention in education cite governments’ failure to meet the goals of such intervention or to ensure sustainable financing. Worse, critics say, the governmental involvement can crowd out potentially efficient and equitable private investment and activity. • Where government should intervene, the key challenge is to find ways to prevent failure due to bureaucratic inefficiency, rent-seeking,10 elite capture,11 and other abuses. Overview of public spending on education The public expenditure review first analyzes public spending on education at the aggregate level. What is the total public expenditure on education? As a percentage of gross domestic product (GDP)? As a share of total public expenditure? • Work with the consolidated budget (central government and local government budgets). • Check for off-budget expenditures. • Work only with executed budgets for the past fiscal years because planned is not executed. If you cannot access recent data for executed budgets and decide to include the preliminary figures, clearly label them as preliminary. • Use real expenditures as opposed to nominal ones if you want to track annual percentage changes in the public funds flowing into the sector. Technical Note 10: Calculating the share of a nation’s resources going to education Example 10: Analysis of total public education expenditure The World Bank Group 23 23 Public spending on education can be broken down by functional, economic, administrative (organization), and program classifications.12 Functional classification. A functional classification of expenses organizes government activities according to the socioeconomic objectives that policymakers want to achieve through various kinds of expenditure. Many, but not all, countries adopt the internationally comparable Classification of Functions of Government (COFOG). This classification follows the level categories of the 1997 International Standard Classification of Education (ISCED-97) of the United Nations Educational, Scientific and Cultural Organization (UNESCO). The UNESCO document breaks down education levels into pre-primary and primary, secondary, postsecondary non-tertiary, tertiary, education not definable by level, subsidiary services to education, R&D education, and education services not-elsewhere- classified (or n.e.c.), such as administration, operation, or support of activities. Note that in countries that have not adopted the Classification of Functions of Government, subsectors may be defined differently. Even in countries that have adopted the classification, the length of a given education program, such as primary education, may vary. • Look for anomalies, such as a disproportionate financing share for tertiary education relative to basic education. Example 11: Analysis of education spending by functional classification • Calculate the public expenditure per student as a percentage of per capita gross domestic product (GDP), by level of education? Functional classifications are also useful in analyzing the allocation of resources among sectors. The review should explore if marginal investments in other sectors can help to further education goals. For example, investing in rural roads or improved access to clean water can help to address barriers preventing girls from attending school, and may have more effect on school enrollment at less cost than direct investments in the education sector. If funding for human capital development—e.g., for technical and vocational education and training programs—is dispersed across sectors, other ministries besides the Ministry of Education may provide better value for money. Economic classification. The economic classification of expense identifies the types of expense incurred according to the economic process involved. It includes compensation of employees (wages and salaries and employers’ social contributions), use of goods and services, consumption of fixed capital, interest, subsidies, grants, social benefits, and other expenses. • Consider how education expenditure is allocated by economic factors, such as capital versus recurrent expenditures; staff compensation versus non-staff recurrent expenditures; expenditures spent on teachers versus non-teaching staff? Example 12: Analysis of education spending by economic classification • Be alert to important complementarities among inputs. The educational effectiveness of certain inputs depends on the simultaneous provision of other inputs. Thus, classrooms need teachers; teachers need textbooks. Not infrequently, wages and benefits for labor in the sector crowd out recurrent expenditures for items such as textbooks and other learning materials or infrastructure 24 Education Public Expenditure Review Guidelines maintenance. Capital investments can be paltry relative to the additional school seats required due to increased enrollment in an education subsector. Capital investments can be misallocated among education subsectors. For example, heavy capital investments to build new university campuses can starve the capital budget needed for pre-tertiary education. For vocational education and training, it is important to check whether the budget covers the costs of regular upgrades of workshops and equipment. ▷ Are salaries crowding out the non-salary recurrent budget? ▷ How much is available for routine maintenance per school? How adequate is this amount, compared with estimates of the costs of routine maintenance for the average school? ▷ Considering the government’s textbook policy, is an adequate amount spent per child on textbooks?13 ▷ Are teachers given incentives tied to performance and services? Administrative (organizational) classification. The administrative classification identifies the entity that is responsible for managing the public funds concerned. Public education budgets may be spent by various ministries, different levels of governments, as well as on- and off-budget. • Figure out which levels of government are responsible for spending resources on education. This question is different from who finances education. For instance, the central government may transfer (finance) education block grants to local governments, but the latter is responsible for spending the block grants. • Check the budgets of all ministries that might have education expenditures for all types and levels of education addressed by the public expenditure review. For example, sometimes the budget for early childhood education falls under a ministry responsible for the welfare of women and families. All, or some, of the budget for vocational education and training may come under the Ministry for Labor. Program classification. The program classification gives detailed costs of every activity or program that is to be carried out together with objectives and expected results of a proposed program. Reports on program-based budget execution provide extremely useful data on the extent to which education policies and plans have been successful, and they can stimulate policy changes and in-year course corrections. However, this type of budgeting requires strong expenditure tracking, monitoring, and accountability arrangements that are often not in place in our client countries. Transitioning to a program-based budgeting system not only requires improvements in the government accounting system, but also a shift in internal controls and a shift in accountability from inputs to program outputs. • Are the expected outputs and outcomes specified for each program? • Is the government accounting system capable of providing data on budget execution by programs and sub-programs? • Are the data on actual outputs and outcomes achieved credible? • Do budget-holders and implementers have the flexibility to reallocate among budgetary items in order to achieve expected outputs and outcomes? 25 Section Financial Accountability 3 Key Question 3: Is the public financial management system set up to enhance financial accountability? Effective systems of public financial management (PFM) can contribute to providing better inputs and improving accountability. The World Development Report 200414 stressed that better public financial management can ease several key challenges facing the education sector, including inequitable access, dysfunctional schools, low quality of instruction, low client responsiveness, and stagnant productivity. More recently, the World Bank Group’s Education Strategy 2020 (2015) recognized the important link between better public financial management and improved service delivery outcomes. For instance, equitable allocation of resources can help improve access to education for the poorest families; performance-based incentives can motivate teachers; and efficient spending can lead to savings that can, in turn, help pay for improvements in the learning environment. While theoretical links between improved public financial management processes and service delivery outcomes are well established, empirical evidence has been more limited in terms of when, and how, such reforms might contribute to better services (Welham et al. 2013; World Bank 2012). A number of country examples show strong links between education outcomes and public financial management improvements, but the evidence remains circumstantial.15 A range of public financial management analyses is available to help users assess whether such a system established for the education sector is successful in ensuring that government-earmarked funds are spent on policies as intended. These analyses include the overall Public Expenditure and Financial Accountability (PEFA) Framework,16 sector-level assessments such as SABER School Finance, Public Expenditure Tracking Surveys (PETS), Quantitative Service Delivery Surveys (QSDS), Service Delivery Indicators (SDI) surveys, project fiduciary assessments, and audit reports. The PEFA Framework is helpful in identifying common public financial management issues across sectors. Sector-level and project fiduciary assessments would provide more details on how issues involving public financial management systems specifically affect the implementation of the education program. Internal and external audit reports of the education sector and a walk-through of public financial management processes—all critical business processes related to management and use of funds and assets—could help identify potential weaknesses in these processes. Technical Note 11: Public expenditure and financial accountability (PEFA) country reports Technical Note 12: Public Expenditure Tracking Surveys (PETS), Quantitative Service Delivery Surveys (QSDS), and Service Delivery Indicators (SDI) 26 Education Public Expenditure Review Guidelines The public expenditure review team should use a three-dimensional approach to analyze potential public financial management weaknesses underlying spending problems, as appropriate . Public financial management laws, policies, and procedures determine and regulate the behavior of public officials and organizations implementing them (Allen et al. 2013). Where laws and procedures are sufficiently appropriate but practices lag, it is also necessary to consider the capacity of actors who implement these laws, and the process through which the actors bargain over the design and implementation of policies within a specific institutional setting (World Bank 2017). A three- dimensional approach to a public financial management analysis considers the following factors: (i) Public financial management legal framework—the adequacy of laws, policies, and procedures. What are the shortcomings and how do these affect education sector spending? Are these laws, policies, and procedures complied with? What are the key reasons for any deviations in both public financial management processes and outputs? Are certain political economy factors at play? Why do oversight agencies fail to pick up these deviations? (ii) Public financial management capacity—capacities at central and decentralized levels. What public financial management training and reform mechanisms are in place at the country level and specifically in the education sector? How well does the Financial Management Information System enable the implementation of the public financial management legal framework? (iii) Institutional setup—who does what and the mapping of public financial management roles of central ministries, decentralized agencies, and other factors. Are financial and transaction authority adequately delegated? What role and impact do civil society, unions, media, and the international community have as pressure groups to influence policy formulation and program implementation? Map the impact of institutional and individual actors’ interests on public financial management. Every spending problem is likely to have a unique set of underlying weaknesses in the public financial management process. Linking such weaknesses to key spending problems can be extremely useful in stimulating necessary reforms. How do public financial management processes, such as budget release, funds flow, and internal controls, apply to each of the main spending problems? What do the links between PFM arrangements and sectoral spending problems imply about the reforms needed? For example, textbook purchases generally require a one-off bulk procurement followed by a logistics process of distributing books from central level to schools. Schools then reuse these books over several years. In contrast, school grants do not require a bulk procurement or logistics involving physical assets. However, the issuance of such grants often entails a long (and often complex) funds-flow from central- or provincial-level governments to schools—with the process then relying more on local accountability mechanisms to ensure the funds’ effective use. Table 3 summarizes major spending problems and potential underlying weaknesses in the public financial management system. Example 13: Issues related to budget formation and execution 27 Table 3: Major spending problems and potential underlying PFM weaknesses Spending problem Potential public financial management weaknesses Disconnect • Financial and human resource data are not used for policy- and decision-making. between the • Sector plans do not define institutional responsibilities for different levels education policy of the government for each programs and subprogram. intent and • The government budgeting and accounting system does not track implementation program and subprogram expenditures due to several system and capacity issues, including challenges in apportioning joint costs over multiple (sub)programs. • The central government (Ministries of Finance and Education) has limited authority over, or capacity to handle, the implementation of sector plans to influence decisions made at subnational and local levels. • There is limited or no incentives or accountability linked to results at subnational and local levels. • Budget-holders who are responsible for the best use of the available budget and implementers have limited flexibility to reallocate the budget among input categories for mid-year course corrections. • Delayed budget releases and mid-year cash rationing with cumbersome funds flow (including of school grants) to lower levels are holding up non- salary and capital expenditures. • Weak procurement and internal controls result in delayed or cancelled implementation of activities. Inadequate • Links are weak among target outputs, outcomes, and the budgets allocated. budget allocation • The multi-year sector plan does not adequately cost out stated education policies or medium-term strategic plans. • Education resources are diverted to other functions due to weaknesses in the chart of accounts or lack of transparency in financial reporting. • The sector has budget arrears. • The sector plan has not fully integrated donor-financed projects. Inequitable • Block grants to subnational and local governments are inadequate for budget allocation equalizing fiscal imbalances among the localities (i.e. horizontal equalization). • Education spending results in disproportionate allocations to poor and vulnerable students. Poor human • Delays in receiving salaries lower the motivation of teachers. resource • Lack of hardship incentives limits teacher deployment to rural or remote areas. management • Weak, or lack of, management system for leave record leads to inadequate docking of salaries for absenteeism. (Table continued on next page) 28 Education Public Expenditure Review Guidelines Spending Problem Potential public financial management weaknesses • Sanctions and other action in response to audit findings are limited, even in cases of repetitive or permanent absenteeism. • Lack of independent oversight of teacher attendance increases the possibility of collusion with school directors and district education staff. • Internal and external auditors, monitors, inspectors, and supervisors apply duplicate and document-based controls in an uncoordinated manner, with no, or minimal, unannounced physical checks. • Personnel and payroll records for public servants in the education sector are not reconciled regularly to account for a possible difference between the two. • School directors or teachers cannot perform their duties in school because they are assigned to administrative functions at the district level or are given political party duties. • Teacher unions are politicized and represent a significant pressure group. Lack of school • School directors’ inability to make independent decisions about hiring autonomy temporary teachers results in an inadequate or excess supply of teachers. • School councils are not sufficiently empowered to participate in decisions on the use of school grants, or to monitor teacher attendance. • School councils have no, or limited, involvement in monitoring school performance. Insufficient • Procurement of textbooks, which is often centralized, is delayed. and/or delayed • Storage and distribution logistics are poorly managed, resulting in availability of pilferage of textbooks. teaching and • Controls and incentives that encourage the reuse of textbooks over learning materials several years are limited. • Teaching materials are often procured locally, but lengthy and cumbersome procurement procedures apply. • Delays take place in the receipt of school grants. Poor school • Budget allocation is inadequate for ongoing or planned construction projects. infrastructure • School construction is not completed as planned, with no return on capital investment, and with exposure to rapid deterioration and cost escalations. • Politicized allocation of funds to schools leads to suboptimal regional coverage. • Inadequate technical supervision by the public works staff leads to poor construction quality. • Funding for school maintenance is inadequate and responsibilities are unclear. • To simplify contracting, authorities use expensive, standard school-building designs countrywide, instead of applying climate-suitable adaptations. (Table continued on next page) 29 Spending Problem Potential public financial management weaknesses Over-reliance on • Policy and capacity to regulate and support private schools (e.g. grants, public provision access to finance) are inadequate, not transparent, and not predictable in a way that would incentivize them to provide good services. • Private schools lack accounting capacity. • Regulatory authorities for private schools engage in limited coordination with the Ministry of Education. 30 Education Public Expenditure Review Guidelines Section Adequacy and Sustainability 4 Key Question 4: Relative to the government’s policies and standards, how much is needed now (adequacy), and what can be afforded in the medium and long term (sustainability)? Although international and regional benchmarks are useful in terms of advocacy and cross-country analyses, the adequacy of the budget for a specific country must be carefully assessed. The Third International Conference on Financing for Development (in Addis Ababa, July 2015)17 set the following international benchmarks for education spending: at least 4 percent to 6 percent of gross domestic product (GDP), and at least 15 percent to 20 percent of total public expenditure to the education sector. Worldwide, the former target has been met, but not the latter: In 2012, countries, on average, allocated 5.0 percent of GDP and 13.7 percent of public expenditure to education (UNESCO 2015c). However, benchmarking should be used cautiously because many factors affect total expenditure in the sector, such as the government’s financial capacity, demography, enrollment rates by subsector, quality and prices of basic inputs, geographical challenges, and policies on public versus private financing. A comparison with neighboring countries may be helpful, but regional neighbors may share the same difficulties as the country in question, and hence, may not be appropriate benchmarks. Two other concepts are useful for measuring the adequacy of education spending. Countries have different education goals and standards, and their cost of achieving them will vary, independent of price differences. In the short term, adequacy can be measured relative to the costs of the inputs required, if the country meets its own provision standards for monetary and non-monetary inputs. These standards include those governing unit cost per student, teacher compensation, and the ratios of students to teachers, classrooms, or textbooks. Detailed descriptions of inputs are discussed below. The second concept for measuring adequacy is obvious shortfalls, as evidenced by arrears in the sector or inadequate expenditures on inputs complementary to teachers, such as teaching and learning materials, school maintenance, and needed capital expenditures. Teacher compensation is typically the largest expense in the education sector, and personnel costs are often met by underfinancing non-salary expenses. Key input indicators Besides macro indicators such as education spending as a percent of GDP and of total public spending, the adequacy of monetary inputs can be measured in two ways. One way is to calculate per student spending as a percentage of per capita gross domestic product (GDP), by level of education for public and private schools. The second way is to measure variations in per student spending among subnational units or schools and between public and private schools. Use comparative data to get some sense of whether, as indicated by costs, the country’s provision of inputs—on average and as they are distributed across schools—is “out of range,” either in terms 31 of apparent over-provision or under-provision. Since country- and sector-specific conditions affect allocations among differently priced inputs, comparisons can only flag if there might be a question to answer. Several different ways are available to compute per student spending, or unit costs. Technical Note 13: Definitions and notes on monetary-input indicators Technical Note 14: Calculating unit costs from aggregate and itemized spending Technical Note 15: Calculating unit costs when the country’s fiscal year and school year do not coincide Example 14: Analysis of per student spending Example 15: Minimum norms and standards for resource allocation Example 16: Analysis of cost of teachers The adequacy of non-monetary inputs (human resources, classrooms, and infrastructure) can be measured by scrutinizing government standards and by assessing their actual provision. Governments, including the Ministry of Education and sometimes parliamentarians, set standards governing inputs to education. These include standards for where to build new schools relative to population settlements, construction designs and standards, student-classroom ratios, student- teacher ratios, student-textbook ratios, annual hours of teacher training, annual instructional time overall and per subject, annual duration of schooling, and so forth. However, actual provision of inputs may or may not meet those standards. Therefore, it is important to examine both standards and actual provision, and identify whether either, or possibly both, may be inadequate to deliver quality educational services. Both are problems that PERs can address, but their solutions differ. The former requires changes in policies that define standards; the latter, changes in how inputs are distributed. Obtain information on input standards by interviewing the Ministry’s policymakers or the senior technical staff, backed by relevant policy documents. Data on actual provision of inputs should be disaggregated by level of education or grade (whichever appropriate) and by region, district, and school (using the smallest unit for which data are available), within a country. They should also be compared over time within the country, and with data for comparator countries. Technical Note 16: Definitions and notes on non-monetary input indicators Technical Note 17: Definitions and notes on research indicators Check these aspects of the three main types of non-monetary inputs: • Human resources ▷ Student-teacher ratios ▷ Percentage of qualified teachers ▷ Ratio of teachers to non-teaching staff ▷ Non-teaching staff per type of school ▷ Organization of teachers’ working time Example 17: Analysis of teacher distribution 32 Education Public Expenditure Review Guidelines • Classroom inputs ▷ Average class size ▷ Student-textbook ratios for core subjects ▷ Percentage of schools meeting minimum standards requirements for educational inputs ▷ Annual instructional time Technical Note 18: Efficiency of the curriculum • Infrastructure ▷ Average school size (number of students per school) ▷ Unit cost of building a classroom ▷ Standards and schedule for infrastructure maintenance ▷ Percentage of schools that run double and triple shifts ▷ Percentage of schools that use multigrade classrooms ▷ Percentage of schools meeting minimum standards requirements for learning conditions Comparing the projected costs of education with the anticipated size of the resource envelope can shed light on whether planned spending levels are realistic. Beyond considering how much is needed now, it is essential to examine what happens to costs in the medium and longer term, given the government’s sectoral goals, demographic projections for the school-age population, assumptions about enrollment rates by level, and macroeconomic and budget forecasts. Projected costs of education provision are often spelled out in the Medium-term Expenditure Framework (MTEF). If the country does not have such a framework in place, a costing exercise will be required. Example 18: Cost projections Example 19: Fiscal sustainability analysis • Sectoral goals. Medium- to long-term education goals and targets may include extending the length of compulsory education, increasing education access for the poor, hiring more qualified and, thus, expensive teachers, converting community teachers to regular teachers, introducing computer-assisted instruction, retrofitting facilities to protect against earthquakes, expanding vocational education and training, and introducing cost recovery for tertiary education. It is essential to examine whether these goals are fiscally feasible. • Country’s demographic structure and trends, including urbanization, and projected enrollment rates by subsectors. Rapid urbanization can increase future enrollment rates. What do the trends in the school-age population and assumptions about trends in enrollment rates by subsector imply about the inputs required now and in future? Technical Note 19: Demographic trends and enrollment projections Example 20: Demographic trends and enrollment projections • Country’s macroeconomic projections and government’s strategic priorities. The former affect the total resource envelope available; the latter, the government’s decision on how to allocate resources among different sectors. 33 If at all possible, do some simple modeling to assess what the joint implications are for projected costs and revenues and, thus, how realistic the government’s plans are for the sector. For instance, UNESCO’s Education Policy and Strategy Simulation Model (EPSSim) is a sector- wide and goals-based generic simulation model that is driven by demographic trends. Enrollment targets are taken as a priori and the simulation calculates the corresponding financial resource implications. The World Bank’s Maquette for MDG Simulations (WB-MAMS) is a Computable General Equilibrium (CGE) model and can help assess the broad, economy-wide effects of alternative education scenarios. Be sure to work closely with the country economist or the macroeconomist on the public expenditure review team. Technical Note 20: UNESCO’s Education Policy and Strategy Simulation Model (EPSSim) Technical Note 21: Simulating the economy-wide effects of alternative education scenarios 34 Education Public Expenditure Review Guidelines Section Efficiency and Effectiveness 5 Key Question 5: Are public resources being used efficiently and effectively? Resources are scarce, and an important purpose of any PER is determining value for money from education investments, regardless of their financing source. The purpose here is similar to the “value for money” (VfM) approach that the United Kingdom’s Department for International Development (DfiD) has adopted for making aid decisions;18 that the World Bank has pursued for decades in its economic analyses of investments; and that studies such as cost-benefit and cost- effectiveness analyses and impact evaluations support. The basic value for money concept is that of obtaining the maximum benefit over time with the resources available. Value for money is high when an optimum balance exists among three elements: when costs of inputs are relatively low, productivity is high (or efficient), and successful outcomes have been achieved (or effective).19 It should be noted that efficiency and effectiveness are not the same. A service or good may be efficiently produced, but not effective. Similarly, it may be effective, but not efficiently produced. For example, a teacher-training program may produce a large number of graduates at a small cost, yielding an efficient cost-output ratio. However, if these teachers’ classroom performances shows no discernible improvement, the training is efficient but ineffective. A student-to-qualified teacher ratio of 15/1 may result in learning gains. However, if increasing the ratio to 25/1 achieves the same learning gains, using qualified teachers is effective, but at the 15/1 ratio, inefficient. Output and outcome data To assess efficiency and effectiveness, we need input, output, and outcome data. Section 4 above defined monetary and non-monetary inputs. The primary output and outcome data needed are the following, including trend data for the PER country and comparative data: Technical Note 22: Definitions and notes on output and outcome indicators • Participation rates ▷ Gross and net enrollment ratios by level of education ▷ Dropout rate by grade ▷ Repetition rate by grade ▷ Completion rate for each educational program • Learning outcomes for core subjects Technical Note 4: Cross-national data • Employment and wage rates by level of education ▷ Labor force participation rate ▷ Employmentandunemploymentrates Several different types of economic efficiency analyses exist, but those most useful for education public expenditure reviews are analyses of allocative efficiency, technical efficiency, internal efficiency, and external efficiency. Technical Note 23: Concepts of effectiveness and efficiency 35 Allocative efficiency Allocative efficiency asks whether the sector is allocating its resources among subsectors “so as to maximize the welfare of the community.” Section 2 asked for an analysis of how public expenditure is allocated among education levels or subsectors. Subsector allocations for the country undergoing a public expenditure review can differ for good or poor reasons from those of countries with which it is being compared. For example, the definitions of education levels can vary across countries. Eliminating school fees usually creates enrollment bulges that require high “catch-up” allocations to the newly “free” subsector. Significant success in obtaining high primary-completion rates usually translates into sharp spikes in demand for secondary school seats that then require major increases in capital budgets for the subsector. However, high and increasing expenditures for tertiary education may crowd out allocations for pre-tertiary-level programs—a trend that ultimately favors the children of wealthier families. Assess the allocative efficiency of the standards themselves, independent of how well the public expenditure review country meets them, to separate inefficient objectives from inefficient implementation. As discussed in Section 4, the sector usually sets standards for inputs, such as student- teacher ratios, teacher compensation schedule, student-textbook ratios, and student-classroom ratios. Comparative data for these variables will reflect, not comparators’ standards, but how comparators implement their own standards. However, these data will give some idea of reasonable ranges for standards in the country undergoing review. Figure 4: Distribution of student-teacher Start with simple histograms to display the distribution ratios for primary schools in a province of an input across regions, schools, or classrooms. Highly variable deployments of inputs always signal allocative inefficiencies. For example, if the sector standard for the Share of schools (%) student-teacher ratio in a primary class is 35 students per teacher, the hypothetical example in Figure 4 shows wide (but not unusual) variances in these ratios between schools— inherently and relative to the 35:1 standard. Completely aside from whether the standard itself is or is not reasonably efficient, the very uneven deployment of a key resource is inefficient Student-teacher ratio (and inequitable). When significantly variable deployments are observed, the review must try to determine the sources of Note: A hypothetical example the variability. Technical efficiency The sector achieves value for money when it gets the best outcomes at least cost. This result can be achieved in either of two ways. The inputs used by a given intervention can be reduced to the minimum required to achieve the outcome sought. Or a different intervention can be used— one with a different combination of resources that achieves, at less cost, the outcome sought, as well as, or better than, the alternatives. Example 21: Technical efficiency of inputs (efficiency indicators) 36 Education Public Expenditure Review Guidelines Analyses of technical efficiency look at costs, input mixes, and results. Cost-effectiveness analysis (CEA) relates monetary inputs and non-monetary outputs and outcomes. Cost-benefit analysis (CBA) is used when both the costs and the outcomes can be monetized. A public expenditure review is not usually expected to conduct either type of analysis. However, the review may use findings of international and in-country cost-effectiveness analysis and cost-benefit analysis to help identify instances of probable low value for money. For instance, in examining six recent systematic reviews or meta-analyses of interventions that improve learning outcomes in low- and middle-income countries, Evans and Popova (2015) observed a wide variation in conclusions across the studies, with much of the variation driven by variation within categories of interventions. Thus, the specific details of the intervention determined its effectiveness. Based on a careful examination of the details of interventions in the six studies, they identified three categories of interventions that were broadly supported across the studies: “(i) pedagogical interventions that match teaching to student’s learning, including through the use of computers or technology; (ii) individualized, long-term teacher training; and (iii) accountability-boosting interventions, such as teacher performance incentives and contract teachers” (Evans and Popova 2015, 3). These findings may help the PER team identify a potentially inefficient area that the sector is strongly advised to investigate. Technical Note 24: Cost-effectiveness analysis Example 22: Analysis of unit costs and outcomes Technical Note 25: Cost-benefit analysis Example 23: Cost-benefit analysis Use cost data to sense whether the sector may be overpaying for an input relative to the value obtained. Especially relevant data are costs of teachers relative to in-country comparators, costs of textbooks relative to comparators, and in-country variations in unit costs for classroom construction. For example, private school teachers may be considerably cheaper than public school teachers. If learning outcomes in private schools are similar to, or better than, those in public schools, controlling on family socioeconomic status (SES), the sector may want to consider shifting more provision of education to the private sector via vouchers or subsidies. Significant in-country variations in unit costs for classroom construction also raise questions about value for money. Theunynck (2009) finds that the most efficient designs are ones that can be mass produced at low cost on the local market, using small- and medium-size enterprises in the formal and informal sectors. He also alerts us to the fact that alternative procurement and contract management arrangements differ in their cost-effectiveness. A relatively complex analytic technique called data envelopment analysis (DEA) can be used to measure the efficiency of multiple service delivery units by comparing the mix and volume of resources used (inputs) and services provided (outcomes) by each unit. DEA is a linear programming methodology. The techniques can be used to define and estimate efficiency as the distance from the observed input-output combinations to an efficient frontier, which is the maximum attainable output for a given input level. It is often used to compare countries relative to an efficiency frontier. For example, for a large panel of countries, total government education expenditures as a percent of GDP (or GNI) can be plotted against different education outcomes, such as primary completion rates, enrollment rates, or learning outcomes.20 Example 24: Data envelopment analysis 37 Internal efficiency21 Internal efficiency measures the percentage of children who complete an educational cycle (e.g., primary education or lower secondary education) as a share of those who start the cycle or as a percentage of those who finish the cycle in the minimum number of years. The first definition allows the calculation of the dropout rate—i.e., the number of those who start minus the number who ultimately complete, as a share of those who start. The second definition measures the dropout rate plus the repetition rate. If the data show either relatively high dropout rates, high repetition rates, or both, the sector has an expensive internal-efficiency problem. Dropping out imposes costs on individuals and countries in the form of unrealized human capital. Repetition imposes costs on the sector in the form of its having to pay double (or triple) the unit cost of a year of school per repeater. In cases of significant internal inefficiency, the difference between unit costs per completer of an education program and unit costs per student who completes the program without interruption are better measures of costs than unit costs. Technical Note 26: Calculating the budget costs of one completer and of one uninterrupted completer Example 25: Internal efficiency indicators External efficiency22 External efficiency measures the returns to individuals, employers, and the country of public and private investments in education. It depends on a match between the type and quality of skills and knowledge that school leavers acquire in school relative to the skills and knowledge needed and paid for by employers. Does education improve the employability and wages of school leavers? Does public investment in education and training contribute to the country’s growth and economic development? Measuring and linking employment and wage returns to education is particularly important for vocational education and training, and tertiary education. Technical Note 27: Estimating private rates of return to education Example 26: Rate of returns to education 38 Education Public Expenditure Review Guidelines Section Equity 6 Key Question 6: Does public spending promote equity? A fundamental responsibility of the state is ensuring equity and managing redistribution. Public policy, including educational finance policy, can help minimize subgroup differences in educational access and achievement. This section explores: (i) how to identify inequity, if any; (ii) whether, and how, the government spends its budget to promote equity in education; and (iii) how households are responding to public policies and filling the funding gap between their needs and public spending. Equity in education financing can be assessed in terms of two main principles: horizontal and vertical equity. Horizontal equity is defined as the equal treatment of equals and is used to justify similar levels of funding across comparable schools or subnational divisions. Accordingly, a mechanism may operate to equalize education spending across subnational divisions to preserve fiscal neutrality, so that the amount of available resources for education is not positively correlated with the wealth of where a student lives. Vertical equity supports the unequal treatment of “unequals” (Underwood 1995). For example, progressive spending may be necessary to provide equivalent education to students whose native language is different than the language of instruction or to students with special education needs. Other examples include targeted support programs (such as conditional cash transfers and scholarships), or student weights to differentiate spending for certain types of students. Vertical and horizontal equity have a special meaning in tertiary education. According to Salmi and Bassett (2012): The vertical dimension looks at who enters tertiary education and who graduates from tertiary education. The horizontal dimension looks at what kind of institution they attend, and what labor market opportunities are offered to graduates with various types of qualifications and levels of degrees. The first analytical step is to diagnose the country’s trends related to equity in education. Disparities in access to education, completion, and learning achievements across different student groups indicate existing problems of inequity in education. 39 • How do school enrollment rates, completion rates, and learning outcomes vary by gender, household income, geographical location, and ethnic or religious group? Example 27: Analysis of inequity • Do the data assembled for Section 4 on the distribution of inputs by school, district, or region indicate substantial variation among geographical areas in terms of children’s opportunities to learn? • What is public spending per student by level of education, subnational division, school type, and student group (e.g., geographical location, income quintile, gender, ethnicity, language, religion, and special needs)? Technical Note 28: Per capita financing Example 28: Analysis of per capita financing The second step is to examine what role the state plays in mitigating or exacerbating inequity. • What demand-side and supply-side financial interventions does the government adopt to promote equity? The most common programs that address equity include: ▷ Demand-oriented interventions, such as conditional cash transfers, school feeding, vouchers, scholarships or student loans, universal and targeted child benefits, and full or partial subsidies for school supplies, transport, and boarding Technical Note 29: Targeting mechanisms, coverage, and depth of programs Example 29: Analysis of cash transfer programs ▷ Supply-oriented interventions, such as an expansion of the education system to reach the poor, and the provision of additional funding for disadvantaged students23 • How progressive or regressive is the state’s financing of education? ▷ What are benefit incidences across different groups of households? Technical Note 30: Benefit incidence analysis Example 30: Benefit incidence analysis • What actions other than financing does the government take to increase parental demand for education? If parents fear for their daughters’ safety during travel to school, does the government implement measures aimed at protecting girls? For families speaking a minority language, does the government publish textbooks and provide teachers who offer instruction in that language?24 • Does the state finance private education through partial or full subsidies to providers or consumers? What are the rules governing these subsidies? For example, what percentage of the estimated per student cost for public schools does a subsidized private school receive? Can a private school receiving public subsidies also charge fees? Depending on how they are designed, public subsidies may implicitly subsidize the wealthy’s preference for private education. Technical Note 31: Subsidies 40 Education Public Expenditure Review Guidelines • Are there corrupt practices that affect access, grades, or graduation—e.g., bribes to university faculty to secure entry into a particular department, or parental “gifts,” such as new computers, given to gain admission to a prestigious secondary school?25 If so, how widespread are these practices? Has the state taken any actions to stop them? Does the state regulate the practice of the student’s teachers, or teachers within the student’s school, providing private lessons to the student? Bribes, gifts, and private lessons penalize the poor. • What public policies govern students’ progression through the educational system? If tertiary enrollments are rationed, examinations during the pre-tertiary years are often used to “weed” students out of the system. Pathways into tertiary education may be highly restricted. Students may have had to complete the academic program at the upper secondary level, and access to this program may be highly restricted, as well. Such policies favor wealthier families. The third step is to analyze private spending. Use household survey data on households’ education expenditures to assess whether educational disadvantages are related to private costs for education (financial barriers). • What do families in different consumption quintiles pay for education by level of education as a percentage of their average consumption? Consider both formal and informal payments. Technical Note 9: School fees and other informal payments Example 31: Analysis of private spending by income quintile TECHNICAL NOTES 42 Education Public Expenditure Review Guidelines Technical Note 1: Government’s official documents Government’s official documents—such as policies, laws, and regulations that specify standards for inputs to, and financial arrangements for, the learning process—are necessary to differentiate policy intent and implementation. For discussion of inputs, see Section 4: Adequacy and Sustainability. The SABER-school finance data collection instrument26 can provide a more detailed list, as needed. Technical Note 2: Country system-level data Data generated by country systems include: (i) government budget documents (central government consolidated accounts; Ministry of Education budgets; state or provincial budgets, if separate from consolidated government accounts; institution-level financing data; medium-term expenditure framework documents); (ii) education management information systems (EMIS); (iii) national learning assessments; and (iv) sector-specific databases, such as school mapping and teacher databases. It should be noted that data generated by country systems tend to be the most problematic and need to be treated with caution. (i) It is not easy to obtain reliable financing data from the government. Wherever possible, try to obtain data from the Ministry of Finance, instead of the Ministry of Education. BOOST, a World Bank initiative launched in 2010, draws detailed government expenditure data from government financial management information systems and creates easy-to-use databases. The program strengthens public-expenditure policy outcomes and accountability by improving the quality of expenditure data, facilitating rigorous expenditure analysis, and improving fiscal transparency. Experience indicates that BOOST is most useful if the raw data from the Ministry of Finance is sufficiently disaggregated. The Governance Global Practice’s BOOST team could support the public expenditure review team by creating an education module in an Excel format that combines expenditure data from the BOOST database with education statistics and other information on public institutions, service delivery, and households.27 The ease of access to, and preparation of, analytical reports supports decision- making for the purposes of planning, budgeting, monitoring, and evaluation. BOOST often does not have expenditures by programs and sub-programs because governments typically do not have relevant disaggregated data. In collaboration with the World Bank, more than 70 countries have developed a BOOST government budget database to date. If the counterpart does not yet have a BOOST, consider working with the Governance Global Practice and Macroeconomics and Fiscal Management (MFM) Global Practice colleagues to encourage the counterpart to develop a BOOST database. If the education public expenditure review is part of a comprehensive expenditure review, the Macroeconomics and Fiscal Management team is most likely to create a BOOST that the education team can then use. However, teams need to be aware that building a BOOST may take a long time. Also, BOOST tends to lack disaggregated data at the local level. If education functions and financing are decentralized, the education team may need to collect disaggregated data separately from local governments. 43 (ii) Especially in low-capacity countries, education management and information system (EMIS) data can be problematic in terms of validity and reliability, especially when it comes to recent data. The team needs to be alert to incentives in the system that encourage misreporting of the numbers. For example, per capita financing creates incentives to inflate enrollment numbers. In such cases, the public expenditure review team will have to triangulate among data sources to estimate values of school-specific variables, enrollment rates, or teacher absenteeism rates. (iii) Designing and administering a good learning assessment is quite technical. Although it is generally preferable on quality grounds to use well-established regional or international learning assessments, national learning assessments can meet technical design standards. Relative to cross-national learning assessments, they can also be better aligned with the country’s curriculum. However, confer with the Bank’s country education team on the technical quality of these assessments. In using the data, flag concerns raised by the Bank’s team. (iv) The quality of other sector-specific databases, such as school mapping and personnel rosters, varies across countries. However, if they are of reasonable quality, they are a good source for determining the distribution of inputs relative to government standards—for example, infrastructure relative to population settlements, or teachers relative to student enrollments. Technical Note 3: National surveys National surveys measure variables important for PERs and can be a goldmine. For example, household consumption surveys are essential to creating poverty maps and to measuring household payments for education. Census data are essential for estimating changes in the size of school-age cohorts that the education system will have to accommodate. Labor force surveys can be used to estimate employment and wage returns to different levels of education or the educational attainment of the working-age population or active labor force. Although the team needs to determine quirks and data-quality problems with any survey, the quality of national surveys tends to be adequate and may be excellent. The statistics unit of a country tends to conduct such surveys, although they may use internationally established frameworks and processes. Data are collected under the same protocol. The staffs of units conducting these surveys benefit from international experience with the design, administration, and analysis of such surveys. Donors also often fund technical assistance to help such units professionalize the conduct of the surveys under their jurisdiction. Census data and reports. Country-specific estimates and projections can be checked with population experts in the Health, Nutrition, and Population Global Practice (HNPGP). Household and other surveys. Household surveys usually collect data on enrollments and completion by level of education, age, gender, and residential location, and private spending on education. Country-specific survey data include the following common and highly developed sources: • Living Standards Measurement Study (LSMS): The LSMS is a household survey program housed within the Survey Unit of the World Bank’s Development Data Group that provides technical assistance to national statistical offices in the design and implementation of multi-topic household 44 Education Public Expenditure Review Guidelines surveys. Since its inception in the early 1980s, the LSMS program has worked with dozens of statistics offices around the world: generating high-quality data, incorporating innovative technologies and improved survey methodologies, and building technical capacity. • The Skills Towards Employability and Productivity Program (STEP) program: The World Bank’s STEP measures skills in low- and middle-income countries. It provides policy-relevant data to enable a better understanding of skill requirements in the labor market; backward linkages among skills acquisition and educational achievement, personality, and social background; and forward linkages among skills acquisition and living standards, reductions in inequality and poverty, social inclusion, and economic growth. The STEP program includes a household-based survey and an employer-based survey. All relevant survey documentation is provided along with the datasets. The “STEP Methodology Note” presents key concepts and describes the STEP survey instruments. It also provides guidance on how to use the data. • Demographic and Health Surveys (DHS): These nationally representative household surveys provide data on a wide range of monitoring and impact evaluation indicators in the areas of population, health, and nutrition. Education is a key background indicator in demographic and health surveys, which help contextualize a country’s health and development situation. • Multiple Indicator Cluster Surveys (MICS): UNICEF supports governments in carrying out these surveys through a global program of methodological research and technical assistance. Cluster survey findings have been used extensively as a basis for policy decisions and program interventions, and for the purpose of influencing public opinion on the situation of children and women around the world. All available results and datasets from these surveys can be accessed on mics.unicef.org. The results from the most recent MICS-5 surveys, carried out in 2012–15, are becoming progressively available. (MICS-6 was launched in October 2016. It will cover five of the household survey-based indicators for Sustainable Development Goal (SDG) 4, Education 2030.)28 • EdStat’s Education Equality: Household surveys can provide detailed information on gender, income, and geographic inequalities in education access, progression, attainment, and expenditures. EdStats gives users access to household survey data through the following tools and resources: ▷ The Education Equality Query holds household survey data from DHS and MICS. Indicator names beginning with the labels “DHS” and “MICS” were generated by EdStats using the ADePT Education tool. Indicator names beginning with “UIS” were generated by the UNESCO Institute for Statistics (UIS) using its stated methodology. Variances may exist in data from differing sources based on methodological differences. ▷ Education Equality Country Profiles are detailed Excel file reports for all available DHS, MICS, and LSMS. They include a series of graphs and a wider variety of indicators than are currently available in the Education Equality Query. ▷ The Education Equality Dashboard is a data visualization tool that allows users to visualize disparities in attendance rates, completion rates, educational attainment, and out-of-school children. 45 ▷ The ADePT Education tool allows users to derive education indicators from household survey data and create customized reports and graphs. The program contains built-in settings for DHS surveys, but also accepts other types of surveys, and determines automatically what tables can be built from the available data. • University of Oxford’s Young Lives survey: This survey includes a household questionnaire with special items focused on the children and a community questionnaire. It covers Ethiopia, India, Peru, and Vietnam. It is often complemented by a school survey and the collection of in-depth, qualitative longitudinal data on some of the children. The household survey collects data similar to the World Bank’s Living Standards Measurement Study. These include information on household composition, livelihood and assets, household expenditure, children’s health and access to basic services, and education. This is supplemented with additional questions about caregiver perceptions, attitudes, and aspirations for the children and the family. The Young Lives survey also collects detailed data on how all family members use their time, information about the children’s weight and height (and similar information for caregivers), and data on children’s learning outcomes (language comprehension and mathematics). It asks the children about their daily activities, their experiences, and attitudes towards work and school, their likes and dislikes, how they feel they are treated by other people, and their hopes and aspirations for the future. The community questionnaire provides background information about the social, economic, and environmental context of each community. It covers topics such as ethnicity, religion, economic activity and employment, infrastructure and services, political representation and community networks, crime, and environmental changes. Technical Note 4: Cross-national data The two main types of cross-national data systems are international learning assessments and statistical databases that can show where a country sits within the range of practice. In addition to national assessments of learning outcomes, check to see whether the country has participated in any of the international or regional learning assessments. EdStats’ Learning Outcome Dashboard/By Country has a table of the assessments in which each country has participated. These assessments usually measure the gender composition and characteristics of the home that can serve as a proxy for socioeconomic status and sometimes other characteristics that indicate subgroup membership. Cross-national learning assessments also provide comparisons and tend to meet higher design standards and better-tested administration and data-cleaning procedures than national learning assessments. 46 Education Public Expenditure Review Guidelines • Program for International Student Assessment (PISA) • Trends in International Mathematics and Science Study (TIMSS) and Progress in International Reading Literacy Study (PIRLS) • OECD’s Programme for the International Assessment of Adult Competencies (PIAAC) and other adult literacy surveys • Early Grade Reading Assessment (EGRA): Applications and Interventions to Improve Basic Literacy and Early Grade Mathematics Assessment (EGMA): A Conceptual Framework Based on Mathematics Skills Development in Children, developed by the Research Triangle Institute (RTI) • Southern Africa Consortium for Monitoring of Education Quality (SACMEQ) • Programme d’Analyse des Systèmes Educatifs des Pays de la CONFEMEN (PASEC), or the CONFEMEN Programme for the Analysis of Education Systems (i.e., conference of education ministers for countries sharing the French language) • Laboratorio Latinoamericano de Evaluación de Calidad de la Educación (LLECE), or Latin American Laboratory for Evaluating the Quality of Education The quality of cross-national statistical databases is mixed. • OECD's annual Education at a Glance: These OECD data are of high quality because of the exceptional processes in place that produce them. • UNESCO Institute of Statistics (UIS): Although the Institute works persistently with governments to help them improve the quality of data generated by their education management information systems, it ultimately has to depend on country-specific data with all of their problems. As a result, the quality of Institute data is uneven. • World Development Indicators (WDI): These are the World Bank’s primary collection of development indicators, compiled from officially recognized international sources. They assemble the most current and accurate global development data available and include national, regional, and global estimates. Six themes are used to organize indicators: world view, people, environment, economy, states and markets, and global links. For education, the World Development Indicators cover five types of variables: education inputs (e.g., government expenditure per student, government expenditure on education as a percentage of GDP and as a percentage of total public expenditure, trained teachers, and student-teacher ratios); participation in education; education efficiency (e.g., cohort survival rates, repetition rates, transition rates); education completion and learning outcomes; and education gaps by income, gender, and area (country and region). The data sources for the first four types of variables are almost entirely from the UNESCO Institute of Statistics. As noted above, these must be used cautiously because the data depend on country- specific education management information systems (EMIS) data. The data for the fifth type (gaps by income and gender) depend heavily on demographic and health surveys, and multiple indicator cluster surveys. 47 • World Bank EdStats (Education Statistics): EdStats All Indicator Query holds more than 4,000 internationally comparable indicators that describe education access, progression, completion, literacy, teachers, population, and expenditures. The indicators cover the education cycle from pre- primary to vocational and tertiary education. Most EdStats data come from the UNESCO Institute of Statistics. EdStats also includes learning outcome data from international and regional learning assessments, equity data from household surveys, and projection and attainment data to 2050. ▷ Education Expenditures: Education-expenditure data reside in two databases on the EdStats website: (i) the EdStats Query–Education Expenditures; and (ii) the World Bank Education Expenditure Database, which has been created using data extracted from World Bank public expenditure review documents. Table TN1 summarizes the differences between the databases. Table TN1: Education-expenditure data base EdStats Query World Bank Education Education Expenditures Expenditure Database Use of Download core expenditure Download detailed expenditure Database indicators and compare data on one country. Data cannot expenditure data across countries. be used to compare countries. Number of Indicators 93 More than 800 Data Source UNESCO Institute for Statistics World Bank PER Documentsa a. The PER documents are available in http://datatopics.worldbank.org/education/wDataQuery/ExpBackground.aspx. ▷ EdStats Dashboards: The Key Education Indicators Dashboard presents a broad portrait of all levels of a selected country’s education system from pre-primary to tertiary education. It includes gender, regional, and income-group comparisons. ▷ The State of Education–Expenditure Dashboard: The State of Education’s Expenditure Dashboard presents a global view of education spending through dynamic maps, charts, and accompanying analysis. It presents not only key expenditure indicators such as government spending on education as a percentage of gross domestic product, but also the correlation between government spending and outputs such as enrollment rate and learning outcomes. 48 Education Public Expenditure Review Guidelines • National Education Accounts (NEA): This tool takes into account multiple education-financing data from different sources, including public, private, and donor funding, in a given country, and seeks to enable international comparisons of education financing (UNESCO 2015a). Drawing a complete picture of education financing in a given country does not necessarily allow international comparisons due to differences in budget classifications. To enable international comparisons, the National Education Accounts methodology (UNESCO Institute for Statistics, UNESCO International Institute for Educational Planning, and IIEP Pôle de Dakar 2016a and 2016b) has been developed on the principles of existing international standards, such as the System of National Accounts (SNA 2008), and the International Standard Classification of Education (ISCED 2011). Development of a National Education Account requires a careful matching of country and international classifications, and only a limited number of countries have implemented this tool to date. A long-term goal for the international education community is to enhance the use of this tool to enable international comparisons of education financing across countries. • World Economic Forum annual Global Competitiveness Report: This report has survey data collected from employers in each country. Respondents identify barriers to doing business, including an inadequately educated labor force, and rate the quantity and quality of a number of human-capital measures. • World Bank's Doing Business Survey: This survey is less useful for education than the World Economic Forum survey or the World Bank’s Enterprise Survey. It sometimes assesses labor market regulation. • World Bank’s Enterprise Survey: An Enterprise Survey is a firm-level survey of a representative sample of an economy’s private sector. The surveys cover a broad range of business-environment topics, including access to finance, corruption, infrastructure, crime, competition, labor, and performance measures. The labor module assesses the characteristics of the firm’s employees (e.g., their educational attainment), the firm’s labor policies (e.g., employee-training programs) and the employer’s views of the extent to which inadequately trained workers constrain their business. Since 2005–06, more than 125,000 interviews in 139 countries have taken place under the Global Methodology. Enterprise Surveys implemented in Eastern Europe and Central Asian countries, conducted jointly with the European Bank for Reconstruction and Development, are also known as Business Environment and Enterprise Performance Surveys (BEEPS). • Transparency International: This organization publishes data on perceived corruption by country. In 2013, it published an analysis of the sources of corruption by level of education, including extensive treatment of corruption in higher education. 49 Technical Note 5: Research reports Before starting a public expenditure review, be sure to scan for relevant research reports published by the World Bank, donor partners, non-governmental organizations (NGOs), or the international research community. The Bank’s education public expenditure reviews tend to underuse the international evaluation literature, such as meta-analyses of rigorous evaluations of the effects of inputs on students’ participation and learning outcomes (Glewwe et al. 2011; Snilstveit et al. 2016). These studies provide important data on the likely effectiveness of different investments and the conditions under which effectiveness occurs. As such, they are useful for public expenditure reviews’ analyses of the efficiency and effectiveness of inputs. To locate studies, use these links: • World Bank research and publications: Look for Systematic Country Diagnostics, Country Economic Memoranda, poverty assessments, earlier public expenditure reviews, Public Expenditure Tracking Surveys (PETS), Quantitative Service Delivery Surveys (QSDS), Service Delivery Indicators (SDI) (see Technical Note 12 for details on PETS, QSDS and SDI), public expenditure and financial accountability (PEFA) country reports (see Technical Note 11 for details on PEFA), and other analytical reports. • Other sources: Check the websites of international donors and non-governmental organizations active in the country in question, websites of the OECD, UNICEF and UNDP, particularly the annual Human Development Report. • External publications, especially academic research and evaluation studies: Type in JOLIS, scroll down, and choose EconLit under “Popular Resources,” then search by topic, author, or title. This source of data varies in its quality. Agencies such as the World Bank and academic journals have standards and processes, such as peer review requirements, that at least create a floor on quality. However, small donor groups, such as non-governmental organizations, may lack such processes. Studies conducted by groups such as these can vary widely in quality, depending on the individual doing the study. The only way to judge data from these studies is to read the original studies. Technical Note 6: New data collection Strongly consider collecting your own data when the topic is important and the data required to assess it are unavailable or unusable. • When new data have to be collected, design decisions by the PER team will define its quality. For example, if sampling is required, the team’s choice of its sampling frame and sampling criteria for selecting the units to be surveyed will determine if the results can be properly generalized to the universe. • When new data collection is essential, the team should, as needed, renegotiate the Bank’s budget for the expenditure review to cover the costs of new data collection or seek substantial trust funding to cover its costs. 50 Education Public Expenditure Review Guidelines • If time and budget do not permit new data collection, consider assembling a panel of experts or a focus group to give you a sense of the shape and magnitude of the issue. Methods certainly exist for increasing the validity and reliability of qualitative data collected through expert panels or focus groups. However, the intent here is to determine whether the PER should recommend new data collection if future operations or analytic work must address the issue in question. • If you can do nothing or little about important data gaps, describe them in the review and flag the need for future work. Finesse the data gaps as best you can. Technical Note 7: Definitions and notes on indicators Core Indicators Definitions / Notes Number of schools • By level, location, type • “School” is a service point (or campus that is part of a larger educational institution) that provides instructional or education- related services to a group of pupils. A school may have a single administrative unit with several service points (or group of branch schools or satellite school or campuses). An administrative unit refers to any school, or group of schools, under a single director or a single administration. A service point refers to any location that provides a service for pupils or students, whether it is a single entity or part of a larger administrative unit (UIS). Number of students • By level, gender, school location, school type in school • Total number of students in the theoretical age group for a given level of education who are enrolled in that level, expressed as a percentage of the total population in that age group (UIS). Number of students in • Number of students enrolled in college and university programs tertiary education in a given academic year, per 100,000 inhabitants (UIS). Net enrollment rate • By level; if net enrollment rate (NER) is not available, gross enrollment rate (GER) • Total number of students in the theoretical age group for a given level of education who are enrolled in that level, expressed as a percentage of the total population in that age group (UIS). Number of out-of- • By level school children • Children in the official primary school age range who are not enrolled in either primary or secondary schools (UIS). 51 Core Indicators Definitions / Notes Dropout rate • Proportion of pupils from a cohort enrolled in a given grade at by grade a given school year who are no longer enrolled in the following school year (UIS) (except for those graduating). Repetition rate • Number of repeaters in a given grade in a given school year, by grade expressed as a percentage of enrollment in that grade the previous school year (UIS). • Administrative data usually calculate the number of dropouts as those individuals who neither transition to the next grade nor are repeaters. But the literature shows that repetition is systematically underestimated, producing overestimates of dropout rates. Significant internal or external migration also poses measurement problems. Students who move are counted as dropouts from their school of origin, but this does not mean that they do not re-enroll in a school at their new destination. Additional Indicators: Private educational • Private educational institutions that are controlled and institutions managed by a non-governmental organization (e.g., a church, a trade union, a business enterprise, or a foreign or international agency), or its governing board consists mostly of members who have not been selected by a public agency (UIS). Net intake rate to Grade • New entrants to Grade 1 of primary education who are of the 1 of primary education official primary school entrance age, expressed as a percentage of the population of the same age (UIS). New entrants • Students who, during the course of the reference school or academic year, enter a program at a given level of education for the first time, irrespective of whether the students enter the program at the beginning or at an advanced stage of the program (UIS). Note: The UIS Glossary includes terms related to education, science, technology and innovation, culture, and communication and information, and can be found at http://uis.unesco.org/glossary. 52 Education Public Expenditure Review Guidelines Technical Note 8: Definition of source of funds Sources of funds include public, private, and international sources (UIS). Private entities include households, firms and business enterprises, and nonprofit organizations (including religious organizations) which, although their principal activity is non-educational, might finance activities in the domain of education (UIS). A useful source for the relative proportions of public and private expenditure on educational institutions by level of education is OECD, Education at a Glance (2016), Table B3.1a. Technical Note 9: School fees and other informal payments The burden on households as a result of fee payments can be significant. For instance, the demand response in countries that have abolished fees at the primary level (e.g., Malawi, Tanzania, Kenya, Uganda, and Cameroon) is strong evidence that tuition fees curtail demand. As Kattan and Burnett (2004) point out, in addition to tuition fees, households frequently face a wide range of user fees for publicly provided primary education, including textbook fees, compulsory uniforms, parent-teacher association (PTA) dues, and various special fees, such as exam fees and community contributions to district education boards. In many countries, private tutoring adds to the household costs of primary education. Typically, the poorer the family, the greater the burden of education spending. In Thailand, for instance, poor households spend 47 percent of their consumption on education, while the average for all households is 16 percent. The World Bank conducted surveys on user fees in 2001 and 2005, with a focus on fees in primary education.29 The 2005 survey included questions about lower secondary education, as well as primary education fees. For the primary school level, the Bank collected data for 79 countries in 2001 and 93 countries in 2005 (out of a total of 144 World Bank client countries). In 2005, it collected data at the lower secondary level for 76 countries. The 2005 survey results showed that user fees were common at the primary level. Of the 93 countries surveyed, only 16 countries had no fees. Five countries had all five types of fees (tuition, textbook, uniforms, financial contributions, and other school-based activity fees): the Dominican Republic, Haiti, Honduras, Indonesia, and Vietnam. In 59 countries, or 62 percent of those surveyed, national policy did not address the elimination of fees. More fees were collected at the lower secondary level (94 percent of surveyed countries) than at the primary level (81 percent of countries). In addition, fee levels were generally significantly higher in lower secondary than at the primary level. Of the 76 countries for which responses were received regarding secondary school fees, 94 percent reported that at least one type of fee was collected. Most countries generally had several types of fees; just 10 countries required only PTA or community contributions. Fifty-one countries had more than one type of fee in lower secondary education, and seven countries had all five types of fees for lower secondary education: Bangladesh, Bolivia, Honduras, Indonesia, Lesotho, Mozambique, and Uganda (Kattan 2006). 53 Informal payments in education are charges for education services or supplies that are meant to be provided for free or are paid “under the table” directly to public officials or teachers to obtain specific favors. These are generally measured as the fraction of survey respondents reporting that they made payments to a public education entity for education services intended to be free of charge. Household surveys and perception surveys of citizens and public officials are the most common sources of information. More detailed surveys may also include data on the average value of payments made, the recipients of the payments, and the specific services for which the payments were made. Types of informal payments include, but are not restricted to, payments for admission, advancement, preferential access to resources, and specific grades. Data on informal payments in education are increasingly being collected, but household surveys vary in whether, and how well, they measure informal payments (Lewis and Pettersson 2009). Technical Note 10: Calculating the share of a nation’s resources going to education Three options are gross domestic product (GDP), gross national product (GNP), and gross national income (GNI). They differ, sometimes noticeably, depending on the country’s economic arrangements. All three measures reflect the national output and income of an economy. The main differences are that gross national product takes into account net income receipts from abroad (gross domestic product + net property income from abroad). In other words, GNP measures the value of all goods and services produced by nationals whether in the country or not. Net income from abroad includes dividends, interest, and profit. Gross national income is based on a principle similar to gross national product; the World Bank defines GNI as “the sum of value added by all resident producers plus any product taxes (minus subsidies) not included in the valuation of output, plus net receipts of primary income (compensation of employees and property income) from abroad” (see http://data.worldbank.org/indicator/NY.GNP.ATLS.CD). The World Bank now uses GNI to classify economies into income groupings. Gross national income is possibly a better metric for the overall economic condition of countries whose economies include substantial foreign investments. However, major comparative databases for education use gross domestic product in their financial measures, such as unit costs. These include OECD’s Education at a Glance and UNESCO’s Institute of Statistics. Although gross national income might be more accurate for certain types of countries (e.g., China), losing comparability to data in the major education databases is of real concern. 54 Education Public Expenditure Review Guidelines Technical Note 11: Public expenditure and financial accountability (PEFA) country reports PEFA is a methodology for assessing public financial management performance. It identifies 94 characteristics (dimensions) across 31 key components of public financial management (indicators) in seven broad areas of activity (pillars). The program provides a framework for assessing and reporting on the strengths and weaknesses of public financial management (PFM) using quantitative indicators to measure performance. The purpose of a good PFM system is to ensure that the policies of governments are implemented as intended and that they achieve their objectives. An open and orderly public expenditure and financial accountability system is one of the enabling elements needed for desirable fiscal and budgetary outcomes. Public expenditure and financial accountability country reports provide a snapshot of a country’s performance in this area at specific points in time using a methodology that can be replicated in successive assessments, giving a summary of changes over time. Technical Note 12: Public Expenditure Tracking Surveys (PETS), Quantitative Service Delivery Surveys (QSDS), and Service Delivery Indicators (SDI) Good public expenditure management requires attention to the level of aggregate spending, allocation of public funds, and actual service delivery. The Bank’s public expenditure reviews focus on the first two issues, but the third tends to receive less attention primarily because of a lack of relevant data. Public spending data, irrespective of category, tends to be a poor proxy for actual service delivery. Lack of information on actual delivery also creates an identification problem when the efficacy of public capital or services needs to be evaluated. The case where public capital or services actually created by public funds are highly productive, but the supply system is not, cannot be distinguished from the case where the supply system is effective, but the goods and services being produced yield few benefits. PETS, QSDS, and SDI are designed to provide the missing information from different tiers of government and frontline service facilities, using the sample survey approach. The former collect data at each tier of government to create a picture of how funds and other resources are flowing and where they may be leaking.30 The latter collect a wide variety of information from schools and other sources to answer a range of questions about service delivery. Methodology As their names imply, both of these tools rely primarily on surveys and require the same technical expertise as that required by any properly conducted survey. This includes skills in constructing sampling; designing and pretesting survey instruments; training and monitoring the performance of data collectors; establishing efficient routines for cleaning the data; and analyzing the data. The QSDS, especially, also uses public accounts sample data, preferably panel data, on government spending and information on outputs of service providers at ministerial, regional, local, and service-provider levels. 55 Public Expenditure Tracking Survey A Public Expenditure Tracking Survey is useful when a resource has to travel from one source to the beneficiary. An example is the money that a secondary student earns after school to help pay his school fees, although even here parents can divert the money to buy goods and services for the household, or even to gamble or drink. Even when the source and beneficiary are close to the same, as when PTAs raise funds for particular uses by their schools, the money can be diverted through corruption or used for unintended purposes. As the chain between the initial source and the ultimate beneficiary lengthens, the chances that the resource will fail to reach the beneficiaries increase. For example, the chain from the procurement of textbooks by the central level to students in schools can be long, with substantial loss through poor storage, lack of distribution from subnational storehouses and school storerooms, or diversion for sale on the private market.31 The first Uganda education Public Expenditure Tracking Survey developed the use of surveys to track resource flows (Reinikka and Svensson 2004). The survey collected data from central ministries, local governments (districts), and schools. It tracked the delivery of capitation grants to cover schools’ non- wage expenditures. Using panel data from a unique survey of primary schools, the study assessed the extent to which the grant actually reached the intended end-user (schools). The survey data revealed that during 1991–95, the schools, on average, received only 13 percent of the grants. Most schools received nothing. The bulk of the school grants were captured by local officials. The data also showed considerable variation in grants received across schools, suggesting that some schools used their bargaining power to secure greater shares of funding, rather than being passive recipients of flows from the government. Specifically, schools in better-off communities managed to claim a higher share of their entitlements. As a result, actual education spending, in contrast to budget allocations, proved regressive. The Public Expenditure Tracking Survey for Peru’s “glass of milk,” or Vaso de Leche, program presents an excellent example of how the flow of complex resources can be traced through multiple levels (World Bank 2002). This program distributed to low-income families milk in any form, milk substitutes, or other foods such as soybean, oatmeal, quinoa, and kiwicha. Its main goal was to improve the nutritional level of infants, small children (school age or younger), and pregnant or breastfeeding mothers, and to improve the quality of life of the poorest segments of the population. The project measured leakages through surveys at each level in the transfer process from the central authority down to the household. Specifically, it traced the flow and leakage of central funds from the top of the chain to the last link at the bottom, by using survey data at the level of the municipality, at the level of the local milk-distribution committees, and, finally, at the level of the beneficiary household. The methodology is very complex, not only because it involved multilevel comparisons, but because the input itself was transformed from cash to commodities as the funds moved from the top to the bottom, and as “the commodity” itself actually became commodities, since the program was not limited to milk or milk products alone. The product was then transformed at the household level, as the food products were mixed with other foods before being served. Yet, despite this complexity, the survey was able to determine the relative magnitude of leakages at each level. 56 Education Public Expenditure Review Guidelines The tracking survey found that targeted beneficiaries received, on average, the equivalent of 29 cents of each dollar initially transferred by the Central Government. This does not mean that 71 cents from each dollar were fully lost in corruption costs. Rather, the diverted resources leaked away through a combination of factors: off-budget administrative costs; expenditure on non-eligible products; in-kind deliveries to non-beneficiaries; fees for overpriced items; and, last but not least, sheer corruption. Surprisingly, the leakages were much higher at the bottom levels (the Vaso de Leche committees and households) than at the top (central government and municipalities). The findings turn on its head the conventional belief that every local body is necessarily more accountable than the national and public authorities. In addition, the Peru case highlights the importance of good program design and of transparent and accountable local organizations. The relationship between the direct beneficiaries and the local Vaso de Leche Committees32 has two important features. First, final beneficiaries had limited information about decisions made by the committees, the amount of resources to which households were entitled, and effective ways to secure the resources. Second, the committees operated without transparency or accountability to either the beneficiaries or the upper echelons of government. In fact, the committees so dominated the running of the program at the local level that they could divert resources from their original purpose, without being held accountable or sanctioned for doing so, since neither the higher authorities nor the intended beneficiaries knew about it. The committees then distributed the resources at their own discretion and sometimes ended up diluting, even unwittingly, the program’s expected effects. These findings highlight the design flaws of the program. The local committees were not accountable to the beneficiaries, and they were frequently dominated by rent-seeking “representatives” of the program beneficiaries. Upon reaching the households, the resources sometimes underwent additional dilution. On average, target beneficiaries only received 41 percent of the ration that arrived at the household (not taking into account all of the losses associated with earlier leakages). This dilution effect was possible because, in most cases, the beneficiaries did not receive their rations directly from the committee; rather, the children received the rations through their mothers (or, in some cases, their fathers), who picked up the total rations allocated to the household. Consistent with evidence from studies of other nutritional assistance programs worldwide, the official distribution criteria are very difficult, if not impossible, to enforce. In most cases, it is impossible to exclude non-targeted members of the household from utilizing the resources. Furthermore, in about 60 percent of the committees visited, the products were distributed in unprepared forms. These unprepared products are frequently mixed into dishes that feed the whole family. 57 Quantitative Service Delivery Surveys A Quantitative Service Delivery Survey examines the efficiency of public spending and incentives, and various dimensions of service delivery in provider organizations, especially at the level of the service facility. It measures the resources delivered, such as the actual hours of instruction per day that teachers provide, the student-textbook ratio at the classroom level, the receipt of fiscal transfers from national and subnational levels of government, the size and use of the school’s own revenues, and so forth. It also quantifies the factors affecting the quality of service, such as incentives, accountability mechanisms, and the relationship between agents and principals. Typically, the facility or frontline service provider is the main unit of analysis, much in the same way that the firm is the unit of observation in enterprise surveys and the household in household surveys. In each case, the surveyor collects quantitative data both through interviews and directly from the service provider’s records. Facility data can be “triangulated” by also surveying local governments, umbrella non-governmental organizations, and private provider associations. The compilation of facility-level quantitative data typically requires much more effort than, say, a perception survey of service users, which makes this type of survey more costly and time- consuming to implement than its qualitative alternatives. The Zambia Public Expenditure Tracking Survey and Quantitative Service Delivery Survey illustrates the types of data (Table TN2) and the range of survey modules (Table TN3) used to produce a Quantitative Service Delivery Survey. Table TN2. Data Source Data Description PETS-QSDS 2014 See sub-section PETS-QSDS 2014 NAS 2014 National assessment of Grade 5 students and teachers LCMS 2010 and 2015, GER, NER, and out-of-school children and  ZDHS 2013-14 ESB 2013 Enrollment and school numbers, repetition and dropout rates, teacher statistics Yellow and Bluebooks Government financial statement C and budget Interviews and Meetings GoZ officials and CPs Source: World Bank 2015, table 2.1, p.16. Note: PETS = Public Expenditure Tracking Survey; QSDS = Quantitative Service Delivery Survey; NAS = National Assessment Survey; LCMS = Living Conditions Measurement Survey; ZDHS = Zambia Demographics and Health Survey; GER = gross enrollment rate; NER = net enrollment rate; ESB = Education Statistical Bulletin; GOZ = Government of Zambia ; CPs = Cooperating Partners. 58 Education Public Expenditure Review Guidelines Table TN3: Survey modules, contents, and respondents Data Primary Respondent/Source Description Teacher List Supervisor based on registry Listing of all teachers (for grades 2, 5, 7, 9 (teacher register) and 11) and their information. Student and Supervisor A list sample students and teachers, and Teacher Selection mapping of students and teachers Module Teacher Head teacher based on Basic teacher information and teacher Attendance I registry (attendance book) attendance for all teachers listed (First Visit) General School, Head teacher based on registry General school information, location, Part A (student and teacher registers facilities, enrollment, and repetition and attendance books) General School, Head teacher/financial School financing, fund flow, expenditure, Part B administrator based on and decision-making registry (accounting books) Head Teacher, Head teacher Head teacher information Part A Head Teacher, Head teacher Head teacher personality and motivation Part B Teacher, Sample teacher Detailed teacher information and Part A (up to three teachers) characteristics Teacher, Sample teacher Teacher personality and motivation Part B (up to three teachers) Student, Sample students Detailed student information and Part A (up to 20 students) characteristics for selected students Student, Sample students Student personality and motivation Part B (up to 20 students) (Table continued on next page) 59 Data Primary Respondent/Source Description Household Parents of sample students Household demographics, education, and economics status Classroom Observer Observation of classroom of Grade 5 Observation sample teachers Teacher Observer Second unannounced school visit to check Attendance II attendance of 10 random, sample teacher (Second Visit) PEO Supervisor PEO office information and PETS DEBS Supervisor DEBS office information and PETS Grade 9 Student Sample students selected for Grade 9 student assessment Assessment interview module 7 Grade 9 Teacher Sample teachers selected for Grade 9 teacher assessment Assessment interview module 6 Source: World Bank 2015, table 2.2, p.16. Note: DEBS = District Education Board Secretaries; PEO= Provincial Education Office. Another good example of a Public Expenditure Tracking Survey and Quantitative Service Delivery Survey is Philippines: Assessing Basic Education Service Delivery in the Philippines: Public Education Expenditure Tracking and Quantitative Service Delivery Study. Service Delivery Indicators (SDI) These data are used to assess the quality and performance of education (and health) services over time. Decision makers can use these data to track progress, and citizens can use them to hold governments accountable for public spending. The indicators are broken down into three categories: (i) provider competence and knowledge; (ii) proxies for effort; and (iii) availability of key infrastructure and inputs. The indicators are quantitative and ordinal in nature (to allow cross-country and country specific comparisons).33 60 Education Public Expenditure Review Guidelines Technical Note 13: Definitions and notes on monetary-input indicators Core Indicators Definitions / Notes Public expenditure OECD, Education at a Glance (2016), Table B4.1 on education Public expenditure on education covers direct public expenditure on educational institutions, as well as public support to households (e.g., scholarships and loans to students for tuition fees and student living costs) and to other private entities for education (e.g., subsidies to companies or labor organizations that operate apprenticeship programs). It includes expenditure by all public entities, including ministries other than ministries of education; local and regional governments; and other public agencies (OECD 2016, 228–9) Public expenditure UIS (EdStats) on education as a OECD, Education at a Glance (2016), Table B4.2 percentage of total public expenditure and As defined above, “public expenditure on education” includes as a percentage of GDP, public subsidies to households for living costs, which are not spent by level of education in educational institutions. Therefore, the figures presented here exceed those on public spending on education institutions found in Table B2.3 below (OECD 2016, 231) Government (i) Average total general government expenditure (current, capital, expenditure per student and transfers) per student in the given level of education, expressed (i) as a percentage of per as a percentage of per capita GDP (UIS) capita GDP; (ii) in U.S. http://data.worldbank.org/indicator/SE.XPD.PRIM.PC.ZS (primary); dollars; and (iii) in local http://data.worldbank.org/indicator/SE.XPD.SECO.PC.ZS?end=201 currency, by level of 1&start=2011&view=chart (secondary); education http://data.worldbank.org/indicator/SE.XPD.TERT.PC.ZS? end=2011&start=2011&view=chart (tertiary) (ii) Average total general government expenditure (current, capital, and transfers) per student in the given level of education, expressed in nominal U.S. dollars at market exchange rates (UIS). Wils (2015) and UNESCO (2015d) project per student spending by country (iii) Average total general government expenditure (current, capital, and transfers) per student in the given level of education, expressed in local currency (UIS) (Table continued on next page) 61 Core Indicators Definitions / Notes Expenditure on UIS (EdStats) educational institutions OECD, Education at a Glance (2016), Table B2.3. as a percentage of GDP, by source of funding (public and private), and by level of education Annual expenditure per OECD, Education at a Glance (2016), Table B1.1. student by educational institutions for all In equivalent U.S. dollars, converted using purchasing power parity services, by level (PPP) for GDP, by level of education, based on full-time equivalents of education Note that this OECD indicator includes private spending on tuition fees that occur at educational institutions. In reality, it might be difficult to include private spending at educational institutions when calculating per student spending Per student spending can be broken down by subnational unit if the budget data and education statistics are available at subnational level. If school-level budget data and education statistics are available, and schools’ locations (typically, urban or rural) can be identified, per student spending by geographical locations can be computed. If schools may be categorized by language of instruction, or religion, breakdown by those categories can be done Annual expenditure per OECD, Education at a Glance (2016), Table B1.4. student by educational institutions for all services, In percentage of per capita GDP relative to per capita GDP, by level of education Relative expenditure per Per student spending can be expressed relatively between levels of student (unit cost) by education. The expenditure for primary is treated as the "base case". level of education The expenditure for each other level/type of education is expressed as a percent of the base case. E.g., primary 1.0; secondary 1.2 (20 percent higher than primary); tertiary 1.5 (50 percent higher than primary). This indicator can show the relative cost of different levels of education easily Public expenditure on OECD, Education at a Glance (2016), education by economic Table B6.1. Share of current and capital expenditure by education level classification Table B6.2. Share of current expenditure by resource category (compensation of teachers, compensation of other staff, and other current expenditure) (Table continued on next page) 62 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Capital expenditure: expenditure for education goods or assets that yield benefits for a period of more than one year. It includes expenditure for construction, renovation and major repairs of buildings, and the purchase of heavy equipment or vehicles. It represents the value of assets acquired or created–i.e., the amount of capital formation–during the year in which the expenditure occurs (UIS). Current expenditure: expenditure for educational goods and services consumed within the current year and which would have to be renewed if needed in the following year (UIS). Ideally, salary data should include all “staff compensation” as defined below: expenditure on teaching staff and non-teaching staff salaries; contributions by employers or public authorities for staff retirement and pension programs, and social insurance; and other allowances and benefits (UIS). Current expenditure other than for staff compensation: expenditure on school books and teaching materials, ancillary services, and administration and other activities (UIS). Teachers or teaching staff: persons employed full-time or part-time in an official capacity to guide and direct the learning experience of pupils and students, irrespective of their qualifications or the delivery mechanism (i.e., face-to-face or distance learning). This definition excludes educational personnel who have no active teaching duties (e.g., headmasters, headmistresses, or principals who do not teach), or who work occasionally or in a voluntary capacity in educational institutions (UIS). Non-teaching staff: persons employed by educational institutions who have no instructional responsibilities. Although the definition can vary from one country to another, non-teaching staff generally include head teachers, principals and other administrators of schools, supervisors, counselors, school psychologists, school health personnel, librarians or educational-media specialists, curriculum developers, inspectors, education administrators at the local, regional, and national level, clerical personnel, building operations and maintenance staff, security personnel, transportation workers, and catering staff (UIS). (Table continued on next page) 63 Core Indicators Definitions / Notes Salary cost of teachers OECD, Education at a Glance (2016), Table B7.1. per student by level of education Teachers’ compensation usually consists of the largest part of education spending, and thus, spending per student. The salary cost of teachers per student is a function of (i) the instruction time of students, (ii) the teaching time of teachers, (iii) teachers’ salaries, and (iv) the number of teachers needed to teach students, which depends on estimated class size (OECD 2015). Differences among countries in these four factors may explain, to a large extent, differences in spending per student. Conversely, a similar level of spending per student may be associated with different combinations of these factors. In other words, governments may be able to improve efficiency by changing combinations of these factors. For a detailed definition of this indicator, see “Box B7.1. Relationship between salary cost of teachers per student and instruction time of students, teaching time of teachers, teachers’ salaries and class size” (OECD 2016, 264). Teacher salary relative OECD, Education at a Glance (2016), Table B7.2. to public sector wages, per capita GDP, and This indicator helps judge the adequacy of teacher salary level those with similar relative to the country’s economic capacity and context. Data from qualifications a labor force or household survey can be used to compare wages among public servants and between similarly qualified individuals in the private and public sectors. It is also important to understand how teacher salaries are set and paid. Teachers’ salary cost per student as a percentage of per capita GDP, by level of education. Teachers’ statutory OECD, Education at a Glance (2016), Table D3.1. salaries, based on typical qualifications, Annual teachers’ salaries, in equivalent U.S. dollars, converted using PPPs at different points in for private consumption. teachers’ careers Teachers' salaries are expressed as statutory salaries, which are scheduled salaries according to official pay scales. The salaries reported are defined as gross salaries (total sum of money that is paid by the employer for the labor supplied) minus the employers' contribution to social security and pension (according to existing salary scales) (OECD). Salaries at starting; after 10 years of experience; after 15 years of experience; at top of scale, by level of education. (Table continued on next page) 64 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Teachers’ actual salaries OECD, Education at a Glance (2016), Table D3.2a. relative to wages of tertiary- educated workers Ratio of salary, using annual average salaries (including bonuses and allowances) of teachers in public institutions relative to the wages of workers with similar educational attainment (weighted average,) and relative to the wages of full-time, full-year workers with tertiary education. Comparison of teachers’ OECD, Education at a Glance (2016), Table D3.3a. statutory salaries, based on typical qualifications Ratio of salaries at different points in teachers’ careers, and salary per hour in U.S. dollars converted using PPPs for private consumption. Ratio of salary at top of scale to starting salary. Salary per hour of net contact (teaching) time after 15 years of experience (in U.S. dollars). Technical Note 14: Calculating unit costs from aggregate and itemized spending Mingat, Tan, and Sosale (2003) identify two basic ways to compute unit costs. Both should yield consistent results. In one approach, unit costs are calculated by dividing aggregate spending, such as that reported in budget documents, by the number of students. This is easy to implement, but the method can have problems. First, the aggregate data may be organized under rubrics that prevent clear-cut attribution of spending by level or type of education. For example, administrative expenditures may appear as one entry, with no distinction by level of education. A second problem is that the aggregate data may be organized according to sources of funds or to the structure of the government bureaucracy. As a result, expenditure for a given level of education may appear in several places in the budget, possibly without any detail regarding functional categories. For example, in some countries, the budget documents show spending supported by external donors separately, even though for analytical purposes, the expenditure may belong in the same category as the government’s own spending. In addition, the data may not distinguish between capital investments and recurrent spending, making it difficult to compute meaningful indicators of costs. Given the potential shortcomings and incompleteness of the foregoing approach, it is useful to check the estimates against those obtained through another approach, namely, by building up from the constituent parts of costs. In this approach, the cost components are identified, evaluated, and then aggregated to obtain the desired estimates. In primary education, for example, teachers and pedagogical materials are two of the main components of costs. Thus, the unit cost of these components is calculated separately and then added together to obtain an estimate of the overall unit cost of primary education. Furthermore, instead of dividing aggregate spending on each component by the number of students, other data can be used to make the estimates. For example, to obtain the unit cost of teacher inputs, we would use data on teacher salaries and pupil-teacher ratios. This approach yields more detailed analysis of education costs and provides a basis for simulating the cost implications of alternative choices in the delivery of education services. 65 Technical Note 15: Calculating unit costs when country's fiscal year and school year do not coincide Although countries can differ in their definition of their fiscal year, it usually runs January 1 to December 31. The school year is almost always split across fiscal years. The unit cost can be calculated for a school year or for a fiscal year. However, the school year is usually the desired estimate. Take this example: The fiscal year runs from January 1 to December 31. The school year is from September to June (where teachers are paid for 10 months), and September to August (where teachers are paid for 12 months). Assume that in Fiscal 2001, total expenditures in primary education are $100; in Fiscal 2002, $110. In School Year 2001/02, the total number of students enrolled in primary education is 15 students. Figure 1 shows how the school year splits across the fiscal years. Figure TN1: Unit costs for one school year and two fiscal years MONTH 1 2 3 4 5 6 7 8 9 10 11 12 FY 2001 expenditures for primary education: US$100 School year 2001/02: primary FY 2002 expenditures for primary education: US$110 Enrollment: 15 students The unit costs for School Year 2001/02 are: - 0.4 of US$100 + 0.6 of US$110)/15 students = US$7.07/per capita (where teachers are paid for 10 months a year) - 0.4 of US$100 + 0.8 of US$110)/15 students = US$8.53/per capita (where teachers are paid for 12 months a year) Technical Note 16: Definitions and notes on non-monetary input indicators Core Indicators Definitions / Notes Human resources inputs Pupil/student-teacher Average number of pupils per teacher at a given level of education, ratio (PTR or STR) by based on headcounts of both pupils and teachers (UIS) level, geographical location, and group Governments may have a policy defining a minimum or maximum of schools student-teacher ratio. If this ratio continues declining, the government may need to require or incentivize subnational governments or schools to reduce the number of teachers to maintain the target size (Table continued on next page) 66 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Percentage of qualified A qualified teacher is one who has the minimum academic teachers qualifications necessary to teach at a specific level of education in a given country. This is usually related to the subject(s) they teach (UIS). Ratio of teachers to non- Whereas PERs tend to focus on efficiency of distribution of teachers, teaching staff that of non-teaching staff can be also important in countries where there is a generous supply of non-teaching staff.  Non-teaching staff per Some countries define norms for non-teaching staff per school size type of school (number of students), school area, type of school, etc. Those norms may be very generous, indicating possible room for efficiency gains. Organization of teachers’ OECD, Education at a Glance (2016), Table D4.1 working time   The proportion of statutory working time spent teaching provides information on the amount of time available for non-teaching activities, such as lesson preparation, correction, in-service training, and staff meetings. A large proportion of statutory working time spent teaching may indicate that less time is devoted to tasks such as assessing students and preparing lessons. It also could indicate that teachers have to perform these tasks on their own time and, hence, to work more hours than required by statutory working time. Actual teaching time is the annual average number of hours that full-time teachers teach a group or class of students, including all extra hours, such as overtime. Statutory teaching time is defined as the scheduled number of 60-minute hours per year that a full-time teacher teaches a group or class of students, as set by policy, teachers’ contracts of employment, or other official documents. Teaching time can be defined on a weekly or annual basis. Annual teaching time is normally calculated as the number of teaching days per year, multiplied by the number of hours a teacher teaches per day (excluding preparation time and periods of time formally allowed for breaks between lessons or groups of lessons). At the primary level, short breaks between lessons are included if the classroom teacher is responsible for the class during these breaks. (Table continued on next page) 67 Core Indicators Definitions / Notes Classroom inputs Average class size The average class size refers to the number of enrolled students divided by the number of classes for the whole country. To ensure comparability among countries, special-needs programs are excluded. Data include only regular programs at primary and lower secondary levels of education and exclude teaching in subgroups outside the regular classroom setting (UIS). Governments may have a policy defining a minimum and maximum for average class size. If this ratio continues declining, the government may need to require or incentivize subnational governments or schools to consolidate schools and classes to maintain the target size. Student-textbook ratio This indicator may be important in countries where textbook supply is an issue. Percentage of schools Minimum standards for educational inputs may include standards on meeting minimum prescribed textbooks, and access to computers for learning purposes.  standards requirements for  educational inputs Annual instructional This consists of the required number of hours of instruction per time year. It is not the same as hours in school because those hours may include lunch and inter-class breaks. What is the required number of instructional hours per week for core subjects by grade? For vocational and educational training, what are theoretical and practical instruction hours?  Infrastructure inputs Average school size This indicator may suggest a possible need to rationalize the school network in some parts of the country. It is often observed in countries where the school-age population is decreasing, or where rapid urbanization is taking place so that rural schools are losing students. Factors that influence a decision on whether to build a school include standards for maximum distance that primary- and secondary-level students should walk to school, and for minimum population in the catchment area required to establish a school. (Table continued on next page) 68 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Unit cost of building The unit cost is primarily determined by construction materials, a classroom school design, and equipment. Vocational education and training programs require occupation-specific workshops and equipment. Standards and schedule for It is essential that the education budget include facility maintenance infrastructure maintenance and is fully disbursed for maintenance purposes. Percentage of schools Multiple-shift schools may be prevalent where there are not enough that run double shifts, school buildings to offer single-shift schools. It is an efficient way triple shifts of using existing infrastructure, but not ideal, particularly if a school runs more than two shifts. Percentage of schools Multi-grade schools may be used to manage small student that use multi-grade populations in rural areas. It is efficient to combine grades, but classrooms teacher training on multi-grade teaching is essential to provide good-quality teaching. Without proper training, such an approach may lead to poor-quality education and outcomes. Percentage of schools Minimum standards for learning conditions may include standards on meeting minimum potable water, functional hygienic facilities, electricity, and libraries. standards requirements for learning conditions Technical Note 17: Definitions and notes on research indicators Core Indicators Definitions / Notes Doctorate productivity The number of doctorate degrees, relative to the number of full-time equivalent (FTE) academic staff. Research publications The number of research publications attributed to the department (absolute numbers) and that are indexed in the Web of Science Core Collection database, where at least one author is affiliated with the source university or higher-education institution. Citation rate The average number of times the department’s research publications from the period 2010–13 are cited in other research (published in 2010–15), adjusted (normalized) at the global level for the field of science, and the year in which a publication appeared. (Table continued on next page) 69 Core Indicators Definitions / Notes Top-cited papers The proportion of the department’s research publications that, compared to other publications in the same field and in the same year, belong to the top 10 percent of most frequently cited papers. Interdisciplinary Percentage of department’s research publications that belong to the field’s publications top 10 percent of publications with the highest interdisciplinary scores. Research orientation The degree to which research in the field informs the education offered of teaching by the institution (based on a survey of students in the program). Postdoctoral positions The number of postdoctoral positions relative to the number of academic staff who work full-time or equivalent. External research income Revenue for research that is not part of a core (or base) grant received from the government. Includes research grants from national and international funding agencies, research councils, research foundations, charities, and other nonprofit organizations. Measured in €1,000s, using Purchasing Power Parities (PPP). Expressed per full-time equivalent academic staff. Research publications The number of research publications indexed in the Web of Science (size-normalized) database, where at least one author is affiliated with the university, relative to the number of students. Publication output Number of all research publications included in the institution’s publications databases, where at least one author is affiliated with the institution (per full-time equivalent academic staff) Art-related output The number of scholarly outputs in the creative and performing arts, relative to the full-time equivalent number of academic staff. Citation rate The average number of times the university’s research publications are cited in other research (during 2011–14); adjusted (normalized) at the global level to take into account differences in publication years and to allow for differences in citation customs across academic fields. (Table continued on next page) 70 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Interdisciplinary Extent to which reference lists of the university’s publications reflect publications cited publications in journals from different scientific disciplines. Research publications The number of university research publications (indexed in the Web (absolute numbers) of Science Core Collections database), where at least one author is affiliated with the source university or higher-education institution. Technical Note 18: Efficiency of the curriculum Focused on a few, versus many, subjects. A curriculum fragmented across multiple subjects increases the number of different subject-matter specialists required. In this situation, teachers are less apt to be deployed efficiently, either because they are assigned classes out of their specialty (competence mismatches), or because specialty teachers teach fewer than the usual hours. A fragmented curriculum also increases the textbook varieties required, and disperses students’ learning instead of focusing it on a few core subjects. Focused on a limited, versus large, number of topics in each subject. A curriculum can cover a large number of topics in each subject, but inevitably, superficially. This approach usually necessitates revisiting the topic over several grades. Or a curriculum can address only a few topics per subject, but in depth, and then leave the topic. Studies show that the latter approach is much more efficient in terms of the use of instructional time and student's learning. Technical Note 19: Demographic trends and enrollment projections A country’s demographic structure and trends significantly affect the sustainability of education spending. When examining the demographics data, it is important to understand what trends in the school-age population may imply about the inputs required now and in future. If the population is trending down, do trend data on the size of the teaching force and number of classrooms in use indicate that the government is downsizing inputs into the system—for example, reducing the number of teachers, or closing schools or classrooms? Population data by single-year or five-year age groups—preferably for male, female, and total between 0 and 29 years of age, for the last five years and as projected for the next decade—will be useful. Population data is usually available in the government’s population census statistics book (typically, General Statistics Office), or UN World Population Prospects. EdStats also has this kind of population data in the Core Indicator Query. 71 Technical Note 20: UNESCO’s Education Policy and Strategy Simulation Model (EPSSim) EPSSim is a sector-wide and goals-based generic model comprising the common features of all modern school systems, together with a set of optional components to be included or excluded as required to achieve a first approximation of a country’s education system development. It is driven by demographic trends; enrollment targets are taken as a priori and the simulation calculates the corresponding financial-resource implications. At the national level, EPSSim can accompany countries through all stages of the strategic planning and management cycle. The model was conceived with a view to providing technical and methodological support to national administrations and specialists in education ministries in their efforts to formulate credible education development plans and programs, including in the context of the Education for All (EFA) goals. The aim of the model is to provide a potentially self-contained package (including self-training features) that can be deployed by country planners without external support and can be adapted within a typical range of structural alternatives with minimal expertise. For instance, the model contains some built-in training modules which provide exercises on the key indicators used in the model and enable users to conduct a simplified simulation using hypothetical data. EPSSim starts by computing the projected intake, enrollment, and flow rates on the basis of population data, enrollment status, and policy objectives. The number of enrollments by level and type of education, combined with the current and future modalities of resource utilization (teaching staff, equipment, infrastructure, etc.), enable the estimation of future requirements for teachers, non- teaching staff, instructional materials, educational facilities, and so forth. These projected requirements, together with cost-related data and hypotheses, provide information on financial requirements and the possible financing gaps associated with certain education policy goals. The EPSSim software can be downloaded from UNESCO’s Inter-Agency Network on Education Simulation Models (INESM) website. For EPSSim file downloads, see: http://archive.is/U6qrO. For user guidance and more information on the model, see: http://unesdoc.unesco.org/images/0022/002201/220198E.pdf. Technical Note 21: Simulating the economy-wide effects of alternative education scenarios Appropriately designed Computable General Equilibrium (CGE) models may be the best tool for assessing the broad effects of alternative education scenarios. If such a model is used, it is essential that the Computable General Equilibrium analysis and the rest of the public expenditure review be closely integrated. One such model is MAMS (Maquette for MDG Simulations), developed at the World Bank to assess strategies for achieving the 2015 Millennium Development Goals (MDG) and currently being broadened for the Sustainable Development Goals (SDGs).34 In most applications, MAMS divides 72 Education Public Expenditure Review Guidelines education into three levels (primary, secondary, and tertiary). It considers the impact of private and government educational services on educational attainment and views financing of the government budget in the context of the projected growth in domestic tax revenues, foreign aid, and non- education demands on government resources (which sometimes may be reduced thanks to improved government efficiency). The level and distribution of labor market entrants across different educational levels influence production in different sectors, wages, trade, and gross domestic product. Such a detailed, economy-wide assessment of educational policies may have particularly high payoffs in low-income countries that are expected to see sharp increases in enrollment at different levels. In those countries, while government spending on education may increase to meet growing demand for teachers and other education workers, the actual supply of such workers may increase with a time lag, potentially resulting in supply-and-demand mismatches. Depending on the country context and data availability, it may alternatively be better to use a more macro-oriented model that needs only data that are available for virtually any country. Such a model, labeled GEM-Education (the General Equilibrium Model for Education), is currently being piloted. The model splits the economy into two sectors (private and public). Then, it assesses the impact of alternative scenarios for enrollment and government-education services—which are translated into changes in government consumption and investment spending—on a set of key economic indicators. These include: household consumption and other final demands, gross domestic product, poverty, and the government budget. The channels through which different education scenarios impact the economy are: (i) labor productivity (mainly by influencing the educational composition of the labor force); and (ii) government spending, which requires adjustments in financing from some combination of domestic and foreign sources (taxes, foreign grants, and foreign and domestic borrowing) (World Bank 2013). 73 Technical Note 22: Definitions and notes on output and outcome indicators Core Indicators Definitions / Notes Gross enrollment rate Number of students enrolled in a given level of education, regardless of age, expressed as a percentage of the official school- age population corresponding to the same level of education. For the tertiary level, the population used is the five-year age group starting from the official secondary school graduation age (UIS). Net enrollment rate Total number of students in the theoretical age group for a given level of education enrolled in that level, expressed as a percentage of the total population in that age group (UIS). Dropout rate by grade Proportion of pupils from a cohort enrolled in a given grade at a given school year who are no longer enrolled in the following school year, except for those graduating (UIS). Repetition rate by grade Number of repeaters in a given grade in a given school year, expressed as a percentage of enrollment in that grade the previous school year (UIS). Completion of an Participation in all components of an educational program educational program (including final exams, if any), irrespective of the result of any potential assessment of achievement of learning objectives (UIS). Education attainment The highest International Standard Classification of Education rate (ISCED) level of education an individual has successfully completed. This is usually measured with respect to the highest educational program successfully completed which is typically certified by a recognized qualification. Recognized intermediate qualifications are classified at a lower level than the program itself (UIS). Attendance rate In contrast to the enrollment rate, this indicator sheds light on the frequency with which students attend school or education programs. Attendance-rate data are collected in household surveys such as Demographic and Health Surveys (DHS), Multiple Indicator Cluster Surveys (MICS), and Living Standards Measurement Study (LSMS). (Table continued on next page) 74 Education Public Expenditure Review Guidelines Core Indicators Definitions / Notes Labor force participation Employment and wage data for recent graduates are often used to shed light on this question. But be aware that these data reflect Employment and supply-and-demand interactions, not necessarily the adequacy of unemployment rates the skills developed by the educational system. External research Research revenue that is not part of a core (or base) grant income received from the government. It includes research grants from national and international funding agencies, research councils, research foundations, charities, and other nonprofit organizations. Measured in €1,000s using Purchasing Power Parities (PPP). Expressed per full-time equivalent academic staff. Technical Note 23: Concepts of effectiveness and efficiency Concept Definition Effectiveness  When something is deemed effective, it has met an intended, desired, or expected outcome. Unlike the concept of efficiency, effectiveness alone is determined without reference to costs. In popular discourse, effectiveness means “doing the right thing,” and efficiency means “doing the thing right.” When effectiveness is combined with cost, “cost-effectiveness” relates monetary inputs and non-monetary outputs and outcomes. Technical and Technical and productive efficiency are two closely related concepts that are productive often lumped together under the term technical efficiency. In both cases, we efficiency are looking for the best outcomes at least cost, a result that may be achieved in either of two ways: (i) by reducing the inputs used by a given intervention to the minimum required to achieve the outcome sought (technical efficiency), or (ii) by selecting a different intervention—i.e., one with a different combination of resources—that achieves at less cost the outcome sought as well as, or better than, the alternatives (productive efficiency). Technical efficiency refers to the physical relation between a set of resources (capital and labor) and an education outcome. A technically efficient position is achieved when the maximum possible improvement in the outcome is obtained from the resource inputs. An intervention is technically inefficient if the same (or greater) outcome could be produced with less of one type of input. For example, if research shows that a student-textbook ratio of 2/1 results in the same learning gains as a ratio of 1/1, the higher ratio is technically more efficient. Technical efficiency cannot directly compare alternative interventions. Productive efficiency, closely related to the concept of technical efficiency, directly compares alternative interventions. It asks about the cost benefit or cost-effectiveness ratios of resource combination A versus alternative resource combinations. For example, if the objective is to improve teachers’ classroom (Table continued on next page) 75 Concept Definition performance, educators can try alternative training regimes. Teachers can attend a week of training at a teacher college. They can participate with their same-grade or same-subject colleagues in weekly, facilitated training sessions at their schools. They can take online courses at home beamed via the government television station. What does each option cost, and how well does it improve teachers’ classroom performances? Since different combinations of inputs are being used, the choice between interventions is based on the relative costs of these different inputs relative to the same outcome (better teacher classroom performance). Productive efficiency enables assessment of the relative value for money of interventions with directly comparable outcomes. It cannot address the impact of reallocating resources at a broader level—for example, from teacher training to infrastructure—because the education outcomes are incommensurate. Allocative Allocative efficiency takes account of the productive efficiency with which efficiency education resources are used to produce education outcomes and how these outcomes are distributed among the community. Such a societal perspective is rooted in welfare economics. The World Bank defines the concept of allocative efficiency as the capacity of government to distribute resources on the basis of how well its public programs meet its strategic objectives (Schick 1998; Shand 2000). This includes the ability to shift resources from old priorities to new ones, and from less- to more-effective programs. Allocative efficiency thus requires that governments establish and prioritize objectives, and that they assess the actual or expected contribution of public expenditures to those objectives. Shand’s (2000) summary of the basic principles of allocative efficiency states that expenditures should be affordable in the medium term and be based on government priorities and the effectiveness of public programs. Internal Internal efficiency measures the percentage of children who complete an efficiency educational cycle (e.g., primary education or lower secondary education) as a share of those who start the cycle or as a percentage of those who finish the cycle in the minimum number of years. The first definition allows the calculation of the dropout rate—i.e., the number of those who start, minus the number who ultimately complete as a share of those who start. The second definition measures the dropout rate plus the repetition rate. Dropping out imposes costs on individuals and countries in the form of unrealized human capital. Repetition imposes costs on the sector in the form of its having to pay double (or triple) the unit cost of a year of school per repeater. External External efficiency measures the returns to individuals, employers, and the efficiency country of public and private investments in education. It depends on a match between the type and quality of skills and knowledge that school leavers acquire in school relative to the skills and knowledge needed by the country and needed and paid for by employers. Does education improve the employability and wages of school leavers? Does public investment in education and training contribute to the country’s growth and economic development? Measuring and linking employment and wage returns to education is particularly important for vocational education and training, and for tertiary education. 76 Education Public Expenditure Review Guidelines Technical Note 24: Cost-effectiveness analysis Cost-effectiveness relates monetary inputs and non-monetary outputs and outcomes. An increasing number of impact evaluations of various interventions have assessed their cost-effectiveness relative to student learning and other outcomes. For instance, to provide comparable cost-effectiveness estimates across different policy options, the Abdul Latif Jameel Poverty Action Lab (J-PAL) has adopted a standard methodology for conducting cost-effectiveness analysis of randomized trials of 29 programs.35 The trials found that the effectiveness and costs of numerous strategies to improve student learning vary considerably. Some programs achieve learning gains with much greater cost- effectiveness than others. Be alert to the quality of cost-effectiveness analyses. The risks of correlations being interpreted as causal relationships rise substantially when the methodology of the study does not use a randomized controlled trial, a difference in differences (DD) regression design, a regression discontinuity design (RDD), or matching methods. For example, teachers with formal degrees may increase their students’ learning more than teachers without such qualifications. However, teachers who obtain formal degrees may differ in important ways from those who do not, and these differences may account more for their students’ learning gains than the teachers’ formal degrees. Most research studies in the development literature do not protect well against bias and cannot be properly used to draw causal inferences (X causes Y). At the same time, if these are the only studies available, flag the potential for bias, triangulate where possible, and be very careful to state conclusions in correlational, not causal, terms. For critical policy questions, suggest that a small, randomized trial be conducted to sort out correlation from causation. Careful meta-analyses36 of the effects of inputs on participation in school and learning find numerous instances where the “common wisdom” about the effects of inputs is either wrong or fails to take into account the conditions that have to be in place for those effects to occur. Take the example of textbooks. The 3ie study reported a relatively consistent pattern of textbooks having no effect on learning outcomes, as measured by math, language, and composite test scores (Snilstveit 2016). However, the study found that many of these programs experienced implementation challenges—e.g., the books did not reach the students because they were locked up for “safekeeping." For their methodologically less-rigorous sample of 79 studies, Glewwe et al. (2011) found that most studies showed positive effects, and that most of these effects were statistically significant. This evidence strongly suggests that textbooks and similar materials (workbooks, exercise books) increased student learning. However, when the analysis was restricted to the methodologically more-rigorous sample of 43 studies, the estimated effects of textbooks were far from unanimous. Slightly less than half of the estimates showed positive effects, and only three of these were significantly positive (and one was significantly negative). Thus, after dropping less-rigorous studies, the evidence that textbooks and similar materials (workbooks, exercise books) increased student learning was quite weak. 77 The one “gold standard” randomized, controlled trial in the Glewwe et al. (2011) sample found no impact of providing textbooks. However, the authors of that study conducted further analyses to determine why textbooks did not raise scores. The results of the study did not appear to be statistical artifacts. The treatment and comparison schools were similar in geographic location, enrollment, and pre-program test scores. Neither selection nor attrition bias appeared to drive the results. They found that the effectiveness of textbooks depended on prior conditions: textbooks improved the scores of students with higher pre-test scores. In other words, there was an interaction effect between pre-test scores and assignment to the textbook program that had a highly significant, positive correlation with post-test scores. Initially better students could benefit from the textbooks much more than initially weak students. The authors also found that the official Kenyan government textbooks were of limited use to many students. English is the medium of instruction in Kenyan schools, but it was the third language for many pupils, including those examined in this study. The study showed that many students could not read the textbooks. Technical Note 25: Cost-benefit analysis Cost-benefit analysis, or CBA, is used when both the costs and outcomes can be monetized.37 Rate-of-return analysis is the type of cost-benefit analysis most frequently applied to education. Rates of return to investments in schooling have been estimated since the late 1950s. George Psacharopoulos’s (1973) findings on the rates of return to education—that primary education ought to be the main focus of national school systems, since its rate of return was found to be the highest among all education levels—continues to play an important role in the formation of significant global educational policies. In the intervening 40-plus years, however, the economies of many developing countries have matured, changing the relative returns to different levels of education. More recent discussions of the rates of return include Psacharopoulos and Patrinos (2004) and Montenegro and Patrinos (2014). 78 Education Public Expenditure Review Guidelines Technical Note 26: Calculating the budget costs of one completer and of one uninterrupted completer This is a hypothetical example, although it uses real data for dropout rates by primary grade and for average repetition rates for primary education. It shows how to calculate what it costs to get one completer, who may, or may not have repeated one or more grades, and what it costs to get one uninterrupted completer who repeats no grades. The example does not calculate the substantial social and individual costs of dropping out of school, although, by making certain assumptions, this could be done. Table TN4: Progression through primary school of Cohort A that starts with 100 grade 1 entrants Grade 1 2 3 4 5 6 7 8 # Number of new entrants to grade 100 67 50 37 27 20 15 10 326 Percentage of cohort who drop 21 12 11 14 11 11 20 14 out during year Number of students who drop 21 8 6 5 3 2 3 3 51 out during year Total in class at end of year 79 59 44 32 24 18 12 7 Using 15% repetition rate, number that do not move to 12 9 7 5 4 3 2 1 43 next grade Total who progress to next grade 67 50 37 27 20 15 10 6 232 Assumptions The per capita cost of a year of primary school in constant dollars is $50. The per capita cost, in constant dollars, is the same across grade and across time. The percentage of students who drop out by grade is taken from real data for a real country. The average percentage of students who repeat a grade are taken from real data for the same country. A student who repeats a grade repeats that grade only once, although he or she may repeat A student who drops out in a year costs $50, regardless of when he or she drops out in the year, because recurrent costs (e.g., number of teachers) are sunk costs by the time the student drops out. (viii) Once a student drops out, the costs of all post-dropout years for that student are as budget savings to the system. 79 Table TN5: Costs of one completer and one uninterrupted completer Cost and Savings U.S. Dollars If no students drop out or repeat and unit costs are $50 per capita in constant dollars, $40,000 then the cost of a primary education (grades 1–8) for Cohort A is 100 × 8 × 50 Per capita cost of 8 years of education (8 × 50) $400 Total costs of those who start each of the eight grades, whether or not $29,100 they subsequently repeat or drop out (326 students) + repeaters (43). The figure for repeaters includes the additional cost of the repeated year + all remaining years of primary education until completion. Costs of repetition alone (43 years repeated × $50 ÷ year) $2,150 Savings from dropouts. Budgetary costs to a system with high dropout rates $13,050 are lower than to a system with no dropout rates if we assume that when a student drops out, the system saves his or her full, $50÷year cost in each post- dropout grade. Social and individual and household costs, not computed here, are obviously substantial. Cost for one completer: 29,100 ÷ 49. Total completers = 49 (6 uninterrupted $594 completers + 43 completers with one repeated grade). The cost is about 50% more than the cost without repetition, which is 8 × 50, or $400. Cost for one completer without interruption ($29,100 ÷ 6 completers who did $4,850 not repeat any grade). The cost of one uninterrupted completion is about eight times the cost of a completion that includes repeaters. Technical Note 27: Estimating private rates of return to education The private rate of return compares the costs and benefits of schooling as incurred and realized by the individual student who undertakes the investment. The models and methods used to calculate private rates of return to education depend on the policy questions of interest and the quality of the available data. For example, allowing for heterogeneous returns to education across individuals with the same level of education--e.g., as a result of variations in cognitive achievements or regional variations in wage structures--is more data-intensive than assuming homogeneous returns. In comparisons between countries of returns to education, Montenegro and Patrinos (2014) provide guidance for calculating private returns to education. Noting that the now-standard method for estimating private returns per year of schooling is the Mincerian earnings function method, they state that this means estimating log earnings equations of the form: Technical Note 1: Estimating private rates of return to education The Technical private rate Technical Note 1: Estimating of return compares Note 1: Estimating private private rates and benefits the costs of return to education rates of return to education of schooling as incurred and realized by the individual student who undertakes the investment. The models and methods used to calculate private rates of Technical Note 1: Estimating private rates return to education depend on the policy return to of ofquestions education interest and the quality of the available The private rate of return compares the costs and benefits of schoolingacross as incurred and realized by the 80 Education PublicThe data. individual For example, Expenditure private rate of return compares student who undertakes allowing Review for heterogeneous Guidelines the costs and benefits returns to of education individuals with the same schooling as incurred and realized by the level of education--e.g., individual student who undertakes as a result the investment. the investment. The models and methods of variations in cognitive achievements The models and methods used to variations or regional calculate private used to calculate private in wage rates The of private structures--is return to education depend on the rate of return compares more data-intensive than assuming homogeneous the policy costs questions and benefits ofof interest schooling returns. and the as quality of the incurred and realized by available the rates of return to education depend on the policy questions of interest and the quality of the available data. individualFor example, student allowing who undertakes for heterogeneous the investment. returns to education The models andacross methods used to with individuals the same calculate private data. For example, allowing for heterogeneous returns to education across individuals with the same level In rates of education--e.g., of return to education depend on the comparisons between as a result countries of of variations returns tocognitive achievements in policy questions Montenegro education, of interest or regional and the variations quality of the in wage available level of education--e.g., as a result of variations in cognitive achievements orand Patrinos regional (2014) variations inprovide wage structures--is data. For guidance for more data-intensive than example, calculating allowing private assuming homogeneous for heterogeneous returns to returns to education. Noting returns. education that across the nowindividuals standard with method the same to structures--is more data-intensive than assuming homogeneous returns. level of education--e.g., estimating private returns asper a resultyear of schooling in of variations cognitive achievements is the Mincerian earnings function method, or regional variations they instate wage In comparisons structures--is that this means between countries of returns more data-intensive than estimating log earnings equations to education, assuming homogeneous of the form: Montenegro returns. and Patrinos (2014) provide In comparisons between countries of returns to education, Montenegro and Patrinos (2014) provide guidance for calculating private returns to education. Noting that the now standard method to guidance for calculating private returns to education. Noting that the now standard method to estimating In comparisons private between returns per year of of countries schooling returns is tothe Mincerian earnings education, Montenegro function method, !and Patrinos (2014) they providestate estimating private returns that this guidance means calculating (1) for estimating log earnings per year private !of = to schooling equations returns + is the of !the education. ! ! + + !Mincerian earnings form: Noting that! the ! + now !standard function method, they state method to that this means estimating log earnings equations of the form: estimating private returns per year of schooling is the Mincerian earnings function method, they state that this means estimating log earnings (1) ! hourly= equations +annual, of !the+ form: ! + data) ! !! ! + ! for the i individual; where ! is the natural (1) ! = or log (of + ! + ! ! ! depending on ! ! + ! !earnings + ! th is years of schooling (as a continuous variable); ! is labor market potential where Ln(wi) is the natural log (of hourly or annual, depending on experience (estimated as ! ! (1) ! = + ! ! + ! ! + ! !!data) + !earnings for the ith individual; Si agei - ! - 6); ! is potential experience-squared; and ! is a random disturbance term reflecting is yearswhere ! is the natural of schooling a continuous (asTherefore, log (of hourly variable); or annual, X is labor market depending on data) earnings potential for the ith individual; (estimated as age experience where is the natural unobserved abilities. log (of ! can hourly or annual, be viewed as depending on the average rate of earnings data) return to for years the ofithschooling to individual; i ! is 2years of !schooling (as a continuous variable); ! is ilabor market potential experience (estimated as - Si - 6); wage age !X is years where ii - is potential employment. of schooling (as - 6); ! ! experience-squared; The list ! the natural is is of control variable); a continuous potential log (of variables hourly experience-squared; or and is! kept annual, isμ and labor depending on i is a random deliberately market is a random disturbance small potential data) to avoid overcorrecting experience (estimated earnings disturbance for term the term i th reflecting reflecting for as individual; unobserved age factors Therefore, ! i - !that - 6); are !!correlated is potential with years of schooling. experience-squared; and This !!is is alsoa random as the “Mincerian” known disturbance term reflecting method abilities. (Mincer 1974). β1can ! is years of schooling (as unobserved abilities. be a continuous Therefore, viewed as variable); ! can be viewed the average as rate ! is labor the average rate of of return market potential return to to years years ofofschooling schooling to as experience (estimated to wage unobserved abilities. !Therefore, ! can be viewed as the average rate of return to years of schooling to wage - ! - 6); !The ageiemployment. is potential list of control variables is kept and experience-squared; ! is a random deliberately small to disturbance term reflecting avoid overcorrecting for employment. wage employment. factorsunobserved abilities. that are The list of Therefore, The correlated control list of control with years variables ! can variables of is kept schooling. be viewed is kept deliberately deliberately small small to to avoid avoid overcorrecting overcorrecting for for factors that The factors earnings that are function correlated method with years can beof used schooling.to as This is also the average rate of estimate This is also returns known known as return at different the to “Mincerian” years of schooling to schooling as the “Mincerian” method levels method by are correlated (Mincer (Mincer wage converting 1974). with employment. 1974). years The of schooling. list of control This variables is also is kept known the continuous years of schooling variable (S) into a series of dummy variables, say Dp, Ds deliberately as the small“Mincerian” to avoid method overcorrecting (Mincer for 1974). factors and Dt (where that are p is correlated primary schooling, with years of schooling. s is secondary This is also schooling and known t is tertiary) as tothe “Mincerian” method denote the fact that a The (Mincer person earnings has 1974). functionthat achieved method level can be used The of schooling. to estimate omitted level returns is peopleat different with no schooling schooling levels and that by The earnings function method can be used to estimate returns at different schooling levels by The earnings converting dummy is converting function the not continuous the in the method continuous years of can equation to avoid matrix be used schooling years of schooling variable to estimate variable singularity. (S) into (S) intoThe a a returns series of dummyat different estimation equation in this series of dummy variables, variables, schooling say is case say Dp Dp , Ds of the , Ds levels by converting and The Dt (where earnings p is primary function schooling, method can s is secondary be used to schooling and estimate t returns is tertiary) at to different denote the fact schooling that levels a by the continuous form: and Dt (where person converting has achieved years p is primary the continuous ofthat schooling, variable schooling levelyears of schooling. s is secondary of schooling The ( S)omitted variable into schooling anda)level (S series into a t isoftertiary) people is series dummyofwith dummy variables, to denote the fact no schooling variables, and say that say Dp Dp, that a Ds and Dt (where , Ds person has achieved that level of schooling. The omitted level is people with no schooling and that p is primary dummy and Dt isschooling, not in the (where s is secondary schooling, s schooling p is equation to avoid matrix primary singularity. is secondary and The schooling andt is tertiary) to t is tertiary) estimation equation in this denote to denote the fact !the case factis ofthat the a that a person has person has(2) ! =level equation to avoid matrix + of ! ! + ! ! + dummy is not in the singularity. The ! is ! + ! !with estimation equation in this + ! ! + case is of the ! and form: achieved that schooling. The omitted level people no schooling that achievedform:that level of schooling. The omitted level is people with no schooling and that dummy is not in dummy is not in the equation to avoid matrix singularity. The estimation equation in this case is of the the equation form: to (2)avoid matrix = singularity. + The + estimation + equation+ in +this ! + is case ! of the form: (2) ! ! ! ! ! ! ! ! ! ! ! ! ! = + ! ! + ! ! + ! ! + ! ! + ! ! + ! After fitting this “extended earnings function” (using the above dummies instead of years!of schooling in (2) ! private the earnings function), the = + of rate ! ! + ! ! + levels return to different ! of ! + ! ! + schooling ! can be ! + ! the derived from following formulas: After fitting After fitting this this “extended earnings “extended earnings function” function” (using the above (using the dummies aboveinstead of dummies years of schooling instead in of years of schooling in After fitting this “extended earnings function” (using the above dummies instead of years of schooling in the earnings function), the private rate of return to different levels of schooling can be derived from the the earnings the earnings following formulas: After fitting function), function), the private this “extended earnings rate the private rate of (3) return to different function” of return ! = ( (using the ! ) / ( to different levels ! above dummies ) of schooling levels can be of schooling instead of derived fromcan bein years of schooling the derived from the following formulas: following earnings function), the private rate of return to different levels of schooling can be derived from the the formulas: following formulas: (4) (3) ! = (=! − !)) ( //(( !)− ! ) (3) ! ! ! = (! )/(! ) ! (3) (4) ! = (!−=(! )/(! ) ! )/(! − ! ) (4) ! = (! ! − ! )/(! − ! ) (5) (4) ! ! = (! ! − ! )/(!! − ! )/( − ! ) !) - where S , S and St stand p s Sp, Ss and where for the St stand for thetotal number total number of years of years of schooling of schooling for each for each successive successive level. level. Care has Care has to to be taken regarding the foregone earnings of primary school-aged children. In the empirical analysis be taken regarding the foregone earnings of primary school-aged children. In the empirical analysis that follows we have assigned only three years of foregone earnings to this group, following tradition that follows, we have (Psacharopoulos assigned only three 2004). = (!of (5) ! years −foregone ! )/(! − earnings ! ) to this group, following tradition (Psacharopoulos 2004). The costs incurred by the individual are her foregone- earnings while studying, plus any tuition fees or where Sp incidental , Ss and Stincurred expenses stand forduring schooling. the total numberSince of yearsschooling is mostly of schooling provided free by for each successive the Care level. state, at to has The costs incurred least be at the basic by taken regarding the education the individual level, then foregone are earnings his of or her foregone in practice the only primary earnings cost in a private rate of school-aged children. while In the studying, return calculation empirical analysisplus is any tuition fees that the foregone follows we earnings. have assignedThe private benefits only three amount years to whatearnings of foregone a more educated individual to this group, earns tradition (after following or incidental taxes), expenses above (Psacharopoulos incurred a comparable 2004). during group of schooling. individuals with less Since schooling.schooling This more isor mostly provided less refers to adjacentfree by the state, at least at the basic education level, then in practice the only cost in a private rate of return calculation is levels of schooling; for example, tertiary versus secondary school graduates. Although convenient because The costs it requires incurred lessbydata, this method is the individual slightly are her foregoneinferior to the full earnings discount while method studying, (Psacharopoulos plus any tuition fees or the foregone 2004); inearnings. incidental fact, it assumes expenses The private flat incurred benefits age-earnings during amount profiles schooling. Since to what for different schooling levels aof more is mostly educated schooling provided individual (Psacharopoulos free by and at the state, earns (after taxes), above a comparable group of individuals with less schooling. This more or less refers to adjacent levels Layard least 1979). at the basic education level, then in practice the only cost in a private rate of return calculation is the foregone earnings. The private benefits amount to what a more educated individual earns (after of schooling; From for equation taxes), example, above (1) the tertiary versus return to potential a comparable group of secondary experience individuals school is given by: with less schooling. This more Although graduates. to adjacent because it or less refersconvenient requires less data, this method is slightly inferior to the full discount method (Psacharopoulos levels of schooling; for example, tertiary versus secondary school graduates. Although convenient 2004); in fact, because it requires less data, this method is (6) ! + slightly 2! to the inferior full discount method (Psacharopoulos it assumes flat age-earnings profiles for different levels of! schooling (Psacharopoulos and Layard 1979). 2004); in fact, it assumes flat age-earnings profiles for different levels of schooling (Psacharopoulos and Layard 1979). From equation From which (1)to the equation needs return (1) the be toat evaluated potential return to potential experience experience a given value of Xi. is For is given by: given by: each sample we use the average years of potential experience as the evaluation point. It is important to stress that in the empirical part when we (6) refer to potential experience we are referring to the ! ! + 2!based on equation (1). estimates which needs to be evaluated at a given value of Xi. For each sample we use the average years of potential which needs to be evaluated at a given value of Xi. For each sample we use the average years of experience as the evaluation point. It is important to stress that in the empirical part when we refer to potential experience as the evaluation point. It is important to stress that in the empirical part when we experience potentialrefer we to potential are referring experience to the to the we are referring estimates based estimates on equation (1). based on equation (1). 81 Technical Note 28: Per capita financing Per capita formula funding is widely used to help achieve both horizontal and vertical equity by adjusting a standard unit cost (horizontal equity) with coefficients to provide more funding for the disadvantaged (vertical equity). Whether per capita financing is in the form of a funding formula or other funding mechanism, it is important to examine how the unit cost is determined, what factors are taken into account to address vertical equity (e.g., special needs), and how funding is indeed implemented.38 Technical Note 29: Targeting mechanisms, coverage, and depth of programs Are there methods in place to identify the needs of disadvantaged students for education transfers or subsidies, such as scholarships and free textbooks? How are eligible students identified—e.g., projection from historical levels, geographic targeting, analysis of household survey data within the last five years, analysis of individual student data? Is there targeting to certain income, gender, or ethnic and religious groups? When educational merit is the basis, the subsidy or transfer will tend to favor wealthier families because of often observed learning gaps at lower levels of education by income quintile. Per student spending on targeted programs to provide various types of subsidies for disadvantaged students, such as schoolfee discounts or waivers, free lunch, or scholarships, can be separately computed. If education spending is allocated based on a formula, we need to assess whether the formula takes into account vertical equity by setting a higher unit cost (or coefficients) for special needs students, rural schools, and other types of disadvantages. Technical Note 30: Benefit incidence analysis Benefit incidence analysis (BIA) is an analytical tool that helps examine whether the benefits of public expenditure are distributed across population groups by wealth or other socioeconomic or geographic characteristics. This type of analysis can be used to (Narayan 2014): (i) inform policymakers about the current incidence of social spending, i.e., the extent to which different segments of population (e.g., the poor or the rich) are benefiting from the current allocation of social spending, and changes in the incidence of spending over time; (ii) establish a benchmark for cross-country comparison of distribution of public spending on social services; (iii) analyze if specific policy reforms in the past may have accounted for the current observed incidence, or changes in the incidence of spending, over time; and (iv) demonstrate whether a pro-poor benefit incidence is actually translated into better social outcomes for the poor). 82 Education Public Expenditure Review Guidelines Minimum practice—known as average (or simple) benefit incidence analysis—is to report average public education financing per child or per household by household consumption (e.g., quintile, decile), or by other socioeconomic or geographic characteristics (urban vs. rural, male vs. female, etc.). In this type of analysis, benefits are calculated, on average, on the basis of unit costs (aggregate expenditure divided by the number of beneficiaries), multiplied by the number of users of the service in a specific group (quintiles or deciles, etc.). In progressive or “pro-poor” public spending, poorer quintiles or deciles get a disproportionately higher share of the total benefit compared to their share in the national income distribution; for example, the bottom 40 percent receive more than 40 percent of the total funds. The choice of quintile definition (household versus population quintiles) may affect the analysis results due to variation in the numbers of individuals in each quintile. When quintiles are defined over the population, the population size of each quintile is defined to be equal. However, the population size of each household quintile varies, depending on the household-size characteristics of the quintile. Because poor households typically have a larger household size, when the household, rather than the population, is used for the analysis, the distribution of spending appears more progressive than it actually is (Demery 2000). Government subsidies for services may vary significantly by region and across groups. Thus, aggregating unit subsidies may mask inequality. The best benefit incidence analysis practice involves more sophisticated analyses as opposed to the average benefit incidence analysis. These include analyses based on disaggregated unit subsidies (as opposed to average unit subsidies), where specific costs for specific interventions are accounted for separately. Demery (2000) gives examples of using disaggregated unit subsidies specified for different geographical areas and education levels. Figure TN2: Concentration curve: government spending on education and various benchmarks Cumulative share of government spending on education (percent) Pro-poor spending 45-degree line - Equality Progressive Lorenz Curve of income or consumption Regressive (poorly targeted) Cumulative share of population (percent) Source: Narayan (2014) 83 Concentration curve: A concentration curve shows the relationship between the cumulative share of government spending on education (y-axis) against the cumulative share of the population, ranked by consumption or income level from the poorest to the richest (x-axis). Lorenz curve: A Lorenz curve shows the relationship between the cumulative distribution of total household expenditures against the cumulative population, ranked by consumption or income level. Targeting: Public spending is pro-poor if the concentration curve for education benefits is above the 45-degree line; otherwise, pro-rich spending. Progressivity: Public spending is progressive if the concentration curve for these benefits is above the Lorenz curve for income or consumption, but below the 45-degree line. The results of a benefit incidence analysis can be presented in two ways: tabular or graphical (Lorenz or Benefit Concentration Curves). The expenditure Lorenz curve is derived from tracking the cumulative distribution of total household expenditures against the cumulative population, ranked by per capita expenditures. It provides a point of comparison against which to judge the distribution of education spending shown in the concentration curves (Demery 2000). Public spending on education tends to be more pro-poor at lower education levels (e.g., primary), but becomes pro-rich for subsequent levels of education. Two main factors account for this phenomenon: (i) Poorer household quintiles tend to have more children than richer quintiles and may receive a disproportionately higher share of public education resources, at least at the primary level, and possibly at lower secondary level; and (ii) any poor children do not progress to secondary or higher levels of education, where per capita spending tends to be higher and where a disproportionately lower share of resources goes. Other types of benefit incidence analysis include marginal benefit incidence analysis, which estimates the incidence of the last (or the next) unit of benefit, and behavioral benefit incidence analysis, which estimates behavioral responses to a policy change. Narayan (2014) and Demery (2003) provide a good basic explanation of some of the issues and concepts in incidence analysis; Mingat, Tan, and Sosale (2003), in Chapter 7, provide a general discussion on calculating the distribution of public subsidies for education; van de Walle (2003) offers a more advanced treatment of some of the methodological approaches. Other useful information on the topic, including manuals, guidelines and a STATA (data analysis and statistical software) program, which includes a module to perform benefit incidence analysis, can be found at the Open Budgets Portal. 84 Education Public Expenditure Review Guidelines Technical Note 31: Subsidies Why do we care about subsidies for private schools? There is no obvious efficiency rationale for fully funding non-public schools unless they are shown to achieve better outcomes than public schools. Paying the full unit cost saves the state nothing, but partial subsidies can create incentives for private provision that save the state money because it does not have to pay the full cost of educating the child. Depending on how they are designed, public subsidies may implicitly subsidize the wealthy’s preference for private education. PART III: EXAMPLES 86 Education Public Expenditure Review Guidelines Process and criteria for selecting examples The World Bank team examined about 80 education and cross-sectoral public expenditure review documents published between 2010 and 2016. A complete PER database, which also includes those published between 2002 and 2009, is available at http://datatopics.worldbank.org/education/ wDataQuery/ByExpenditures.aspx. The team adopted the following criteria to identify good examples of different topics of reviews: Analysis that is relevant, detailed, and comprehensive Analysis that provides a methodology Analysis that includes international and regional comparisons and benchmarks Analysis based on disaggregated data (if available), including subnational or school-level data, or data on specific student groups Analysis with clear and original presentation of results (graphs, tables, diagrams, etc.) Most recent analysis For good examples of policy recommendations, the team selected reviews that provide clear, prioritized, specific, and implementable, and costed recommendations. The team selected several country public expenditure review documents for more than one example because they show a relatively better treatment of a specific topic, compared to other country reviews. This section will be updated as more good examples become available. Table of Contents Examples No. Topic Example Page Albania (2014); Ethiopia 1 Policy recommendations 88 (2015) Myanmar (2015); Samoa 2 Data sources 96 PER Notes (2014) 3 Analysis of sources of funds Nigeria (2015) 99 4 Analysis of revenue sources Nigeria (2015) 100 5 Analysis of decentralized financing Sudan (2014) 103 Analysis of education financing by level of 6 Indonesia (2013) 104 government and intergovernmental fiscal transfers 7 Analysis of school budget by financing source Philippines (2013) 108 8 Analysis of household surveys on private spending Mali (2016) 112 9 Analysis of donor funding Mali (2016) 115 10 Analysis of total public education spending Albania (2014) 117 Honduras (2015); 11 Analysis of functional classification 118 Kenya (2014) (Table of Contents: Examples continued) 87 No. Topic Example Page Analysis of education spending by economic 12 Honduras (2015) 120 classification 13 Issues related to budget formation and execution Solomon Islands (2011) 121 14 Analysis of per student spending Armenia (2011) 124 15 Minimum norms and standards for resource allocation Belarus (2013); 126 Kosovo (2014); Jordan 16 Analysis of cost of teachers 130 (2016) 17 Analysis of teacher distribution Indonesia (2013) 140 Tajikistan (2013); Jordan 18 Cost projections (2016); Democratic 144 Republic of Congo (2015) Albania Volume II (2014); 19 Fiscal sustainability analysis 149 Madagascar (2015) Guinea (2015); Belarus 20 Demographic trends and enrollment projections 154 (2013); Georgia (2015) 21 Technical efficiency of inputs (efficiency indicators) Belarus (2013) 161 Albania (2014); Belarus 22 Analysis of unit costs and outcomes (2013); Indonesia (2013); 163 Madagascar (2015) Ethiopia Policy Research 23 Cost-benefit analysis 173 Working Paper (2008) Kenya (2013); 24 Data envelopment analysis Democratic Republic of 176 Congo (2015) 25 Internal efficiency indicators Bangladesh (2010) 182 26 Rate of returns to education Sri Lanka (2011) 183 27 Analysis of inequity Costa Rica (2015) 185 Tajikistan (2013); 28 Analysis of per capita financing 188 Mauritania (2016) Costa Rica (2015); 29 Analysis of cash transfer programs 193 Indonesia (2012) Democratic Republic of 30 Benefit incidence analysis 197 Congo (2015) 31 Analysis of private spending by income quintile Madagascar (2015) 199 88 Education Public Expenditure Review Guidelines Example 1: Policy Recommendations Overall comment: The Albania and Ethiopia public expenditure reviews meet most of the criteria for examples of good policy recommendations: clear, prioritized, specific, implementable, and ideally, costed. In the Albania example, the recommendations are clear, specific, and prioritized (have a time frame for short- and medium-term policy recommendations), and at least part of these recommendations are costed. In the Ethiopia example, most recommendations are clear, quite specific, and somewhat implementable, but not prioritized, which is a drawback, given their number. Democratic Republic of Congo PER (2015) also offers good policy recommendations, but it is not included in this section because of its length. It provides a separate table that lays out a proposed time frame and identifies the ministry (ies) within whose jurisdiction the policy recommendation falls. Bad examples of policy recommendations include: • Recommendations that are not prioritized or sequenced • Vague imperatives that provide policymakers with little or no guidance about available options for fixing the problem • Vague imperatives that fail to acknowledge the fiscal costs, technical complexity, political costs, or time frame required to try to fix the problem • Policy implications or policy directions masquerading as policy recommendations For example: • “In a context of a rapid fall in learning outcomes, there is a need for additional public spending targeted directly at improving quality.” • “Improve the efficiency of public funding in the education sector.” • “Invest more in education, in line with other countries.” • “Increase the allocation of the public expenditures in Technical and vocational education and training (TEVET) as a share of total government education expenditure because the (private) rate of returns is high and the supply of the TEVET graduates is welcomed by the market but remains limited.” 89 Albania PER (2014) Expand Access to Pre-primary Education One priority is the expansion of access to and quality of preprimary education; the cost can be offset by savings from declining student numbers in higher grades (short-term). By 2015, Albania’s preprimary population is likely to stabilize at current levels but the numbers of students in basic, upper secondary, and higher education will drop significantly. Assuming there will be a 95 percent enrollment rate for preprimary and universal enrollment in basic and upper secondary by 2025, fewer basic and upper secondary education places will be needed relative to 2013. On the other hand, Albania’s investments in preprimary and primary education are far behind international benchmarks, especially given studies that confirm robust returns to investment in children’s early years compared to equivalent investments later in life. Preprimary education will therefore need to be expanded by approximately 36,000 places. Based on current unit costs, creating the 36,000 places preprimary education will need in 2025 would cost about 5 percent of the education budget. Using the same criteria to quantify savings, about 9 percent of the education budget can be saved from the decline in numbers of basic and upper secondary students, which can redeployed to finance the expansion in preprimary. Furthermore, since preprimary classes are mainly housed within primary schools, the places freed by basic education could be used to expand access to preprimary. In particular, the school mapping data collected by the Council of Europe Development Bank project (ALB-IPF-TA-10) should be better utilized. Rural to urban migration and expansion of private education should also be taken into account. For these interventions to produce improvements in quality of preprimary education, continued teacher training and capacity building is needed. Improve Governance, Efficiency and Equity Improve management and governance in the education sector (short-term). Sector planning and coordination across different levels of government could be improved to increase transparency and accountability in the use of resources and also enhance policy making and education delivery. Current financing mechanisms and accountability arrangements for pre-university education do not create incentives for schools and local governments to rationalize spending. Limited capacity and lack of a clear monitoring framework exacerbate sector inefficiency. To bring more transparency and equity to education financing, Albania could consider introducing per capita financing (PCF) to fund pre-university education (short-term). Currently, expenditures per pupil vary substantially by region; and it does not seem that regions with poorer results or a higher incidence of poverty receive more funding. Furthermore, schools depend on the REDs and local governments for financing major inputs and meeting maintenance needs, which impairs their ability to sustain a high-quality teaching environment. The introduction of a well-designed PCF mechanism could not only improve transparency around what regions receive and how that relates to student population, but also efficiency and accountability. It could also help provide additional funding to the disadvantaged and hard-to-reach populations. As PCF is phased in, in the medium term both school autonomy and accountability should be heightened (short-term). Currently, schools have only minimal financial autonomy. The PISA 2012 90 Education Public Expenditure Review Guidelines school questionnaire shows that only 20 percent of principals reported formulating their school budgets and less than 10 percent said they had autonomy over teacher hiring, firing, and salaries. Albania should consider giving schools more autonomy on both HR and financial management; but this shift will need to be accompanied by investment in building capacity at the local level and greater accountability for results. While this shift is unlikely to have significant budgetary implications, it will require a cultural shift whereby schools, REDs, and EOs will have to perform new roles. In a more decentralized setting, for example, REDs and EOs should act as advisors to schools. Cost-effectiveness of sub-sector investments could be examined as a basis for reallocating funds. Vocational education students cost the government three times as much per year as general education students, but there is no evidence that there is a concomitant return on the investment in terms of greater learning. Given the tight fiscal space, investment decisions should be driven by benefits that can accrue to both individual children and to society as a whole. For example, preprimary investments can be leveraged to produce positive effects on female labor participation, reach marginalized populations, and reduce the intergenerational transfer of poverty. Consider Increasing Education Spending over the Medium Term In the short term, given the lack of fiscal space and the fiscal consolidation plan in place for the period 2014-16, Albania should carry out reforms to make the sector more efficient and do more with the same level of budgetary resources. In the medium to long term, particularly after 2016, Albania should consider increasing public spending on education. Albania’s small budget envelope for the education sector should be raised gradually from about 3 percent of GDP to 4.0 percent of GDP, closer to the Eastern European average of 4.6 percent of GDP. While there is agreement that more public spending will not guarantee better education quality, Albania ranks at the bottom with respect to both learning outcomes and public spending on education as a share of per capita GDP. Even after exploiting all efficiency gains, there is likely an additional need for budgetary resources to increase the quality of education and learning outcomes over the long term. The additional public spending could be channeled to several areas in which Albania still needs investment, such as: (i) teachers’ professional development; (ii) learning materials and school supplies; (iii) quality of school facilities; and (iv) more time on tasks and activities in schools. While the cost of basic investment in pre-primary education can to some extent be offset by the savings from the declining student population, it also requires well prepared educators, learning materials and school supplies for which additional resources are needed. Improve Quality of Education The government could ensure that tertiary education is not expanded at the expense of quality (short-term). The government should strengthen the regulation guiding the expansion of higher education—largely through the implementation of the 2010 law on tertiary education— both to assure quality and to guarantee alignment with population trends. In the last decade, Albania has seen much higher enrollments in tertiary education, both public and private, in line with the government’s commitment to providing access to all who wish to continue beyond secondary education. Now it is necessary to direct attention to quality—with a view to ensuring that standards articulated in the 2010 Tertiary Education Law, itself aligned with the Bologna Process and overall 91 principles of Europe 2020 Strategy—are complied with. It also means that financing has to keep pace with the increase in the number of students, so quality is not jeopardized. Quality data must be generated and disseminated to the wider public to build evidence and inform policy making (short-term). Currently, there is little evidence on which policies might improve student learning in basic education. Unfortunately, the unreliable household data from PISA does not allow for analysis of determinants of learning outcomes, and causal links have not yet been established between various policy reforms and student achievement. To avoid making policy decisions in the dark, Albania should immediately finalize the design of the EMIS and, once it is in place, use the data in making decisions. This will also help strengthen school accountability. Albania should also seek evidence on policy reforms that have improved student learning in other countries, such as investments in preprimary education, improving teacher and principal effectiveness, and increasing school autonomy and accountability. Ethiopia PER (2015) Resource reallocation to improve equity If the role of public finance is to counteract the inequality in access to education, resulting from unequal income distribution, the government ought to be directing more than 20 percent of public resources towards the lowest income quintile, whereas it is currently directing only 13 percent towards the poorest quintile, less than even the 14 percent share of out-of- pocket expenditure on education contributed by this quintile. At the other pole, 39 percent of the benefits of public education spending goes to the highest income quintile, which contributes 31 percent of out-of-pocket spending. Correcting for this anomaly calls for a reallocation of public education spending, from higher education (which caters largely to the top quintile) to lower primary education (which caters to all classes and disproportionately more to the poorest quintile). The case for such a reallocation of resources from the highest to the lowest level in the education ladder is made stronger by the following facts: (i) more than half the recurring public expenditure on higher education is on provision of free food and lodging to all students in residence, many of who can afford to bear at least part of this cost; (ii) serious classroom and teacher shortages exist in primary schools in particular regions and districts, addressing which will have high marginal impact on efficiency and school performance; and (iii) improved efficiency at lower primary level will increase the access of pupils from poorer households to higher levels, thereby also contributing to improved equity in the distribution of benefits. The provision of free higher education services is based on the rationale that the beneficiaries will pay back after graduation and finding employment, through the “graduate tax”. However, actual level of cost recovery through this instrument is negligible and is not even being reported in any official publication. Moreover, while the graduate tax is in theory a suitable instrument for recovering the academic cost of providing higher education, the non-academic recurring costs need to be, and can be, recovered more quickly. While it may not be politically feasible to withdraw or cut down on subsidies being provided to university students, it should be possible to at least freeze the aggregate amount of subsidy in nominal Birr, and gradually shift a part of the non-academic recurring cost onto the students, complemented by financial assistance to the few students from lower income households who gain admission in higher educational institutions. 92 Education Public Expenditure Review Guidelines Resource reallocation to improve efficiency Additional resources for non-salary recurring inputs, provided through GEQIP, have not had a visible impact on average efficiency in primary education, even though efficiency has improved in better equipped woredas and schools. The average has been pulled down by a lagging quarter of the woredas, where efficiency has declined in spite of additional non-salary funding, in the absence of addressing the binding constraint, i.e., teacher and classroom shortages. The persistence of acute shortages of teachers and classrooms in lower primary education in the case of the lagging quartile of woredas shows that the existing structure of intergovernmental grants is unable to address such shortages, which are concentrated in a few regions and woredas. Formula-based general purpose grants from federal to regional and from regional to woreda governments cannot address such concentrated shortages. Nor can the existing GEQIP school grant in its present design, being distributed to all schools proportional to enrolment. Some mechanism for transferring teachers from one woreda to another has to be worked out. It should ideally be done in a way that makes it gainful for both parties involved in the transfer, and at minimum additional cost to the public exchequer per transfer. An exchange system could be created wherein woredas that have large teacher shortfalls can post their needs/demands—and those that have extra teachers can choose to offer some on transfer. Suppose a rule is established whereby the donor woreda is required to transfer 90 percent of the budget for Remuneration of Transferred Teachers to the recipient woreda (it could be a time bound contract), on the condition that the latter allocates 10 percent and commits to take on all future increases in salary and allowances. For the Donor, the benefit is the 10 percent immediate saving, plus further saving in the future since the continuing commitment towards the lent-out teachers remains fixed in nominal Birr. For the Recipient, the benefit is the ability to hire additional teachers with minimal initial cost and longer time to mobilise internal financing. For the nation as a whole, the benefit is better utilization of available teacher capacity and hence improvement in efficiency at lower cost (than if all woredas simply keep going after desired PTR targets, as fast as their financing capacity permits). Ensuring essential inputs and processes to improve quality Minimum school resources including infrastructure (electricity, water, sanitation), learning resources (textbooks and reference materials) and discretionary funds should be ensured at every school as these are factors that have positive impacts on student learning. Schools should monitor the teaching/ contact time and reduce the time teachers being in school but not teaching. Teachers’ increased efforts in monitoring student attendance will reduce their absenteeism. Teacher professional development should address teachers’ pedagogical challenges and increase their knowledge. Strengthening the system quality assurance capacity will likely bring long-term improvement in learning. Leveraging additional resources for education Ethiopia’s ambitions to become a middle-income country (MIC) by 2025 and its education sector development targets (ESDPV) are admirable, but the achievement of these targets depends on the commitment of all stakeholders: government, families, communities, schools, 93 educators, and administrators. The resources required are very large, necessitating that all key financiers of the system: the government, families and development partners to step up their efforts. Given that a steady 20 percent share of the government budget has been allocated to education over the past decade and the medium-term fiscal framework implies a constrained envelope for government spending as a whole, public resources for education are unlikely to rise above the current level of around 4 percent of GDP. However, the relatively low share of education in private household expenditures indicates that additional resources could be leveraged from private sources. Such potential could be tapped in the case of secondary, TVET and higher education. Analysis of the results of the 2011 Household Income, Consumption and Expenditure Survey shows that a majority of households who send their children to public secondary school have the same spending power as households that send their children to private secondary schools. Yet non-government providers account for less than 5 percent of enrolment in grade 9, and the growth of aggregate supply has failed to keep pace with the growth in demand for general secondary (grades 9-10) education. These facts suggest that it is worthwhile to tap the unused potential for expanding private provision of general secondary education, so as to fulfil the unmet demand with least additional public spending. Higher enrolment of pupils from higher and middle income households in private secondary schools will enable public resources to benefit larger numbers from lower income families. A subsector where demand has apparently declined due to unfulfilled expectations is TVET, where employment and earnings prospects do not seem worth the investment to many households. Low external efficiency, meaning too few among TVET graduates finding the jobs they aspired for, has been the main reason for the recent decline in enrolment. Developing a partnership with the potential employers could perhaps be a more effective way to address technical training needs and at the same time leverage additional resources from the private sector. A public-private partnership approach is also an option for further expansion of higher education in the future, having already created a significant number of publicly funded universities in the country. Strengthening credibility of EMIS data The grade-specific enrolment, repeater and readmitted numbers reported by many of the woredas do not fulfil even a simple consistency criterion, namely, that the drop-out rates implied by the reported data must not be negative. Of a total of over 800 woredas for which EMIS data is available for the five successive years, 2008/09 to 2012/13 (EC01 to EC05), only 37 percent of all woredas in the country are credible and can be used for analysing output efficiency of schools and its determinants. Data submitted by the remaining 63 percent of woredas are not suited for such analysis. Data submitted by schools and woredas need to be checked for consistency and reliability. Realistic planning and target setting Education sector goals and targets need to take into account not only the constraints on supply but also on the demand for education. For instance, the fact that enrolment drops as one moves from lower to upper primary education is, to a significant extent, influenced by the opportunity cost of sending 10-14 year old children to school when they could be doing some work and contributing 94 Education Public Expenditure Review Guidelines significantly to family income. While the first milestone of getting all children (or at least 95 percent) to enter primary education in time is within reach in the majority of regions, this is not the case with respect to the second milestone of getting more than 95 percent to complete the primary cycle. Achieving the latter requires not only improvements in service delivery but also easing of the economic constraints on poor households. Additional primary teacher recruitment needs to be carefully regulated to ensure that it is targeted at those schools/woredas where teacher shortage is clearly a binding constraint to improving efficiency and effectiveness. It is possible to use EMIS data to generate simple indicators of adequacy/inadequacy of (i) teachers and (ii) classrooms, and establish simple guidelines that woreda councils could use while allocating scarce resources. The ratio of Teachers per Section (TPS) = PSR / PTR is a useful indicator. If PTR is much worse than desired target (in a school or woreda) AND TPS is way below 1, then it clearly means that teacher shortage is a binding constraint and of HIGH priority to address. On the other hand, if PSR is much worse than target and TPS is way higher than 1, classrooms/ sections are the most binding constraint and of top priority. Reform options The following are some reform options for addressing the problems highlighted above: • Safeguard the financing for the educationally disadvantaged areas and groups to improve their access to education. This should cover: ▷ ECD/pre-schools (to improve the school readiness) ▷ Grade 1-4 (increasing enrolment of the last 10% and reducing drop-outs) ▷ Grade 5-8 (increasing the supply and addressing the demand constraints for the bottom half ) ▷ Grade 9-10 (increasing the supply and addressing the demand constraints in rural areas) ▷ Preparatory schools (Grade 11-12), universities and TVET programs: ensure that financial aids are available and used to increase their access to these types and levels of education • Introduce a cap on the amount of subsidy to cover non-academic costs (on food and lodging) incurred by higher educational institutions • Reallocate savings on higher education recurring expenditure to initiate a new Special Grant from federal via regions to woredas, targeted at those woredas with most acute shortages of primary teachers and classrooms’ • Establish an appropriate mechanism for transfer of teachers across woredas to reduce the wide inter-woreda variance in PTR • Establish and publicize guidelines for regions/woredas to ensure that additional teacher recruitment is targeted at those woredas/schools where teacher shortage is clearly a binding constraint • Within secondary education, shift the emphasis of additional investments from teacher recruitment to the creation of additional classroom space • Conduct an in-depth study of the demand and supply of TVET training and explore the possibility of a public-private partnership approach to leverage additional resources as well as to improve external efficiency • Encourage families to invest in learning materials (in addition to their current investment in uniforms, bags and transport) 95 • Develop a conducive policy framework for expansion of non-government provision in secondary and higher education, thereby leveraging additional resources • Ensure that data reported by schools and woredas through the Education Management Information System (EMIS) are subjected to adequate and effective consistency checks • Monitor and analyse the sector financing using unit cost approach. Items within the recurrent expenditures that are deemed essential for the efficiency/quality of delivery should gradually be funded domestically to ensure sustainability • Build capacity building to analyse several sources of data (administrative reports, census, surveys and in-depth studies) to analyse, triangulate and interpret the data and identify system challenges and potential solutions • Strengthen the system quality assurance including ▷ assessing students regularly (classroom assessment) and use examination and national (and international) assessment data to provide feedback to schools and policy makers ▷ strengthening teacher training programs (both pre- and in-service) and developing teacher competency validation processes (accreditation, licensing and career development) ▷ using school inspection data to inform school improvement planning, teacher and school leaders’ development; and ▷ conducting school effectiveness analysis using student/teacher/school characteristics 96 Education Public Expenditure Review Guidelines Example 2: Data sources Overall comment: Myanmar (2015) flags challenges associated with data availability and quality for different types of data. Few public expenditure reviews address such data challenges in depth. Samoa PER Notes (2014) has an annex dedicated to detailed information on budget-data sources. Myanmar PER (2015) The government would benefit from investing early in better data and analysis on coverage, quality and equity to prioritize spending towards the 10 Points Policy. Data gaps pose challenges for effective policy-making, especially because most important decisions are made at the Union level rather than at the subnational level where officials would be better informed of local needs and issues. Establishing a solid framework for data compilation and analysis that feeds into policy-making is a long term endeavor (Box 4.2). In the short to medium-term however, the MoE could enhance the capacity for policy analysis by maintaining three critical databases: (i) Education Management Information System to electronically record administrative data, which is currently collected in paper format; (ii) student learning outcomes, and (iii) budget and expenditure data. Box 4.2: Data Challenges • Education Management Information System: MoE [Myanmar] currently collects monthly and annual administrative data (e.g. number and type of schools, teachers and students), compiled in a Statistical Yearbook published with 12 months delay. An MIS would improve quality, timeliness, and accessibility of this important information. • Household surveys: These are important sources of estimates on enrollment, drop-out rates, household spending on education, and returns to education among other things. Existing surveys are dated (2009/10) and do not cover the entire country. Census data and new household survey data will become available in 2015. These data provide invaluable information on enrollments and dropouts. To be useful, though, staff in MOE need training to be able interpret and use these data for policy making. • Accounting for different education levels: Many schools that integrate multiple levels (i.e. primary, middle and high school) do not account for spending at each tier (e.g. nearly all middle schools are also primary). Going forward it will be important to separate cost of each tier to accurately assess costs per student at each level. 97 • Assessment of student learning outcomes: There are currently no regular reviews of whether schooling translates into learning at different stages of the education cycle. The Early Grade Reading Assessments (EGRA) and an Early Grade Mathematics Assessments (EGMA) could be instituted to start the process for grades 1-3. • Government spending: Key to monitoring the effectiveness of links between policy and government spending is timely, relevant, and accurate fiscal data. This will involve review of budget classification and broader FMIS. As part of the writing of this PER, data were compiled in a consolidated BOOST database to facilitate analysis. However, this database need to be updated and actively used by budget staff in MOE – a process which will require training and experience. Samoa PER Notes (2014) Data Annex The PER is primarily based on analysis of disaggregated public financial data covering the seven year period between FY06 and FY12, taking advantage of the FinanceOne FMIS system that has been in place since FY06. This analysis represents the first time that the FinanceOne data has been used to produce a consistent public expenditure analysis. To facilitate the analysis of expenditure based on consistent functional and economic classifications some adjustment to the data were made for the purposes of the analysis. While in more recent years, FinanceOne has full coverage based on GFS classifications, these were not fully in place in FY06 and FY07, so the team remapped expenditure by line item using GFS86 as a guide. While the data from FinanceOne has full coverage of domestic expenditure, externally-funded expenditure was recorded most consistently in other systems held by the Ministry of Finance, so data on estimated grant and loan utilization by ministry was added separately to the analysis. Because this information was only available at a higher level, an analysis of donor-funded expenditure by economic classification or detailed functional classification was not possible. Some further adjustments were made to the FinanceOne data to treat expenditure classifications consistently over the period, to help to illustrate the underlying trends. Firstly, tax expenditure incurred by government ministries that was subsequently rebated was netted off both expenditure and revenue, as it’s treatment in the financial accounts changed over the period which would otherwise give the impression both expenditure and revenue has risen. Secondly, some capital items that were recorded above the line were moved into financing in the budget frame. Thirdly, and most significantly, newly incorporated public beneficial bodies, the National Health Service and the Land Transport Authority were reincorporated into the public accounts. This was necessarily since their new status as corporate bodies meant that halfway through the period, their accounts were no longer consolidated in the public accounts ledger, but instead expenditure was in the form of a public grant. The team used the bodies’ corporate accounts to reintegrate disaggregate expenditure trends for the period since they became corporatized. The South Pacific Games Authority was also reincorporated into the public accounts to smooth out a temporary spike in transfers to that agency relating to the hosting of the South Pacific games. 98 Education Public Expenditure Review Guidelines The analysis of the public wage bill combines FMIS data with data from the payroll system. It also incorporates payroll data from the accounts of the largest public agencies to present approximate estimates of payroll trends for the whole government for the first time. Similarly, the health analysis represents the first time that unified data for the public healthcare sector has been presented since the creation of the National Health Service as an autonomous agency. For the education analysis, the data was adjusted for the spike in trends relating to the South Pacific Games in FY07 and for lumpy tax expenditures to establish the underlying structural expenditure trends. The outcome data are mostly based on the Ministry of Education’s statistical database. Changes are presented in real terms taking FY06 as the base unless otherwise stated. In some cases, additional data has been collected from official government sources including the annual Budget Statement, the World Bank’s World Development Indicators are used for international comparisons and outcome data and other documents. 99 Propoor Spending 45 Degree Line – Equality Example 3: Analysis of source funds Progressive Lorenz Curve of Income or Consumption Nigeria PER (2015) Regressive (Poorly Targeted) Overall comment: The Nigeria example provides a comprehensive breakdown of education-sector financing by source, 100% including public sources by level of government, private sources, and donor funding. In Nigeria, 40 percent of the education sector is funded by private households’ out of pocket contributions while local government constitutes the second highest share (25 percent). Figure 8 presents the sources of finance by origin. Overall, in 2013, the total cost of the education sector (all levels of education) in Nigeria amounted to 2,329.4 billion Naira (14.6 billion US$). The breakdown of the education sector finance was as follows: federal government (18 percent), state government (13 percent), Local Government Authority (LGA) (25 percent), household out-of-pocket payment (40 percent), Universal Basic Education Commission (UBEC) initiative (3 percent), and donors: the remaining 0.4 percent. FINANCE, 2013 Figure 8:FIGURE 8:education Sources of SOURCES finance, 2013 SECTOR FINANCE, 2013 OF EDUCATION sector 2/13 Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 9 | 1% 426 | 18% LGA Federal Government GA Household State Household 936 | 40% UBEC LGA 295 | 13% UBEC State UBEC tate Household Federal Gov’t Donors Donors ederal Gov’t 587 | 25% Donors 76 | 3% Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 ON FINANCING dependent Devp’t venue Partners Federal Ministry of Finance Federal 0% 20% 40% 60% 80% 100% CUMULATIVE SHARE OF POPULATION 100 Education Public Expenditure Review Guidelines Example 4: Analysis of revenue sources FIGURE 8: SOURCES OF EDUCATION SECTOR FINANCE, 2013 Nigeria PER (2015) Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 9 | 1% Overall comment: This example provides a detailed and comprehensive summary of the sources of revenues at all 426 | 18% 936 | 40% administrative levels in Nigeria and explains the revenue-sharing formula. LGA Household 13% the enactment of the 2004 295 |since Figure 12 shows the structure of basic education financing UBEC Universal Basic Education Act (UBE). The law preserves the constitutional responsibility State of states and local governments in Nigeria to provide basic education and expands the federal government’s responsibility in ensuring it is free and compulsory. Federal Gov’t 587 | 25% Donors The proceeds of the Federation Account are shared among the federal, state, and local governments, 76 | 3% in accordance with a revenue-sharing formula and the funding is tracked from the source to the service delivery point. The current formula for dividing up total revenues to government allocates 52.68 percent to the federal government, 26.72 percent to state governments, and 20.6 percent to local governments (Figure 13). The UBE Act was developed based on constitutional mandates, and clearly demarked the financing sources between salary and non-salary. FIGURE 12: THE STRUCTURE OF BASIC EDUCATION FINANCING Figure 12: The structure of basic education financing Internally Federal VAT Independent Devp’t Fund Generated Account & Sources Revenue Derivatives Pool Revenue Partners Federal Ministry of Finance LGA SJGA Federal State Con. Rev. Responsibility Federal Budget UBFC (2%) FMF Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Education Finances Primary Lower Lower Unity School Secondary Secondary College SUBEB Source: Author’s sketch following funding allocation arrangements in UBE following UBE Act of 2004 Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 Note: Each Note: Each StateState shallamaintain shall maintain special accountatospecial account be called “State toGovernment Joint Local be called “State Account Joint Local Government Account (SJLGA) “into which shall (SJLGA) be whichall paid “into allocations shall to the be paid all allocations local to the government local government councils councils of ofthe the state from the state from the Federation Account and from the Government of Federation Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). the State” (Section 162 [6], 1999 Constitution of Nigeria). 101 Education finance depends to a large extent on federal revenues and the ability of the states to finance education expenditures is directly linked to the availability of federal revenues. Figure 13 shows a summary of the sources of revenue for the three tiers of government. Internally generated revenues for LGAs stand at 1.6 percent while allocation from the state government to LGAs only represent about 0.7 percent of their total revenue. This implies that about 95.3 percent of the LGAs’ total revenue comes from statutory allocations, excluding the grants and revenues from the stabilization fund which accounts for the remaining 2.3 percent. Given that salaries are automatically deducted from the statutory transfers at source, this implies that LGAs have, in reality, very little say in education finance, and their role is more symbolic than anything else, given that there is no financial planning or budgeting on their part for the basic education level. At the state level, internally generated revenue (IGR) represents 19 percent of total revenues indicating some potential fiscal space for spending on education based on IGR. In particular, given that the states are responsible for capital and non-salary spending, states’ ability to generate more revenue may suggest variations in resource availability across states for basic education spending. Figure 13: Sources of overall revenues at all administrative levels, 2013 Federal Government State Governments Local Grand Source Governments Total FG’s Share PCT Sub-Total States 13% Sub-Total Statutory Allocation 2,777.4 53.5 2,830.8 1,435.8 615.0 2,050.9 1,107.0 5,988.7 Augmentation 1/ 351.3 6.8 358.1 181.6 101.6 283.2 140.0 781.3 Share from Excess Crude 208.7 4.0 212.7 107.9 60.3 168.2 83.2 464.2 NNPC Refunds - - 0.0 44.9 11.9 56.8 34.6 91.4 SURE-P 191.8 3.7 195.5 99.2 55.5 154.6 76.5 426.6 Share of VAT 106.9 7.6 114.6 381.9 - 381.9 267.3 763.8 FG Independent Revenue 274.4 - 274.4 - - 0.0 - 274.4 Internally-Generated Revenue - 11.0 11.0 574.9 - 574.9 29.3 615.2 Less State Allocation to LG - - 0.0 12.8 - 12.8 - 12.8 Net Internally-Generated Revenue - 11.0 11.0 562.2 - 562.2 29.3 602.4 Grants - - 0.0 69.7 69.7 43.0 112.7 Share of Stabilization Fund - - 0.0 1.3 1.3 16.4 17.7 State Allocation to LG - - 0.0 - - 0.0 12.8 12.8 Others 45.7 - 45.7 8.7 8.7 - 54.4 Total 3,956.2 86.6 4,042.8 2,893.2 844.3 3,737.5 1,810.0 9,590.3 Includes share of the difference between provisional distribution and actual budget. Note IGR is noted for 12 billion naira 1/ and FCT included in state level which make some difference. Source: Cited from CBN annual report with source from “Federal Ministry of Finance (FMF), Office of the Account-General of the Federation (OAGF), and Fiscal returns from state and lovely governments Survey UBFC (2%) FMF Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Education Finances Lower 102 Primary Education Public Expenditure Review Guidelines School Secondary SUBEB Lower Secondary Unity College Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) “into which shall be paid all allocations to the local government councils of the state from the Federation Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). Given that the vast majority of basic education salaries come from the federal allocation, which in turn heavily depends on oil, factors that affect oil revenue also directly affect basic education finance. The share of IGR varies greatly by state, and some tend to depend entirely on federal allocation due to low IGR levels. Figure 14 shows (i) total revenue breakdown by IGR; (ii) revenue other than IGR; (iii) the share of internally generated revenues out of total revenue; and (iv) the per capita allocation of non-IGR. The figure shows that share of IGR revenue varies from a low of 1 percent in Benue state (2 % in Borne state) to a high of 41 percent in Lagos. Overall, only four states including Lagos have an IGR share of revenue more than 15 percent of their total revenue—Rivers (22%), Ogun (21%) and Kano (20%). Edo ranks a distant fifth with 14 percent. In addition to Lagos and Kano, five of the 9 Niger Delta states have higher revenue. In general, revenues across states hover around 100 billion Naira, except for the 6 states where it is substantially higher (Lagos, Kano, Rivers, Delta, Akwa-Ibom and Bayelsa). However, since Nigeria has developed an allocation formula justified by rights enshrined in the constitution such as the right of the Niger Delta states to receive 13 percent of oil revenue prior to allocation; resource availability at the state level clearly depends not only on IGR but also on what is being allocated from the federal level. Figure FIGURE14: Sources 14: of OF SOURCES educational EDUCATIONsector finance, SECTOR 2013 FINANCE, 2013 Source: Calculated from CBN Annual report 2013 and per capital allocation from monthly shares of distribution from The Federation Account by State, Nigeria Economic Report, The World Bank Group (2013) Federal Allocation From All Sources 450 100 Internal Generated Revenue (IGR) 400 Share of IGR (right axis) 90 Per Capita Revenue (right axis) 80 350 70 300 60 250 50 200 40 150 30 100 20 50 10 0 0 Borno Taraba Adamawa Bauchi Yobe Gombe Jigawa Zamfara Kaduna Katsina Kebbi Sokoto Kano Benue Niger Kogi Nasarawa Plateau FTC Abuja Kwara Enugu Imo Anambra Ebonyi Abia Bayelsa Akwa Ibom Delta Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun Lagos North East North West North Central South East South South South West Source: Calculated from CBN Annual report 2013 and Per capital allocation from Monthly Shares of Distribution from The Federation Account By State, Nigeria Economic Report, World Bank Group (2013) FIGURE 3: AN OVERVIEW OF THE COMPLEXITY OF TRANSFERS & FUND FLOWS, 2012 LEVEL BUDGET NATIONAL MINISTRY OF FINANCE MINISTRY OF EDUCATION CULTURE MOF MOEC Adjustment Fund Tugas Central DAU DAK BOS Dekon Pembantuan Functions Own PROVINCIAL Source District Budget Revenue DAU/SDA Provincial Education Central Government O ce Agencies in the Regions PAD 103 Example 5: Analysis of decentralized financing Sudan PER (2014) Sudan State-level Public Expenditure Review Meeting the Challenges of Poverty Reduction and Basic Service Delivery Synthesis Report, Vol. 1, Summary for Policymakers, May 2014 Sudan State-level Public Expenditure Review Meeting the Challenges of Poverty Reduction and Basic Service Delivery, Vol. 2, Background Papers, May 2014. Overall comment: These volumes focus on the strengths and challenges of Sudan’s decentralization arrangements. They do not focus on education per se, although this function is decentralized to subnational units and enters into analyses of decentralization processes. As is often the case in studies of decentralization, these studies had to be based on new data collection in the form of case studies of four of Sudan’s 17 states. Sudan highlights several lessons to which education PERs need to be alert. 1. PEFA issues become particularly salient under decentralization because subnational governments tend to vary significantly in the quality of their public financial management. The opportunities for corruption expand accordingly. 2. Subnational governments usually vary, sometimes significantly, in their capacities to raise own revenues. This reality leads to varying gaps between financing responsibilities and resources and the potential for substantial horizontal imbalances between subnational governments that central government should but may not address. 3. Subnational governments often try to collect cost-ineffective taxes—ones that raise little or no revenue, are costly in terms of tax administration, or that are virtually impossible to enforce. 104 Education Public Expenditure Review Guidelines Example 6: Analysis of education financing by level of government and intergovernmental fiscal transfers Indonesia PER (2015) Overall comment: The Indonesia example provides a focused and detailed analysis of the intergovernmental transfer system and an overview of the education-funds flows in the country. In Indonesia, district governments are responsible for managing the two main assets at the primary and secondary education levels: schools and teachers. Legally, primary and secondary schools are owned by district governments. In fact, when it comes to budgets, the school’s legal status is similar to that of a district government department. Similarly, civil service teachers are legally district government employees, although the hiring process, like that of other civil servants, depends on a number of central government ministries. Provincial governments have very limited authority when it comes to schools, mostly coordinating districts at the basic and secondary levels of education, including with regard to staff development and the provision for education facilities. The central government formulates policy, issues regulations/guidelines and standards at the national level, and still directly controls higher education. Schools have considerable autonomy over operational, budgetary and programmatic decisions. Since 2003, School Based Management (SBM) has applied to all stages of formal education. A degree of decision-making power and management have thereby devolved to the school level, taking account of local norms and encouraging community involvement. The funding system for the education sector is complex, involving multiple sources and transfers across various levels of government. Expenditures for education come from central government funds, transfers to subnational governments, subnational governments’ own-source revenues, and central government spending at the subnational level that is not recorded in subnational budgets. Currently, schools receive funds from eight different sources and four different budgets, including the national, provincial, district and school budgets (Figure 3). 105 Figure 3: An overview of the complexity of transfers and fund flows, 2012 Fund Originated from MoD Fund Originated from MoEC Sectoral Budget Fund Originated from Provincial Budget Fund Originated from District Budget Source: Elaboration based on Permendagri 13/2006 on Guidelines of Subnational Financial Management, World Bank (2009) and Law 22/2011 on AP3N 2012 Note: Adjustment Fund also includes the local incentive grant (Dana Insentif Daerah, or DID) Central government transfers are the main source of revenue for district government budgets (APBD). Central government transfers to subnational governments have more than doubled in real terms since decentralization, accounting for 88 percent of district budgets and 44 percent of provincial budgets in 2009. While the majority of transfers are not earmarked – making it impossible to determine exactly what they are spent on – transfers are estimated to finance about 90 percent of subnational spending on education, and 60 percent of the total national education budget. Subnational governments receive many types of transfers for education spending. The main transfer to subnational governments is the General Allocation Fund (Dana Alokasi Umum, DAU) block grant, which provides funding for the salaries of district civil servants, including civil service (PNS) teachers. DAU transfers represented about 60 percent of district and 20 percent of provincial budgets in 2009. The DAU is allocated through a two-part formula consisting of the “Basic Allocation” and the “Fiscal Gap” (See Box 1 for details on each transfer type). The Basic Allocation, which is calculated largely based on the salary bill for civil servants in the district or province, implicitly incentivizes civil service hiring. Covering about 72 percent of the salary bill, it accounts for about 45 percent of the total DAU. Fund Fund Adjustment Adjustment Tugas Tugas Central Dekon Central DAU DAU DAK DAK BOS BOS Pembantuan Dekon Pembantuan Functions Functions Own Own PROVINCIAL PROVINCIAL SourceSource BudgetBudget DistrictDistrict Revenue Provincial Provincial Education Education Central Central Government Government Revenue DAU/SDA DAU/SDA O ce O ce Agencies Agencies in the Regions in the Regions PAD PAD 106 Education Public Expenditure Review Guidelines Own Own BudgetBudget DistrictDistrict DISTRICT Source DISTRICT Source DAU/SDA DAU/SDA Revenue PAD PAD District District Education Education Revenue O ce O ce DAK DAK Fund Fund from from Originated Originated MoF MoF Fund Fund from from Originated Originated MoEC MoEC Sectoral Sectoral Budget Budget Fund Fund from from Originated Originated Provincial Provincial Budget Budget Fund Fund from from Originated Originated District District BudgetBudget In 2009, districts accounted for 50 percent of total PRIVATE national PRIVATE SCHOOLS PUBLIC SCHOOLS education PUBLIC SCHOOLS expenditures, while provincial SCHOOLS governments accounted for only 8 percent. Salaries are a major component of district spending. When salaries are excluded, education spending is still largely centralized. The central government based on based Source: Elaboration Source: Elaboration Fund alsoFund Note: Adjustment Note: Adjustment on Permendagri Permendagri alsothe includes 13/2006 on includes 13/2006 local incentive on Guidelines Guidelines the local incentive grant grant (Dana (Dana Insentif Financial Financial of Sub-National of Sub-National Insentifor Daerah, Daerah, Management, Management, DID) or DID) World World Bank Bank (2009) (2009) and and Lawon Law 22/2011 22/2011 on AP3N 2012 AP3N 2012 controls the majority of the non-salary budget at all levels of education, from the 70 percent in ECD to 99 percent at the university level. Even in basic education, almost 90 percent of non-salary spending still occurs at the central level. What brings down the overall average to 67 percent is unclassified spending at the district and provincial levels. Figure FIGURE FIGURE 12:12: 12: BYby Spending SPENDING SPENDING BYlevel LEVEL LEVEL OFof OFgovernment, GOVERNMENT, GOVERNMENT, 2001-2009 2001 2001 2009 2009 250 250 4 47 76 66 65 55 54 46 68 8 100 100 90 90 200 200 80 80 70 70 IDR TRILLIONS 2009=100 IDR TRILLIONS 2009=100 65 65 66 66 59 59 65 65 55 55 48 48 55 55 59 59 50 50 150 150 60 60 PERCENT PERCENT 50 50 100 100 40 40 30 30 50 50 20 20 10 10 31 28 35 29 40 47 41 34 41 31 28 35 29 40 47 41 34 41 0 0 0 0 2003 2003 2001 2001 2002 2002 2005 2005 2004 2004 2007 2007 2006 2006 2009 2009 2008 2008 2002 2002 2001 2001 2003 2003 2004 2004 2005 2005 2006 2006 2008 2008 2007 2007 2009 2009 CentralCentral District Source: Source: World World Bank sta Bank based on based sta estimates estimates on state state budget and data budget data Central Regional Financial and Financial Regional information information system data system data District Province District (Sistem Informasi (Sistem Informasi Keuangan Keuangan Daerah, Daerah, SIKD), SIKD),of Ministry Ministry Financeof Finance Province Province Source: World Bank staff estimates based on state budget data and Regional Financial information system data (Sistern Informasi Keuangan Daerah, SIKD), Ministry of Finance Figure FIGURE FIGURE 13: 13:13: NON NON SALARY SALARY Non-salary EDUCATION EDUCATION education EXPENDITURE EXPENDITURE expenditure BY PROGRAMS BY PROGRAMS by programs and of AND level AND OF OF LEVEL LEVEL GOVERNMENT, GOVERNMENT, 2009 2009 government, 100 100 2009 100 100 2 2 10 10 11 11 12 12 90 90 23 23 9 9 80 80 15 15 80 80 6 6 70 70 60 44 44 60 60 60 PERCENT PERCENT PERCENT PERCENT 100 100 51 51 50 50 40 40 75 75 40 40 73 73 30 30 20 20 47 47 20 20 25 25 10 10 0 0 0 0 EARLY EARLY SECONDARY BASIC BASIC SECONDARY UNIVERSITIESOTHERS UNIVERSITIES OTHERS 2008 2008 2009 2009 2008 2008 2009 2009 CHILDHOOD CHILDHOOD EDUCATION EDUCATION EDUCATION EDUCATION EDUCATION EDUCATION TOTALTOTAL NON SALARY NON SALARY Central Central Central Central Central-80S Central-80S District District District District Province Province Province Province Central Central-80S Central District Province Sources: MoF District Province 107 Box 1: An Overview of Current Transfer Mechanisms in the Indonesian Education System From the Central to Subnational Governments This box provides a brief description of the objectives and means by which the various transfer mechanisms from the central government to subnational governments within Indonesia are determined. These transfers represent the major source of financing for subnational governments and thus, to a large extent, explain the level and composition of their spending. General Allocation Fund (Dana Alokasi Umum, DAU) The DAU, according to Law No. 33/2004 Article 1 (21), is a discretionary block grant sourced from the Central Budget (APBN) and aims to equalize the fiscal capacities of subnational governments. It is transferred monthly and directly from central to subnational governments. The DAU is allocated based on a national formula and is the sum of a basic allocation (a portion of the subnational budget for public servant salaries) and the”fiscal gap” (the difference between the estimated fiscal needs and fiscal capacity) of the subnational government. The basic allocation accounted for about 45.5 percent of the DAU in 2010. Fiscal needs are based on regional variables such as population, area, GDP per capita, and the human development index. Fiscal capacity is measured by a region’s own-source revenue and a fraction of total revenue-sharing. Based on Government Regulation No.55/2005, provinces only receive 10 percent of the total DAU, while districts receive 90 percent. Specific Allocation Fund (Dana Alokasi Khusus, DAK) DAK is an earmarked grant allocated to finance specific investment expenditures that are aligned with national priorities and carried out under the jurisdiction of subnational governments. The DAK cannot be used for research, training, administration, or official travel. In 2011, 19 economic sectors received DAK allocations including education, health, agriculture, forestry, trade and various infrastructure sectors (road, irrigation, water, sanitation, rural electricity, housing and local government and remote areas infrastructure). Education is a key priority for DAK spending, with about 40 percent of DAK transfers allocated for education and used primarily for school rehabilitation and quality improvement. The DAK allocation has a formula component that takes into account the fiscal gap and has a 10 percent matching requirement. DAK is transferred in three tranches: the first is allocated after the budget is submitted to the central government; the next two depend on the depletion of the previous tranche. Although DAK is earmarked to fund capital spending, the government allowed some routine maintenance expenditures. Revenue Sharing Fund (Dana Bagi Hasil, DBH) Unlike DAU, which is a horizontal equalization grant, DBH is a vertical equalization grant which consists of revenue sharing from natural resources and taxes. Local governments are obliged to use 0.5 percent of their receipts from the natural resources part of DBH on basic education.” DBH represented approximately 20 percent of total subnational government revenues in 2009. Special Autonomy and Adjustment Funds Special Autonomy Funds include specific grants for Papua, Papua Barat and Aceh (Dana Otsus) and Special Adjustment Funds (Dana Penyesuaian) which include additional allowances for teachers, such as professional benefits for certified teachers and for uncertified civil service teachers, a School Operational Assistance program (Bantuan Operasional Sekolah, or BOS), and local incentive grants (Dana Insentif Daerah, or DID) for education. Central government spending at the subnational level not recorded in subnational budgets (APBD) De-concentration (Dekon) and Co-Administered Tasks (Tugas Pembantuan, TP) Dekon and TP funds originate from the central government’s budget (APBN), and are administered by the provincial Dinas. The funds cover a variety of projects and activities, including school and classroom reconstruction and school quality improvements, social assistance programs (which included BOS until 2011) and capacity building programs for civil servants. Sources: MoF and various laws 108 Education Public Expenditure Review Guidelines Example 7: Analysis of school budget by financing source Philippines PER (2015) Overall comment: This PER investigated the effects of school-based management (SBM) in basic education in the Philippines. Among other questions, it used survey data from three provinces to assess whether SBM alters the resources available at the school level. Using the school-level financing data generated by the survey, the analysis addressed these questions: How has the resource situation at the school level changed in recent years? What is the financial resource situation of schools and what use is made by schools of SBM grants? What resources are schools able to mobilize in addition to Department of Education (DepED) transfers? Does the SBM grant act as a catalyst for the school to access other resources? What uses are made by the school from existing sources? What is the variation in pattern of uses from different sources? What is the relationship between resource allocation and school performance, and what can be done to improve the efficiency and equity of resource allocation? What has been the resource allocation trend in the past few years? What do we know about the equity of resource allocation across schools? What factors contribute to inequality in resource distribution? What is the role of Local Government Units (LGUs) in supplementing the financial resources available to schools? Conclusions assembled from the PER The analysis of school survey data collected in three provinces shows several interesting observations related to the improved availability of financing at the school level. First, the proportion of budget raised by LGUs and communities has been increasing in our sample schools, which is consistent with the expansion of SBMs. There does appear to be some evidence of the so- called ‘fly – paper effect’ where central grants stick to the recipient and overall resource position of the school improves. Parent-Teacher-Community Association (PTCA) funds appear to be higher in the year following a school receiving a SBM grant. Second, the gap between planned and actual budgets looks smaller among schools that perform well in SBM. Financial decentralization to schools has doubled, but at PHP 450 per pupil per year for elementary schools and PHP 965 per year for high schools, school level funds account for only about 5% of overall basic education spending. The sample survey data from 2010 indicates that high schools received about PHP 500,000 from various sources and elementary schools received an average of PHP 134,000 pesos. About 60-70% of these resources come from DepED in the shape of annual capitation grants for maintenance and operational expenses and occasional SBM grants. DepED should consider accelerating the pace of resources transferred to the school level. 109 Analysis behind the conclusions Survey data reveals that school level resources have increased indicating growing school financial empowerment but not substantial school level financial decentralization. An increased share of resources is managed at the school level compared to total national government spending on basic education in the last five years. Funds managed at the school level have grown in absolute and per pupil terms during the last five years. School level resources were almost double among the sample high schools at PHP 965 per student in 2010 when compared with only PHP 449 per student among the sample elementary schools (in nominal terms). Average real per student school-level managed funds doubled from just under PHP 200 in 2007 to nearly 400 pesos in 2010 in the survey sample schools (in constant 2005 prices). Nevertheless, compared to the national average spending per pupil, school level managed resources remain a small proportion of 5.4 percent in 2010. In the sample schools, the overall size of school managed funds has grown from 2006 to 2010 even as Department of Education (DepED) share of school level resources has not increased substantively in the survey schools. It is important to understand how much of these increases in school level managed resources represent a deliberate trend towards financial decentralization by DepED versus a voluntary increase in contributions towards capital expenditures, teacher salaries and various Maintenance and Other Operating Expenses (MOOE) by local government units (LGUs), Parent Teacher Community Association (PTCA) and the community. If this trend were to represent the results of a deliberate strategy by DepED to give increased financial autonomy to schools, the role of DepED grants including school MOOE, School Based Management (SBM) and School Based Repair and Maintenance Scheme (SBRMS) grants would increase over time. Table 3.3 illustrates that DepED resources have hovered around 70% of total school level resources. While the share of resources the school is able to mobilize from local and community sources as compared to what it receives from DepED in terms of MOOE, SBM and SBRMS grants has fluctuated slightly in the last five years, it has always been between 25 to 30 percent of total school resources (except 2010). Table 3.3 Sources of school level funds (mean values in constant 2005 PHP) 2007 2008 2009 2010 2011* DepED 53,000 70% 71,000 76% 98,000 73% 124,000 59% 119,000 73% PTCA 15,000 20% 15,000 16% 21,000 16% 31,000 15% 21,000 13% LGU 2,000 3% 3,000 3% 6,000 4% 28,000 13% 12,000 7% Community - - 1,000 1% 2,000 1% 7,000 3% 6,000 4% Others 5,000 7% 2,000 2% 9,000 7% 20,000 10% 4,000 2% Total (%) 76,000 (100) 93,000 (100) 135,000 (100) 209,000 (100) 162,000 (100) Source: 3D-SFSD. *2011 data not complete. ** Figures have been rounded off to the nearest ‘000 110 Education Public Expenditure Review Guidelines The financial role of LGUs in basic education has been growing in keeping with DepED policies to enable deeper partnerships between DepED and local government units. Table 3.3 indicates that in 2010, the share of non-DepED school level funds rose to a high of 41 percent of total school managed funds. This was driven by LGUs contributing a substantially greater amount of 8.9 percent of per pupil school level funds (up from 6.6 percent in 2007). A possible hypothesis to explain this is that 2010 was an election year, causing local governments to spend more on basic education as a strategy to win votes. It remains to be seen whether LGU funding for basic education as a whole and as a share of total basic education funding continues to increase in real terms post 2010. Of the four major sources of funds that the school can manage, funds contributed by the parent-teacher association (PTCA) form the highest share every year, although the percentage varies between 33 to almost 75 percent of school mobilized non-DepED funds. Nevertheless, about one-third of the sample schools either did not receive any PTCA funds or did not maintain records of these funds in 2010. Further investigation is required to determine if this is an effect of the no- collection policy issued in 2009 which was strictly enforced by DepED from 2010. The surveyed schools did receive some in kind resources in 2011 but the estimated value of these resources was very small at only 0.7% of total funds received by the school. Data on in-kind resources were collected only for the year 2011 and the analysis shows that the only source from which in kind resources form a bigger portion of their contribution was donors, where in kind resources were 18% of total donor contributions. When data about municipal per capita income is included, we find that the richest municipalities received significantly more at 4.5 times the amount in terms of the value of in kind resources as compared with the schools in the poorest municipalities. However, the value of in kind resources received by schools with higher SBM levels of implementation was not significantly different from that received by schools with lower SBM levels of implementation. The proportion of schools that receive the SBM grant annually is growing but still remains small. In 2011 only 20 of the 150 survey schools received the grant (only 2 schools received it in 2007). Three of the 150 schools received the SBM grant in two years. No sample school received the grant more than twice. The percentage of funds represented by SBM grants for those schools that did receive the grant has remained at about one-third of total school level funds in the last three years. The number of schools that received school MOOE grants has increased substantially in the last five years from merely 13 in 2007 to 115 schools in 2011. The percentage of funds represented by school MOOE grants for those schools that did receive the grant has remained at about 60 percent of total school level funds in the last three years. All the schools have school MOOE allocation, and the allocation per school is posted in the DepED website. However, during school visits in the implementation support for National Program Support for Basic Education (NPSBE2-SPHERE), it was observed that there were still schools which: a) opted not to get their MOOE allocation but instead requested their Division Office to provide the Division-procured supplies based on their list of requirements or requested Division to pay their utility bills directly; b) did not request their allocation because of large unliquidated cash advance; or c) the Division decided not to provide the allocation for the same reason. Central-80S District District Province Province 111 An analysis of the sources and uses of funds data for 2011 shows that a greater proportion of school level resources for survey schools that ever received a SBM grant was from LGUs and PTCAs. We can see from Figure 3.2 below that schools that ever received the SBM grant received an average of 26,331 pesos from LGUs in 2011. The average LGU funds in 2011 received by schools that were never SBM grant recipients were much lower at about 7,000 pesos. SBM grant recipient schools were also significantly more likely to raise higher resources from PTCAs in 2011- an average of 49,700 pesos compared with an average of 17,000 pesos for schools that never received the SBM grant. However, SBM grantees mobilized fewer resources from community and other sources compared with non-SBM grantees. FIGURE 3.2: SOURCES OF FUNDS: 2010 Figures 3.2: Sources of funds: 2010 100 6.9 16 10 80 15 29 60 20 40 20 48 54 0 Non SBM Grantee SBM Grantee PTCA PTCA LGU Source: 3D-SFSD LGU Others Community Others Community APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES Share of total household expenditures, by region Source: EMOP 2014 2% 15% Bamako Koulkoro Koyos 11% Gilo 90 80 70 60 PERCENT 50 40 30 20 10 112 Education Public Expenditure Review Guidelines 0 2008 2009 2008 2009 TOTAL NON SALARY Central District Province Example 8: Analysis of household surveys on private spending Mali PER (2016) Overall comment: This example provides a comprehensive and broad overview of household spending on education. It includes information on the share of household spending as part of total education expenditure in Mali; international comparisons on private spending in education (as a share of gross domestic product) based on UNESCO Institute of Statistics data; share of household budget spent on education by region; and the composition of household spending on education by region and socioeconomic 6.9 group. The analysis at subnational level reveals significant variance in how much households spend 10 on education and how they allocate this budget. 29 Household data suggests that in 2014, households in Mali spent approximately CFA 28 to 31 billion on education expenditures, or approximately 8 to 10 percent of all education expenditures.a Private spending on education, estimated in this manner, is roughly equivalent to 0.6 percent of Malian GDP 54 (and this estimate is identical to UIS estimate from 2009). To put this number in context, UIS reports private sources account for 2.4 percent of GDP in Benin, 2 percent in Burkina Faso, 1 percent in Burundi, antee 0.3 to 0.4 percent in Malawi and Niger and 1 percent in Guinea. Households generally spend about 1 percent of their budgets on education, and 65 percent of this expenditure goes to pay for tuition and school fees. How much households spend on education and how they allocate this budget varies greatly from region to region, reflecting both income disparities and disparities in access. Households in Bamako collectively account for half the private spending across entire Mali; households in Mopti, Gao, and Tombouctou incur only 6 percent of all private expenditure, when combined together (Appendix Figure 29). This skewed spending is the result of many disparities including more children, more schools and more income in Bamako. The L HOUSEHOLD EXPENDITURES disparities are also apparent APPENDIX in the share 29: FIGURE of household spending SHARE OF TOTAL on education. HOUSEHOLD In Bamako, households EXPENDITURES allocate 1.75 percent of their budgets to education (twice the national average) whereas in Mopti and Share of total household Source: EMOP 2014 expenditures, by region Tombouctou, this share is only about one tenth of a percent. 2% Appendix Figure 29: Share of total household Bamako 15% Koulkoro expenditures, by region 50% Bamako Bamako 2% Mopti Koyos Koulkoro Gilo 2% Gao Koyos 11% Segou 50% 11% Segou 50% Gilo 10% Kayes 15% Sikasso Mepti Segou Silcasco 2% 10% Koulikoro Mepti 2% Tombouctou Tombouctou 10% Silcasco Tombouctou 10% Source: EMOP 2014 2% USEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 10% 2% 113 FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 Source: EMOP 2014 Figure 52: Share of household budget spent on education, by regions, 2014 TOMBOUCIOU .13% GAO .41% KOULIKORO .51% MOPTI .11% KAYES .50% SEGOU .63% BAMAKO 1.75% SIKASSO .79% Source: EMOP 2014 Tuition and fees constitute the largest share of household education expenditures. In 2014, this category accounted for 62 percent of all education expenditure. Books follow at 29 percent. All other education expenditures (tutoring, supplies, transportation, etc.) account for less than 10 percent. Books and supplies are a much higher share of household expenditures in rural areas FIGURE (where school fees are lower and private schools are rare. Across rural Mali, 53: DISTRIBUTION households allocate OF HOUSEHOLD nearly half of their budgets to books and supplies whereas the similar share in urban areas is only Source: EMOP 2014 26 percent (Figure 53). Not only urban Malians spent nearly twice as much on education (in nominal terms) they also allocated a much larger share of their income on tuition and fees. Once again, thisMAIL is an indicator of not just household capabilities, but also availability of schooling. The urban households include 4% 29% 5% 44% 62% SIKASSO .79% GAO .41%Koyos 5% Koulkoro 50 IDR TRILLIONS 2 10% PERCEN 10% 11% Gilo FIGURE 12: SPENDING BY LEVEL OF GOVERNMENT, 2001 2009 100 Koyos 40 50% Segou GAO .41% 62% 11% 250 100 50% 450 7 6 6 5 5 Gilo 4 6 8 30 20 71% 10 Mepti 90 Segou 31 28 35 29 40 47 41 34 41 2% 200 80 0 0 Silcasco 2% 2001 2002 2003 2004 Mepti 2005 2006 2007 2008 2009 2001 2002 2003 2004 2005 2006 2007 2008 2009 2% 70 IDR TRILLIONS 2009=100 10% Tombouctou 150 10% 60 65 66 59 65 55 48 55 59 Silcasco 50 Central 2% PERCENT KOULIKORO .51% 50 FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 Tombouctou District Source: World Bank sta estimates based on state budget data and Regional Financial information system data 100 40 (Sistem Informasi Keuangan Daerah, SIKD), Ministry of Finance KOULIKORO .51% Province MOPTI .11% 10%Source: EMOP 2014 50 10% 30 20 4% MOPTI .11% 10 31 28 35 29 40 47 41 34 41 FIGURE 67: SHARE OF TEACH 0 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: own calculations using NUPTIK Data, 2010 2% 2001 2002 2003 2004 2005 2006 2007 2008 2009 FIGURE 13: NON SALARY EDUCATION EXPENDITURE BY PROGRAMS AND LEVEL OF GOVERNMENT, 2009 2% Central 2 District Source: World Bank sta estimates based on state budget data and Regional Financial information system data 100 (Sistem Informasi Keuangan Daerah, SIKD), Ministry of Finance 100 80 SEGOU .63% Province % OF TEACHERS WITH SENIOR SECONDARY 10 11 OR BELOW AS HIGHEST EDUCATION LEVEL 12 90 BAMAKO KAYES .50% SEGOU .63% 80 15 9 23 80 BAMAKO 6 60 1.75% 70 1.75% 60 44 60 FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BYFIGURE REGIONS, 2014 EXPENDITURE BY PROGRAMS AND LEVEL OF GOVERNMENT, 2009 PERCENT PERCENT Tuition 100 ON & Fees EDUCATION 13: NON SALARY EDUCATION 51 50 ITEMS & Other Tutoring 40 114 FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING VARIOUS 40 75 40 FIGURE 52: SHARE OF HOUSEHOLDTOMBOUCIOU Education Public Expenditure Review Guidelines BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 73 2 100 30 Source: EMOP 2014 .13% 100 10 11 20 47 20 Source: EMOP 2014 12 23 90 25 20 Books 9 10 80 15 80 Supplies 6 0 Source: EMOP 2014 70 0 EARLY BASIC SECONDARY UNIVERSITIES OTHERS 60 44 60 CHILDHOOD EDUCATION EDUCATION 2008 2009 2008 2009 0 PERCENT PERCENT 100 EDUCATION TOTAL NON SALARY 51 50 40 75 40 73 30 Central Central 20 Central-80S District 47 20 SIKASSO .79% SIKASSO .79% GAO .41% 0 25 10 District Province Province 0 F EARLY BASIC SECONDARY UNIVERSITIES OTHERS 2008 2009 2008 2009 MAIL RURAL URBAN TOMBOUCIOU .13% CHILDHOOD EDUCATION EDUCATION EDUCATION TOTAL NON SALARY S Central Central KOULIKORO .51% Central-80S District District Province MOPTI .11% Province GAO .41% 4% 3% KAYES .50% SEGOU .63% .13% TOMBOUCIOU KOULIKORO BAMAKO .51% 1.75% MOPTI .11% 8% FIGURE 67: SHARE OF TEA KAYES .50% BAMAKO SEGOU .63% GAO .41% 29% 21% Source: own calculations using NUPTIK Data, 201 private schools as well as upper secondary and tertiary schools that tend to have higher fees. The 1.75% 80 % OF TEACHERS WITH SENIOR SECONDARY FIGURE 3.2: SOURCES OF FUNDS: 2010 OR BELOW AS HIGHEST EDUCATION LEVEL KOULIKORO .51% 44% MOPTI .11% 5% 100 60 FIGURE same thing holds true 53: DISTRIBUTION for households OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS under the poverty line (they spend half their education 6.9 SIKASSO .79% TOMBOUCIOU .13% 16 10 KAYES .50% BAMAKO SEGOU .63% 44% SIKASSO .79% 80 40 1.75% 5% FIGURE 3.2: SOURCES OF FUNDS: 2010 15 62% 71% 60 29 20 budget on tuition, and other half on books and supplies) compared to households above poverty Source: EMOP 2014 100 20 6.9 16 SIKASSO .79% 80 10 40 0 GAO .41% 15 line (for 53:tuition which, and feesOF are 64 percent, see AppendixON Figure 30). 4% 60 29 20 48 54 FIGURE FIGURE 53: DISTRIBUTION DISTRIBUTION OF HOUSEHOLD HOUSEHOLD SPENDINGSPENDING ON VARIOUS VARIOUS EDUCATION ITEMS ITEMS EDUCATION 20 40 0 KOULIKORO .51% Non SBM Grantee SBM Grantee MOPTI .11% MAIL RURAL URBAN 20 48 54 PTCA LGU Source: EMOP 2014 Source: EMOP 2014 Others FIGURE 53: DISTRIBUTIONTuition & Fees SPENDING Tutoring & Other 0 Figure 53: Distribution of household Supplieson various education items spending Community KAYES .50% BAMAKO SEGOU .63% OF HOUSEHOLD Non SBM Grantee ON VARIOUS EDUCATION ITEMS SBM Grantee 1.75% Books PTCA FIGURE 53: DISTRIBUTION OF LGU Source: EMOP 2014 HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Others FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Source: EMOP 2014 Community APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES FIGURE Source: EMOP 2014 MAIL RURAL URBAN Share of total household expenditures, by region Share of total household expenditures, by region EDUCA MAIL RURAL Urban URBAN SIKASSO .79% MAIL Mali RURAL Rural URBAN Source: EMOP 2014 Source: EMOP 2014 MAIL RURAL URBAN Source: own 2% 2% FIGURE 67: APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES Source: own calculati Share of total household expenditures, by region Share of total household expenditures, by region 4% 4% Source: EMOP 2014 3% 15% Source: EMOP 2014 3% Bamako 15% 8% 8% TOTAL SALARY SPENDING FOR BASIC RP. BILLIONS 2% Koulkoro2% Bamako MAIL RURAL URBAN Koyos Koulkoro 29% 29% 11% 21% Gilo 21% Koyos 50% 11% 15% Segou 50% Gilo TEACHER RATIO Bamako Mepti Segou 3% 2% 15% 4% 44% 5% 3% Koulkoro Silcasco Bamako 2% Mepti 4% FIGURE 53: 44% 5% TEACHERS DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Koyos 10% Tombouctou Koulkoro 10% Silcasco 11% 44% Gilo 11% Koyos Tombouctou 5% 50% 8% Segou 50% Gilo 62% 44% 71% 10% 10% 5% 8% 2% Mepti Segou 3% STUDENT 4% Source: EMOP 2014 62% Silcasco 2% Mepti EDUCATION 71% 10% Tombouctou 2% 10% Silcasco 2% Tombouctou 8% 4% 29% 29% 10% 10% 21% 21% 4% 29% 21% FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS 2% 2% FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 Tuition & Fees Tutoring & Other MAIL RURAL URBAN Source: EMOP 2014 Source: EMOP 2014 Books Supplies 44% 44% 5% 5% 5% Tuition & Fees FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY 44% & Other REGIONS, Tutoring 2014 Source: EMOP 2014 Books 44% Supplies 5% 44% 62% 5% 44% 62% 5% MAIL RURAL 71% URBAN FIGURE 67: 62% 71% 71% Source: own calculatio 4%& Fees Tuition TOMBOUCIOU .13% FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Source: EMOP 2014 Tutoring & Other GAO .41% Books 3% TOMBOUCIOU .13% 4% 4% Supplies TOTAL SALARY SPENDING FOR BASIC STUDENT TEACHER RATIO KOULIKORO .51% FIG 8% 4% MOPTI .11% GAO .41% FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING EDU URBAN ON VARIOUS EDUCATION ITEMS KAYES .50% SEGOU .63% BAMAKO MAIL RURAL 1.75% 29% 21% KOULIKORO .51% MOPTI .11% Sourc Source: EMOP 2014 Tuition & Fees Tutoring & Other Books & Fees Tuition Tutoring & Other KAYES .50% SEGOU .63% BAMAKO 1.75% SIKASSO .79% SIKASSO .79% Supplies 44% 5% MAIL 5% Books Tuition 44% RURAL & Fees Supplies URBAN Tutoring & Other Source: EMOP 2014 62% Tuition & Fees Tutoring & Other 71% Books FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Books Tuition & Fees Supplies Source: EMOP 2014 Supplies Tutoring & Other FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Source: EMOP 2014 4% Books Supplies MAIL RURAL URBAN FIGURE 68: ESTIM EDUCATION TEAC MAIL RURAL 4% URBAN 3% Source: own calculations based o 8% Tuition & Fees Tutoring & Other 29% 21% Appendix Figure 30: Household expenditures in education, size and share, different 4% 3% Tuition & Fees 80k Tutoring & Other Books 8% TOTAL SALARY SPENDING FOR BASIC EDUCATION TEACHERS RP. BILLIONS Supplies 44% 5% APPENDIX 30: HOUSEHOLD FIGUREFIGURE 70k EXPENDITURES IN EDUCATION, 29% 21% 53: DISTRIBUTION OF HOUSEHOLD Books SPENDING ON VARIOUS EDUCATION ITEMS 5% 44% 60k Supplies 44% 62% 5% 71% 50k socioeconomic SIZE,AND groups SHARE, DIFFERENT SOCIOECONOMIC GROUPS 44% 5% 40k Source: EMOP 2014 62% 4% 71% 30k 4% 20k 10k Tuition & Fees Tutoring & Other 0 Books Supplies Tuition & Fees 2009 ratio = 18 Tutoring & Other Tuition & Fees Tutoring & Other Books Supplies Books Supplies Below Poverty APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES APPENDIX FIGURE SIZE,AND SHARE, Level DIFFERENT MAIL 30: SOCIOECONOMIC HOUSEHOLD Above Poverty Level IN EDUCATION, EXPENDITURES IN EDUCATION, GROUPS RURAL URBAN FIGURE SIZE,AND SHARE,53: DISTRIBUTION DIFFERENT OF SOCIOECONOMIC HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS GROUPS FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS FIGURE BELOW 53: DISTRIBUTION POVERTY LEVEL OF HOUSEHOLD ABOVE POVERTYSPENDING LEVEL ON VARIOUS EDUCATION ITEMS Source: EMOP 2014 FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS 2% Source: EMOP 2014 Source: POVERTY BELOW LEVEL EMOP 2014 FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ABOVE POVERTY LEVEL MAILITEMS RURAL URBAN FIGURE 68: ES EDUCATION T Source: EMOP 2014 43% Source: EMOP 2014 5% 5% MAIL RURAL URBAN Source: own calculations b BELOW POVERTY LEVEL ABOVE POVERTYEXPENDITURES LEVEL 26% Tuition & Fees Tuition & Fees 80k TOTAL SALARY SPENDING FOR BASIC EDUCATION TEACHERS RP. BILLIONS APPENDIX FIGURE 30: HOUSEHOLD Books IN EDUCATION, 70k 26% Other Tutoring &GROUPS Tuition & Fees FIGURE 45: IMPORTANCE SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC Books 5% 60k 64% International Donor Funds, 2015 CFA 50k 5% Supplies MAIL MAIL RURAL RURAL URBAN URBAN 40k 52% Books 30k Tutoring & Other 2006 2 MAIL 4% RURAL URBAN 20k 26% Tuition & Fees 10k BELOW POVERTY LEVEL 64% POVERTY LEVEL ABOVE 5% Books Tuition & Fees Tutoring & Other Tutoring & Other Tuition & Fees Books Tutoring & Other Supplies CFA 53B 0 CFA 61B Supplies Supplies Books Supplies CFA 158B NO EDUCATION COMPLETE PRIMARY 64% 5%5% COMPLETE SECONDARY COMPLETE POST SECONDARY Tutoring & Other 26% Supplies Tuition & Fees Domestic Funding External Funding Books APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 64% 5% APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, Tutoring & Other SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS Tuition & Fees SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS Supplies Tutoring & Other Books BELOW POVERTY LEVEL ABOVE POVERTY LEVEL Supplies BELOW POVERTY LEVEL ABOVE POVERTY LEVEL No Education Complete Primary Complete Secondary Complete Post-Secondary 5% 5% 26% Tuition & Fees 26% Tuition & Fees Books Books NO EDUCATION COMPLETE PRIMARY COMPLETE SECONDARY 5% Tutoring & Other COMPLETE POST SECONDARY Tutoring & Other 64% 5% FIGURE 14: PUBLIC SPENDING ON EDUCATION LEVEL, SELECTED FIGURE COUNTRIES SORTED 45: IMPORTANCE OF DONOR 64% Supplies International Donor Funds, 2015 CFA Supplies Honduras Social Expenditure and Institutional Review NO EDUCATION 6% NO EDUCATION COMPLETE PRIMARY COMPLETE PRIMARY COMPLETE SECONDARY COMPLETE SECONDARY COMPLETE POST SECONDARY COMPLETE POST SECONDARY Source: World Bank/ICEFI social spending database for CA countries; EdStats for the rest of the countries 2006 2007 APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 3% 100 CFA 53B 1 CFA 61B C SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS 33% 26% Tuition & Fees 2% 1% 90 13 7 CFA 158B CFA 172B NO EDUCATION Tutoring & Other NO EDUCATION COMPLETE PRIMARY COMPLETE SECONDARY COMPLETE COMPLETE PRIMARY POST SECONDARY COMPLETE SECONDARY COMPLETE POST SECONDARY 16 19 19 12 21 Domestic Funding 19 15 BELOW POVERTY LEVEL ABOVE POVERTY LEVEL Books Supplies 12% 809% 12 10 External Funding 39 31 9 70 3% 5% 27 12 32 26 5% 60 26 PERCENT 36 APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 50 8 71% 19 GROUPS SIZE,AND SHARE, DIFFERENT SOCIOECONOMICTuition 77% 40 22 FIGURE 14: PUBLIC SPENDING ON EDUCATION LEVEL, SELECTED COUNTRIES SORTED BY GDP P 66% FIGURE 14: PUBLIC SPEN & Fees Honduras Social Expenditure and Institutional Review 69 Honduras Social Expenditure and Institutiona 56% 5% 30 61 18 Source: World Bank/ICEFI social spending database for CA countries; EdStats for the rest of the countries Source: World Bank/ICEFI social spending data Books 51 Tuition & Fees APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 48 100 48 46 1 FIGURE 100 14: PU7 Tutoring & Other Tutoring & Other APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 42 SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS 20 90 13 7 16 12 33 13 Supplies 19 19 21 90 19 22 Honduras Social Expenditur 10 23 15 80 10 15 Tuition & Fees 12 80 Books Tutoring & Other 39 BELOW POVERTY LEVEL ABOVE POVERTY LEVEL Source: World Bank/ICEFI so Supplies BELOW POVERTY LEVEL ABOVE POVERTY LEVEL 31 9 70 70 31 0 26 12 27 32 60 26 32 This estimate is based a BELOW POVERTY LEVEL on estimated the ABOVE POVERTYaverage LEVEL household expenditure on education from the most recent Enquête Modulaire PERCENT 27 36 60 PERCENT 100 Books Nicaragua Honduras Guatemala Georgia Paraguay El Salvador Colombia Costa Rica Panama APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 50 COMPLETE POST Supplies 8 APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 19 50 8 NO EDUCATION SIZE,AND SHARE, DIFFERENT COMPLETE PRIMARY SOCIOECONOMIC GROUPS COMPLETE SECONDARY SECONDARY 40 22 Permanente etSIZE,AND SHARE, Auprès DIFFERENT Ménages 5% desSOCIOECONOMIC (EMOP) data (2014 last quarter), which covers all parts of the country except for Kidal. GROUPS Tuition & Fees Books Tutoring & Other Tuition & Fees Books 30 48 51 61 48 46 69 42 40 18 30 90 46 51 20 APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 80 Tutoring & Other 33 48 Supplies 20 23 Supplies 10 BELOW POVERTY LEVEL ABOVE POVERTY LEVEL 10 SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS 26% Tuition & Fees 0 0 70 Nicaragua Honduras Guatemala Georgia Paraguay El Salvador Colombia Costa Rica Panama Chile Books Pre-primary and Primary Nicaragua NO EDUCATION COMPLETE PRIMARY COMPLETE SECONDARY COMPLETE POST SECONDARY 60 PERCENT NO EDUCATION COMPLETE PRIMARY COMPLETE SECONDARY COMPLETE POST SECONDARY Secondary BELOW POVERTY LEVEL ABOVE 64% 5% POVERTY LEVEL Tutoring & Other Tertiary and Post-Secondary 50 BELOW POVERTY LEVEL ABOVE POVERTY LEVEL Other 5% Supplies Pre-primary and Primary Secondary 40 Pre-primary and Tuition & Fees Tertiary and Post-Secondary Secondary Books Other 30 Tertiary and Post APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES 26% IN EDUCATION, Tuition Tutoring & Fees & Other Other 20 SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS Books Supplies 10 5% Tutoring & Other APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, Tuition & Fees Supplies 64% 0 115 Example 9: Analysis of donor funding Mali PER (2016) Overall comment: This provides a good example of how to address financial sustainability issues related to donor funding fluctuations. It provides analysis of donor funding before and after the political crisis in the country, examines the role of donor funding (donor activities), and mentions issues related to finding good international or regional comparisons on this source of funding, due to lack of a consistent definition of official development aid (ODA). Mali relies on funds provided by international donors, especially for capital investments. Through 2010, international funding had accounted for nearly a quarter of all budgeted education expenditures. Several international and bilateral donors and NGOs supported the education sector, and the focus of support was largely on school construction, school canteens, teacher training and quality improvements in general, with a strong focus on the south (Table 9). Table 9: Summary of main pre-crisis donor activities Agency Domain of Intervention Geographic Area The Netherlands, CIDA Quality improvement (reading Kati, Koulikoro, Segou, Mopti (Through NGOs and Firms) and writing; textbooks) UNICEF Teacher training, early childhood Bamako, Segou, Mopti, San, development (ECD) support, Kati, Kayes School construction Save the Children, Plan Mali, School canteens, support to Countrywide but inadequate Handicap International, Aga Khan CGS, school construction, ECD to meet all the needs Foundation, Right to Play, Islamic Relief, CRS, BIT, GARDL, JICA World Food Program (WFP) School health and feeding Bamako, Segou, Mopti, San, program Kati, Koulikoro USAID Scholarships Tombouctou, Gao, Kidal Source: Emergency Basic Education Program PAD 116 Education Public Expenditure Review Guidelines Since 2004, donor funding, on average, equaled 0.8 percent of Mali’s GDP. Because there is no consistent definition of official development aid (ODA), it is hard to provide international comparisons. One study puts the average official development aid for education at 5.6 percent of total government expenditures in education across Sub-Saharan Africa but the proportion of ODA in public education resources varies greatly across the region (UNESCO, 2013). For example, in 2008, 72 percent of total education funding in Liberia was financed externally; the comparable share was 42 percent in Ethiopia, 35 percent in Eritrea and Burundi, 26 percent in Niger, 24 percent in Central African Republic, 9 percent in Togo, and 3 percent in Democratic Republic of Congo (all very low income countries like Mali). Donor funding could be unreliable. It completely disappeared in 2012 due the political crisis. Before the crisis, donors had committed upwards of CFA 350 million to PISE III, but according to budget records, only CFA 262 billion, or 85 percent of these funds were received (See Box 4 on the importance of external funds in Mali’s capital improvement plan). After the coup d’état and suspension of aid, donors support to the sector was channeled mainly through UNICEF, NGOs and other direct financing interventions for short-term humanitarian and recovery response. Donor support in education has resumed since the end of 2013, but is not back to normal yet. 2014 and 2015 budgets once again include donor funding, which is expected to pay for approximately 3 to 6 percent of budgeted expenditures (Figure 45). Figure 45: Importance of donor finding, over time Domestic Funding External Funding Source: International Donor Funds, in 2015 CFA 117 Example 10: Analysis of total public education spending Albania PER (2014) Overall comment: This analysis of education expenditure as a share of gross domestic product, total government spending, and per capita spending, provides good international benchmarks based on similar demographics, such as similar shares of school-age populations. Public spending on education is particularly low considering the composition of Albania’s population. About one-third of Albanians are 19 or younger. Table 2.7 shows that Albania compares unfavorably with most other countries with similar shares of school-age population. Even poorer countries like Morocco and Vietnam invest more of their income on education. Public spending per student is among the lowest in Europe even taking into account Albania’s income level. Table 2.7: School-age population and public spending on education Share of Public expenditures on Public expenditures on education GDP per Country population below education as share of as share of total Government capita (PPP) 19 years old (%) GDP (%) expenditures (%) Albania 32.5 2.9 9.5 9,403 Europe and Central Asia Kazakhstan 33.5 3.1 - 13,667 Azerbaijan 32.7 2.4 7.2 10,125 Armenia 29.1 3.1 13.7 8,417 FYR Macedonia 25.0 - - 11,834 Serbia 23.7 4.4 10.6 11,801 Croatia 21.0 4.9 9.9 20,964 Mexico 39.7 5.3 19.4 16,734 Chile 30.8 4.1 19.4 22,363 New Zealand 27.8 7.3 18.7 32,219 OECD Ireland 27.5 6.5 9.8 43,683 United States 26.9 5.6 12.7 51,749 Peru 39.8 2.6 18.1 10,765 Morocco 38.3 5.4 - 5,220 Colombia 38.2 4.4 15.8 10,436 Countries Other Brazil 33.9 5.8 18.1 11,716 Vietnam 33.6 6.8 20.9 3,787 Source: World Development Indicators, UN DESA Population Projections and UNESCO Institute for Statistics Note: Population figure are 2010 estimates. Expenditures are 2010 or latest available and GDP per capita is from 2012. BELOW POVERTY LEVEL ABOVE POVERTY LEVEL 2% 5% Tuition & Fees 26% 43% Books Tutoring & Other 118 Education Public Expenditure Review Guidelines 52% 64% 5% Supplies 4% NO EDUCATION COMPLETE PRIMARY COMPLETE SECONDARY COMPLETE POST SECO 2% 1% Example 11: Analysis of functional classification 6% 3% 9% 12% 33% 26% 5% 5% Honduras PER (2015) 56% 5% 66% 71% 77% Overall comment: This analysis examines education expenditures by functional classification--i.e., level of education or subsector. It provides comparisons with countries at similar stages of economic development and provides international benchmarking for changes in the composition of spending by level of education as the country develops. In Honduras, primary education takes up the of public bulk 45: FIGURE spending IMPORTANCE onFINDING, OF DONOR education, but this share OVERTIME is expected to fall as the country develops. Pre-primary and primary educationa alone represents 51 International Donor Funds, 2015 CFA percent of all educational expenditures, while, on average, a quarter of the budget is devoted 2006 2007 2008 2009 2010 2011 to secondary education and around 15 percent to tertiary education. This distribution is in line with CFA 53B CFA 61B CFA 49B CFA 56B CFA 21B CFA 39B countries at similar stages of economic development, such as Guatemala and Nicaragua (Figure 14). CFA 158B CFA 172B CFA 168B CFA 192B CFA 213B CFA 248B However, countries with a higher level of economic development tend to show more balanced Domestic Funding spending across levels, with their spending on primary education ranging between a minimum External Funding of 20 percent and a maximum of 35 percent. Therefore, as Honduras develops, international benchmarking suggests that public spending on primary education should decrease in relative importance compared to other educational levels, even if public spending on primary education increases in absolute terms. FIGURE 45: IMPOR International Donor Funds, 2015 CFA Figure FIGURE 14: 14: Public PUBLIC spending SPENDING onLEVEL, ON EDUCATION education by level, SELECTED COUNTRIES selected SORTED countries BY GDP PER CAPITA sorted by GDP per capita (circa 2013) Honduras Social Expenditure and Institutional Review Source: World Bank/ICEFI social spending database for CA countries; EdStats for the rest of the countries 2006 100 1 CFA 53B 13 7 10 Pre-primary and 90 16 19 19 12 21 19 22 Primary CFA 1 15 80 12 10 39 17 32 Secondary 31 9 70 26 Tertiary and Post- 2011 12 60 27 32 26 32 Secondary PERCENT 36 CFA 39B 50 8 19 38 Other 42 CFA 40 69 22 61 18 30 51 48 48 46 46 42 Domestic Fu 20 33 35 26 External Fun 10 23 0 a p al Re nd a ile ua Co bia ica a as ay Co or m gi m , ad ur Ch g gu la te ea aR or m na ra Fin nd ua or lv ra lo Ge Pa ca st Sa Pa K Ho G Ni El Source: World Bank/ICEFI social spending database for CA countries; EdStats for the rest of the countries. Pre-primary and Primary Secondary Tertiary and Post-Secondary Other 119 Kenya PER (2014) Overall comment: A functional analysis shows that Kenya’s university subsector receives a disproportionally large share of the current sectoral budget in contrast to primary education. These relative shares raise serious questions about equity and almost certainly about efficiency. Intra-sectoral composition in the education sector can benefit from rationalization in order to enhance efficiency and equity. The education budget allocation in 2014/15 is skewed in favor of tertiary education at over 40 percent of the total sector budget, which compares unfavorably to 26 percent allocated to primary education. A functional analysis shows that Kenya’s university sub-sector receives 39 percent of the current sectoral budget in contrast to 26 percent for primary education (grades 1-8). These relative shares raise serious questions about equity and almost certainly about efficiency. Figure 1.21: Current expenditure composition in education could undermine efficiency and equity Total education sector budget, 2014/15 Primary Education Secondary Education University Education TIVET Other Source: Staff computation based on the National Treasury data a In this report, due to reasons of consistency and comparability, primary education includes the first two cycles of basic education, and secondary education includes the third cycle of basic education (lower secondary education) and high school (upper secondary education). The Honduran educational system is structured according to the following levels: pre-primary education (ages 3 to 5), basic education (1st cycle, ages 6 to 8; 2nd cycle, ages 9 to 11; 3rd cycle, ages 12 to 14), high school education(ages 15 to 17) and higher education (ages 18 to 22). 120 Education Public Expenditure Review Guidelines Example 12: Analysis of education spending by economic classification Honduras PER (2015) Overall comment: This example shows analyses of education expenditure by economic classification (with a particular focus on spending on wages) and provides regional and international benchmarking on the wage bill. The wage bill, accounting for almost 90 percent of the total public spending on education, is strikingly high in Honduras when compared to similar countries. A large wage bill is partly attributed to the high average level of teachers’ salaries, especially after the significant increase in the minimum wage in 2009 (by 63 percent). Figure 16 shows that Honduras spends considerably more on the wage bill than its neighboring CA countries or even other LAC countries with higher income, such as Chile or Colombia. The share of expenditures going to salaries is also much higher than countries with top-class education systems such as Finland and Korea. In 2012, only 2 percent of the total public spending on education went to construction, renovation, rehabilitation and/or non-routine maintenance of the facilities. Other recurrent expenditures accounted for the remaining 8 percent. This picture is quite similar for higher education. Between 2008 and 2011, the share of wages averaged 83 percent of total higher education expenditures. Nevertheless, universities devote a larger share of its budget to capital expenditures, averaging 7 percent for the same years. FIGURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION SPENDING, CIRCA 2013 Figure 16: Wage bill as a percentage of public education spending, circa 2013 Source: SEDUC (2012). O cial data for El Salvador (2011). EdStats for the rest of the countries. 100 90 88 87 82 81 80 80 79 79 % PUBLIC EDUCATION SPENDING 73 70 68 60 57 50 49 49 40 30 20 10 0 Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep Source: SEDUC (2012). Official data for El Salvador (2011). EdStats for the rest of the countries. 121 Example 13: Issues related to budget formation and execution Solomon Islands PER (2011) Overall comment: This example describes and analyzes issues of budget formation and execution in the Solomon Islands. It assesses accountability for the use of resources and results achieved, and recommends options for addressing the identified problems. The PER team found three main issues related to budget formation: a) policy priorities, plans, and budgets are not well linked or integrated; b) budgetary allocations are made with little consultation with the line ministries or citizens and without a feed-back process to facilitate corrections; and c) there is a general lack of consistency over time in allocations to public services. It is not clear how budgets support SIG policy and plans. The Ministry of Planning requests that each ministry develop a four year corporate plan. Ministries are also asked to prepare an annual or operational plan for each year of the corporate plan. These plans lay out in more detail, what will be achieved from the implementation of each activity. These plans typically include a large number of goals, often far more than could feasibly be achieved under present human and budget constraints. National plans set targets for outputs and/or outcomes but these are not generally accompanied by cost estimates. By contrast, the national budgets are organized in terms of inputs and do not include clear expectations for service delivery targets. National plans and budgets are not linked: the budget does not show clearly and consistently what activities are to be delivered and how spending contributes to government policy priorities. Consultations with line ministries have been poor. The only portions of a ministry budget that are systematically deliberated upon are a small fraction of the recurrent budget set aside for new activities,a and the development budget.b Decisions regarding most of each ministry budget are largely incremental, meaning that the bulk of the allocation decisions are percentage increases over the previous year’s baseline allocations – decisions are made with insufficient regard for what ministries should achieve or how much their goals would cost. There are few mechanisms for consultations with citizens. Consultations with citizens and provincial governments are not used well in forming and prioritizing national plans and line ministry corporate plans. The budget process includes no formal processes for hearing citizen input and encouraging debate regarding allocations of resources. Unlike in many other poor countries, there is very limited discussion within media or civil society regarding resource allocation decisions included in the budget. There may be a need to better align budgetary allocations with ministry requirements. There is a general lack of consistency over time in allocations to large, priority public services such as policing, health, or education. 122 Education Public Expenditure Review Guidelines Actual spending does not resemble approved allocations. While execution across some Ministries has been relatively close to what was approved by Parliament, there are more ministries where actual expenditure has very substantially exceeded or fallen short of budgets. This is shown in Table 3 below. In 2009 for example, 6 of 30 budget heads were spent within plus or minus 10 percent of their original approved recurrent allocations. Execution of the development budget is very weak. While SIG contributes only a fraction of the funding for the development budget, Table 3 shows that fraction has been underspent in most ministries. In 2006-09, for example, no more than two ministries spent within plus or minus 10 percent of their approved allocation from the consolidated portion of the development budget. The majority of ministry allocations were under-spent. This is partially the consequence of SIG placing a higher priority on recurrent obligations (notably the wage bill) but it may also a reflection of generally weak capacity for project implementation throughout SIG. Table 3: Budget execution, 2006-09 2006 2007 2008 2009 Deviation in SIG Controlled Resources (% of Budgeted Amount) 0.3 4.8 -11.9 -13.6 Recurrent Spending a Good Execution (Number of budget heads withing ±10%) 10 14 16 6 Poor Execution (Number of budget heads) 5 15 14 24 Over +10% of approved allocation 2 8 2 4 Under -10% of approved allocation 3 7 12 20 Spent without allocation 0 0 0 0 Total Budget Heads with Allocations or Expenditures 15 29 30 30 Development Spending, Consolidated Good Execution (Number of budget heads withing ±10%) 0 2 1 0 Poor Execution (Number of budget heads) 9 13 25 26 Over +10% of approved allocation 1 1 0 1 Under -10% of approved allocation 8 12 25 25 Spent without allocation 0 2 0 0 Total Budget Heads with Allocations or Expenditures 9 17 26 26 Development Spending, Non-appropriatedb Good Execution (Number of budget heads withing ±10%) 3 10 4 5 Poor Execution (Number of budget heads) 14 13 23 20 Over +10% of approved allocation 6 6 10 3 Under -10% of approved allocation 8 7 13 17 Spent without allocation 7 2 1 2 Total Budget Heads with Allocations or Expenditures 24 25 28 27 a Excludes sector budget support for health and education. b The outcome in 2009 reflects changes in the accounting of support for police and justice programs as well as some under- counting in other ministry programs rather than actual under-spending. Sources: MOFT and World Bank staff calculations using most recent database from January 2011. 123 There has been little accountability for the use of resources and results achieved. The public has not been able to see what activities and programs the government was using money for, nor the results achieved, because of the way the budget is presented and because of delays in reporting actual outcomes. Line ministries have not been held to account by Ministers, in part because there is little information about what activities they will deliver or results they will achieve. It is important to deal with the formation, execution, and accountability problems together because they are mutually reinforcing. The incentive to budget well increases when three conditions are met: i) line ministry officials have confidence that they will be consulted; ii) ministry officials have confidence that allocations provided will match what was agreed during consultations; and iii) the approved budget is the final word -- ministers cannot press for changes except in clear cases of emergency. Similarly, the incentive to execute each budget faithfully improves when line ministry officials believe it meets their needs and when they are held accountable for their decisions. a The process also makes it is difficult to know with certainty whether bids for alleged new activities represent genuine new activities or instead represent inflated costs for ongoing activities. b The public expenditure review team received mixed reports as to how much scrutiny development budget expenditures get. 124 Education Public Expenditure Review Guidelines Example 14: Analysis of per student spending Armenia PER (2011) Overall comment: This offers a good example of examining the distribution of educational inputs (including per student spending, average class size, average school size, and student-teacher ratio) by school location and community type. The analysis points to inefficiency in the utilization of educational inputs across schools. The main contrast in educational efficiency within Armenia is between small rural schools and large urban ones. In 2009/10, rural schools had an average of 174 students, with an average class of 13.6 and a student-teacher ratio of 8.0. In contrast, urban schools educated an average of 450 students, grouping them 21.8 per class and 11.4 per teacher. Rural schools have been able to flourish with smaller classes because they have ample funding (rural schools receive 53 percent more per student than urban ones) and do not run into infrastructure constraints (nationwide, rural schools utilize less than half of their building capacity). Similar trends are evident when comparing schools in mountainous/highly mountainous locations to non-mountainous ones, as well as the only schools in small communities versus all other ones (Figure 4.5). The pattern of potential inefficiency is even more evident when viewed through the lens of school size. Nearly two-thirds of Armenia‘s schools have 300 or fewer students. These institutions receive 42 percent of all PCF funding and employ 43 percent of staff in general secondary education. Yet the utilization of educational inputs in these schools is strikingly inefficient. For example, among the 369 schools with 100 or fewer students (27 percent of all schools in Armenia), the average class size is 5.6 with 3.7 students per teacher. These schools utilize only 27 percent of their available building capacity, compared to 56 percent nationwide. 0 E Hond N N Pan Korea,Korea Fin K K Jo El Salv Costa Costa Guate Para Nicar Colo El El Honduras Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland rep FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 125 Source: World Bank calculations based on data from the National Statistics Service (NSS), The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF MPACTS ACTS BY COMMUNITY BY COMMUNITY TYPE,TYPE, 2009/10 FIGURE 4.5: DISTRIBUTION2009/10 OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 Republic of Ameri Source: World Bank calculations based on (a) Average data PCF from the National Allocation Statistics Service (NSS), per Student (b) Average School Size S), The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF (000s AMD) (Number of Students) nter TC) and (ATC) MOFand MOF FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 Source: World Bank calculations 300 Figure 4.5: Distribution of educational inputs by community type, 2009/10 (a) Average based on PCF data from the National Allocation Statistics Service (NSS), per Student 500 (b) Average School Size Republic of Ameri 250 The National Center for Education Technology (000s AMD)(NaCET), the Assessment and Testing Center (ATC) 400MOF and (Number of Students) (b) Average (b) Average SchoolSchool 200 Size Size Republic of Armenia 300 Republic Republic of America of America 150 Republic of Ameri (Number (Number of Students) of Students) 300 (a) Average PCF Allocation per Student 100 250 500 200 (b) Average School Size 50 (000s AMD) 400 100 (Number of Students) 200 0 300 0 150 l s ra ou l us 500 500 n ra ta s 300 0 m y s en ity 200 500 n nt us ou 0 m y us en ity Ru no ou ba 40 om nit 40 Com nit 40 Com nit 100 in Ru ba 40 om nit 40 Com nit 40 Com nit ud n ts ai ou ino a no m tain ud n ts Ur in u St u t Ur u St u 250 t m n ai m n n- nta 50 ou 100 400 n w ol i om w ol i om w ol i om ou n w ol i om w ol i om w ol i om No ou ou 400 400 No ou 200 M ho n C M M ho n C ly M 0 y m m ith n a 300 l m ith n a h 0 - Sc s l i h n Sc s l i l ig ly ou oo 150 s ig u < ly nou oo ra H s u < 300 300 ou alH Ontai Sch ou n s 0 m y en ity Ontai Sch ur an n ly us 200 u y en ity Ru ba in ou n o 100 in ud n ts ou n o b a R m t O ain ud n ts Ur n y St u t a m O in Ur St u t l n n 0 m n- Not nta t ou No n 50 100 ou n 200 200 No ou No ou yM ho n C M M ho n C l M y m ith n a 0 l m ith n a gh - Sc s l i 0 h n Sc s l i i ly ou oo ig < lynou oo lH s 100 100 < ra ou lH us in h n ra s ai ch 0 m y en ity ta S c an nt ly Sus un nlyou y en ity Ru no ba in Ru ud n ts i ou n o b ta in ud n ts Ur u St u ta m O in Ur u St u un s y mo t Oa un 0 m n- Notnta t On 0 0 on Noun On o o No ou yM o < aC o nC yM M < aC ho in C us us l M al al gh hl - Sct ol i ith n ur ur n n ith n s nt us Hi ty s ho 4 n Co uni s ty oy o o g l Hi oo Mou nou nlSc ol in ou nou ly itholoin mm nou ba ba niiN t t in in nle ch i R R o On Sch < a C de un Sn hs ta ta (c) Average Class Size (d) Student-Teacher Ratio t Ur Ur in ith n tu mu St nlyu n yn Sc i i i un un w l i S mm 0 t Om a a y nl cho n-m nta a nl nt ly ount Od m om tO o o u ou 40 No o o yM yM No Sc < l i a C C M m l l 0 12 0 gh gh n- Hi Hi No o 25 10 (c) Average Class Size (d) Student-Teacher Ratio yS N On w cho o y Sho 20 8 6 nl 15 y tO tO 12 NoO 10 4 No 25 10 5 20 (c) Average Class Size 2 8 (d) Student-Teacher Ratio 0 0 15 6 s l us ra ou n l nt us 12 0 m y us en ity ra Ru ba 40 om nit 40 Com nit 40 Com nit in n 4 nt us 0 m y us en ity no ou no no ud n ts 10 Ru ba ta 40 om nit 40 Com nit 40 Com nit Ur u St u i ou ino no ud n ts 25 -m tai ai ta 10 m un Ur u St u 2 ai m un n w ol i om w ol i om w ol i om n- nta 5 o No ou w ol i om w ol i om w ol i om 20 o 8 yM No ou ho n C M M 0 hl m ith n a ho n C M y m Sc s l i 0 hl n m ith n a lHig ly nou oo 6 Sc s l i 15 s u < ig ra ou ly ou oo s On ai ch u < n l nt ly Sus 0 m y en ity ra H ou Ru ba in Ontai Sch n n ly us y en ity ou n no 4 ud n ts Ru ba 10 n ta Ur St u ai ou n o ud n ts -m t O ai un m O in Ur n St u t t (d) Student-Teacher (d) Student-Teacher RatioRation 0 m No n n- Not nta 2 o ou No ou 5 M No ou ho n C M yM ly m ith n a ho n C 0 M h Sc s l i hl n m ith n a 0 ig lynou oo Sc s l i ig lH s < ly ou oo ra ou Onai ch < s n lH nt ly Sus 0 m y en ity ra ou Ru ba ai ch in n nt ly Sus y en ity ou n no ud n ts Ru ba n ta Ur u St u ai ou n o ud n ts 12 12 -m t Oai un m O in Ur n u St u nt t 0 m Noun n- Notnta o On ou yM o ou < aC ho n C M 10 10 yM l < aC ho in C M gh Sc ol i hl ith n n Hi No ith n g o l Hi oo No On ch 8 8 On Sch yS (e) Non-Teaching Sta as % of All School Sta (f) School Capacity Utilization Sc nl ly y tO nl ly 6 6 tO (Student Enrollment as % of No No 4 4 Total Builing Capacity) 33% 2 2 (e) Non-Teaching Sta as % of All School Sta (f) School Capacity Utilization 32% 70% (Student Enrollment as % of 0 0 31% s 60% Total Builing Capacity) al al ou us 30% ur ur n n s s 33% m y 0 m s ty 50% us Sc < l i a C mu us y en ity no29% Mou nou nl c ol in ou nou ba ba t 40 om ntniit in i (e) Non-Teaching Sta as % of All School Sta (f) School Capacity Utilization ni R R aiino o < a C de un ud n ts ta a 32% Ur Ur chol Comtain u St u t 40% i 70% m un un n- nta (Student Enrollment as % of nl cho n-m nta ount 28% o 00 om m n o o 31% 30% ith l in Stu ou 60% ho 4 n Co M M Total Builing Capacity) M 27% m y y hl hl 30% 20% 50% 33% ly ithooin g g Hi Hi 26% No o 29% 10% yS N 32% 40% y Sho l s 70% ra ou 0% an nt us 28% 0 m y us en ity On w w Ru 40 om nit 40 Com nit 40 Com nit 31% in 30% ou ino b yS o ud n ts ta 60% l s Ur n u St u nl ra ou ai ith n mm ith l in mm ith l in mm un n- nta n 27% nt us 0 m y us en ity Ru tO tO ba 20% 40 om nit 40 Com nit 40 Com nit 30% o in o no ud n ts No ou 50% NoO ta m tain o Ur u St u yM ho n C ai No m un M 26% m hl u < a 10% n w ol i om w ol i om w ol i om 29% o Sc us l i No ou 40% ou ig ly o oo s M alH ho n C M o ou ly On ai ch ur 0% n m ith n a nt ly Sus 0 m y en ity 28% h - ba Sc s l i n n ai 30% ig ou n o R w ud n ts ly nou oo s m O in u < alH Ur n St u nt ou n- ot nta Ontai Sch ur n n ly us 0 m y en ity 27% ou ba 20% in ou n o R ud n ts No ou ta m t O in o Ur St u yM ho n C un N a M 26% hl t < a 10% No n o Scus l i No ou ig ly o oo yM ho n C alH s M o u l Onai ch m ith n a ur 0% n nt ly Sus 0 m y en ity no gh - Sc s l i n ba w ai lHi ou n o R lynou oo ud n ts < s m O in Ur n u St u nt Source: World Bank calculations ra based on data from the National Statistics Service ou National Center for (NSS),inthe ai ch n- otnta n nt ly Sus 0 m y en ity ou Ru ba ou n o ud n ts No ou ta m t O in o Ur u St u Education Technology M (NaCET), the Assessment and Testing Center (ATC) and MOF. ou < aC ho in C N n a M y hl t On No n No ou g l Hi w li oo M < aC ho in C M o ly (f) School (f) School Capacity Capacity Utilization Utilization On ch gh - ith n n yS l Sc Hi oo nl ly On Sch (Student (Student Enrollment Enrollment of % of as % as tO Sc y No nl ly tO Builing TotalTotal Capacity) Builing Capacity) No 126 Education Public Expenditure Review Guidelines Example 15: Minimum norms and standards for resource allocation Overall comment: The Belarus and Bosnia public expenditure reviews provide examples of education sector spending norms and standards for resource allocations. In Belarus, these include minimum, per student allocation by level of education, facility conditions, and equipment, etc. The Bosnia example provides specific norms and standards related to staffing. Belarus (2013) In Belarus, several sectoral spending norms and standards govern resource allocation. In particular, services of education institutions should meet approved social standards (Box 4). In addition, the MOE issues certain standards and spending norms. While some of these norms are linked to the number of students, some are related to inputs. For example, in schools with indoor swimming pools, an additional janitor (0.5 of full-time equivalent) is introduced for each 250 square meters of the area of swimming pool subject to cleaning, irrespective of the number of students in the school. These norms further constrain discretion in spending decision at the facility level, while not necessarily leading to better learning outcomes, as shown in the subsequent sections of the PER. Box 4. Social Standards in Education In 2003 the Government adopted social standards for services, including in the education sector. These define the minimal requirements that should be met by educational institutions. Compliance remains an issue in some areas. For example, the number of seats in the pre-school institutions in Minsk rayon of Minsk oblast is still below the standard. Resolution of the Council of Ministers 724 of May 30, 2003, and revision 47 of January 13, 2013 specify the following standards in education: • Number of seats in pre-school institutions to the number of children of pre-school age (85 percent) • Net enrollment rate of 5 year-old children in pre-school education (100 percent) • Minimum per student allocation in pre-school education (2,050,000 BYR per year) • Minimum per student allocation in general secondary education of all types (1,370,000 BYR per year) 127 • Minimum per student allocation in special education institutions for children with disabilities (5.5 million BYR per year) • Minimum per student allocation in vocational schools (3.5 million BYR per year) • Minimum per student allocation in out-of-school training institutions (200,000 BYR per year) • Minimum area of general educational institution per student (8 sq. m) • Areas equipped with facilities for sports (1.62 sq. m per student) • Premises for sports activities (0.5 sq. m per student) • Minimum number of personal computers in general secondary, special and vocational schools (one computer per 30 students or at least one computer lab per school) Bosnia (2012) The Republika Srpska (RS) and each canton in Federation of Bosnia and Herzegovina (FBH) set standards and norms on staffing, defining the minimum, optimal, and maximum class sizes, the number of teaching hours, the number of support staff primarily based on the number of students, recurrent expenses by different teaching system. Table 6.9 shows the staffing standards and norms, which in turn define school budgets in RS. If these norms are strictly applied, for instance, schools with 60 students are expected to form only two classes with 30 students each, instead of three classes with 20 students each. However, the guidelines also note that in exceptional cases, the Ministry can permit a class with less than 18 pupils, if it is the only class in one grade and a combined class has more than the norms. Without detailed analysis of school-level data, it is difficult to form a clear picture as to how these norms are applied in practice, but given that the average school size for RS is very small at 144 (or 16 students per grade on average) (see Table 6.9 below), it is likely that many schools are given an exceptional status. Table 6.9: Factors determining school budget for primary schools 1 No. of classes 2. No. of puplis 2 No. of pupils 1/ 1. No. of classes a No. of pupils in classes 1 18-32 2 33-60 3 61-90 4 90-120 5 121-150 6 151-180 7 181-210 8 211-240 9 241-279 b No. of pupils in combined classes 1 merging two grades up to 18 2 merging three grades up to 12 c No. of pupils in special education classes 1 up to 10 (Table continued on next page) 128 Education Public Expenditure Review Guidelines 3 No. of lessons in a class in accordance with the No. of teachers = Class hours (curriculum) / Teacher’s curriculum weekly work hours (40 hours) 4 No. of lessons planned per class 5 No. of working hours of non-teaching staff in line Coefficients with the Pedagogue 1 for school with 16 or more classes 0.05 per class < 0.5 for school with 16 or less classes Psychologist 1 for school with 24 or more classes May be combined with special education teacher / logo- pedist, or social worker 0.05 per class <0.5 for school with 24 or less classes Assistant principal 0.025 per class exceeding 24 classes or 8 or more branch classes 0.5 for school with 2 or more nine-grade branch classes Librarian 1 for school with 16-32 classes and 5,000 units of literary and non-literary 0.05 per class <0.5 for school with 16 or less classes Assistant librarian 0.05 per class exceeding 32 classes and 10,000 units of literary and non-literary Secretary 1 for school with 16 or more classes 0.05 per class <0.5 for school with 16 or less classes Accountant 1 for school with 16 or more classes 0.05 per class <0.5 for school with 16 or less classes Administrative-finance worker 1 for school with 24 or more classes 0.05 per class exceeding 24 classes Janitor 1 for school with 16 or more classes and an ara between 2,000-5,000m2 0.05 per 100m2 or per each class below 16<0.5 or less than 2,000m2 0.05 per 100m2 or per each class exceeding 32<1 or more than 5,000m2 Night watchman 1 per school facility Heating maintenance mechanic 1 for school with central heating for up to 2,000m2 2 for school with central heating for more than 4,000m2 Transportation 1 for school with no organized transportation Hygiene 0.25 per class Maintenance 1 per school facility 6 Years of past work experience of employees 7 Salary coefficients 8 Labor cost 9 Increase according to working conditions Source: Ministry of Education and Culture of the Republika Srpska, Rules on Primary School Funding. Notes: 1/ In exceptional cases, the Ministry can permit a class with less than 18 pupils if it is the only class in one grade and a combined class has more than the norms. 2/ Schools in mountainous areas and extremely undeveloped municipalities may form smaller class sizes if formed in the “most cost-effective manner”. 129 Table 6.10 summarizes the minimum, optimum, and maximum class sizes and teaching loads by subject in six cantons in FBH. There are three important issues to be addressed. First, it is interesting that all cantons (where data are available) specify the optimum class size, in addition to the minimum and maximum sizes. This custom seems to discourage schools from forming classes beyond the optimum size even if they are below the maximum size. Since there is no clear evidence to suggest that these optimum sizes result in better education than anywhere between the optimum and maximum sizes, by removing the optimum sizes, schools may form slightly larger classes efficiently without harming quality. Second, there is no clear rationale for the wide variations in the minimum class size from 16 (Bosnian Podrinje Canton) to 22 (Tuzla Canton). Cantons which set the minimum class size lower than others could increase it, unless there is clear justification such as geographical constraints. Third, teaching loads are almost the same for all cantons except for Tuzla Canton. By increasing the teaching hours by one or two hours, other cantons would be able to reduce the number of teachers, and therefore the wage bills for teachers, by 5-10 percent. In order to analyze the correlations between costs and quality, more detailed school-level data such as per student spending, wage bills, and student performance are needed. Table 6.10: Standards for class sizes and teaching loads by canton, 2011 Teaching loads (hours), Class size (regular with no combined classes) excl. preparation, correction, evaluation hours Foreign History, National languages, Biology, Min. Optimum Max. geography, languages math, informatics music, etc. chemistry Una-Sana 18 27 35 18 19 20 21 Posavina 17 25 33 18 19 19 20 Tuzla 22 28 34 20 20 21 22 Zenica-Doboj 18 26 36 18 19 20 21 Bosnian Podrinje 16 24 32 - - - - Central Bosnia - - - - - - - Herz.-Neretva - - - - - - - West Herz. - - - - - - - Sarajevo 18 24 32 18 19 20 21 Canton 10 - - - - - - - Source: Ministry of Education of the respective cantons. Notes: - indicates that data are not available. 8 6 4 2 0 l s ra ou an nt us 0 m y us en ity Ru 40 om nit in 0 m y en ity ou ino b no ud n ts ta 40 om nit Ur u St u ud n ts ai m un n- nta u St u m w ol i om o No ou 130 w ol i om M Education Public Expenditure Review Guidelines < aC ho n C M y m hl < aC o nC Sc ol i i th n ig i i th n o H On Sch ch y nl ly tO No f All School Sta Example 16: Analysis of cost of teachers (f) School Capacity Utilization (Student Enrollment as % of Total Builing Capacity) 70% 60% 50% 40% Overall comment: 30% 20% 10% The Kosovo example provides analyses and policy recommendations related to the government 0% 0 m y s en ity reform s on differentiated teacher pay scales, which aimed to improve teacher quality in Kosovo. The 40 om nit ud n ts al u St u ou m ur n nt us 0 m y us en ity ba 40 om nit in w ol i om ou ino R no ud n ts ta Jordan case presents a comprehensive analysis of teacher compensation system and provides good Ur u St u < C ho in C ai m n n- nta ou a w l i om ith n No ou l yM < aC ho n C M m hl examples of different ways to assess whether teacher pay is adequate. Sc ol i ith n g Sc Hi o o On Sch y nl ly tO No Kosovo PER (2014) In Kosovo, spending on wages under the education budget increased by over 25 percent in real terms between 2009 and 2012, taking wages from 85 percent of total spending on basic education (grades 0-9) in 2009 to 92 percent by 2012. Spending on non-salary recurrent items was low in 2012 compared to OECD or regional countries. On average, OECD countries spent 22 percent of education budget on non-salary recurrent items, and about 8.7 percent on capital expenditures.a In Europe, Slovenia spent 19 percent of total expenditures on non-salary items and 8 percent on capital expenses, while Bulgaria and Romania spent 26 percent on non-salary items, and 6 and 4 percent on capital expenditures respectively.b In Kosovo, increases granted to the education sector have been devoted almost entirely to salary increases, and for the most part have not been directed to other quality enhancement investments. GORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP Source: Kosovo BOOST Figure 5.11: Education expenditures by economic category, % GDP 4.5% 4.0% 0.3% 0.2% 0.1% 0.1% 3.5% 0.3% 0.4% 0.1% 0.1% 0.4% 0.4% 0.2% 0.1% 3.0% 0.1% 0.4% 0.5% Capital Outlays 0.1% Capital Outlays 0.5% ubsides & Transfers 2.5% 0.5% Subsides & Transfers Utilities Utilities 2.0% Goods & Services Goods & Services Wages & Salaries 1.5% 3.3% 3.3% Wages & Salaries 2.9% 2.8% 1.0% 2.5% 2.6% 0.5% 0.0% 2007 2008 2009 2010 2011 2012 Source: Kosovo BOOST. 131 Recent changes to the teacher salary structure aimed to improve teacher quality. In October 2008, the pre-university “teachers’ differentiated salary system” came into effect. This was part of a comprehensive teacher licensing and professional development effort intended to improve the teaching system’s ability to attract and retain qualified staff. Prior to the reform, teacher salaries were uniform, differentiated only by the grade taught (i.e., with different salaries for teachers at pre-primary, primary and secondary grades). Under the reform, differentiated pay scales were introduced based on qualifications, grade level and experience. As a result, all teaching staff, other than those who were unqualified and those with less than one year of experience, received a pay increase.c The reform of teachers’ career development and remuneration was a welcome development; however, the politically-motivated increases that followed in 2011 distorted the reform. In keeping with electoral promises, the government increased the “base salaries”d of all teachers by 50 percent in 2011. While the differentiated salary structure was kept, the decision was not aligned with the reform principles of future salary increases being linked with performance and professional development. Moreover, increases were granted to unqualified or beginner teachers, which was not the case with the 2008 increases. As a result, the difference in base salary for an unqualified/beginner teacher and a teacher with bachelor’s degreee was reduced to 18 percent compared with 27 percent in 2008. Table A.19. Teacher salary structure teachers KS (euro, monthly) Current base salary with Actual % Base salary based on % Increase Base salary (in 2007) 50% increase based on Increase qualifications (in 2008) based on qualifications (in 2011) based on qualifications qualifications Pre- Pre- Pre- Primary Secondary Primary Secondary (in 2008) Primary Secondary primary primary primary (in 2011) Unqualified/ 201 216 236 201 216 236 0% 302* 324 352 0% Beginner235 5-year secondary 201 216 236 221 238 260 10% 322** 346 377 7% teacher school 2-year Higher 201 216 236 239 257 281 19% 340 365 398 13% Pedagogical School BA (3-year & 4-year) 201 216 236 255 274 300 27% 356 382 417 18% degrees Masters and Ph D 201 216 236 271 292 319 35% 372 400 436 23% degrees Source: Ministry of Public Administration, Teachers’ Payroll Data July 2013; *Payroll data suggest that unqualified/beginner teachers’ salaries were also raised by 50 percent; ** The 50 percent increase and the percentage increase for each qualification were applied to 2007 base salary. In the short run though, it appears that teachers were incentivized to upgrade their qualifications, albeit through a government subsidized program. When comparing the composition of the teaching force by qualifications in 2008 with data available in 2013, it appears that teachers have invested in upgrading their pre-service qualifications (Figure 5.12). Over this time, there was a large reduction in teachers with Higher Pedagogical School (HPS) education and a corresponding increase in those with a Bachelor degree. This shift can be attributed to the initiative to fund the upgrade of pre-service qualifications of all teachers with HPS to a Bachelor degree by the end of 2015. 132 Education Public Expenditure Review Guidelines FIGURE 5.12: DISTRIBUTION OF TEACHERS BY PRE SERVICE QUALIFICATIONS, 2008 AND 2013 Figure 5.12: Distribution of teachers by pre-service qualifications, 2008 and 2013 12k 10k 8k 2008 6k 2013 4k 2k 0 UNQUALIFIED 5 YEAR 2 YEAR BA 3 YEAR MASTERS AND /BEGINNER SECONDARY HIGHER & 4 YEAR PH.D DEGREES TEACHERS TEACHER PEDAGOGICAL DEGREES SCHOOL SCHOOL Source: Kosovo EMIS Source: Kosovo EMIS Following the reform process and the 2011 increases, teachers’ salaries became comparable with other 2008 2013average teacher’s net salary was €347 per month. This was only about 5.7 percent below the sectors. The f average net salary in the public sector, and about 3 percent below the average salary in health sector.g The best use for any additional funds for teacher salaries would be to provide incentives for improved performance and professional development rather than politically motivated increases.h The National Teacher Licensing Council has developed a professional development and performance evaluation mechanism that will provide teachers with an opportunity to strengthen their qualifications and move up the career ladder. The system, which is already in place, grants a temporary license and a regular license to teachers. Those on temporary licenses (about 14 percent of the current teaching force that are beginner or unqualified) FIGURE 16: TEACHER will need to meet SALARIES qualification RELATIVE TO PER and CAPITA training GDP, 2012 criteria to receive a regular license or risk losing the right on Expenditure to teach. Education At the same time, teachers on a regular license can be promoted through five career grades. i Teachers’ Salaries and Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 Teachers need to take training and receive at least one positive performance evaluation in a period of five years to move from one grade to another.j Additionally, the teacher career reform has made it mandatory for teachers to attend AND 70 at least COUNTRIES percent ECONOMICS WITH of “core trainings”. Linking future COUNTRIES salary AND increases WITH to training and ECONOMICS PER CAPITA GDP OVER USD 20 000 PER CAPITA GDP LESS USD 20 000 education of the teaching force would reinforce these reforms designed to enhance quality. Granting CUMULATIVE EXPENDITURE PER STUDENT IN THOUSAND USD, PPPS 200 2.5 politically-motivated across-the-board increases, on the other hand, would risk undermining them. 180 TEACHERS’ SALARIES RELATIVE TO GDPICAPIA % a OECD, Education at a Glance 2012, Table B6.2. 160 2.0 b World Bank EdStats. c The base salary 140increased by 10 percent to 35 percent depending on qualifications: teachers with (i) 5-year secondary teacher school received a 10 increase; (ii) 2-year Higher Pedagogical School received 19 percent; (iii) Bachelor degree holders received 27 percent; and (iv) Master’s or Ph.D. degree holders received 35 percent. 120 1.5 Additionally, for each year of working experience teachers received 0.003 percent of their base salary. d “Base salaries” are 100the salaries that teachers received in 2007, before any adjustments for inflation or other increases followed. Salary adjustments are not always made with relation to teachers’ current salaries but to their base salary. 80 1.0 e As per the Law on Pre-University Education (2011), all new teachers must now complete at minimum a Bachelor’s Degree to be able to join the teaching force. f This was in 2013 before 60 the wage increase of 25 percent granted in April 2014. g Kosovo Agency of Statistics, General Statistics, Quarterly Bulletin, Average Monthly Paid Net Wages in the Budget sector by year, April 2013. 40 0.5 h In fall 2013, the government had promised another wage increase by 50 percent. In March 2014 GoK decided to increase wages by 25 percent applicable from April 1st 2014, and the decision20 has had a negative impact in implementation of reform for career development of teachers. Furthermore, because the wage increase was done in a similar way as in 2011, the wage structure has pressed further the difference by qualifications pre-service. 0 0 decided to increase wages of all civil and public servants by 25 percent, applicable from April 1st 2014. As noted earlier, i In March 2014, GoK due to time and data constrains, further analysis on the implications of these increases could not be included in this report. Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania j According to Administrative Instruction No. 5/2010, a teacher career license is valid for 5 years. To extend the license, teachers need, at minimum, to have a satisfactory performance evaluation and to have completed at least 100 training hours, of which 70 percent in core in-service training programs and 30 percent in optional courses, in optional courses, over five years. The criteria for advancement to a higher license require at least 300 hours of teacher training over five years. Cumulative Expenditure by Educational Institutions per Students Aged 6 – 15 Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP 133 Jordan PER (2016)a Compensation captures a high share of recurrent spending in the education sector in Jordan, leaving few resources for non-wage inputs. The teacher pay scale and allowances in Jordan reward initial qualifications, seniority, and personal teacher attributes, as opposed to being an instrument for policy makers to incentivize better teaching performance or other desired education sector outcomes. Teachers in Jordan are relatively well-paid, based on comparisons with other tertiary-educated workers in the economy, relative to per capita GDP, and taking teaching time into account. A closer look at recurrent spending under the Ministry of Education (see Table 19) shows that fully 92.3 percent of total recurrent spending was dedicated to worker compensation in 2013, with nonwage recurrent spending amounting to 7.7 percent. The non-wage recurrent spending includes spending on rent, utilities, maintenance, and cleaning – i.e. basic operations of educational institutions – as well as direct teaching inputs such as stationery, learning materials, and textbooks. By contrast, the 2011 OECD average share of compensation in total recurrent spending was 78.9 percent, leaving 21.1 percent for non-wage inputs.b Table 19: Ministry of education recurrent spending, by program (2013) Adminis- Vocational Educa- Special Early Basic Secondary Eradicating Sub- tration and Education tional, educa- childhood education education illiteracy total support social & tion education services physical activities Worker 4.5 2.8 0.0 0.3 0.3 76.4 7.9 0.0 92.3 compensation 4.3 2.6 0.0 0.3 0.3 71.4 6.7 0.0 85.6 Salaries, wages, 0.2 0.2 0.0 0.0 5.0 1.2 6.6 allowances 0.6 0.1 0.1 0.1 0.0 1.8 3.2 0.0 5.8 Social security 0.0 0.0 contributions 0.0 0.0 Goods and Services 1.6 0.0 0.3 0.0 1.9 Assistance Assistance/stipends Other Sub-total 6.7 2.9 0.1 0.4 0.3 78.4 11.1 0.1 100.0 Source: Ministry of Finance. June 2014. General Budget Final Accounts of Fiscal Year 2013, pages 158-174 Looking at the breakdown of worker compensations (see Table 20), overall – across all levels of education – additional allowances constitute the largest share at 36.6 percent, followed by wages of non-classified employees (26.8 percent), and the cost of living personal allowance (24.1 percent). However, since employees’ wages appear divided into three categories – classified, non-classified, and contract employees – arguably they should count together and so constitute 36.2 percent of total compensation. The additional allowance reflects the decision to double the base salary of all education sector employees, beginning in 2013. In effect, since it is an automatic doubling, it is no longer an allowance but constitutes part of the base salary. The largest allowance is therefore the cost of living personal allowance. 134 Education Public Expenditure Review Guidelines The cost of living allowance is an allowance allocated to all staff without preconditions. Other allowances depend on the particular circumstances of the individual, e.g. whether they are married or single; teaching in a remote area; or had to relocate from their home district to another district, governorate, or even region of Jordan. This approach to allowances blunts their usefulness in terms of achieving certain desired results, such as location of teachers in certain areas, or teaching of certain subjects, or motivating certain behaviors of teachers. Table 20: Ministry of education worker compensation, by program (2013) Adminis- Vocational Educa- Special Early Basic Secondary Eradi- Total tration and Education tional, educa- child- education education cating support social & tion hood illiteracy services physical educ- activities ation Worker compensation Salaries, wages, allow- ances Classified employees 17.4 12.7 6.8 8.4 13.8 9.3 Non-classified employees 16.0 23.8 35.0 29.6 27.7 25.1 26.8 Contract employees 1.2 Cost of living personal 18.0 22.1 25.5 27.1 24.7 22.3 0.1 allowance Cost of family allowance 2.1 1.7 0.7 0.1 1.3 1.5 24.1 Additional work 0.7 0.8 55.0 1.0 3.7 1.4 allowance 1.2 Additional allowance 36.4 38.9 37.2 36.4 36.8 33.6 36.6 Other allowances 1.5 0.1 Transportation allowance 4.2 0.2 Transfer compensation 1.0 0.1 Field allowance 0.6 0.0 Employee bonus 0.9 0.0 45.0 1.6 0.0 100.00 0.1 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Finance. June 2014. General Budget Final Accounts of Fiscal Year 2013, pages 158-174 Public education employees fall into three categories, depending on whether they hold a BA (Category 1), a diploma (Category 2), or are technical support staff, e.g. drivers (Category 3). Focusing on Categories 1 and 2, under which teachers would fall, the salary progression takes place both in terms of steps within the same category (seven steps in Category 1 and nine steps in Category 2) as well as by years of service (up to a maximum of fifteen years at the highest step in each of the categories). As a result, the lowest base monthly pay for an entry level Category 1 employee (Step 7) is 150 JD, while the highest pay for a Category 1 employee (Special) with 15 years of service is 593 JD, meaning almost four times as much pay. Similarly, an entry level Category 2 employee (Step 9) earns 125 JD per month, compared to a Category 2 Step 1 employee with 15 years of service who earns 353 JD, i.e. almost three times as much (see Table 22). In many countries, and in most OECD countries, teachers’ salaries increase with the level of education taught.c This is not the case in Jordan. Instead, the initial qualification (BA or diploma) determines the starting salary. 135 Table 22: Education sector staff basic salary scale, 2012 (JD/month) Years at grade level Category Level Grade 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Special 425 437 449 461 473 485 497 509 521 533 545 557 569 581 593 First 1 302 312 322 332 342 352 362 372 382 392 2 260 268 276 284 292 3 228 234 240 246 252 One Second 4 202 207 212 217 222 5 181 185 189 193 197 6 165 168 171 174 177 Third 7 150 153 156 159 162 1 269 275 281 287 293 299 305 311 317 323 329 335 341 347 353 First 2 243 248 253 258 263 3 218 223 228 233 238 4 197 201 205 209 213 Two Second 5 177 181 185 189 193 6 161 164 167 170 173 7 146 149 152 155 158 Third 8 135 137 139 141 143 9 125 127 129 131 133 Three 120 123 126 129 132 135 138 141 144 147 150 153 156 159 162 Source: MOE Human Resources Department In all school systems, teachers’ salaries rise during the course of a career, although the rate of change differs greatly. Since Jordan participated in the 2012 PISA, it is possible to compare its salary scale with other participating countries. Thus, Jordan is among the countries where salaries at the top of the scale are considerably higher than starting salaries – on average for this group, which includes Korea, Shanghai-China, Malaysia, Singapore, Romania, and Jordan, salaries at the top of the scale are 2.5 times higher than starting salaries and it takes between 20 and 40 years to reach the top salary.d By contrast, in Denmark, Iceland, Norway, Slovenia, Sweden, Finland, Germany, the Slovak Republic, the Czech Republic, Spain, Peru, Montenegro, and Croatia, teachers’ salaries at the top of the scale are at most 1.4 times higher than starting salaries. Jordan’s approach to teacher compensation therefore encourages longevity in the sector, since rewards with seniority are substantial. Additional payments based on teachers’ qualifications, training, and performance are also common in OECD countries. In other words, while Jordan uses teachers’ qualifications to distinguish between the base salary for those with a diploma versus a BA, OECD countries tend to use an allowance to reward an initial education qualification that is higher than the minimum requirement. In this manner, the initial qualification is rewarded but its importance in teacher total pay recedes with time as the growth potential for all teachers is the same. Moreover, among the OECD countries, 21 countries offer an additional payment to teachers for outstanding performance, and in 17 of those countries, the decision to award the additional payments is made by the school principal (OECD 2014). 136 Education Public Expenditure Review Guidelines One comparison that is often employed to assess relative attractiveness of teacher pay is between teacher salaries and the earnings of tertiary-educated workers in the economy. Ideally, teachers’ salaries here either refers to actual salary, including bonuses and allowances, for teachers aged 25-64 or to statutory salary after 15 years of experience and minimum training. The comparison is made then relative to full-time, full-year workers with tertiary education. For the OECD countries, teachers in pre-primary, primary, lower secondary, and upper secondary education earned on average 0.80, 0.85, 0.88, and 0.92 times the amount earned by full-time, full- year workers with tertiary education. The country with the highest relative teacher pay was Korea, where pre-primary teachers earned 1.32 times the earning of other tertiary-educated workers in the economy, and all other teachers earned 1.36 times as much. In the case of Jordan, using data from the Jordan Employment and Unemployment Survey and Household Income and Expenditure Survey, Assaad et al. (2014) provide first-job wages for individuals between the ages of 25 and 40 in 2012 who graduated in either commerce/business or information technology (IT) from a four-year higher education institution and live in urban areas. The average monthly wage is found to be JD 342, rising to JD 561 five years later (where the individual did not go on to further higher education). This translates to JD 4,104 in 2012 JDs, or JD 4,334 in 2013.e By comparison, a starting teacher with BA earned 5,460 JD, or 1.26 times the earnings of a commerce or IT graduate working in the private sector. While the comparison is not completely parallel with the OECD comparison described above, it does provide an indication that teachers are relatively well-paid in Jordan. Another method used often to assess whether teacher pay is adequate or not is to compare teacher pay to a country’s per capita GDP. Thus, per capita GDP in Jordan in 2013 was 3,653 JD, while the minimum starting salary for a teacher was 4,860 JD, i.e. 1.3 times the per capita GDP. However, this comparison is more typically done not for starting teachers but for teachers with 15 years of experience and minimum training – separately for lower secondary and upper secondary education. Using the 2012 PISA findings, the OECD average for lower and upper secondary is 1.24 and 1.29, respectively. For Jordan, the ratio is the same for both levels of education and stands at 2.15 – the highest ratio amongst all countries participating in PISA in 2012 (see Figure 16). In other words, a teacher with 15 years of experience in Jordan is earning more than twice the per capita GDP according to PISA data. 137 Figure 16: Teacher salaries relative to per capita GDP, 2012 Expenditure on education and teachers’ salaries Cumulative Expenditure by Educational Institutions per Students Aged 6-15 Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP) Upper Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP) Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), page 96. The above analysis of teacher salaries has focused on statutory pay using available data on the pay scale. However, actual average teacher pay in Jordan may well be quite different, as the average teacher may not have 15 years of experience and possess minimum qualifications. Table 25 provides data on total salaries and allowances at the governorate level. The wage data are not available by type of education sector employee, i.e. whether they are teachers, principals, supervisors, etc. If there are significant differences across governorates in the breakdown across the different types of education sector employees, then taking a simple average will make the comparison across governorates very tenuous. Therefore, 6,800 JD probably represents a good estimate of the average pay for education sector staff in Jordan in 2015 – although not for teachers specifically. Therefore, given that Jordan’s per capita GDP was 3,811 JD in 2014, average education sector pay equals 1.78 of per capita GDP, i.e. relatively high, but not as high as the ratio reported in comparison to other PISA countries. This, in turn, indicates that the average education sector staff in Jordan has fewer than 15 years of experience on the job. 138 Education Public Expenditure Review Guidelines Table 25: Average education sector staff pay, by governorate (2015) Governate Total governate wages Total Staff Average Wage and allowances Amman (incl. MOE) 162,220,932 23,774 6,823 Balqa 43,917,540 6,453 6,806 Zarqa 69,781,632 10,359 6,736 Madaba 21,924,624 3,239 6,769 Irbid 127,378,668 18,241 6,983 Mafraq 58,181,928 8,873 6,557 Jarash 25,154,568 3,672 6,850 Ajloun 20,585,724 3,017 6,823 Karak 47,644,260 6,999 6,807 Tafilah 19,880,508 2,966 6,703 Maan 24,860,952 3,812 6,522 Aqaba 13,679,316 2,066 6,621 Total 635,210,652 93,471 6,796 Source: MOE Human Resources Department Yet another angle to assessing teacher pay in Jordan is the following: in addition to class size, student-teacher ratio, and teachers’ salaries, the number of hours of student instruction and the amount of time teachers spend teaching also affect the financial resources countries need to allocate to education. In Jordan, students spend on average 6 hours per day, 5 days a week in school – adding up to 30 hours of instruction per week. For teachers, the average workload is 24 lessons a week, although there is some variation: for basic education, teachers’ workload is 24-26 lessons a week, each lesson 45 minutes long; and for secondary, the weekly workload is 18-20 lessons, each 55 minutes long. The relationship between class size, student-teacher ratio, student instruction time, and teaching time can be described as: Class size = student-teacher ratio x student instruction time / teaching time per teacher Using available data on average class size and student-teacher ratio, this relationship holds for Jordan using the average student instruction time of 30 hours per week and teaching time per teacher of 18 hours per week (24 lessons each 45 minutes long). Given that the school year in Jordan consists of 195 days (slightly above the OECD average of 180-183 depending on the level of education), the annual teaching workload is provided in Table 27 for basic and secondary levels of education, using the different weekly teaching workloads (and keeping in mind that lessons in basic education are 45 minutes long, whereas they are 55 minutes long in secondary education). As Table 27 shows, annual teaching hours are roughly comparable with averages observed in OECD countriesf. In other words, teachers in Jordan are not teaching above average annual hours so that above average pay is warranted. 139 In most countries, teachers are formally required to work a specified number of hours per week, including teaching and non-teaching time, to earn their full-time salary. Some countries also regulate the time that a teacher has to be present in the school. In fact, more than half of OECD countries specify the time during which teachers are required to be available at school, for both teaching and nonteaching activities, at one or various levels of education. Although teaching time is a substantial component of teachers’ workloads, assessing students, preparing lessons, correcting students’ work, in-service training and staff meetings should also be taken into account when analyzing the demands placed on teachers in different countries. The amount of time available for these non-teaching activities varies across countries, and a large proportion of statutory working time spent teaching may indicate that less time is devoted to activities such as assessing students and preparing lessons. Table 27: Teaching hours per year Level of education Jordan teaching workload OECD average Weekly Annual Annual Basic 26 761 782 (primary) 24 702 694 (lower secondary) Secondary 20 715 655 (upper secondary) 18 644 655 (upper secondary) Source: World Bank SABER 2010; Ministry of Education; OECD 2014 page 474 The Ministry of Education determines teachers’ working time, stipulating that teachers spend the school day on the school premises. Since the school year consists of 195 days with 6 hours of school per day, this amounts to 1,170 hours of teacher working time annually. While such a definition is more favorable than limiting working time only to hours spent directly in the classroom, it does not go far enough in recognizing that lesson planning and grading may take place outside of official school hours. Nonetheless, the resulting teacher working time required at school in Jordan is almost identical to the OECD average for lower secondary (1,173 hours), below the OECD average for primary (1,200 hours) and above the OECD average for upper secondary (1,142 hours). Therefore again, teachers in Jordan are working average numbers of hours annually and receiving relatively high pay. a The external link to the document is expected to become available in June 2017. b OECD, 2014, Education at a Glance 2014, Indicator B6, Table B6.2. “Expenditure by Educational Institutions, by Resource Category and Level of Education (2011)”. c OECD, 2014. Education at a Glance 2014, page 456. d OECD, 2013. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), Programme for International Student Assessment, OECD Publishing, page 95. e The 2013 inflation rate of 5.6 percent is based on Department of Statistics data, accessed at http://www.dos.gov.jo/dos_home_a/jorfig/2013/1.pdf. f OECD, 2014, Education at a Glance 2014, page 474. 180 TEACHERS’ SALARIES RELATIVE TO GDPICAPIA % 160 2.0 140 120 140 Education Public Expenditure Review Guidelines 1.5 100 80 1.0 60 40 0.5 Example 17: Analysis of teacher distribution 20 0 0 Indonesia PER (2013) Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania Overall comment: Cumulative Expenditure by Educational Institutions per Students Aged 6 – 15 The Indonesia public expenditure review displays the magnitude of the misallocation of teachers; 15 years the estimates Lower Secondary Teachers’ Salaries (after of size of the gains that might be realized by a rationalized distribution; and provides policy experience/minimum training relative to per capita GDP options to address these issues. It demonstrates how improving the distribution of teachers can have Lower Secondary Teachers’ Salaries (after 15 years of important effects on efficiency, equity, and quality of education. experience/minimum training relative to per capita GDP Despite the low average STR at the national level, there are vast differences in the availability and qualification of teachers across schools. Student-teacher ratios are very unequal, as are the levels of teacher qualifications. For example, in 2010, student-teacher ratios in primary schools ranged from fewer than 10 to greater than 60 students per teacher. Equally important are the differences in qualification. Wealthier urban areas have a higher concentration of more qualified and experienced teachers. The regional differences in the distribution of teachers by education level are very sharp: richer districts, especially those in Java and Bali, have access to more educated teachers. The share of teachers with a senior secondary or below education is under 20 percent in all districts in Java, whereas in some districts in Papua or Sulawesi, it reaches 60 percent. FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 Figure 66: Share of teachers with senior secondary or less as their highest Source: own calculations using NUPTIK Data, 2010 education by province, 2010 80 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL 60 40 20 0 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Source: Own calculations using NUPTK Data, 2010 DI. Yoguaun Jamalu DI. Yoguaun Jamalu DI. Yoguaun Jamalu DI. Yoguaun Jamalu Dou Jau Dou Jau Dou Jau Dou Jau Jama Tim Jama Tim Jama Tim Jama Tim B B B B Awa Tung Awa Tung Awa Tung Awa Tung Nuggoe Acoh Dulosala Nuggoe Acoh Dulosala Nuggoe Acoh Dulosala Bant Bant Bant Sunatraba Sunatraba Sunatraba 141 Improving the distribution of teachers is a matter of efficiency, equity and quality of education. Making the distribution of teachers more equitable by ensuring that poor and remote schools have an equal share of qualified and experienced teachers might raise overall levels of learning and narrow learning disparities. The student-teacher ratio is the main factor when it comes to district spending on education. Districts with low STRs spend significantly more per student than districts with higher STRs. At the school level the relationship is even stronger. As implied by the trend in Figure 67, an increase in the student-teacher ratio of 5 students per teacher reduces spending per student by about one-third. FIGURE 67: PER STUDENT SPENDING AND STUDENT TEACHER RATIOS AT THE SCHOOL LEVEL, 2010 Source: own calculations using School Based Management Figure 67: Per student spending and student-teacher ratios at the school level, 2010 50 45 40 STUDENT TEACHER RATIO 35 30 25 20 15 10 5 0 0 1 2 3 4 5 6 7 BUDGET FOR STUDENT IN RUPIAH Source: Own calculations using School Based Management The potential for efficiency savings is large. The exact amount of potential efficiency savings from raising student-teacher ratios is hard to calculate. However, using information on the current levels of students and teachers and average teacher salaries, it is possible to estimate the teacher salary bill for different levels of the student-teacher ratio. Holding the number of students and the average teacher salary constant, Figure 68 shows the effect on the total salary bill of raising the student-teacher ratio. Raising the student-teacher ratio FIGURE to67:23, PER for instance, STUDENT a level that existed SPENDING AND the early 2000s, inSTUDENT would TEACHER reduceAT RATIOS the overall THE teacher SCHOOL salary 2010 bill by IDR 15 LEVEL, trillion or Source: own calculations 22 using percent. School Raising the student-teacher ratio to 28 students, a level similar to other lower-middle- Based Management income countries, would reduce the overall salary bill by 31 percent, equivalent to the total amount currently being spent 50 by the Ministry of Education and Culture (MoEC) on university education (IDR 22 trillion). 45 40 STUDENT TEACHER RATIO 35 30 25 20 15 10 5 142 Education Public Expenditure Review Guidelines Figure 68: Estimates of government spending on basic education teachers for student-teacher ratios, 2009 education teachers (Rp. billions) Teacher salary spending as % of Total salary spending for basic 2009 actual value Student teacher ratio Source: Own calculations based on APBDN (2009) The Government of Indonesia (GoI) has made significant efforts to improve efficiency and equity in the allocation of teachers. The GoI has issued various regulations over the last five years that have set standards for school staffing levels. However, the staffing norms associated with each regulation are different and provide different guidance on standards. This has caused some confusion and difficulty in interpreting the regulations and has complicated the monitoring of compliance at the school, district and provincial levels. As the hiring institution, districts also issue their own regulations on staffing norms and these can go beyond the minimum staffing levels implied by national standards. We can use the latest guidelines (the joint decree) to identify understaffed and overstaffed schools and estimate the extent of the reallocation needed for all schools to meet the guidelines. The magnitude of this reallocation can be interpreted as a measure of the inequality in teacher distribution. The guidelines clearly define the minimum number of teachers required in a school according to its characteristics. Using the latest school-level information on teachers and student numbers, it is then possible to identify which schools have insufficient teachers and which have too many. Under the technical guidelines of the joint decree, the number of teachers necessary for current levels of provision is smaller than the existing teaching force. There are large mismatches in the existing stock of teachers in primary and junior secondary schools. At the primary level, the number of teachers required is approximately 100,000 less than existing levels. The magnitude of the reallocation needed to make the distribution of teachers more equitable is massive – 340,000 teachers or about 17 percent of the total teaching force would have to be transferred. Most of this redistribution would involve moving teachers within districts. However, approximately 74,000 teachers would need to be moved from districts with excess teachers to deficit districts in the same province (Figure 70). 143 FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD FIGURE 70: PERCENTAGE Figure NEED 70: Percentage TO BE TRANSFERRED of primary TO COMPY school WITH JOINT teachers that would need to be DECREE NEED TO BE TRANSFERRE transferred to comply with joint decree # of existing teachers to move within districts 30k # of existing teachers to move between districts in same province 30k NUMBER OF PRIMARY SCHOOL TEACHERS 25k NUMBER OF PRIMARY SCHOOL TEACHERS 25k 20k 20k 15k 15k 10k 10k 5k 5k 0 0 Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Riau Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Kalimantan Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung Note: The estimates show the number of teachers currently in schools with excess teachers (according to the joint decree) that could be transferred to take up teaching in schools with deficits in their staffing levels. Class-based, sport and local content teachers are included in the estimates. The estimates include both PNS and non-PNS teachers. Source: MoEC school data (2010) and NUPTK Data (2010) In order to contain the growth of the teaching force, local governments must face the true cost of hiring teachers. The joint decree provides a reporting mechanism to contain teacher hiring excesses. Combining this reporting mechanism with a transparent and improved system of setting quotas for civil service hiring could go some way to improving teacher hiring decisions.a A more direct approach to eliminating incentives for over-hiring would be to break the link between intergovernmental transfers and teacher hiring. The current system for hiring civil servant (PNS) teachers creates strong incentives for local governments to continue to increase the size of their teaching force and accelerate the conversion of contract teachers to PNS status. At present, intergovernmental resource transfers are partly determined by the size of a local government’s payroll. Districts with larger numbers of civil servants receive more from the transfer system. Key to addressing this issue would be the elimination of the link between the size of the civil service and the size of a local government’s General Allocation Fund grant (DAU) allocation. a For example, quotas for teachers could be based on national staffing standards and district school-age populations rather than local government assessments of teacher need. 144 Education Public Expenditure Review Guidelines Example 18: Cost projections Tajikistan PER (2013) Overall comment: This review provides an example of projecting the education costs that would be needed to pay for the implementation of various national policy options, including expanding preschool education, increasing teacher salaries, and accommodating projected enrollment increases in Tajikistan. It estimates the additional spending needs associated with each option as a share of the country’s gross domestic product. The government intends to further increase educational spending. In the short-term, the planned increase will be focused on increasing teachers’ and principals’ wages and investing in school infrastructure. In the long-run, as stated in the National Education Development Strategy Up To 2020, the government aims to increase education spending—up to 6 percent of GDP by 2015 and not less than 7 percent of GDP by 2020. To accommodate various investment needs, higher spending on general secondary education may be warranted. Table 11 shows public spending on general secondary education in 2009 to 2011, which remained stable at around 2.7 to 2.8 percent of GDP. Using 2011 spending as a baseline, Table 12 illustrates potential additional public spending to (i) accommodate projected enrollment increases in the next decade, (ii) expand one-year preschool for all six year-olds, (iii) increase salaries for educational personnel, and (iv) increase instruction hours. This may drive the general secondary education expenditures up for additional 1.2 to 2.7 percent of GDP (in 2011 prices). Any increase in education spending should be considered within the overall government budget envelope with the identification of priority policy interventions and assessment of their efficiency and costs. Sustaining high growth rates would allow for higher per pupil spending without sizable increase in educational spending as percentage of GDP. Table 11: Public spending on general education, 2009-2011 Million TJS Million US dollars 1/ Percentage of GDP 2009 2010 2011 2009 2010 2011 2009 2010 2011 Personnel costs 390.5 451.5 536.9 94.3 103.1 116.5 1.89 1.83 1.79 Goods and services 87.3 109.1 122.7 21.1 24.9 26.6 0.42 0.44 0.41 Other rec. expenditures 1.0 1.3 1.7 0.2 0.3 0.4 0.00 0.01 0.01 Capital expenditures 106.7 113.0 163.3 25.8 25.8 35.4 0.52 0.46 0.54 Total 585.5 675.0 824.6 141.3 154.1 178.9 2.84 2.73 2.74 Source: Tajikistan BOOST v0.4 government expenditure database. Note: 1/ Exchange rates at 1US$ - TJS 4.14 in 2009, TJS 4.30 in 2010, and 4.61 in 2011. 145 Table 12: Additional annual spending needs for various policy options at 2011 prices Million TJS1/ Million US Percentage Dollars1/ of GDP2/ Accommodating projected enrollment increases during the 143-250 31-54 0.47-0.89 next decade New construction of additional 550 classrooms per year 88-158 19-34 0.29-0.53 Renovation of 400-700 classrooms per year 46-83 10-18 0.15-0.28 Recurrent cost to accommodate additional stu- 9.4 2.0 0.03 dents (27,500) per year Expansion of one-year pre-school for six year-olds 98-155 21.4-33.6 0.32-0.51 New construction of 600 classrooms per year over 6 years 97 21 0.32 (3,550 classrooms in total) for 177,000 six year-olds New construction of 250 classrooms per year over 6 years 40 8.8 0.13 (if 2,000 classrooms are available for pre-school) Additional recurrent cost per year of enrolling all 6 year-olds 58 12.6 0.19 (i.e. 177,000 additional children) Increasing salaries 60-270 13-59 0.2-0.9 Increasing salaries for all education personnel by 30 percent 270 3/ 59 0.90 Increasing salaries only for teachers by 30 percent 4/ 180 39 0.60 Increasing salaries for lower categories by 20 percent 5/ 60 13 0.20 Increasing instruction hours 60-120 13-26 0.2-0.4 Increasing instruction hours by 10 percent (i.e. increasing 60 13 0.20 teacher salaries by 10 percent)4/ Increasing instruction hours by 20 percent (i.e. increasing 120 26 0.40 teacher salaries by 20 percent)4/ Source: World Bank Staff estimates. Notes: 1/ Exchange rate at 1 US$ = TJS 4.61 2/ Based on GDP In 2011 (TJS 30.1 billion or US$ 6.52 billion). 3/ The data on personnel costs for all educational staff are available only up to 2011, and no breakdown by type of staff (teachers, school administrators, non-teaching staff, and government administrators) is available. While teacher salaries increased substantially in September 2011 (by 30 percent) and September 2012 (by 60 percent), that was not the case for other staff. Hence, it is assumed that total personnel costs increased by 20 percent (two-thirds of the teacher salary increase) between 2011 and 2012, and another 40 percent (two-thirds of the teacher salary increase) between 2012 and 2013. Based on this assumption, total personnel cost in 2013 is assumed to be TJS 902 million (TJS 537 million or 1.15 x 1.3), which is used as the baseline. 4/ Assuming that personnel costs for teachers include two-thirds of total personnel costs for general education. 5/ Assuming that one-half of teachers will be subject to this increase. 146 Education Public Expenditure Review Guidelines Jordan PER (2016)a Overall comment: The Jordan Response Plan 2016-2018 case provides a good example of projecting education costs to meet specific, national education goals and targets. The analysis provides details on indicators used in the education sector vulnerability assessment and costing analysis in accordance with national norms and standards on class size, number of classes per school, and student-teacher ratio. However, the Jordan Response Plan approach goes beyond abiding by national standards on core education-system indicators. It proposes a number of projects and sector-specific objectives aimed at improving access to quality and inclusive education for Syrian refugees and vulnerable Jordanians. The Jordan Response Plan 2016-2018 estimates education sector needs by carrying out an education sector vulnerability assessment at the district level by using the three indicators of school size, class size, and student-teacher ratio. It defines the national standard for class size at 27, thereby identifying vulnerability to crowding in classes to be most severe in seven districts of the four governorates with high concentrations of Syrian refugees, i.e. Amman, Zarqa, Irbid, and Mafraq. The assessment further finds that 150 new schools would be needed to meet a national standard of 19 classes per school, which translates to a school size of 513 (assuming an average class size of 27). These schools would be located primarily in the same four governorates with highest concentrations of Syrian refugees. Finally, the assessment finds that an additional 8,600 teachers would be needed to meet a national standard of 17 students per teacher. In other words, the MOE is aiming to absorb the Syrian students while maintaining its current class size and student-teacher ratio. In addition, new schools to be constructed are to accommodate above 500 students, which is greater than the current average school size. Beyond abiding by national standards on core education system indicators, the Jordan Response Plan 2016-2018 aims to apply lessons learned from past refugee crises and enhance the Government’s ability to respond to emergencies while at the same time strengthening the education system’s resilience. This resilience implies that the education system is able to adapt and maintain quality in the face of potential new crisis scenarios. Rather than relying on the unit cost approach, the Plan proposes a number of projects that aim to improve access to quality and inclusive education for Syrian refugees and vulnerable Jordanians, boosting the capacity of the public education system with additional learning spaces, remedial/catch-up classes for those children who have missed out on weeks or months of schooling, and access to improved and diversified certified alternative learning opportunities for children and youth. Projects are also to deliver capacity building of teachers to safeguard the quality of education. The resulting three-year Plan starts with a baseline number of 156,663 Syrian children enrolled in education services (whether formal, non-formal, or informal), and targets increases in enrolment to 222,000 in 2016; 248,000 in 2017; and 272,800 in 2018. The increased enrolment is not expected to result from influx of additional Syrian refugees into Jordan, but rather from increased enrolment rates of Syrian children already in Jordan. 147 A range of projects are proposed in the Plan, spanning improving the capacity of education authorities to ensure the continuous delivery of quality inclusive education services; enhancing access to safe and protective learning spaces; and increasing provision of adequate, protective, and safe learning spaces and facilities. In terms of associated costs, over US$850 million are estimated over the 2016-2018 period, with over US$470 million dedicated to facilities, i.e. construction of 150 new schools and additional classrooms or renovation of 450 prioritized schools. The fact that attention is being paid not only to construction of new schools but also to adding classrooms to existing schools is warranted given the relatively small average school size in Jordan. In other words, the Plan recognizes the necessity for a dual approach that is based on detailed analysis of the needs on the ground in specific locations in the country. The next largest project at over US$180 million includes the hiring and pay of teachers. a The external link to the document is expected to become available in June 2017. Democratic Republic of Congo PER (2015) Overall comment: The DRC case briefly analyzes the government’s projection of the budget required to implement the medium-term education-sector strategy. It identifies strengths and weaknesses in the way that the government conducted its cost analysis, and explains potential challenges to the projected medium- term strategy. The medium-term outlook of the education sector strategy does not address the current challenges arising from the financing of the education sector. From the recently adopted 2016-2025 sector strategy, the projected budget still shows high dependence on donors. It was also planned with a significant financing gap, which has not been addressed. However, the projection of the costs based on the new sector strategy reveals both good and bad news. The good news is that the functional classification by level of education is well-crafted which is very promising for sectorial analysis at the monitoring and evaluation stages. And the strategy planning is based on the focused and measurable targets, which is also important. On the negative side, there are three core areas of concern: (i) the ministry of budget needs to establish a clear budget line for each level of education and properly plan according to the budget lines- the lack of clear and consistent budget nomenclature is one of the drawbacks for this analysis, (ii), the projected strategy has not taken private provision of schooling into account both in terms of cost and the human resources needs, and (iii), projected scenarios are missing the demographic aspect of the unit cost, which currently is projected to increase over time in US dollars. The unit cost calculation is also based on the expected funds from outside resources, which may or may not be realized. 148 Education Public Expenditure Review Guidelines FIGURE 40: MEDIUM TERM OUTLOOK OF PUBLIC SPENDING ON EDUCATION BY SOURCES AND UNIT COST PROJECTION Source: Education Sector Strategy; 2016-2025, January 2015 719 120% 709 Figure 40: Medium-term outlook of public spending on education by sources and 676 100% unit cost projection, 2016, 2025 DIUM TERM OUTLOOK OF PUBLIC SPENDING ON EDUCATION 562 529 522 120% 80% ND UNIT COST PROJECTION 455 Strategy; 2016-2025, January 2015 100% 2016 120% 427 428 60% 420 411 394 719 709 80% 2017 100% 314676 40% 80% 60% 2018 279 244 241 562 20% 230 2016 60% 529 2021 522 40% 204 185 165 2017 40% 0% 455 20% 2025 428 427 411 2018 420 394 20% 93 96 71 72 74 0% 37 52 2016 -20% 62 2012 2017 2018 2021 2025 27 33 0% 314 15 13 11 11 9 -20% 2016 2017 2018 2021 2025 2025 279 241 2016 2017 2018 2021 2025 244 230 -20% Preschool Primary Non-Formal 1st Cycle 2nd Cycle Technical Higher 204 185 165 Secondary Secondary and Education 74 93 96 School Internal Internal Resources Resources School Vocational 62 72 Internal 15 13 11 11 9 Resources 27 33 37 52 71 Internal Resources 2018 2021 2025 Development Preschool Partners Non-Formal Primary 1st Cycle 2nd Cycle Technical Higher Development Development Development Partners Partners Partners Secondary Secondary and Education Shortage of Planned Budget School School Vocational Shortage Shortage of Planned Shortage of of Planned Planned Budget Budget Budget rces Partners Source: Education Sector Strategy, 2016-2025, January 2015 anned Budget Based on the projected GDP growth, the government could afford to increase the budget for the education sector especially given that education is one of the top five priority sectors designated by the authorities. Education spending as share of GDP is projected to reach only 3.4 percent by 2025 (Figure 41) which is still below the current SSA average (5.0 percent) and the minimum suggested rate of 4.7 percent. Just as with the unit cost analysis performed in the last sector strategy, the main concern with the projected estimate for the new sector strategy is that the capital spending still heavily depends on external sources (about 44 percent annually) especially given the recent history of low execution rate for external resources. FIGURE 41: MEDIUM TERM OUTLOOK OF PUBLIC SPENDING ON EDUCATION BY SOURCES Table COST PROJECTION; AND UNIT outlook 41: Medium-term 2016, 2025 of public spending on education by sources and Source: Education Sector Strategy; 2016-2025, January 2015 unit cost projection, 2016, 2025 30% 4.0% 3.0% 20% 2.0% 10% 1.0% 0% 0.0% 2016 2017 2018 2021 2025 Education spending as share of total public spending Capital spending as share of total education spending Education spending as share of GDP Source: Education Sector Strategy, 2016-2025, January 2015 149 Example 19: Fiscal sustainability analysis Albania PER (2014) Overall comment: The Albania public expenditure review conducted a long-term fiscal sustainability analysis with specific policy recommendations. Based on the assumption that the country’s fiscal consolidation program will be implemented as planned, it examines the impact of the proposed increases in public education spending on the public debt. In the short to medium term, given the lack of fiscal space and the fiscal consolidation plan in place for the period 2014-16, Albania should carry out reforms to make the social sectors more efficient and do more with the same level of budgetary resources. In the medium to long term, particularly starting with 2017, Albania could consider increasing public spending on education, with a clear recognition of the trade-offs – that increased public spending will lead to a higher public debt-to- GDP ratio (relative to the baseline) but that given the needs in the sector such investments may be worthwhile. The key public spending recommendations with implications for fiscal sustainability related to education are as follows: Increase public spending on education, as fiscal space opens starting from 2017, from the current 3 percent to about 4 percent of GDP, bringing Albania closer to its regional peers in terms of the level of public spending on education. The additional public spending could be channeled to several areas in which Albania still needs investment, such as teachers’ professional development; learning materials and school supplies; quality of school facilities; and, more time on tasks and activities in schools. The assumptions behind a baseline scenario for Albania’s fiscal sustainability are as follows: the current fiscal consolidation program for the period 2014-2016, both on the revenue and on the expenditure side, will remain in place and will be implemented as planned. Under this scenario, Albania’s public debt-to-GDP ratio will fall from 72 percent in 2014 to 69.7 in 2016, including the clearance of arrears of 5.3 percent of GDP over three years. Beyond 2016, the baseline scenario assumes that the fiscal consolidation continues between 2017-2022, with the expenditure to GDP ratio declining from 28.7 percent in 2016 to 25.2 percent in 2022, while revenues staying at 25.2 percent of GDP. Public debt-to-GDP reaches 49.7 percent at the end of 2022 under the baseline. Relative to the baseline scenario, the impacts of the proposed increases in public spending on education and health (from 2017) on the public debt-to-GDP ratio are presented in Figure 6.1. 0% 0% 0.0% 2016 2017 2018 2016 2021 2017 2025 2018 2021 150 Education Public Expenditure Review Guidelines Education Education spending as share of total public spending as share of total public spending spending Capital Capital spending as share of total education spending as share of total education spending spending Education spending as share of GDP Education spending as share of GDP An education spending increase scenario assumes an increase in education public spending from the current 3 percent of GDP to 4 percent of GDP starting in 2017 and continuing throughout the considered period. The effect of this will be to increase Albania’s public debt-to-GDP ratio by 5.2 percent of GDP by 2022. FIGURE Figure 6.1: 6.1: ALBANIA Albania PUBLIC public debt to FIGURE DEBT 6.1: TO GDP GDP ratioRATIO ALBANIA underUNDER INCREASED PUBLIC increased DEBT TO GDP education andRATIO UNDER INCREASED health EDUCATION AND HEALTH sector spending, 2017-2022 SECTOR SPENDING, EDUCATION AND 2017 2022 HEALTH SECTOR SPENDING, 2017 2022 Source: World Bank Sta Calculations Source: World Bank Sta Calculations 75 75 70 70 65 65 63.1 63.1 60 60 56.9 56.9 55 55 54.9 54.9 50 50 49.7 49.7 45 45 40 40 2008 2008 2009 2010 2011 2012 2013 2014 2015 2009 2016 2010 2017 2018 2011 2019 2012 2013 2020 2014 2021 2015 2016 2017 2018 2019 2020 2021 2022 2022 Combine Combine Increase in Education and Health Increase in Education and Health and Spending and Spending Health Spending Increases from 2.6 to 4 percent Health of GDP Increases from 2.6 to 4 percent of GDP Spending Education Spending Increases from 3 to 4 percent Spending Education of GDP Increases from 3 to 4 percent of GDP Baseline Baseline Source: World Bank staff calculation FIGURE 15: EVOLUTION FIGURE 15: EVOLUTION OF REAL EDUCATION OF REAL EXPENDITURES EDUCA BY BROAD E CATEGORY SOURCE: DATA FROM MFB/SI CATEGORY SOURCE: DATA FROM MFB/SIGFP. 700 700 600 600 ANT 2013 AR. ANT 2013 AR. 500 500 400 400 0% 0.0% 0% 2016 2017 2018 2021 2025 2016 2017 2018 2021 Education spending as share of total public spending Education spending as share of total public spending Capital spending as share of total education spending Capital spending as share of total education spending Education spending as share of GDP Education spending as share of GDP 151 Madagascar PER (2015) FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED EDUCATION AND HEALTH SECTOR SPENDING, 2017 2022 Source: World Bank Sta Calculations 75 Overall comment 70 The wage bill in the education sector has to be carefully monitored for its fiscal sustainability and for the 65 presence of “fiscal space” to cover non-wage, recurrent expenditures and capital expenditures required to 63.1 60 complement teachers. 56.9 55 This example from Madagascar assesses the trajectory of the current wage bill and simulates the fiscal 54.9 consequences of a planned Government policy that would integrate community teachers into the civil 50 service. Community teachers are locally hired and employed by parents’ associations. They consist of 49.7 two groups: (i) subsidized teachers, who receive salaries from the State, and also, depending on the local 45 context, additional funding from parents; and (ii) non-subsidized teachers, who receive salaries solely from parents. Integrating these teachers into the civil service shifts their entire wage payment to the central 40 government. 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Combine Increase in Education andThe share Health of regular teachers’ salaries in the budget has been fast increasing, and the wage bill has and Spending Health Spending Increases from 2.6 to 4 percent of GDP reached unsustainable levels. Looking at changes in regular salaries is informative to address issues Education Spending Increases from 3 to 4 percent of GDP Baseline of sustainability because regular salaries are not flexible downward. Using more recent data from the MFB/SIGFP shows that the share of regular salaries is following a fast increasing trend, reaching over three fourths of the education executed budget in 2013 (Figure 15). Although this is still within reasonable limits, the trend is not sustainable. The recent announcement that 10,000 community teachers will be absorbed into the civil service in 2014, and another 10,000 in 2015, are in that sense worrying (see Box 7 for more details). FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC Figure Evolution 15:SOURCE: CATEGORY of real DATA FROM education expenditures by broad economic category MFB/SIGFP. 700 600 BILLION CONSTANT 2013 AR. 500 400 300 200 100 0 2006 2008 2009 2010 2011 2012 2013 Capital Expenditures Capital Expenditures Other Recurrent Expenditure Regular Salaries Other Recurrent Expenditure Regular Source: Data fromSalaries MFB-SIGFP. S: SIMULATING THE IMPACT OF THE ON OF COMMUNITY TEACHERS 600 600 HYPOTHESIS OF 586 RESOURCE NEEDS 573 152 Education Public Expenditure Review Guidelines Box 7. The Impact of the Regularization of Community Teachers as Civil Servants Recent political decisions in the sector include the progressive integration of community teachers in the civil service. Analyses carried out under this PER have aimed at estimating the potential impact on the MFB budget of such integration, using various scenarios. Projections for 2020 show that the number of teachers would reach 96,100 for an expected 5.2 million children in primary school. This would translate into the recruitment of an additional 68,000 civil servants compared with 2014. An analysis of the potential impact of this decision on public spending on education is presented below. The evolution of the macroeconomic context was simulated using two different scenarios, in line with the recent scenarios elaborated by the IMF. More details on the methodology and scenarios for the macro context are provided in the case study developed for this PER. In addition, the analysis examined two different hypotheses for integrating community teachers in the civil service: • Hypothesis 1: all community teachers are integrated in the civil service as early as 2016, and new teachers are hired as civil servants. • Hypothesis 2: community teachers are progressively integrated in the civil service to ensure that all teachers are civil servants by 2020. More specifically, this hypothesis assumes the integration of 10,000 community teachers in 2016, 15,000 in 2017, 15,000 in 2018, 16,000 in 2019 and 16,960 in 2020. The results of projections show that Hypothesis 1 is unsustainable even in the most favorable case Simulating the impact of the of the evolution of the macroeconomic context. integration of community teachers Indeed, from 2016 onwards, the salary needs for primary schools would amount to 550 billion MGA. This amount exceeds the global amount of the 2014 MFB budget, which was 541 billion MGA in 2014, and is 10 times larger than the current amount allocated to community teachers’ subsidies. In the case of Hypothesis 2, the results of projections show that the integration of 10,000 community teachers in 2016 and as many in 2017 would result in salary costs equivalent to about 300 billion Ar in 2016 and as much as 350 billion Ar in 2017. This compares with 286 billion Ar and 318 billion Ar projected for salary costs in 2016 and 2017 respectively. Under both scenarios 1 and 2, the wage bill would amount to 600 billion Ar by 2020, far above the projected resources available 153 for salaries under scenario 1 (504 billion Ar) and scenario 2 (414 billion Ar). This scenario seems hardly sustainable even in the case of favorable economic growth (Scenario 1), and completely unrealistic in case of a slower economic growth (Scenario 2). The results of these simulations show that the integration of community teachers as planned will have large and unsustainable consequences on the MoE budget. To improve its feasibility, it would seem essential to envisage one of the following options: (a) dramatically slow down the progression of integration, or (b) allocate more resources to education. Even in the latter case, this simulation shows that the integration of community teachers into the public service, even if progressive, will result in further increasing the already high weight of salaries in the MoE budget. The interventions aiming to improve the quality of education, such as the improvement of infrastructure, but also teacher training and the improvement of the availability of learning material, could therefore not be financed by the State budget. 154 Education Public Expenditure Review Guidelines Example 20: Demographic trends and enrollment projections Guinea PER (2015) Overall comment: This analysis shows the projected enrollment and education-funding gap based on demographic trends and different gross domestic product growth scenarios. It concludes that the current level of funding is a real constraint given the projected population growth in the country. Guinea has one of the highest levels of population growth in the world: 52 percent of the Guinean population is composed of women, and fertility rates are among the highest in the region at 5.1 live births per woman, and even higher in rural areas at 5.8 live births. The school- age population (ages 3 through 25) comprises almost half the population in Guinea, and grew at about 2.8 percent annually between 2008 and 2013. The current level of funding is a real constraint given the projected population growth. Enrollments have expanded, but the capacity to serve students lags behind what is needed given Guinea’s large and growing school age population. If the education sector continues to receive less than 3 percent of GDP, and 13 percent of public resources—and the country does not improve its resource use—an increasing number of children and youth will be out of school. To illustrate this point, assuming that current gross enrollment ratios will hold steady–9 percent at the pre-primary, 98 percent at the primary, 45 percent at the lower secondary, and 25 percent at the upper secondary levels - enrollments across all levels would have to increase by 620,000 students by 2020 just to accommodate the population growth. If the public sector continues to serve the same share of students, this would mean an additional 414,000 students enrolled in public schools. To support that kind of growth, under current resource use patterns, Guinea’s public education spending on operating costs only must grow by one fourth in real terms, which is only possible under current resource allocation patterns if the economy grows by 3 to 4 percent annually— faster than the population. If growth rates averaged to 2 percent, for example, by 2020, another 124,000 children, or 2 percent of the projected school age population in 2020, would be excluded from education (left panel of the figure below). To serve this additional population—that is, just to keep the out-of-school youth rate at its 2012 level—the funding for operating expenditures would have to grow by GNF 1.1 billion in real terms. That would mean the education sector would have to double or receive 2 percent more of the country’s GDP. 155 Projected enrollment gap and operating expenditure gap due to population growth Enrollment gap Funding gap GDP Growth - 1 percent Operating costs with current funding and GER GDP Growth - 2 percent GDP Growth - 1 percent Projected Public Enrollment GDP Growth - 2 percent GDP Growth - 3 percent GDP Growth - 3 percent GDP Growth - 4 percent GDP Growth - 4 percent Source: World Bank Staff calculations using population data from U.S. Census international databases, expenditure and enrollment data from MEF, and dropout rates from ELEP 2012. 156 Education Public Expenditure Review Guidelines Belarus PER (2013) Overall comment: The Belarus analysis examines the implications of the declining number of students in the country, particularly in rural areas, for the size of the education workforce and school network. Although Belarus has made progress in consolidating its school network in response to demographic trends, it has not seen a commensurate adjustment in the number of teachers, which has remained relatively stable. Student-teacher ratios have declined, particularly in rural areas, with significant implications for per student costs. The education system must adjust to a dramatic decline in the number of students. Belarus’ population is projected to decrease to 8 million in 2050 from more than 10 million in 1995. The school-age population has decreased dramatically during the last 40 years: from 1.9 million in 1970 to 0.9 in 2010 (Figure 75). Just in the last four years, the number of students in general secondary schools declined by 14.1 percent. At the same time, the number of people living in cities has almost doubled during the last 30 years (from 4 million in 1970 to 7.2 million in 2011). As a result, the demand for education, particularly in rural areas, collapsed. Based on current fertility and internal migration rates, population dynamics among school-aged cohorts in urban areas are expected to stabilize over next two decade, while the decline in rural areas is expected to continue. Figure 75: Number of students has declined, especially in rural areas Demographic trends 1970-2030 (thousands) All Ages Age 5-14 Age 5-24 Rural Urban Source: UN, Beistat. In response to these trends, the Government has begun to right-size the school network. Accommodating such large shifts in student enrollment is tremendously challenging for a school system because it involves closing down schools, a politically difficult task in any country. Nevertheless, Belarus has made progress consolidating its school network. Since the early 1990s more than 1,200 schools were closed, equal to about 30 percent of the existing schools at the beginning 511 511 S1 S1 1.2M 428 428 S2 S2 300 300 3.0% FUNDING GAP 350 350 229 229 1.1M 200 200 2.0% 1M 900k 1.0% 100 100 800k 0.0% 2016 2017 2018 2019 0 2020 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 157 (B MGA) (B MGA) 2017 2017 2016 2016 2019 2019 2018 2018 2020 2020 h Current Funding and GER PROJECTED of the period (Figure 76). School closures accelerated in the 2000s, and every year since 2002 more ENROLLMENT PROJECTED EXPENDITURE GAP AND ENROLLMENT GAP DUE TO DUE OPERATING GAP POPULATION AND OPERATING GROWTH EXPENDITURE GAP TO POPULATION GROWTH Source: World than 100 school have closed annually. Impressively, school closures have kept pace with the declining Bank Sta Source: Calculations World Bank Sta Calculations number of student. 300k 300k 1.4M 1.4M 200k 200k 1.3M 1.3M 4.0% 4.0% Figure FIGURE76: Consolidation 76: CONSOLIDATION of school OF SCHOOL network NETWORK ENROLLMENT GAP ENROLLMENT GAP 100k 100k 1.2M 1.2M 3.0% 3.0% FUNDING GAP FUNDING GAP Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). 1991=100 - 1991=100 - 1.1M 1.1M 2.0% 2.0% All Ages (100k) (100k) 1M 1M Rural Age 5-14 130 (200k) (200k) 900k 900k 1.0% 1.0% Urban Age 5-24 (300K) (300K) 800k 800k 0.0% 0.0% Number of Teachers 2014 2013 2012 2013 2012 2016 2015 2015 2014 2017 2016 2018 2017 2019 2018 2020 2019 2020 2014 2013 2012 2013 2012 2015 2014 2016 2015 2018 2017 2017 2016 2020 2019 2019 2018 2020 110 urce: UN, Beistat Operating Costs with Operating Current Costs withFunding Current and GER and GER Funding GDP Growth: GDP1% GDP Growth: GDP2% Growth: 1% Growth: 2% Number of Schools GDP3% 90GDP Growth: Growth: 3% GDP Growth: GDP4%Growth: 4% Number of Students 70 50 ER OF STUDENTS : NUMBER HAS DECLINED, OF STUDENTS ESPECIALLY HAS DECLINED, ESPECIALLY 76: CONSOLIDATION FIGURE FIGURE OF SCHOOL 76: CONSOLIDATION NETWORK OF SCHOOL NETWORK DEMOGRAPHIC TRENDS: AREAS; DEMOGRAPHIC 1970 2030 TRENDS: 1970 2030 “Statistics Source: CISSTAT.Source: of the“Statistics Countries of CISSTAT. the Countries of the (2011). CIS”, MOE of the CIS”, MOE (2011). 1991=100 1991=100 All AgesAll Ages Rural Rural 1991 Age 5-14 1993 Age 19955-14 130 130 Urban1997 Urban 1999 2001 2003 2005 2007 2009 Age 5-24Age 5-24 Source: CISSTAT. “Statistics of the Countries of the 110 CI5,” MOE (2011). 110 Source: UN, Beistat Source: UN, Beistat Number of Schools Number of Students 2030 However, the size of the Number of Teachers education workforce 90 90has not seen a commensurate adjustment, and consequently student-teacher ratios have 70 declined, 70 with significant implications for per student costs. The number of teachers has remained relatively stable despite the significant decline both in the number of students and schools (Figure 50 76). 50 HER RATIOS ARE AMONG THE LOWEST IN THE REGION Student-teacher ratios at both primary and 1991 secondary 1993 1991 have 1995 1993 continued 1997 1995 1999 to 1997 fall and 2001 1999 2003are 2001 among 2005 2003 2005 the 2007 lowest 2009 2007 2009 A region (1999-2009) in the region (Figure 78). Secondary Since school teacher student-teacher salaries ratios in the are of the ECA region (1999-2009) Number Number single Schools of Schools largest cost item, this increase in the 980 975 1985 1980 1990 1985 1995 1990 2000 1995 2005 2000 2010 2005 number 2015 2010 2020 of teachers 2025 2015 2020 2030 2030 2025 per student has been associated of Students Number Number with of a significant increase in per student costs. Students 18 of Teachers Number Number of Teachers FIGURE 1.21: CURRENT EXPENDITURE FIGURE 1.21:COMPOSITION IN EDUCATION 16 CURRENT EXPENDITURE COMPOSITION IN EDUCATION COULD & COULD UNDERMINE EFFICIENCY EQUITY UNDERMINE EFFICIENCY & EQUITY Total Education Sector Budget, 2014/15 Total Education Sector Budget, 2014/15 FIGURE 1.21: CURRENT EXPENDITURE FIGURE 1.21:COMPOSITION IN EDUCATION CURRENT EXPENDITURE COMPOSITION IN EDUCATION COULD & COULD UNDERMINE EFFICIENCY EQUITY UNDERMINE EFFICIENCY & EQUITY Total Education Sector Budget, 2014/15 14 Total Education Sector Budget, 2014/15 10% 10% Figure 78: Student-teacher ratios are among the lowest in the region 4% 4% 26% 26% FIGURE 70: PERCENTAGE FIGURE OF PRIMARY SCHOOL TEACHERS 70: PERCENTAGE THATSCHOOL WOULD TEACHERS THAT WOULD OF PRIMARY FIGURE 70: PERCENTAGE FIGURE OF PRIMARY SCHOOL TEACHERS 70: PERCENTAGE THATSCHOOL OF PRIMARY WOULD TEACHERS THAT WOUL 10% 10% Primary Education NEED TO BE TRANSFERRED TO COMPY NEED WITH JOINT DECREE TO BE TRANSFERRED TO COMPY WITH JOINT DECREE NEED TO BE TRANSFERRED TO COMPY NEED WITH JOINT DECREE TO BE TRANSFERRED TO COMPY WITH JOINT DECREE 4% Primary Education 4% Secondary Education 26% Secondary Education FIGURE 70: PERCENTAGE FIGURE OF PRIMARY SCHOOL TEACHERS THATSCHOOL WOULD TEACHERS THAT WOULD FIGURE 70: PERCENTAGE FIGURE OF PRIMARY SCHOOL TEACHERS WOULD TEACHERS THAT WOULD THATSCHOOL 26% 70: PERCENTAGE OF PRIMARY 70: PERCENTAGE OF PRIMARY 12 # of existing teachers to move within districts University Education # of existing teachers to move within districts # of existing teachers to move Primary Education University Education NEED TO BE TRANSFERRED TO COMPY NEED WITH JOINT DECREE TO BE TRANSFERRED TO COMPY WITH JOINT DECREE NEED # of existing teachers to move TO between # of existing teachers to move within districts BE districtsTRANSFERRED provinceto move in sameteachers TO NEED COMPY TO WITH JOINT 30k BE TRANSFERRED DECREE TO COMPY WITH JOINT 30k DECREE # of existing teachers to move between districts in sameteachers # of existing provinceto move Primary Education TIVET 30k # of existing between districts in same province TIVET 30k NUMBER OF PRIMARY SCHOOL TEACHERS 25k NUMBER OF PRIMARY SCHOOL TEACHERS Secondary Education Secondary Education 25k Other Other NUMBER OF PRIMARY SCHOOL TEACHERS 25k NUMBER OF PRIMARY SCHOOL TEACHERS # of existing teachers to move within districts University Education University Education 25k # of existing teachers to move within districts # of existing teachers to move within districts 30k 30k 20k # of existing teachers to move within districts 20k in same province Primary school student-teacher AREratios in theINECA region Secondary school student-teacher ratios in the ECA region # of existing teachers to move between districts in sameteachers provinceto move between districts in same province # of existing teachers to move between districts # of existing teachers to move between districts in same province 39% TIVET 21% 39% 21% 30k 20k # of existing TIVET 30k NUMBER OF PRIMARY SCHOOL TEACHERS 20k 25k NUMBER OF PRIMARY SCHOOL TEACHERS 15k 10 ARE 25k 15k Other FIGURE 78: STUDENT TEACHER RATIOS AMONG THE LOWEST THE REGION Other FIGURE 78: NUMBER OF PRIMARY SCHOOL TEACHERS STUDENT TEACHER RATIOS AMONG THE LOWEST IN THE REGION 25k NUMBER OF PRIMARY SCHOOL TEACHERS 25k 15k 15k 20k 20k 10k 10k 39% 21% 39% 21% 20k 10k 20k 10k 15k 15k 5k 5k 1999-2009 1999-2009 15k 15k 5k 5k Source: EdStats database. 10k 0 Source: EdStats database. 10k 0 8 10k 10k 0 Selatan Jawa Barat Sulawesi Tenggara Jawa Timur Jawa Tengah Sumatera Utara Selatan Kalimantan Barat Lampung Sulawesi Selatan Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Tenggara Timur Bali Dki Jakarta Timur Kalimantan Tengah Aceh Darussalam Tengah Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung 0 Kalimantan Selatan Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Selatan Barat KalimantanBanten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Kalimantan Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo 5k 5k Selatan Jawa Barat MalukuJawa Timur Jawa Tengah Sumatera Utara Selatan Kalimantan Barat Lampung Sulawesi Selatan Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Tenggara Timur Bali Dki Jakarta Timur Kalimantan Tengah Aceh Darussalam Tengah Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung Kalimantan Selatan Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Selatan Barat KalimantanBanten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Kalimantan Selatan Maluku Jawa BaratPapua Jawa Timur Bali Tengah Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Banten Barat Papua Barat 5k Belitung 5k Sulawesi Kalimantan 0 Nusa Tenggara 0 Kepulauan Sulawesi Sumatera Nanggroe Sulawesi Sulawesi Kalimantan 0 Nusa Tenggara Sulawesi Sulawesi Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Aceh Darussalam Tengah Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung 0 Primary school student-teacher ratios in the ECA region (1999-2009) Secondary school student-teacher ratios in the ECA region (1999-2009) Kepulauan Sulawesi Sumatera Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Selatan Barat KalimantanBanten Jambi Sumatera Barat Riau Nusa Tenggara Timur Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Maluku Papua Utara Di. Yogyakarta Gorontalo Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung Primary school student-teacher ratios in the ECA region (1999-2009) Secondary school student-teacher ratios in the ECA region (1999-2009) Sulawesi Bangka Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Aceh Darussalam Tengah Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung Nusa Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Selatan Barat KalimantanBanten Sulawesi Tenggara Jambi Sumatera Barat Riau Nusa Tenggara Timur Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Maluku Papua Utara Di. Yogyakarta Gorontalo Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung Nusa Tenggara Jawa NusaSulawesi 6 FIGURE 16: WAGE BILL AS A PERCENTAGE OF AS AEDUCATION PUBLIC SPENDING, CIRCA 2013 SPENDING, CIRCA 2013 Nusa Dki Nanggroe Sulawesi Nusa FIGURE 16: WAGE BILL PERCENTAGE OF PUBLIC EDUCATION Nanggroe Tenggara NusaSulawesi Dki Nanggroe Sulawesi FIGURE 16: WAGE BILL AS A PERCENTAGE FIGURE OF 16: WAGE PUBLIC BILL AS AEDUCATION PERCENTAGESPENDING, OF PUBLICCIRCA 2013 SPENDING, CIRCA 2013 EDUCATION 100 100 30 30 90 88 18 18 4 87 90 88 87 82 81 100 80 80 79 82 79 81 80 79 79 % PUBLIC EDUCATION SPENDING 100 80 % PUBLIC EDUCATION SPENDING 73 73 88 68 8770 16 90 87 90 88 70 68 80 82 81 80 6079 82 79 81 80 60 79 79 57 16 % PUBLIC EDUCATION SPENDING 80 57 % PUBLIC EDUCATION SPENDING 73 2 73 49 49 25 50 70 68 50 49 49 25 70 68 FIGURE 2: SIX TYPES OF DATA CAN BE USED FOR PER ANALYSIS THAT FIGURE 2: SIX TYPES OF DATA THAT CAN BE USED FOR PER ANALYSIS 60 50 60 50 40 30 40 57 30 49 49 57 49 49 14 14 FIGURE 40: FIGURE 40: R analysis FIGURE 2: SIX TYPES OF DATA THAT FIGURE 2: SIX TYPES 40 40 20 20 BY SOURCES AND UNIT COST PROJECTION BY SOURCES AND UNIT COST PROJECTION 1 OF DATA THAT 0 of data that can be used for PER analysis 1 20 20 12 Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 2016-2025, January 2015 12 719 (ie: policies, laws, and regulations) 10 Source: Education 709 CAN BE USED FOR PER ANALYSIS 30 719 (ie: policies, laws, and regulations) 10 709 CAN BE USED FOR PER ANALYSIS 30 676 FIGURE 40: 676 FIGURE 40: 2 ANALYSIS Country System-Level Data (Budget (ie: BOOST), 20 20 0 0 BY SOURCES AND UNIT COST PROJECTION BY SOURCES AND UNIT COST PROJECTION 2 562 FIGURE 2: SIX TYPES OF DATA 2: CAN THAT BE USED 1 FOR PER ANALYSIS Country System-Level Data (Budget (ie: BOOST), 562 Cross-na FIGURE SIX TYPES OF DATA THAT CAN BE USED FOR PER 529 1 sectoral data (ie: EMIS, national assessment) 522 529 sectoral data (ie: EMIS, national assessment) Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 719 120% 2016-2025, January 2015 522 Honduras Costa Rica Guatemala Panama Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep (ie: policies, laws, and regulations) 10 Source: Education 709 surveys 719 data (interna 120% 2016 Honduras Costa Rica Guatemala Panama Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep (ie: policies, laws, and regulations) 10 709 onal 2016 427 676 HUN 10 HUN 455 427 676 MDA MDA nsus, household UZB UZB 10 LTU ROU assessment, cross RQU 455 AZE AZE LVA CZE LVA 100% KAZ KAZ BUR SVN KGZ SVN POL POL 428 TJK MKD LTA MKD 2017 420 100% 411 BLR 428 SVK SVK UKR UKR surveys) 2017 420 na onal sta s cal 1 4 3 BGR 411 394 Cross-National Data (International assessment,Cross-National Data (International assessment, 1 (ie: policies, laws,2 4 2 National Surveys 3 0 394 Research databases) (ie: policies, laws, and regulations) Country statistical cross national System-Level Data (Budget (ie: BOOST), databases) National Surveys 0 80% 2018 562 (Census, household surveys) 2018 TYPES DATA OF(donors, 2: CAN THAT BE reports USED FOR Researc h PER ANALYSIS and regulations) Country statistical cross national System-Level Data (Budget (ie: BOOST), databases) 80% 562 (Census, household surveys) FIGURE SIX TYPES DATA OF(donors, THAT CAN BE USED FOR PER ANALYSIS 529 sectoral data (ie: EMIS, national assessment) 314 522 reports 529 sectoral data (ie: EMIS, national assessment) 314 120% 522 2012 Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Finland Korea, rep 120% 60% 2016 2012 Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Finland Korea, rep 279 15 60% 2016 279 academic research 15 244 427 455 241 academic research 411 230 244 2025 455 241 100% 2 5 3 papers) 230 427 428 Country System-Level Data (Budget (ie: BOOST), Research Data Reports 2017 2025 420 100% 2 5 3 4 Cross-National papers) 40% 204 411 428 Country System-Level (Budget (ie: BOOST), Research Reports Data (International assessment,Cross-National Data (International assessment, 2017 420 4 4 cross national statistical databases) 40% 204 394 185 Cross-National data (ie: EMIS,assessment, Data (International 8 1 4 sectoral national assessment) National (Donors, Surveys academic research papers) 394 185 Cross-National Data (International assessment, 165 cross national statistical databases) 8 sectoral data (ie: EMIS, national assessment) National (Donors, Surveys academic research papers) 165 , and regulations) cross (ie: policies, laws, and national statistical databases) regulations) cross national statistical databases) (Census, household surveys) (Census, household surveys) 80% 80% 20% 2018 2018 20% 314 314 93 96 60% 74 2012 93 96 2012 279 71 62 72 60% 0% 74 279 71 62 72 3 6 52 New data 244 0% 241 data Surveys National 37 3 6 5 52 New 27 33 230 244 Research Reports 27 2025 241 National Surveys New Data Collection 33 37 5 5 15 13 11 11 230 c (Budget (ie: BOOST), Level Data Research Reports New Data Collection Research Reports 2025 2 5 4 40% 9 204 Country System-Level Data (ie:household (Census, BOOST), surveys) Cross-National Data (International assessment, 15 13 11 11 c (Budget Research Reports 2016 2017 2018 2021 2025 4 (Donors, academic research papers) 40% 9 204 (Census, household surveys) -20% 185 EMIS, national assessment) (Donors, academic research papers) Cross-National Data (International assessment, (Donors, academic research papers) -20% 2016 2017 2018 2021 2025 185 165 PER sectoral data (ie: EMIS, national assessment) (Donors, academic research papers) cross national statistical databases) Preschool Primary Preschool1st Cycle Non-Formal Primary2nd Cycle 1st Cycle Higher Technical 165 cross national statistical databases) FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL FIGURE IMPACTS 4.5: DISTRIBUTION OFBY COMMUNITYIMPACTS EDUCATIONAL TYPE, 2009/10 BY COMMUNITY TYPE, 2009/10 20% 20% Secondary Non-Formal Secondary andSecondary 2nd Cycle Education Secondary Technical and Higher Education 10 93 96 10 6 71 74 93 96 62 72 6 6 Source: World Bank calculations based on data Source: from the National Statistics Service (NSS), 0% 71 62 72 74 School School Vocational 6 6 52 New Data Collection World Bank calculations based on data from the National Statistics Service (NSS), 0% 33 37 School School Vocational 3 6 5 52 National Surveys New Data Collection Research Reports New Data Collection Internal 27 Resources 33 37 5 15 13 11 11 old surveys) (Census, household surveys) New Data Collection (Donors, academic research papers) Research Reports The National Center for Education Technology (NaCET), the Assessment The National and Testing Center for Education Center (ATC) Technology and MOF (NaCET), the Assessment and Testing Center (ATC) and MOF -20% 2016 2017 2018 2021 2025 2016 2017 2018 9 2021 2025 15 13 11 11 9 Internal 27 Resources (Donors, academic research papers) -20% Development Preschool Partners Primary Non-Formal 1st Cycle Development 2nd Cycle Partners 1st Cycle Higher Technical FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL FIGURE IMPACTS 4.5: DISTRIBUTION OFBY TYPE, 2009/10 COMMUNITYIMPACTS EDUCATIONAL BY COMMUNITY TYPE, 2009/10 Republic of America Preschool Shortage of Planned Budget Primary Secondary Non-Formal Secondary andSecondary 2nd Cycle Education Technical Higher (a) Average PCF Allocation per Student (a) Average PCF Allocation per Student (b) Average School Size (b) Average School Size Republic of America Secondary School of Planned Shortage Budget Vocational and Education 6 New Data Collection 6 New Data Collection Source: World Bank calculations based on data Source: from the National World BankStatistics Service calculations (NSS), based on data from the National Statistics Service (000s (NSS), AMD) (000s AMD) (Number of Students) (Number of Students) Internal Resources School School School Vocational 4 The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF Internal Resources 4 The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF Development Partners Development Partners 300 FIGURE 4: FIGURE 4: FIGURE 4: FIGURE 4: (a) Average PCF Allocation per Student 250 (b) Average School Size 300 500 Republic of America 500 Republic of America Shortage of Planned Budget Shortage of Planned Budget (a) Average PCF Allocation per Student 250 (b) Average School Size 400 FOR PRIMARY EDUCATION RELATIVE SCHOOL SIZE TO EDUCATION FOR PRIMARY EDUCATION RELATIVE SCHOOL SIZE TO EDUCATION 5 400 FOR PRIMARY RELATIVE TO SCHOOL SIZE FOR PRIMARY RELATIVE TO SCHOOL SIZE (000s AMD) (Number of Students)200 5 (000s200 AMD) (Number of Students) 300 300 150 150 200 2 100 200 2 300 500 100 4: FIGURE 4: 30 FIGURE 30 4: FIGURE 4: 30 30 250 300 250 50 400 500 50 100 100 MARY EDUCATION RELATIVE FOR 25 SIZE SCHOOL TO EDUCATION PRIMARY RELATIVE TO SCHOOL FOR PRIMARY EDUCATION RELATIVE TO EDUCATION 25 SIZE SCHOOL 0 400 0 25 SIZE FOR PRIMARY RELATIVE TO SCHOOL 25 SIZE 200 200 0 0 300 l s s l s ra 300 l n ain s ou s s 40 m ity en ity 150 ra l en ityn ain s ou s 40 m ity ou en ity ba ou ra n ain s ou ou s 40 m ity St unba ou 150 Ru ra n ain s ou < Co un s 40 m ity St un ts ou en ity ba ou ain Ru ou n- ntain < Co un St un ts ba ou Ur 200 Ru ain < Co un n- ntain ts 0 m Ur ain Ru n- ntain < Co un St un ts nt m 0 m Ur 100 200 ain n- ntain nt m 0 m Ur 20 20 nt ud nt m wi ol in m 100 nt ud nt m 0 m ou wi ol in m 20 20 nt No ou ud ou ou wi ol in m nt Co No ou ud 30 ou ou wi ol in m M Co No ou ou ou 50 100 M Co No ou M 30 30 ou m M ho in Co ly M m 100 M th a 50 ou Sc ol in 0 ly M m th a Sc ol in ly M m th a ou Sc ol in 0 gh ly th a 0 gh 15 15 0 gh 0 Hi nt On s Scho gh 15 15 0 ol Hi On Scho 0 s ho Hi nt On s Scho 25 0 ho Hi ho s s ho l 25 25 l s l s Sc ra n ain s ouly ou s s 40 m ity en ity ra l en ityn ly ou 40 m ity ou en ity ba ou Sc ra n ain ly ain s t On ouly s 40 m ity ou St unba ra 10 10 Ru n ly t On ly ou < Co un 40 m ity St un en ity ts ou ba ou ain Ru ain ly t On n- ntain < Co un St un ts ba Ur 10 10 Ru ain ly t On < Co un n- ntain ts 0 m Ur ain Ru n- ntain < Co un St un ts nt m 0 m Ur ain On No ou No n- ntain nt m 0 m Ur 20 nt No ud nt m wi ol in m FIGURE 41: No ou No ud nt m 0 m ou wi ol in m 20 20 nt No No ou ud ou ou wi ol in m FIGURE 41: Co FIGURE 41: ud ou ou wi ol in m M Co No ou ou ou M Co 5 5 FIGURE 41: M ou m M Co ly M m M th a Sc ol in ly 5 5 M m th a Sc ol in ly M m th a Sc ol in ly gh BY SOURCES AND UNIT COST AND 2016, PROJECTION; 2025 PROJEC th a Sc ol in gh 15 Hi gh BY SOURCES AND UNIT COST PROJECTION; AND 2016, 2025 PROJECTION; 2016, 2025 BY SOURCES UNIT COST On Scho gh 15 15 HUN HUN Hi On Scho BY SOURCES UNIT COST MDA MDA ho Hi On Scho BLR UZB LTU UZB ROU RQU HUN CVE AZE ho HUN AZE Hi On Scho LVA CZE LVA KAZ KAZ KGZ SVN KGZ SVN POL MDA POL MDA TJK TJK MKD LTA ho MKD 0 0 BLR UZB LTU UZB ROU RQU ARM CVE AZE ho AZE LVA CZE LVA BLR KAZ KAZ SVK SVK KGZ RUS SVN UKR KGZ UKR SVN POL POL TJK TJK MKD LTA MKD 0 0 BGR ly BGR ARM BLR SVK ly SVK RUS UKR UKR ly BGR t On ly BGR 10 ly t On ly Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 2016-2025, January 2015 Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 2016-2025, January 2015 ly t On 10 10 ly t On Source: Education Source: Education No No 0 100 200 300 100 400 200 400 0 100 200 0 300 100 400 200 FIGURE 41: No No 50 300 300 400 FIGURE 41: FIGURE 41: FIGURE 41: 5 5 (c) Average Class Size (c) Average Class Size (d) Student-Teacher Ratio (d) Student-Teacher Ratio 30% BY SOURCES 4.0% AND UNIT COST PROJECTION; 2016, 30% 2025 PROJECTION; SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS BY SOURCES AND UNIT COST PROJECTION; BY SOURCES AND 2016, 2025 PROJECTION; UNIT COST 30% 2016, 2025 BY SOURCES 4.0% AND UNIT COST 2016, 2025 30% 0 0 SCHOOL SIZE BY NUMBER OF STUDENTS 0 SCHOOL SIZE BY NUMBER OF STUDENTS 12 Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 2016-2025, January 2015 Source: Education Sector Strategy; 2016-2025, January 2015 Sector Strategy; 2016-2025, January 2015 12 Source: Education Source: Education 25 100 200 0 300 100 400 200 300 400 0 100 200 0 300 100 400 200 300 400 25 10 10 3.0% 3.0% (c) Average Class Size (c) Average Class Size (d) Student-Teacher Ratio 20 20 8 (d) Student-Teacher Ratio 8 20% 20% SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS 15 6 6 30% 30% 20% 4.0% 4.0% 30% 30% 20% 15 4 10 12 10 12 4 2.0% 2.0% 25 5 10 2 2 25 5 10 0 3.0% 3.0% 20 20 08 08 s 0 20% 10% 10% 20% 10% 10% l s ra l l s 20% 1.0% 20% n ain s ou s 40 m ity 1.0% en ity 6 ou 15 ba ou l s ra n ain s ra n s 40 m ity ain s ou en ity s 40 m ity ou 6 en ity Ru ou 15 < Co un St un ts ou ba ou ba ou ra ain n n- ntain ain s 40 m ity < Co un Ur s 40 m ity ou en ity Ru Ru < Co un St un ts ou 40 m ity < Co un St un ts ba ou ain 2009 2004 ain n- ntain nt m 0 m Ur n- ntain 4 Ur Ru 2.0% St un M ts 10 2009 nt 2009 ain ud nt m 0 m n- ntain nt m 0 m 4 Ur wi ol in m 10 ou 2.0% nt No ou ud nt ou ud nt m 0 m wi ol in m m ho in Co ou M ou No ou nt ou No ou ud 2 ou m ho in Co ho in Co m 5 M ou ly M No ou 2 ou th a M ho in Co m M m 5 ly M ly 0.0% th a < Co un th in a gh 0% 0% M ly hool m 0 ly 0.0% < Co un th in a gh 0% 0% ol gh Hi lyous hool 0 0 gh 10% 10% ol Hi ho lyous hool l Hi 10% 10% Sc 0 s ol l Hi 1.0% ON GOVERNMENTCURVE CONCENTRATION CURVE CONCENTRATION SPENDING Sc Sc 2016 2017 2018 2017 2021 2018 2025 2016 2017 2016 2018 2017 2021 ain Sc l s ra s n nt ly s m wi ly ou 1.0% en ity ON GOVERNMENT SPENDING Sc 2016 2021 2025 Sc ou On ou ain Sc ba l s ra n nt ly s ly ra m wi t On n ain s ou en ity ou s 40 m ity en ity Ru Sc ou On ou St un ts ou ba ba ou ain On ly ra t On n n- Nontain ain s < Co un Ur ou s 40 m ity en ity Ru Ru St un ts ou < Co un St un ts ba ou ain On ain On n- Nontain nt 0 m Ur m ain Ur ON EDUCATION AND VARIOUS BENCHMARKS Ru No St un M ts ain On ud nt 0 m n- ntain m t nt m 0 m Ur wi ol in m ON EDUCATION AND VARIOUS BENCHMARKS No ou nt No ou ud nt m t ud nt m 0 m wi ol in m wi ol in m Co ou M ou ou nt ou ud ou wi ol in m Co Co 2004 M ou ly M ou ou th a Sc ol in M Co M 2004 ly M ly 0.0% th a Sc ol in th a 0% 0% Sc ol in gh M m 0.0% n- ly th a 0% 0% Sc ol in gh Hi gh On Scho No No gh ho Hi On Scho Hi On Scho No ho ho Hi On Scho ON GOVERNMENTCURVE TION CURVE CONCENTRATION SPENDING 100% 2016 2017 Education spending as 2017 2021 share 2025 2016 2018 of total public spending 2016 2017 2016 2018 Education spending as 2017 2021 share of total public spending ho ly ON GOVERNMENT SPENDING 100% 2018 Education spending 2021 as share 2025of total public spending 2018 Education spending 2021 as share of total ly t On ly ly ly t On ly t On ly ly t On ON AND VARIOUS BENCHMARKS No ON EDUCATION AND VARIOUS BENCHMARKS Capital spending as share of total education spending Capital spending as share of total education spending No No Capital spending as share of total education spending Capital spending as share of total edu No CUMULATIVE SHARE OF GOVERNMENT SPENDING ON EDUCATION (f) School Capacity Utilization Education spending as share of GDP Education spending as share of GDP CUMULATIVE SHARE OF GOVERNMENT SPENDING ON EDUCATION (f) School Capacity Utilization Education spending as share of GDP Education spending as share of GDP (Student Enrollment as % of (Student Enrollment as % of Education spending as share of total public spending Education spending as share of total public spending 100% 80% 80% Total Builing Capacity) Education spending as share of total public spending Education spending as share of total public spending Total Builing Capacity) Capital spending as share of total education spending Capital spending as share of total education spending 33% 33% Capital spending as share of total education spending Capital spending as share of total education spending 32% (f) School Capacity Utilization Education spending as share of GDP Education spending as share of GDP CUMULATIVE SHARE OF GOVERNMENT SPENDING ON EDUCATION 32% 70% Utilization (f) School Capacity 70% Education spending as share of GDP Education spending as share of GDP 31% (Student Enrollment as % of 31% (Student Enrollment 60% as % of 60% 80% 60% 30% Total Builing Capacity) 60% 30% Total Builing 50%Capacity) 50% 33% 33% 29% 29% 40% 32% 28% 40% 2025 2025 32% 70% 28% 70% 30% Propoor Spending Propoor Spending 31% 27% 30% Source: EdStats database. 31% 60% 27% 60% 20% 20% 30% 30% 26% 50% 10% 60% 40% 26% 50% 10% 40% 29% 29% 40% ra l s l s n 0% ain s Progressive ou s 40 m ity en ity ou ba ou 40% ra n 0% ain s Progressive ity ou s en ity Ru ou < Co un St un ts ba ou ain s n- ntain l Ur 28% Ru < Co un St mun ts 30% ain s FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED n- ntain nt m 0 m l Ur 28% ra ud n ain s 40 m ity ou s en ity FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED ou nt Propoor Spending 30% 0 mba ou ud nt m ra n wi ol in m ain s ou s 40 m ity en ity Propoor Spending Lorenz Curve of Income or Consumption ou Ru ou nt < Co un St un ts ba ou No ou ou wi ol in m ain n- ntain ho in Co 27% 40 Ur Lorenz Curve of Income or Consumption M ou Ru < Co un St un ts No ou ou 20% ain ain ho in Co nt m 0 m 27% Ur M m ly 20% M th a nt ud nt m 0 m M wi ol in m m ly ou nt th a nt No ou ud ou gh wi ol in m s Sc ol Co 26% ou Regressive (Poorly Targeted) Regressive (Poorly Targeted) M ou ou 10% gh s Sc ol ho in Co 26% Hi ou ly ho M m 10% ly M th a Sc ol in Hi ou ly ho M m ain sOn ly Sc 40% 20% s ly th a l gh n- ain sOn ly Sc 20% ra l s n 0% gh ain s Progressive ol ou s 40 m ity en ity Hi On Scho No ou ba ou ra n 0% ain s ou t On Progressive ou s ity en ity ho Hi ho Ru ou < Co un St un ts ba ou ain s ou t On n- ntain l Ur Ru < Co un St mun ts Sc ain s FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED ain nt m 0 m l Ur ra ud n ly 40 m ity n- ntain No ou en ity Sc FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED nt 0 mba ud nt m ra n wi ol in m ly t On ly 40 m ity n- ntain No ou en ity Lorenz Curve of Income or Consumption ou Ru nt nt un St un ts No ou ba 75 ou wi ol in m ain ly t On Co 40 Ur Lorenz Curve of Income or Consumption M ou Ru < Co un St un ts ou 75 ou ain On Co nt m 0 m Ur M m M No ly th a Sc ol in nt ud nt m 0 m M wi ol in m m No ly ou th a Sc ol in nt No ou ud ou gh wi ol in m Co < Co n- ou Regressive (Poorly Targeted) Regressive (Poorly Targeted) M No ou ou gh ho in Co Hi On Scho M No m ly M th a Sc ol in ho Hi ho M m ly th a ho gh Sc 20% 0% gh ly ol Hi On Scho 0% ly t On ly ho Hi ho ly t On Sc On ly No Sc ly t On ly No 0% 20% 40% 60% 20% 80% 40% 100% 60% 75 70 ly t On 0% 80% 100% 75 70 On No No CUMULATIVE SHARE OF POPULATION 0% CUMULATIVE SHARE OF POPULATION Number of Students per Primary School 20% 40% 0% 60% 20% 80% 40% 100% 60% 80% 100% 70 70 65 65 CUMULATIVE SHARE OF POPULATION 63.1 63.1 CUMULATIVE SHARE OF POPULATION 65 60 60 65 63.1 78: CLASS FIGURE FIGURE 78: AND SCHOOL SIZES ARE SMALL 63.1 56.9 CLASS AND SCHOOL SIZES ARE SMALL 56.9 60 60 55 55 54.9 54.9 FIGURE 8: SOURCES OF EDUCATION SECTOR FINANCE, 2013 SECTOR FINANCE, 2013 56.9 500 FIGURE 8: SOURCES OF EDUCATION 56.9 FIGURE 5.11: EDUCATION FIGURE 50 50 EXPENDITURES BY ECONOMIC CATEGORY, GDP %ECONOMIC % 5.11: GDP EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, %ECONOMIC GDP 55 49.7 Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 FIGURE 5.11: EDUCATION EXPENDITURES BY CATEGORY, FIGURE 5.11: EDUCATION EXPENDITURES BY CATEGORY, % GDP 55 54.9 54.9 49.7 Source: Kosovo BOOST Source: Kosovo BOOST Source: Kosovo BOOST Source: Kosovo BOOST GURE 8: SOURCES OF EDUCATION SECTOR FINANCE, FIGURE 8: SOURCES 2013 SECTOR OF EDUCATION 9 | 1% FINANCE, 2013 e: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 Number of Students Number per of Classroom Students per Classroom 9 | 1% FIGURE 5.11: EDUCATION EXPENDITURES FIGURE BY ECONOMIC 5.11: EDUCATION CATEGORY, EXPENDITURES GDP %ECONOMIC BY FIGURE CATEGORY, % 5.11: GDP EDUCATION EXPENDITURES FIGURE BY ECONOMIC 5.11: EDUCATION CATEGORY, EXPENDITURES %ECONOMIC BY GDP CATEGORY, % GDP Number of Students Number per of Primary StudentsSchool per Primary School 50 50 45 45 49.7 49.7 4.5% 4.5% 4.5% 4.5% Source: Kosovo BOOST Source: Kosovo BOOST Source: Kosovo BOOST Source: Kosovo BOOST 426 | 18% 426 | 18% 936 | 40% 9 | 1% 936 | 40% LGA 4.0% 4.0% 9 | 1% LGA 4.0% 4.0% 45 45 40 40 Household Household 4.5% 4.5% 4.5% 4.5% 2008 2009 2010 2011 2012 2013 2014 2008 2015 2009 2016 2010 2017 2011 2018 2012 2019 2013 2020 2014 2021 2015 2022 2016 2017 2018 2019 2020 2021 2022 295 | 13% 3.5% 3.5% 3.5% 3.5% UBEC 500 295 | 13% 426 | 18% UBEC 40 500 936 | 40% 426 | 18% 40 936 | 40% LGA 4.0% 4.0% LGA State 4.0% 4.0% 40 400 State 3.0% 3.0% Capital Outlays Capital Outlays 3.0% 3.0% Capital Outlays Capital Outlays 40 Combine Increase Education in 2021 and Health Combine and Spending Household Household Federal Gov’t Federal Gov’t 3.5% 3.5% 2008 2009 2010 2011 2012 2013 2014 2008 2015 2009 2016 2010 2017 2011 2018 2012 2019 2013 2020 2014 2015 2022 2016 2017 2018 Increase 2019 2020in 2021 Education 2022 and Health and Spending 295 | 13% UBEC 3.5% 2.5% Subsides & Transfers 3.5% Subsides 2.5% & Transfers Subsides & Transfers Subsides & Transfers Health Spending Increases from 2.6 to 4 percent of GDP from 2.6 to 4 percent of GDP 295 | 13% UBEC 587 | 25% Donors 587 | 25% Donors 2.5% 2.5% Health Spending Increases State 3.0% 3.0% Utilities Utilities Utilities Utilities from 3 to 4 Education Spending IncreasesEducation percent of Spending GDP from 3 to 4 percent of GDP Increases State 3.0% Capital Outlays Capital Outlays 2.0% 76 | 3% 2.0% 2.0% Capital Outlays Goods & Services 3.0% 2.0% Capital Outlays Goods & Services Combine Increase in Education and Health Combine and Spending Increase in Education and Health and Spending Baseline 35 76 | 3% Federal Gov’t Goods & Services Goods & Services Baseline 35 Federal Gov’t Subsides & Transfers Subsides & Transfers 587 | 25% Donors 587 | 25% 2.5% 2.5% Subsides 2.5% & Transfers Wages & Salaries 2.5% Subsides & Transfers Wages & Salaries Health Spending Increases from 2.6 to Health 4 percent Spending of GDP from 2.6 to 4 percent of GDP Increases Donors 1.5% 1.5% Wages & Salaries 1.5% 1.5% Wages & Salaries Utilities Utilities Utilities Utilities from 3 to 4 Education Spending IncreasesEducation percent of Spending GDP from 3 to 4 percent of GDP Increases 400 2.0% 2.0% 400 76 | 3% 76 | 3% 2.0% 1.0% Goods & Services 1.0% Goods & Services 2.0% 1.0% Goods & Services1.0% Goods & Services Baseline Baseline FIGURE 12: THE STRUCTURE OF BASIC FIGURE EDUCATION 12: THE FINANCING STRUCTURE OF BASIC EDUCATION FINANCING Wages & Salaries Wages & Salaries 1.5% 1.5% Wages & Salaries 1.5% 1.5% Wages & Salaries 30 0.5% 0.5% 30 0.5% 0.5% Internally Federal 300 Fund Internally VAT Federal Independent Devp’t VAT Independent Devp’t 1.0% 1.0% Generated Account & Fund Generated Account & 1.0% 1.0% Pool Revenue Partners 0.0% 0.0% URE 12: THE STRUCTURE OF BASIC FIGURE EDUCATION 12: THE Sources FINANCING STRUCTURE Revenue OF Derivatives BASIC EDUCATION Sources FINANCING Revenue Derivatives Pool Revenue Partners 0.0% 0.0% 0.5% 2007 2008 2009 2010 2012 2011 2008 2007 0.5% 2009 2010 2011 2012 2007 2008 2009 2010 2011 2008 2007 2012 2009 2010 2011 2012 Federal 0.5% 0.5% Internally Federal Federal Internally VAT Federal Independent Devp’t VAT Ministry of FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC BY BROAD ECONOMIC 25 Generated Account & Independent Devp’t Ministry of Fund FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES 25 Generated Pool Account & Revenue Partners Finance Finance 0.0% 0.0% Revenue Derivatives Pool Revenue Partners 0.0% 0.0% 300 Sources Revenue Derivatives 300 Federal LGA SJGA State LGA SJGA Con. Rev. State Federal 2007 2008 2009 2010 2012 2011 2008 2007 2008 2009 2010 2011 2008 2012 CATEGORY SOURCE: DATA FROM MFB/SIGFP. CATEGORY SOURCE: DATA FROM MFB/SIGFP. Con. Rev. 2007 2009 2010 2011 2012 2007 2009 2010 2011 2012 Federal Federal Responsibility Ministry of Federal Federal Responsibility Finance Ministry of Finance Budget Budget FIGURE 15: EVOLUTION OF REAL EDUCATION FIGURE EXPENDITURES 15: EVOLUTION BY BROAD OF REAL EDUCATION ECONOMIC BY BROAD ECONOMIC EXPENDITURES LGA SJGA State LGA Federal SJGA Con. Rev. State Federal UBFC (2%) UBFC (2%) CATEGORY SOURCE: DATA FROM MFB/SIGFP. CATEGORY SOURCE:700 DATA FROM MFB/SIGFP. 700 20 Con. Rev. sibility Responsibility Federal Budget Federal 20 FMF FMF 600 600 200 Budget 700 700 BILLION CONSTANT 2013 AR. BILLION CONSTANT 2013 AR. UBFC (2%) UBFC (2%) Salary Non-Salary Salary Non-Salary Salary Non-Salary 500 500 200 Salary Non-Salary Salary Non-Salary Salary 200 Non-Salary Salary Non-Salary Salary Non-Salary Education 600 FMF Education Lower Finances Primary FMF Lower Primary Lower Unity 600 15 Finances Secondary Lower Unity 15 School Secondary Secondary College School Secondary College 400 BILLION CONSTANT 2013 AR. SUBEB 400 BILLION CONSTANT 2013 AR. SUBEB on Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary Education Lower Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 Salary Non-Salary Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 500 500 s Lower Note: Lower Unity 300 Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) Primary Finances 300 Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) Secondary Primary “into Secondary Lower which shall be paid all allocations to the local government councils of the state from the Federation Unity School Secondary College “into which shall be paid all allocations to the local government councils of the state from the Federation SUBEB School Secondary College 400 Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). SUBEB FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 400 Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). FIGURE 5.12: DISTRIBUTION OF TEACHERS BY ketch following funding allocation arrangements in UBE following UBE Act of 2004 Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 hall maintain a special account to be called “State Joint Local Government Account (SJLGA) Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) 10 10 300 300 200 200 100 be paid all allocations to the local government councils of the state from the Federation 100 100 “into which shall be paid all allocations to the local government councils of the state from the Federation the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 100 Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 100 12k 12k 200 200 0 NCHMARKS 5 R i nt 30% BY SOURCES AND UNIT COST PROJECTION; 3.0%2025 2016, BY SOURCES 4.0% AND UNIT COST PROJECTION; 2016, 20 30% e nta u St u No en U n-m N nt O ou ou i nt ud i ou t B m 0 m (c) Average Class Size (d) Student-Teacher Ratio 8 n-m nta w ol in om 40 om 20 10k nta U U K P o K P 20% 20% S S K M M H H ou No ou ud S S R M M w ol in om 40 om St M 12k 30% 4.0% 30% R No ou < aC C R B B M 2004 M hly 0 12 Source: Education Sector Strategy; 2016-2025, January 2015 Source: Education Sector Strategy; 2016-2025, January 2015 Sc ol in < aC C 15 6 M hly 0% Sc ol in Hig o 25 10 Hig 4 3.0% On Sch 2.0% ith ho o 10 12 On Sch ith ho nly 2 (c) Average Class Size 20 (d) Student-Teacher Ratio 8 20% 20% ly nly 25 5 10 30% 30% tO 3.0% 4.0% ly tO 15 6 No 0 10% 10% No 20 08 20% 8k 20% 0 0 l s 4 l s ra ou 10 12 10k 1.0% 2.0% an nta us s 0 m y en ity 15 6 ou u ou 40 om nit ra an nta us in (f) School Capacity Utilization s 0 m y 40 om nity en ity ou ino 25 R Urb ud n ts Ru 2 ou in u ta St u in ou ino 5 Urb 10 ud n ts un 3.0% m n-m nta 4 in nta 2.0% u St u 10 (Student Enrollment as % of m 0 m n-m nta w ol in om o 0 ou Ed No ou 10% 10% m wit l in om 20 M 08 20% 20% No ou < aC C 2 M M l s 5 hly Total Builing Capacity) 1.0% Sch ol in 40 om nit < a C C ou M s l ra 0.0% an nta us hly s 0 m y en ity 6 0% 0% Sch l in 15 ou Ru ou 40 om nit Hig ra an nta us in Ca s 0 m y 40 om nity en ity 0 ou ino Urb ud n ts Ru o ou Hig 10% 10% ly us oo in u nta St u 0 in ou ino Urb 33% On Sch ith ud n o ts m oo n-m nta 4 6k in nta 2.0% u St u l s Onino Sch u h 10 ou 1.0% m 0 m n-m nta w ol in om l s ra ou 8k Utilization 2018 2021 2025 2016 2016 2017 2018 2017 2021 an nta nlyus ou No ou nly en ity m wit l in om ou Ru M 32% (f) School Capacity Ed ly ra an No ou < aC nta us C in 2 s 0 m y en ity ou t Oino Urb M ud n ts Ru 70% tO ou M 40 om nit 5 hly Sc ol in 40 om nit < a C ho in C nta St u in ou ino Urb M 0.0% ud n ts hly n-m N nta 0% 0% in nta u St u No ou 0 Hig 31% (Student Enrollment as % of m n-m nta w ol in om o o l ig 60% 10% 10% ly us oo ou No ou 0 On Sch ith w ol in om ho M l H o No ou < aC C s Onino Sch u h M M Total Builing Capacity) 1.0% CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK hly DNK SVK ou 30% 2016 2017 2018 2021 2025 Sch ol in < aC C l s ra an nta nlyus nly M en ity 50% 0.0% Sc hly ou Ru 0% 0% Sch ol in ly ra an nta us in 0 m y Hig s en ity ou t Oino Urb ud n ts Ru tO ou 40 om nit o nta St u in ou ino Hig Urb 33% ud n ts 29% On Sch ith o n-m N nta in o ta u St u No ou 40% 4k2018 On Sch ith un m n-m nta w ol in om o o 2016 30% 2017 Education spending as share 2021 2025 of total public spending 2016 2017 2018 Education spending as share of total public spending 2021 ou 6k w ol in om nly o M 32% ou < aC 28% C ly M nly M hly 70% tO Sc ol in < aC C ly M Propoor Spending hly 0% 0.0% 0% Sc ol in tO Hig No No 31% 27% Capital spending as share of total education spending Capital spending as share of total education spending o Hig 60% No No 20% On Sch ith ho o On Sch ith ho 30% 26% 2016 2017 2018 Education spending as share 2021 2025of total public spending 2016 2017 2018 Educ nly 50% 10% ly nly (f) School Capacity Utilization Education spending as share of GDP Education spending as share of GDP tO ly 29% l s tO ou No 40% ra an 0% nta us Progressive Capital spendingas share of asshare total public of total education spending Capit 0 m y s en ity Ru No ou (Student Enrollment as % of 40 om nit in l shareuof Education spendinguraas 2k total public spending Education spending spending ou ino Urb ud n ts s in 28% 4k u nta St u 30% m n-m nta an nta us s 0 m y o en ity ding Total Builing Capacity) ou ou 40 om nit w ol in om in Lorenz Curve of Income or Consumption ou ino (f) School Capacity Utilization R Urb Education spending share asof of GDP Educ ud n ts No ou 27% in nta u St u M Capital spending as share ou of total education spending Capital spending as share total education spending < aC ho in C 20% m n-m nta M 33% hly w ol in om (Student Enrollment as % of OURCES OF EDUCATION SECTOR FINANCE, 2013 Education spending as share of total public spending 2025 spending Education spending as share of total public No ou s Sc ol 26% Hig M Regressive (Poorly Targeted) < aC 10% C ou nly ho M hly ith Sc ol in 32% (f) School Capacity Utilization Total Builing Capacity) Education spending as Hig share of GDP Education spending as share of GDP Sc 70% l s ra ou an 0% nta us nta us nly Capital spending as share of total education spending Capital spending as share of total education spending s 0 m y en ity o Ru t 2013 and per capital allocation from monthly shares of distribution from ou 40 om nit On Sch ith ho (Student Enrollment as % of in ou ino Urb 31% 33% ud n ts ou ino t O l s 0 in 60% u nta St u in O ou FIGURE 6.1: ALBA conomic Report, The World Bank Group (2013) 2k m n-m nta ra an nly 0 m y o en ity ou Ru 40 om nit w ol in om n-m nta N ly of Income or Consumption Total Builing Capacity) (f) School Capacity Utilization in Urb 30% 32% Education spending as share of GDP Education spending as share of GDP ud n ts No ou tO 50% 70% nta u St u M < aC C m M No hly ou Sc ol in w ol in om 33% 29% 31% (Student Enrollment as % of MASTERS AND No ou orly Targeted) 40% 60% Hig M UNQUALIFIED < aC C o eral Allocation From All Sources M 2025 2013 hly On Sch ith Sc ol in ho 32% 28% 70% 30% 30% Total Builing Capacity) 50% Hig /BEGINNER SECONDARY HIGHER PH.D DEGREES 100 nly o ly On Sch ith TEACHERS TEACHER PEDAGOGICAL DEGREES ho 33% 29% tO rnal Generated Revenue (IGR) 31% 27% 60% 20% 40% nly No 0 SCHOOL SCHOOL ly 60% 80% 100% 32% 28% 75 tO 30% 26% 50% 10% 70% 30% re of IGR (right axis) 90 No 158 29% l s 31% 27% 60% 40% ra ou 20% an 0% nta us s 0 m y en ity Education (Ministry ofExpenditure Public Review Guidelines Ru ou UNQUALIFIED Source: Kosovo EMIS MASTERS AND 40 om nit in ou ino Urb ud n ts 30% Capita Revenue (right axis) l us 26% in 28% u nta St u 50% 10% OF POPULATION 30% HIGHER 6.1: ALBANIA PUBLIC DEBT FIGURE TO GDP RATIO UNDER INCREASED m n-m nta ra an nta us 80 /BEGINNER SECONDARY PH.D DEGREES s 00 om ity o en ity ou 29% Ru ou w ol in om in s ou ino l Urb 100 un ud n ts No ou Source: NSI except Poland (BOOST data), Slovakia Education), statistics authorities of CIS countries. 27% ou in nta 40% St mu M ra an 0% nta us 0 m y s en ity < aC TEACHERS TEACHER PEDAGOGICAL DEGREES C 20% ou wit l in mm Ru n-m nta ou M 40 om nit hly in ou ino Urb ou ch l in ud n ts l s 70 in 28% u nta St u 2008 No ou ou FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED Co 26% Hig 30% m SCHOOL SCHOOL n-m nta ra o an M nta us 0 m y s en ity < C 10% ou Ru ou ou nly cho 40 om nit 70 M w ol in om hly in ou ino Urb ith h a in ud n ts o No ou 27% in nta u St u 90 l s M 4 < aC ho in C S 20% ou ino Sch ol m n-m nta M ra ou Hig hly an 0% s S ou nta us nta us nly s 0 m y en ity o 2013 w ol in om oo Ru ou 40 om nit n-m nta On Sch No ou in ou ino Urb s Sc ol 26% Hig ud n ts ou ino t O l s M < aC in 10% C u nta St u Source: Kosovo EMIS ou nly ho M in O ou FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED “Western Europe” includes other members of the European Union not mentioned in the gure, less Malta and hly m n-m nta ith ra an Sc ol in nly 0 m y o en ity ou 60 Sc Ru 40 om nit s n-m nta N ly w ol in om in l Urb ou Hig ud n ts 80 No ou 75 tO ura an 0% nta us nta us nly s 0 m y en ity o nta u St u M < aC C 40 om nit On Sch m ith M ho R in Urb No 65 ud n hly ts ou ino t O ou Sch ol in l s in w ol in om u nta St u in O ou FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED No ou m ra an nly 0 m y o en ity Hig M ou Ru 40 om nit < aC C w ol in om n-m nta N ly o M in 2008 Urb 50 ud n ts ou hly 75 tO On Sch ith Sch ol in o nta u St u M < aC C 70 m M No Hig hly ou Sc ol in FIGURE 78: CLASS AND SCHOOL SIZES ARE SMALL w ol in om nly o Cyprus, plus Norway, 40 Iceland and Switzerland. OECD at a Glance, 2011. ou ly On Sch ith o Hig No M tO < aC C 2013 o M hly On Sch ith Sc ol in ho nly No ly Hig No 70 nly 75 tO 60 o 60 ly On Sch ith ho tO No nly No ly 75 70 tO 30 Notes: Year of reference 2008. Public institutions only (including for Belarus). No 50 EDUCATION SECTOR FINANCE, 2013 al Government Budget, and General Household Survey Panel 2012/13 40 20 FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, 70 % GDP FIGURE 5.11:65 EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 55 10 Source: Kosovo BOOST Source: Kosovo BOOST 70 63.1 65 9 | 1% Number of Students 30 20 per Classroom 0 FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: 65EDUCATION EXPENDITURES BY ECONOMIC60 CATEGORY, % GDP Number of Students per Primary School 63.1 50 4.5% 4.5% 60 Source: Kosovo BOOST Source: Kosovo BOOST 63.1 65 Yobe Gombe Jigawa Delta Zamfara Kaduna Katsina Kebbi Sokoto Kano Benue Niger Kogi Lagos Nasarawa Plateau FTC Abuja Kwara Enugu Imo Anambra Ebonyi Abia Bayelsa Akwa Ibom Delta Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun Lagos 426 | 18% 10 56.9 4.0% 4.0% 63.1 ALL LGA 56.9 45 60 55 54.9 Household 0 4.5% 4.5% FIGURE 16: TEACHER SALARIES60 RELATIVE TO PER CAPITA GDP, 2012 55 295 | 13% 3.5% 3.5% 54.9 UBEC Expenditure on Education and Teachers’ Salaries 500 Abia Bayelsa Akwa Ibom Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun 56.9 40 4.0% 4.0% ast North West State South5.11: North Central FIGURE East EDUCATION South SouthEXPENDITURES South West BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 3.0% 55 Capital Outlays 50 Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 Capital Outlays 49.7 3.0% 56.9 40 54.9 50 Federal Gov’t Source: Kosovo BOOST FIGURE 5.11: Source: Kosovo BOOSTEDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES 3.5% BY ECONOMIC CATEGORY, % GDP FIGURE 3.5% 16: TEACHER Subsides SALARIES RELATIVE TO PER CAPITA GDP, 2012 & Transfers 55 Subsides & Transfers 54.9 49.7 2008 2009 201 Number of Students per Primary School 587 | 25% Donors Source: Kosovo BOOST Source: Kosovo BOOST 2.5% 2.5% Expenditure on Education and Teachers’ Salaries Utilities Utilities RE t 5.11: EDUCATION South South EXPENDITURES South West BY ECONOMIC 4.5% CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC FIGURE Source: Kosovo5.11: CATEGORY, % GDP 4.5% CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC 3.0% 2.0% Capital Outlaysof Students per Primary School Number Source:50 3.0% OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Goods & Services 2.0% 45 Policies Capital Outlays and Practices COUNTRIES (Volume IV), pg. 96 49.7 AND ECONOMICS WITH 50 OVER USD Goods & Services 45 COUNTRIES AND ECONOMICS WITH Combine Incr 35 BOOST EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP PER CAPITA GDP 20 000 49.7 PER CAPITA GDP LESS USD 20 000 ovo BOOST 4.5% 4.5% 2.5% Subsides & Transfers 2.5% Subsides & Transfers Health Spend Source: Kosovo BOOST Source: Kosovo BOOST 1.5% Wages & Salaries 1.5% Wages & Salaries 4.0% 4.0% Utilities 45 40 Utilities 200 2.5 Education Sp 4.0% 4.0% As a result of demographic change400 FIGURE 3: AN OVERVIEW OF THE COMPLEXITY OF TRANSFERS &4.5% FUND FLOWS, 2012 2.0% Goods & Services 2.0% COUNTRIES AND ECONOMICS WITHGoods & Services 201445 2010 2011 2012 2013 COUNTRIES AND ECONOMICS WITH 40 Schools in rural areas are particularly resource intensive. and 4.5% 1.0% 1.0% 2008 2009 2015 2016 2017 2018 2019 2020 2021 2022 Baseline RE OF BASIC EDUCATION FINANCING 3.5% 4.5% 3.5% 4.5% PER CAPITA GDP OVER USD 20 000 180 PER CAPITA GDP LESS USD 20 000 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 3.5% 3.5% Wages & Salaries Wages & Salaries 500 1.5% 1.5% 30 500 4.0% 4.0% 0.5% 0.5% 3.0% 4.0% 3.0% 4.0% 40200 160 2.0 deral LEVEL BUDGET Capital Outlays Capital Outlays Combine 40 2.5 VAT Independent Devp’t 3.0% Capital Outlays 3.0% 1.0% Capital Outlays 1.0% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Increase 2019 2020in 2021 Education 2022 and Health and Spending OF THE COMPLEXITY count & 3.5% Pool erivatives OF TRANSFERS Revenue Partners FUND FLOWS, 2012 & 2.5% 3.5% Subsides & Transfers 2.5% Subsides & Transfers 0.0% 0.0% Health Combine 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 140 Spending Increases from 2.6 to 4 percent of GDP 2018 Increase 2019 2020in 2021 Education 2022 and Health and Spending MINISTRY OF FINANCE 3.5% MINISTRY OF EDUCATION CULTURE 3.5% Subsides & Transfers Subsides & Transfers 180 Health Spending Increases from 2.6 to 4 percent of GDP urbanization, rural areas have seen the sharpest declines in population and student numbers. While NATIONAL 2.5% 2.5% 2007 2008 2009 2010 2011 2012 2007 2008 2009 2010 2011 2012 MOF MOEC Utilities Utilities 0.5% 0.5% 3.0% Federal 3.0% Utilities Utilities Education Spending Increases from 3 to 4 percent of GDP 2.0% Capital Outlays 3.0% 2.0% Outlays CapitalOutlays 3.0% 120 160 Combine Increase in Education and Health and Spending 2.0 Education Spending Increases from 1.5 3 to 4 percent of GDP BUDGET Ministry of Goods & Services Capital Goods Capital Outlays && Services Baseline 25 Adjustment Fund 2.0% 2.0% Combine Increase in Education and Health and Spending Finance Subsides & Transfers Subsides& &Transfers Transfers Goods Services 0.0% Goods & Services 0.0% Baseline 300 Tugas Central 2.5% Federal Dekon 2.5% Wages & Salaries Subsides Wages Subsides & Transfers Health Spending Increases from 2.6 to 4 percent of GDP && Salaries State DAU DAK BOS 100 MINISTRY OF FINANCE Con. Rev. 1.5% MINISTRY OF EDUCATION CULTURE Pembantuan 2.5% Functions 1.5% 2.5% Wages Salaries 2007 2008 2010 2009Wages 2011 & Salaries 2012 140 2007 2008 2009 2010 2011 2012 Health Spending Increases from 2.6 to 4 percent of GDP Utilities 1.5% Utilities Utilities 1.5% Utilities Education Spending Increases from 3 to 4 percent of GDP MOF Federal MOEC Education Spending Increases from 3 to 4 percent of GDP rural schools tend to be 400 2.0% 2.0% 1.0 400 continues to be a quite substantial misalignment in smaller everywhere, there Budget 1.0% Goods & Services 2.0% 1.0% Goods& Services &Services 2.0% 120 Baseline 80 1.5 Adjustment Fund Own 1.0% Goods 1.0% Goods & Services Baseline Central 1.5% UBFC (2%) Tugas Dekon Source Wages & Salaries 1.5% Wages& Wages &Salaries Salaries Wages & Salaries 60 20 U DAK BOS PROVINCIAL District Budget Pembantuan Revenue 0.5% Functions DAU/SDA Provincial Education 1.5% Central Government 0.5% 1.5% 100 Agencies in the Regions 0.5% 0.5% FMF PAD 1.0% 1.0% 1.0% 1.0% 80 40 1.0 0.5 0.0% 0.0% 0.0% 0.0% the demand and supply of education services between rural and urban areas (Figure 79). 200 y Non-Salary Salary 0.5% Non-Salary District Budget Salary Non-Salary 2007 2008 2009 2010 2011 2012 0.5% 2007 2008 2007 2009 2010 2008 2009 20112012 2010 2011 2012 2007 2008 2009 2010 2011 2012 60 20 DAU/SDA Provincial Education Own Central Government District Budget 0.5% 0.5% ower Lower Unity Agencies in the Regions 15 PAD DISTRICT Source DAU/SDA econdary Secondary College Revenue PAD District Education 40 0 FIGURE 15: EVOLUTION0.5 0 ECONOMIC OF REAL EDUCATION EXPENDITURES BY BROAD FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC 0.0% UBEB 0.0% 0.0% 0.0% 300 DAK 300 2007 2008 2009 2010 2011 2012 2007 2007 Fund Originated 2008 2008 2009 2010 2009 2010MoF from 2011 2012 2011 2012 2007 2008 2009 2010 2011 2012 CATEGORY SOURCE: DATA FROM MFB/SIGFP. CATEGORY SOURCE: DATA FROM MFB/SIGFP. Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania g UBE Act of 2004 ernment Account (SJLGA) Fund Originated from MoEC Sectoral Budget 20 District Budget FIGURE 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS e from the Federation Fund Originated from Provincial Budget FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC of Nigeria). DAU/SDA District Education 0 0 FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC PAD Fund Originated from District Budget FIGURE 5.12: DISTRIBUTION OF TEACHERS BY CATEGORY SOURCE: DATA FROM MFB/SIGFP. 700 CATEGORY SOURCE: DATA FROM MFB/SIGFP. 700 10 DAK Fund SCHOOLS from MoF Originated PRIVATE PUBLIC SCHOOLS Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania Fund Originated from MoEC Sectoral Budget 600 Fund Originated from Provincial Budget FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 700 600 100 700 BILLION CONSTANT 2013 AR. BILLION CONSTANT 2013 AR. Source: Elaboration based on Permendagri 13/2006 on Guidelines of Sub-National Financial Management, World Bank (2009) and Law 22/2011 on AP3N 2012 Fund Originated from District Budget 12k 500 Note: Adjustment Fund also includes the local incentive grant (Dana Insentif Daerah, or DID) Source: Belstat, MOE. 200 500 Figure 79: Most students are in urban area while most schools are in rural areas PUBLIC SCHOOLS Cumulative Expenditure by Educational 200 PRIVATE SCHOOLS 5 600 600 Institutions per Students Aged 6 – 15 400 BILLION CONSTANT 2013 AR. Lower Secondary Teachers’ Salaries (after 400 BILLION CONSTANT 2013 AR. ational Financial Management, World Bank (2009) and Law 22/2011 on AP3N 2012 12k 10k 15 years of Daerah, or DID) Cumulative Expenditure by Educational 500 experience/minimum training relative to per capita GDP 500 Institutions per Students Aged 6 – 15 Lower Secondary Teachers’ Salaries (after 300 15 years of 300 8k 400 0 experience/minimum training relative to per capita GDP FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 400 0 10k FIGURE 5.12: DISTRIBUTION OF TEACHERS BY Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP 200 200 300 300 100 Lower Secondary Teachers’ Salaries (after 15 years of 100 6k FIGURE 12: FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 8k 100 100 ROU LTU BGR LVA EST RUS UKR EUROPE MDA SVN SVK FIGURE 5.12: DISTRIBUTION OF TEACHERS BY experience/minimum training relative to per capita GDP 12k 200 200 CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK DNK SVK 250 12k 4 7 6 6 5 5 4 6 8 100 6k 4k 0 0 12 200 90 10k 100 100 100 80 12k 12k 10k 2006 2008 2009 2010 2011 20 4 7 6 6 5 5 4 6 8 70 4k 2k 2006 2008 0 2009 2010 2011 2012 2013 150 100 60 65 66 59 65 55 48 55 59 50 0 TOR FINANCE, 2013 90 8k 0 10k Capital Expenditures PERCENT 8k 50 0 10k 80 Capital Expenditures 2006 2008 2009 2010 2011 2012 2013 40 0 2006 2008 2009 2010 2011 2012 2013 90 100 70 2k Other Recurrent Expenditure HYPOTHESIS: SIMULATING THE IMPA Other Recurrent Expenditure 22.6 30 UNQUALIFIED MASTERS AND INTEGRATION OF COMMUNITY TEAC 10 60 65 66 59 65 55 48 55 59 50 8k 6k /BEGINNER SECONDARY HIGHER PH.D DEGREES Regular Salaries 20 Capital Expenditures Regular Salaries ROU LTU BGR LVA EST RUS 100 6k WEST. EUROPE MDA SVN BLR GEO HUN HRV SVK WEST. 50 8k TEACHERS TEACHER PEDAGOGICAL DEGREES PERCENT 50 Capital Expenditures ROU LTU BGR LVA EST RUS WEST. EUROPE MDA SVN BLR GEO HUN UKR HRV SVK 10 0 SCHOOL SCHOOL Other Recurrent FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY Expenditure AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 DNK SVK 40 90 31 28 35 29 40 47 41 34 41 600 Other Recurrent Expenditure MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR DNK SVK 0 80 4k 35 UNQUALIFIED Source: Kosovo EMIS MASTERS AND 0 30 80 4k 2009 2001 2002 2003 2004 2005 2006 6k 2007 2008 2009 /BEGINNER SECONDARY HIGHER PH.D DEGREES Source: own calculations using NUPTIK Data, 2010 Regular Salaries 100 2001 6k 2002 2003 2004 2005 2006 2007 2008 Regular Salaries 573 Source: NSI except Poland (BOOST data), Slovakia (Ministry of Education), statistics 90 20 10 authorities of CIS countries. 70 TEACHERS TEACHER SCHOOL PEDAGOGICAL SCHOOL 2008 DEGREES 550 561 2013 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIRHYPOTHESIS: HIGHEST EDUCATION PROVINCE, BY THE IMPACT 2010 500 60.4 Central 31 28 35 29 40 47 41 34 41 4k 2k Source: Kosovo EMIS SIMULATING OF THE “Western Europe” includes other members of the European Union not mentioned in the gure, less Malta and 80 District 0 60 2k HYPOTHESIS: SIMULATING THE IMPACT OF THE 80 8 4k INTEGRATION OF COMMUNITY TEACHERS 70 (Sistem Informasi Keuangan Daerah, SIKD), Ministry of Finance Province 2001 2003 2002 2004 2005 2006 2007 2008 2009 Source: own calculations using NUPTIK Data, 2010 % OF TEACHERS WITH SENIOR SECONDARY 2007 2008 2009 OR BELOW AS HIGHEST EDUCATION LEVEL 50 2008 INTEGRATION OF COMMUNITY TEACHERS 70 Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. 60 40 0 2k 0 2013 HYPOTHESIS: SIMULATING THE IMPACT OF THE 600 HYPOTHESIS OF 400 2k 30 UNQUALIFIED /BEGINNER SECONDARY HIGHER MASTERS AND PH.D DEGREES HYPOTHESIS: SIMULATING THE IMPACT OF 600THE 80 INTEGRATION OF COMMUNITY 60 TEACHERS HYPOTHESIS OF 586 600 600 RESOURCE NEEDS Notes: Year of reference 2008. Public institutions only (including for Belarus). 50 UNQUALIFIED MASTERS AND INTEGRATION OF COMMUNITY TEACHERS 600 600 573 % OF TEACHERS WITH SENIOR SECONDARY TEACHERS TEACHER PEDAGOGICAL DEGREES 60 OR BELOW AS HIGHEST EDUCATION LEVEL /BEGINNER SECONDARY HIGHER PH.D DEGREES 586 RESOURCE NEEDS 561 20 0 SCHOOL SCHOOL 573 550 H1 40 FIGURE 13: TEACHERS TEACHER PEDAGOGICAL DEGREES 561 600 500 300 0 SCHOOL SCHOOL UNQUALIFIED Source: Kosovo EMIS MASTERS AND 550 40 H1 600 600 HYPOTHESIS OF H2 350 60060 10 600 586 500 RESOURCE NEEDS 6 30 /BEGINNER SECONDARY HIGHER PH.D DEGREES 229 HYPOTHESIS OF 573 H2 es of CIS countries. 600 UNQUALIFIED Source: Kosovo EMIS MASTERS AND TEACHERS TEACHER PEDAGOGICAL DEGREES 2008 586 561 20 100 2 /BEGINNER SECONDARY 0 HIGHER 100 PH.D DEGREES SCHOOL SCHOOL 573 RESOURCE NEEDS 550 H1 MOBILIZATION 200 50 TEACHERS TEACHER PEDAGOGICAL DEGREES 10 12 11 2008 90 2013 561 500 H1 20 400 Plateau FTC Abuja Kwara Enugu Imo Anambra Ebonyi Abia Bayelsa Akwa Ibom Delta Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun Lagos SCHOOL SCHOOL POTENTIALSCENARIO Source: Kosovo EMIS 550 MOBILIZATION H2 e, less Malta and 40 23 10 9 80 15 Source: Kosovo EMIS 6 2013 80 500 400 H2 POTENTIALSCENARIO S1 0 70 2008 FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 MOBILIZATION 511 100 S1 S2 60 2008 44 60 2013 Expenditure on Education and Teachers’ Salaries 400 511 0 300 428 POTENTIALSCENARIO 100 20 MOBILIZATION 40 90 PERCENT 350 PERCENT entral South East 23 2013 South West South South 100 50 Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 400 428 S2 51 300 POTENTIALSCENARIO S1 77.4 4 80 229 FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat 350 511 40 75 40 70 73 511 229 S1 428 S2 0 Expenditure on Education and Teachers’ Salaries 300 200 30 0 Jama 60 COUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH 350 Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 428 S2 PERCENT 20 (B MGA) 47 20 300PER CAPITA GDP OVER USD 20350 200 2016 2017 20 30 100 000 PER CAPITA GDP LESS USD 20 000 229 65 51 50 Nuggoe Acoh 25 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta DI. Yoguaunta 10 40 200 229 2.5 0 0 200 100 30 OF TRANSFERS AN OVERVIEW OF THE COMPLEXITY EARLY BASIC SECONDARY & FUND FLOWS, UNIVERSITIES OTHERS 2012 2008 2009 2008 2009 COUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH PER CAPITA GDP OVER USD 20 000 180 PER CAPITA GDP LESS USD 20 000 CHILDHOOD EDUCATION 20 EDUCATION 200 100 2 POL EDUCATION FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 20 25 TOTAL LEVEL 10 BUDGET 200 160 2.5 2.0 100 0 POL 39.6 Expenditure on Education and Teachers’ Salaries ANSFERS & FUND FLOWS, 2012 0 FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 140 UNIVERSITIES ONAL Central MINISTRY OTHERS OF FINANCE MINISTRY OF EDUCATION CULTURE Central Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 180 MOF Central-80S 2008 2009 MOEC 2008 2009 Expenditure on Education and Teachers’ Salaries District 100 0 (B MGA) 2016 2017 2018 2019 2020 FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 160 120 2.0 1.5 DGET District TOTAL Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 Province (B MGA) 2018 0 Adjustment Fund 10 DAUProvince DAK FIGURE 16: BOS TEACHER SALARIES Pembantuan RELATIVE Dekon TO PER CAPITA GDP, 2012 Functions Tugas Central Expenditure on Education and Teachers’ Salaries 140 100 2016 2017 2019 2020 NISTRY OF EDUCATION CULTURE COUNTRIES Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH Central (B MGA) MOEC Expenditure on Education and Teachers’ Salaries PER CAPITA GDP OVER USD 20 000 PER CAPITA GDP LESS USD 20 000 0 1.0 2016 2017 2018 2019 2020 0 District 120 80 1.5 COUNTRIES Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH Own (B MGA) s VINCIAL bantuan Dekon Source Central Functions District Budget Province Provincial Education PER CAPITA GDP OVER USD 20 000 Central Government PER CAPITA GDP LESS USD 20 000 COUNTRIES AND ECONOMICS WITH 200 COUNTRIES AND ECONOMICS WITH 2.5 100 60 2016 2017 2018 2019 2020 0 Revenue DAU/SDA Agencies in the Regions PAD 200 PER CAPITA GDP OVER USD 20 000 180 PER CAPITA GDP LESS USD 20 000 0.5 2.5 80 40 1.0 COUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH 200 160 2.5 2.0 ovincial Education Central Government PER CAPITA GDP OVER USD 20 000 180 PER CAPITA GDP LESS USD 20 000 60 20 Own District Budget Agencies in the Regions 140 Students Teachers Schools STR TRICT Source DAU/SDA 180 0 Revenue 200 PAD DAK District Education 160 2.5 2.0 40 0 0.5 PROJECTED FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS ENROLLMENT AS THEIR GAP AND OPERATING HIGHEST EDUCATION BY PROVINCE, 2010 Fund Originated from MoF 160 120 2.0 1.5 EXPENDITURE GAP DUE TO POPULATION GROWTH Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania 180 140 Fund Originated from MoEC Sectoral Budget 20 Source: own calculations using NUPTIK Data, 2010 100 PROJECTED ENROLLMENT GAP AND OPERATING FIGURE 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS Fund Originated from Provincial Budget 140 120 Fund Originated from District Budget 1.5 0 PROJECTED ENROLLMENT GAP AND OPERATING District Education 160 2.0 0 EXPENDITURE GAP DUE TO POPULATION GROWTH Fund Originated PRIVATE SCHOOLS from MoF PUBLIC SCHOOLS 120 80 1.5 1.0 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS EXPENDITURE THEIR HIGHEST EDUCATION GAP BY PROVINCE, DUE TO POPULATION 2010 GROWTH Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania 140 100 300k 80 Fund Originated from MoEC Sectoral Budget 1.4M 60 PROJECTED using NUPTIK Data, GAP ENROLLMENT Source: own calculations 2010 AND OPERATING % OF TEACHERS WITH SENIOR SECONDARY Fund Originated from Provincial Budget 100 OR BELOW AS HIGHEST EDUCATION LEVEL mendagri 13/2006 on Guidelines of Sub-National Financial 120 80 1.5 1.0 es the local incentive grant (Dana Insentif Daerah, or DID) Management, World Bank (2009) and Law 22/2011 on AP3N 2012 Fund Originated from District Budget 0.5 EXPENDITURE GAP DUE TO POPULATION GROWTH 300k 1.4M 200k 1.3M 40 1.0 Source: Belstat, MOE. SCHOOLS 80 Cumulative Expenditure by Educational ENROLLMENT GAP 100 60 60 300k 1.4M Institutions per Students Aged 6 – 15 100k 1.2M 60 20 80 200k 1.3M 4.0 FUNDING GAP 1.0 0.5 Lower Secondary Teachers’ Salaries (after 15 years of % OF TEACHERS WITH SENIOR SECONDARY 80 40 OR BELOW AS HIGHEST EDUCATION LEVEL 2 200k 1.3M 0 ENROLLMENT GAP experience/minimum training relative to per capita GDP - Rural City of Minsk Vite 40 0 0.5 300k 1.4M 1.1M Urban Cumulative Expenditure by Educational Rural 100k 1.2M ENROLLMENT GAP 20 60 Lower Secondary Teachers’ Salaries (after 15 years of 40 3.0 FUNDING GAP Institutions per Students Aged 6 – 15 100k 1.2M Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania 20 experience/minimum training relative to per capita GDP 200k 60 1.3M (100k) 1M FUNDING GAP 40 0 0.5 0 Lower Secondary Teachers’ Salaries (after 15 years of - 1.1M 4.0% FIGURE 3.2: SOURCES OF FUNDS: 2010 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS 0 experience/minimum training relative to per capita GDP ENROLLMENT GAP - 1.1M 0 100k 1.2M (200k) 900k 2.0 Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania 20 Lower Secondary Teachers’ Salaries (after 15 years of (100k) 20 1M 3.0% FUNDING GAP Urban Grodno Oblast Mog 40 Hong Kong-China Germany Korea Portugal Slovak Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania experience/minimum training relative to per capita GDP (100k) 1M 100 (300K) N URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS 800k 0 6.9 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 - 2012 2013 2014 2015 2016 2017 2018 2019 2020 1.0 (200k) 1.1M 900k RCES OF FUNDS: 2010 4 7 6 6 5 5 4 6 8 100 16 900k 2.0% (200k) 12 0 Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania at, MOE. 90 10 Cumulative Expenditure by Educational (100k) 1M 0.0 (300K) 800k 80 2017 2018Costs Operating 2020 Current Funding and GER 2019 with 20 100 80 Institutions per Students Aged 6 – 15 2012 2013 2014 2015 2016 2018 (300K) 2019 2020 2012 2013 2014 2015 2016 800k 2017 15 2012 2013 2014 2015 2016 2017 2018 2019 2020 GDP Growth: 1% 2012 2013 2014 2015 2016 2017 2018 2019 2020 1.0% Brest Oblast Gom DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat 70 (200k) 900k 4 7 6 6 5 5 4 6.9 6 8 Lower Secondary Teachers’ Salaries (after 15 years of 60 65 66 59 65 55 48 55 Cumulative Expenditure 29 59 50 by Educational experience/minimum training relative to per capita GDP GDP Growth: 2% Cumulative Expenditure by Educational 800k with Current Funding and GER Operating Costs 0 PERCENT 60 10 50 Institutions per Students Aged 6 – 15 Lower Secondary Teachers’ Salaries (after 15 years of (300K) Operating Costs with Current Funding and GER GDP Growth: 3% 0.0% FIGURE 75: NUMBER OF STU 20 40 Institutions per Students Aged 6 – 15 2012 2013 2014 2015 2016 2017 2018 2019 2020 GDP Growth: 1% 2012 2013 2014 2015 2016 2017 2018 2019 2020 90 Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP GDP Growth: 1% GDP Growth: 4% Nuggoe Acoh Lower Secondary Teachers’ Salaries (after 15 years of 22.6 30 experience/minimum training relative to per capita GDP GDP Growth: 2% Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Student-teacher ratio in secondary schools 10 65 66 59 65 55 48 55 59 50 Cumulative Expenditure by20 Educational experience/minimum training relative to per capita GDP GDP Growth: 2% Minsk Oblast Bela 40 29 Institutions per Students Aged Lower Secondary Teachers’ Salaries (after 15 years of Operating Costs with Current Funding and GER GDP Growth: 3% 10 6 – 15 Lower Secondary Teachers’ Salaries (after 15 years of GDP Growth: 3% FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY 31 Lower Secondary Teachers’ Salaries 28 35 29 (after 15 years of 40 experience/minimum 47 41 34 41 training relative to per capita GDP experience/minimum training relative to per capita GDP FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY GDP Growth:OR LESS AS THEIR HIGHEST EDUCATION 1% GDPBY PROVINCE, Growth: 4% 2010 0 GDP Growth: 4% 80 35 2004 2005 2006 experience/minimum 2007 2008 2009 training relative to per 2001 2002 2003 capita GDP 2006 2004 2005 2007 2008 2009 Source: own calculations using NUPTIK Data, 2010 GDP Growth: 2% 20 Lower Secondary48 Teachers’ Salaries (after 15 years of 54 GDP Growth: 3% experience/minimum training relative to per capita GDP 12 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION 60.4 BY PROVINCE, 2010 31 28 35 29 40 47 41 34 41 GDP Growth: 4% 12k 80 70 8 ngan Daerah, SIKD), Ministry of Finance 2001 2002 2003 2004 2005 2006 2007 2008 2009 Source: own calculations using NUPTIK Data, 2010 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL 12 0 54 Non SBM Grantee SBM Grantee 60 FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY FIGURE 76:10k CONSOLIDATION OF SCHOO 80 12k Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). % OF TEACHERS WITH SENIOR SECONDARY 60 OR BELOW AS HIGHEST EDUCATION LEVEL 10 PTCA 1991=100 LGU 40 FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, FIGURE ESPECIALLY 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY FIGURE 76: CONSOLIDATION All OFAgesSCHOOL FIGURE 76: NETWORKCONSOLIDATION OF SCHOOL NETWORK 8k FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 60 10k Rural 6 SBM Grantee Others Age 5-14 Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). 130 35 100 Urban 10 Source: own calculations using NUPTIK Data, 2010 1991=100 50 1991=100 11 Community 23 Age 5-24 12 90 20 All Ages 6k 9 80 40 FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY FIGURE 76: CONSOLIDATION All Rural OFAgesSCHOOL NETWORK Rural 8k FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS FIGURE 67: SHARE OFAS THEIR HIGHEST TEACHERS EDUCATION WITH SENIOR BY PROVINCE, SECONDARY OR LESS2010 AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 60.4 6 70 12k Age 5-14 Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE Age 5-14 (2011). 130 130 110 00 80 Urban Source: UN, Beistat 8 44 Source: own calculations using NUPTIK Data, 2010 Source: own calculations using NUPTIK Data, 2010 0 1991=100 Urban % OF TEACHERS WITH SENIOR SECONDARY 60 Age 5-24 Age 5-24 OR BELOW AS HIGHEST EDUCATION LEVEL 20 4k 40 90 PERCENT 100 51 50 All Ages 6k 77.4 4 80 Rural DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 FIGURE 40 60.4 60 70 73 12k 12k Age 5-14 10k 130 110 110 90 80 80 Source: UN, Beistat Source: UN, Beistat 8 30 Source: own calculations using NUPTIK Data, 2010 0 Urban 2k % OF TEACHERS WITH SENIOR SECONDARY 60 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL Age 5-24 4k OR BELOW AS HIGHEST EDUCATION LEVEL 47 20 30 65 50 Nuggoe Acoh 25 40 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta 10 40 60 10k 8k 90 70 60 10k FIGURE 67: 110 90 6 30 2008 29: 80 APPENDIX FIGURE SHARE2008 OF TOTAL 0 HOUSEHOLD EXPENDITURES 12k Source: UN, Beistat 0 BASIC SECONDARY UNIVERSITIES OTHERS 2009 2009 2k % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL UCATION EDUCATION 20 Source: own calculations using School Based Management 2 Share of total household expenditures, by region 20 20 TOTAL 10 40 6k 1970 1975 1980 19 Source: EMOP 2014 60 40 10k 8k 8k 70 50 39.6 90 70 6 0 Central FIGURE 67: 0 E 29:2008 SHARE OF TOTAL 2009 2008 HOUSEHOLD 2009 EXPENDITURES District 0 50 tures, by region TOTAL Province 2% 20 20 6k 6k 4k Source: own calculations using School Based Management 1970 1975 1980 1985 1990 1995 2000 2005 2010 10 40 4 8k 50 DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Central 70 45 50 1991 1993 1995 1997 1 0 0 District 0 4k 2k Province 20 6k 4k 50 40 65 2% 15% Nuggoe Acoh STUDENT TEACHER RATIO 0 Jama Timur Dou Jauta Awa Tungah Bali Yoguaunta Jamaluat Banton Sunatrabarat Dulosalami Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta 4 50 DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Timur Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Bamako 45 0 35 1991 1993 1995 1997 2001 of Schools 1999 Number 2003 2005 2007 Jama Koulkoro 2k 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 0 Dou Jama 4k 2k Number of Students 2 Koyos DI. 40 Nuggoe Acoh 11% 1990 1995 2000 2005 1985 30 1970 1975 1980 2010 2015 2020 2025 2030 Students Teachers Schools STR Urban 15% Nuggoe Acoh Number of Teachers STUDENT TEACHER RATIO FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Gilo 39.6 Bamako 50% Number of Schools Segou Source: own calculations using NUPTIK Data, 2010 0 35 199125 1993 1995 1997 1999 2001 of Schools Number 2003 2005 2007 2009 S Koulkoro 2k 0 Number of Students Mepti 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 Number of Students P 2 2% Koyos 11% FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY 1970 OR 1975 LESS 1980 AS THEIR 1985 HIGHEST1990 1995 EDUCATION 2000 2005 BY PROVINCE, 30 2010 2015 2020 2025 2030 2010 Number of Teachers Gilo 10% Silcasco 20 Number of Teachers 80 39.6 50% Tombouctou Source: own calculations using NUPTIK Data, 2010 Number of Schools % OF TEACHERS WITH SENIOR SECONDARY 0 0 Segou OR BELOW AS HIGHEST EDUCATION LEVEL 25 15 of Students Number Mepti 10% 1970 1975 1980 1985 1990 1995 2005 2000 60 2010 2015 2020 2025 2030 Number of Teachers Silcasco 20 FIGURE 78: 10% 80 10 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL Tombouctou Source: EdStats database. 2% Rural 0 40 City of Minsk 15 5 Vitebsk Oblast Primary school student-teacher ratios in the ECA region (1999-2009) Secondary school student-teacher ratio ts Teachers Schools STR Urban Rural 10% FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 60 FIGURE 78: GURE 3.2: SOURCES OF FUNDS: 2010 10 FIGURE 78: Source: own calculations using NUPTIK Data, 2010 0 Source: EdStats database. 30 18 20 Source: EdStats database. Urban Grodno Oblast Mogiev Oblast 2% 40 5 0 1 Primary school student-teacher ratios in the ECA region (1999-2009) 2 3 4 5 6 7 Secondary school student-teacher ratios in the ECA region (1999-2009) 16 6.9 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 Primary school student-teacher ratios in the ECA region (1999-2009) 25 Secondary school student-teacher ratios in the ECA region (1999-2009) Schools STR Urban Rural 16 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS2010 80 THEIR HIGHEST EDUCATION BY PROVINCE, 2010 FIGURE 78: 14 10 Source: own calculations using NUPTIK Data, 0 0 30 18 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL Source: own calculations using NUPTIK Data, 2010 20 Source: EdStats database. 30 20 18 12 15 FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 BUDGET FOR STUDENT IN RUPIAH 16 Brest Oblast Gomel Oblast DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat 29 60 Primary school student-teacher ratios in the ECA region (1999-2009) 0 1 2 3 25 4 5 6 7 Secondary school student-teacher ratios in the ECA region (1999-2009) 16 10 FIGURE 67: Source: EMOP 2014 80 SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 14 0 15 % OF TEACHERS WITH SENIOR SECONDARY 25 8 OR BELOW AS HIGHEST EDUCATION LEVEL 20 80 14 City of Minsk Vitebsk Oblast Source: own calculations using NUPTIK Data, 2010 30 20 18 12 % OF TEACHERS WITH SENIOR SECONDARY Nuggoe Acoh OR BELOW AS HIGHEST EDUCATION LEVEL 40 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta BUDGET FOR STUDENT IN RUPIAH 10 6 URE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 20 12 Brest oblast Vitebsk oblast Gomel oblast 60 City of Minsk Grodno 10 Minsk Oblast Belarus 16 15 4 OP 2014 80 60 25 14 5 10 8 20 15 2 % OF TEACHERS WITH SENIOR SECONDARY Oblast Mogiev Oblast 8 OR BELOW AS HIGHEST EDUCATION LEVEL 40 10 6 City of Minsk Vitebsk Oblast 48 54 20 12 0 0 40 10 4.14: FISCAL IMPACT FIGURE 4.14: FISCAL IMPACT OF PARENTAL F 6 4 OF PARENTAL FEE ELIMINATION BY REGION 60 0 10 HUN MDA UZB ROU RQU CVE AZE LVA KAZ KGZ SVN KGZ POL TJK TJK MKD LTA ARM 5 BLR SVK RUS UKR SVK BGR 20 15 8 4 2 Grodno oblast Mogiev BrestMinsk oblast Mogiev oblast Belarus 5 Oblast Gomel Oblast DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat M Grantee 20 SBM Grantee 0 2 0 Grodno Oblast Oblast 40 10 6 0 2009 KGZHUN HUN MDA MDA AZEUZB POLROU UZB UKRRQU CVE TJK AZE RQU LVA CZE HUNKAZ KAZ KGZ LTASVN POLKGZ SVN SVK POL POL TJK MKD LTA CZE TJK MKD ARM BLR SVK UKRRUS UKR SVK UKR BGR BGR 0 0 4 ost as percent of GDP Additional cost as percent of GDP 0 FIGURE 67: 2004 HUN Nuggoe Acoh MDA MDA BLR UZB ROU UZB LTU CVE AZE LVA LVA KAZ KAZ KGZ SVN SVN TJK MKD MKD ARM 5 BLR SVK RUS BGR BGR Jama Timur Dou Jauta Awa Tungah Nuggoe Acoh Dulosalami Bali Yoguaunta Jamaluat Banton Sunatrabarat Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta 20 2 Oblast Minsk Oblast Belarus FIGURE 78: CLASS AND SCHO Source: own calculations using School Based Management 2009 Brest Oblast Gomel DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat 0 0 TOMBOUCIOU .13% Source: Belstat, MOE 2004 DI. 0 2009 HUN HUN Number of Students per Classroom MDA MDA BLR UZB ROU UZB LTU RQU CVE AZE AZE LVA CZE LVA KAZ KAZ KGZ SVN KGZ Jama SVN POL POL TJK LTA TJK MKD MKD FIGURE 67: ARM BLR SVK RUS UKR SVK UKR BGR BGR 50 Nuggoe 2004 Nuggoe Acoh Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Source: own calculations using School Based Management 40 Minsk Oblast Belarus 2009 45 TOMBOUCIOU .13% GAO .41% 35 2004 50 40 FIGURE 78: CLASS AND SCHOOL SIZES ARE SMALL 30 KOULIKORO .51% FIGURE 67: STUDENT TEACHER RATIO PENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES MOPTI .11% 25 of total household expenditures, by region GAO .41% Source: own calculations using School Based Management 45 35 Number of Students per Classroom Number o 20 MOP 2014 KAYES .50% BAMAKO SEGOU .63% FIGURE 78: CLASS AND SCHOOL SIZES ARE SMALL USEHOLD EXPENDITURES KOULIKORO .51% 1.75% FIGURE 67: 40 30 40 15 500 STUDENT TEACHER RATIO 2% MOPTI .11% Source: own calculations using School Based Management 50 FIGURE 78: CLASS AND SCHOOL SIZES ARE SMALL Number of Students per Classroom Number of Students per Primary School 35 35 25 10 400 KAYES .50% SEGOU .63% 45 Number of Students per Classroom 30 Number of Students per Primary School 500 40 While the majority of students (about 77.4 percent) now reside in cities, 60.4 percent o schoo s and BAMAKO 5 15% 1.75% SIKASSO .79% 50 FIGURE 78: CLASS AND40 SCHOOL SIZES ARE SMALL 30 20 25 35 300 STUDENT TEACHER RATIO Bamako 40 500 0 45 400 Koulkoro Number of Students per Classroom 35 25 30 15 20 Number of Students per Primary School CHN CHL JPN ISR IDN PRK Koyos FIGURE 67: 35 200 40 30 15 400 Tbilisi Tbilisi 11% SIKASSO .79% 20 25 10 300 STUDENT TEACHER RATIO 35 percent o teachers were ocated n rura areas Equa y the student-teacher rat o s substant a y Gilo 30 Bamako 50% Source: own calculations using School Based Management 40 500 Koulkoro Segou 35 25 20 10 100 2% Koyos Mepti FIGURE 67: 35 25 15 5 5 300 200 Source: NSI except Poland (BOOST data), Slovakia (Mini 30 15 400 Gilo 10% Silcasco 50 30 20 20 Cyprus, plus Norway, Iceland and Switzerland. OECD at 0% Segou Tombouctou FIGURE 67: Source: own calculations using School Based Management 10 10 0 0 0 Notes: Year of reference 2008. Public institutions only (in 45 25 15 200 100 ower n rura areas or as ow as 4 6 n rura areas o V tebsk ob ast Wh e rura schoo s are mportant Mepti 25 15 300 CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK DNK SVK Silcasco 10% Source: own calculations using School Based Management 50 20 5 5 0 1 2 3 4 5 6 7 40 20 10 10 FIGURE 53: DISTRIBUTION OF HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS 100 STUDENT TEACHER RATIO Tombouctou 0 0 FIGURE 67: 45 15 5 5 0 200 2% 35 15 BUDGET FOR STUDENT IN RUPIAH ROU LTU BGR HUN LVA EST RUS WEST. EUROPE MDA SVN BLR SVK 50 IRL CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DNK DEU AUT ESP ITA HUN POL PRT BEL CZE ISL MEX FIN RUS CHE AUT ITA POL EST ISL GRC RUS BLR PRK BLR DNK SVK Source: EMOP 2014 Source: own calculations using School Based Management 40 10 0 0 1 2 3 4 5 6 7 Source: NSI except Poland (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. 0 n prov d ng access to educat on the resource ntens ty dr ves up per student costs Schoo - eve data 30 10 0 STUDENT TEACHER RATIO 45 SPENDING ON VARIOUS EDUCATION ITEMS FIGURE 53: DISTRIBUTION OF HOUSEHOLD 100 ROU LTU BGR LVA EST RUS Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. GEO EUROPE MDA SVN BLR GEO UKR HRV SVK CHN CHL JPN ISR IDN ARG TUR BRA GBR AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE EST GRC PRK SVK 35 25 5 5 0 1 2 3 4 5 6 7 Notes: Year of reference 2008. Public institutions only (including for Belarus). 50 Source: EMOP 2014 40 BUDGET FOR STUDENT IN RUPIAH Source: NSI except Poland (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. STUDENT TEACHER RATIO MAIL 30 RURAL URBAN 0 0 0 Shida Kartli Shida Kartli 45 20 35 BUDGET FOR STUDENT IN RUPIAH Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. ROU LTU BGR LVA EST RUS WEST. WEST. EUROPE MDA SVN BLR HUN UKR FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON HRV SVK EDUCATION, BY REGIONS, 2014 0 1 2 3 4 5 6 7 CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK DNK SVK wou d be necessary to understand whether such resource d fferent a s are ust fied and how they Notes: Year of reference 2008. Public institutions only (including for Belarus). 25 15 Source: NSI except Poland (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. 40 Source: EMOP 2014 30 STUDENT TEACHER RATIO MAIL RURAL 4% 20 URBAN 10 3% Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. OLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 35 25 8% BUDGET FOR STUDENT IN RUPIAH Notes: Year of reference 2008. Public institutions only (including for Belarus). 15 5 21% Source: NSI except Poland (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. 30 20 29% 100 d be cou BY opt m zed e ther by w th n-schoo reorgan zat on or schoo conso dat on OF PARENTAL FEE ELIMINATION BY REGION POL Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. 10 3% 0 25 4% 44% Notes: Year of reference 2008. Public institutions only (including for Belarus). 90 15 CAL IMPACT OF PARENTAL FEE ELIMINATION REGION FIGURE 4.14: FISCAL IMPACT 8% 5 21% 0 1 2 5% 3 4 5 6 7 FIGURE 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS 80 20 10 POL 29% Source: Belstat, MOE. 70 5% 0 44% 60 15 FIGURE 68: ESTIMATES OF GOVERNMENT SPENDING ONIN BASIC 5 44% BUDGET FOR STUDENT IN RUPIAH 62% 0 1 2 3 4 5 6 7 FIGURE 79: MOST STUDENTS ARE URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS 50 ent of GDP GDP Additional cost as percent of FIGURE 5% 71% FIGURE 67: 10 0 Source: Belstat, MOE. 100 12 40 Samtskhe-Javakheti Samtskhe-Javakheti Source: own calculations using School Based Management 5% 4% Source: own calculations based on APBDN (2009) 30 5 44% 0 1 2 3 4 5 BUDGET FOR 7STUDENT IN RUPIAH 6 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS 90 22.6 TOMBOUCIOU .13% FIGURE 68: ESTIMATES OF GOVERNMENT SPENDING ON BASIC 10 20 62% FIGURE 67: 80 35 Source: Belstat, MOE. 0 71% 50 100 12 60.4 10 Source: own calculations using School Based Management 70 8 4% BUDGET FOR STUDENT IN RUPIAH 90 0 0 1 2 3 4 5 6 7 Tuition & Fees Source: own79: FIGURE MOST STUDENTS ARE IN URBAN AREA WHILE 80k MOST SCHOOLS ARE IN RURAL AREAS 22.6 60 TEACHER SALARY SPENDING AS % OF calculations based on APBDN (2009) 10 TOMBOUCIOU .13% 45 100 TOTAL SALARY SPENDING FOR BASIC FIGURE 4.14: FISCAL IMPACT OF PARENTAL FEE ELIMINATION BY REGION GAO .41% EGION Tutoring & Other 80 35 50 6 50 Source: Belstat, MOE. 100 70k 12 60.4 90 Books Supplies 40 70 40 8 KOULIKORO .51% 77.4 4 STUDENT TEACHER RATIO BUDGET FOR STUDENT IN RUPIAH 90 60k 80 GAO .41% MOPTI .11% 45 22.6 60 30 10 65 35 TEACHER SALARY SPENDING AS % OF 80k 80 100 70 6 TOTAL SALARY SPENDING FOR BASIC Tuition & Fees Tutoring & Other 50 35 20 2 Additional cost as percent of GDP KAYES .50% SEGOU .63% FIGURE 67: 50k 12 60.4 39.6 30 100 60 BAMAKO 1.75% 40 70k 70 40 90 10 8 Books Supplies 77.4 4 2009 ACTUAL VALUE 51% STUDENT TEACHER RATIO MOPTI .11% Source: own calculations using School Based Management 90 22.6 60k 60 40k 30 80 50 0 10 65 0 FIGURE 4.14: FISCAL IMPACT OF PARENTAL FEE ELIMINATION BY REGION 35 25 80 35 50 6 40 30k 20 70 Students Teachers 2 Schools STR Samegrelo-Zemo Svaneti SIZE,AND Samegrelo-Zemo Svaneti SEGOU .63% FIGURE 67: 60.4 39.6 APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES IN EDUCATION, 50 30 20 70 50k 40 10 8 60 30 77.4 4 SIKASSO .79% 20k 0 SHARE, DIFFERENT SOCIOECONOMIC GROUPS 2009 ACTUAL VALUE Source: own calculations using School Based Management 60 40k 30 0 65 50 20 45 25 15 50 10k 6 Rural City of Minsk V 20 Students 40 Teachers 2 Schools 10 STR Urban Additional cost as percent of GDP FIGURE 67: 30k 39.6 Urban Grodno Oblast M Tbilisi APPENDIX FIGURE 30: HOUSEHOLD EXPENDITURES Tbilisi O .79% IN EDUCATION, 50 40 20 10 40 77.4 10 0 4 30 0 20k 0 STUDENT TEACHER RATIO Source: own calculations using School Based Management 30 Brest Oblast G SIZE,AND SHARE, DIFFERENT SOCIOECONOMIC GROUPS 65 0 Rural 19 20 21 20 22 23 24 25 26 27 28 29 30 2009 ratio = 18 45 35 15 5 20 City of Minsk Vitebsk Oblast FIGURE 67: 10k Students 39.6 Teachers Urban 2 10 Schools STRCity of Minsk Grodno Urban Grodno Oblast Oblast Brest Oblast Mogiev Oblast Minsk Minsk Rural Oblast Oblast B BELOW POVERTY LEVEL 50 ABOVE POVERTY LEVEL 40 30 10 0 10 0 0 Oblast Oblast Vitebsk Oblast BrestMogiev Goeml Oblast Belarus UDENT TEACHER RATIO Source: own calculations using School Based Management 0 Gomel Oblast 0 0 1 2 3 STUDENT TEACHER RATIO 35 5 19 420 Rural 5 23 24 21 22 6 25 26 7 27 28 29 30 2009 ratio = 18 45 25 Students Teachers Schools STR Urban City of Minsk Vitebsk Oblast Minsk Oblast Rural Belarus 5% FIGURE 53: DISTRIBUTION OF 2% HOUSEHOLD SPENDING ON VARIOUS EDUCATION ITEMS Urban Grodno Oblast Mogiev Oblast BELOW POVERTY LEVEL 50 ABOVE POVERTY LEVEL 40 30 20 0 BUDGET FOR STUDENT IN RUPIAH echkhumi and Kvemo Svaneti Racha-Lechkhumi and Kvemo Svaneti Tuition & Fees FIGURE CHER RATIO Source: EMOP 2014 Brest Oblast Gomel Oblast 45 26% 25 0 1 2 3 4 Rural 5 6 7 STUDENT TEACHER RATIO City of Minsk Vitebsk Oblast 35 Books 15 City of Minsk Grodno Grodno Oblast Brest Oblast Oblast MinskOblast Minsk Belarus Additiona FIGURE 53: DISTRIBUTION 2% OF HOUSEHOLD SPENDING ON 5% VARIOUS EDUCATION 43%ITEMS Urban Oblast Mogiev Oblast 40 30 20 Tutoring & Other 10 BUDGET FOR STUDENT IN RUPIAH FIGURE 4.14: FISCAL IMPACT OF PARENTAL FEE E Vitebsk Oblast BrestMogiev Oblast Oblast Goeml Oblast Belarus TIO Gomel Oblast Source: EMOP 2014 MAIL 52% Tuition & Fees RURAL 5% URBAN 159 Georgia PER (2015) Overall comment This review analyzes the fiscal and equity implications of a policy that provided free preschool. The policy is expected to increase enrollments, but reduce local government revenues, as a result of the elimination of fees. The analysis provides four different scenarios, based on different assumptions about enrollment trends, and explores the impact on equity (the impact of school-fee elimination policy on interregional differences in spending and inequality in the provision of preschool services). Making preschool education free will also encourage enrollment, which will raise the costs for local governments. Before the fees were eliminated, enrollment rates were much higher among rich than among poor households, ranging from 30 percent in the poorest households to more than 50 percent in the richest. Without fees enrollment rates for less well-off families are expected to rise and enrollment rate gaps are likely to diminish. Regions more heavily reliant on parental fees as a source of preschool funding are generally associated with lower enrollments, except for Tbilisi. This means that with parental fees eliminated, regions that were more reliant on parental fees will be confronted by a more acute increase in costs because of two shocks: (1) higher preschool spending to compensate for the higher parental fee, and Table 4.3: Additional fiscal cost of eliminating (2) a bigger increase in enrollment due to the preschool fees, million GEL lower enrollment rate before the change. Scenarios Estimated Fiscal Impact In different scenarios the immediate increase in preschool spending after the parental fee Fee Reform Scenario 1 elimination is estimated to total GEL24–39 Fixed Enrollment 24 million. Four scenarios are considered (Table Fee Reform Scenario 2 4.3). In Scenario 1 enrollment remains the same Fixed Supply 33 as pre-reform and the central government Fee Reform Scenario 3 compensates for the amount used to be paid Enrollment Equal to Average 27 in fees by parents. In this scenario the cost of in Top Three Consumption the reform will be GEL24 million. In Scenario 2, Quintiles enrollment increases so that preschool places Fee Reform Scenario 4 currently available are totally filled. In that Enrollment Equal to Average in 39 scenario enrollment will increase by about Top Consumption Quintile 10 percent to max out preschool capacity, Source: Staff calculations leading to a higher cost of GEL33 million. Scenario 3 corresponds to the case where enrollment increases up to the average enrollment rate for households in the top three quintiles of the consumption distribution. In that case, enrollment will increase by 4.3 percent, at a cost of GEL27 million. In Scenario 4, enrollment is assumed to increase up to the average enrollment rate for households in the top quintile of the consumption distribution, bringing a 19 percent increase in enrollment and a much larger fiscal cost of GEL39 million. 160 Education Public Expenditure Review Guidelines Interregional differences in spending are likely to exacerbate inequality in the provision of preschool services. For some regions the implied fiscal impact is much heavier than for others, driven by differences in parental fees, enrollment rates, and per child preschool spending (Figure 4.14). Before the parental fee was eliminated there was already a large disparity in the quality of preschool services: for instance, Tbilisi provides much better preschool services and its cost per child is nearly 30 percent higher than the rest of the local governments. After the fee exemption, more developed local governments like Tbilisi are likely to be in a better position to curb the fiscal impact; others where parental contributions were high and enrollment rates low may not be able to fully compensate for the amounts parents used to pay. This will lead to further deterioration of the quality of preschool education in these localities and result in more unequal preschool services across the country. Figure 4.14: Fiscal impact of parental fee elimination by region (Additional cost as percent of GDP) Sources: 2011 WMS and staff calculations. Enrollment Equal Fixed Supply to Average in Top Fixed Enrollment Consumption Quintile Enrollment Equal to Average in Top Three Consumption Quintile H GDP Growth: 1% GDP Growth: 2% GDP Growth: 3% GDP Growth: 4% 1.4M 1.3M 4.0% 1.2M 3.0% 161 FUNDING GAP 1.1M 2.0% 1M UDENTS HAS DECLINED, ESPECIALLY 900k FIGURE 76: CONSOLIDATION 1.0% OF SCHOOL NETWORK APHIC TRENDS: 1970 2030 Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). 1991=100 800k 0.0% 2012 2013 2014 2015 2016 2017 2018 2019 2020 All Ages Rural Example 21: Technical efficiency of inputs (efficiency indicators) Age 5-14 130 Urban Age 5-24 110 Source: UN, Beistat Belarus PER (2013) 90 70 Overall comment: 50 FIGURE 76: CONSOLIDATION This publicOF SCHOOL NETWORK expenditure review analyzes the technical efficiency of inputs (student-teacher ratio, and Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). 1991=100 class and school size) in Belarus, comparing them with regional and international benchmarks. It finds 130 inefficiencies in the provision of these education inputs, with the1991 1993 1995 student-teacher 1997 ratio among 1999 the 2001 2003 2005 2 lowest in the region, and small school and class sizes. Number of Schools 110 Number of Students 985 1990 1995 2000 2005 2010 2015 2020 2025 2030 Number of Teachers FIGURE 1.21: CURRENT EXPENDITURE COMPOSITION IN EDUCATION COULD UNDERMINE EFFICIENCY & EQUITY 90 In Belarus, student-teacher ratios at both primary and secondary have continued to fall and are among Total Education Sector Budget, 2014/15 FIGURE 1.21: CURRENT EXPENDITURE COMPOSITION IN EDUCATION COULD UNDERMINE EFFICIENCY & EQUITY Total Education Sector Budget, 2014/15 10% the lowest in the region (Figure 78). 4% 26% 10% Primary Education 4% Secondary Education 70 26% FIGURE 70: PERCENTAGE OF PRIMARY S University Education Primary Education NEED TO BE TRANSFERRED TO COMPY W TIVET Secondary Education Other University Education 39% TIVET 21% 30k Figure 78: Student-teacher ratios are among the REGION in the region lowest Other FIGURE 78: STUDENT TEACHER RATIOS ARE AMONG THE LOWEST IN THE NUMBER OF PRIMARY SCHOOL TEACHERS 25k 39% 21% 20k 15k 50 Source: EdStats database. 10k Primary school student-teacher ratios in the ECA region 1999-2009 5k 0 Primary school student-teacher ratios in the ECA region (1999-2009) Secondary school student-teacher ratios in the ECA region (1999-2009) Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau FIGURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION SPENDING, CIRCA 2013 FIGURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION SPENDING, CIRCA 2013 100 30 90 88 87 82 81 80 79 79 18 100 80 % PUBLIC EDUCATION SPENDING 1991 1993 1995 1997 1999 2001 2003 2005 90 80 88 2007 87 82 81 70 80 6079 2009 79 73 68 57 16 % PUBLIC EDUCATION SPENDING 73 49 49 25 50 70 68 FIGURE 2: SIX TYPES OF DATA THAT CAN BE USED FOR PER ANALYSIS 60 50 40 30 57 49 49 14 FIGURE 2: SIX TYPES OF DATA THAT 40 20 1 Number of Schools 20 12 (ie: policies, laws, and regulations) CAN BE USED FOR PER ANALYSIS 30 10 FIGURE 40: 2 Country System-Level Data (Budget (ie: BOOST), 20 0 BY SOURCES AND UN GURE 2: SIX TYPES OF DATA THAT CAN1 BE USED FOR PER ANALYSIS sectoral data (ie: EMIS, national assessment) Source: Education Sector Strategy; Honduras Costa Rica Guatemala Panama Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep (ie: policies, laws, and regulations) Number of Students 10 1 (ie: policies, laws, and regulations) USED FOR PER ANALYSIS 4 2 Cross-National Data (International assessment, Country statistical cross national System-Level Data (Budget (ie: BOOST), databases) sectoral data (ie: EMIS, national assessment) 3 National Surveys (Census, household surveys) 0 10 120% Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Finland Korea, rep 2 Country System-Level Data (Budget (ie: BOOST), -National Data (International data (ie: EMIS,assessment, 5 3 Research Reports 4 15 Number of Teachers Cross-National Data (International assessment, 100% 8 sectoral national assessment) National (Donors, Surveys academic research papers) cross national statistical databases) national statistical databases) (Census, household surveys) 80% 60% 3 National Surveys rch Reports 6 New Data Collection 5 Research Reports 4 (Census, household surveys) Cross-National Data (International assessment, 40% ors, academic research papers) (Donors, academic research papers) cross national statistical databases) FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 20% FIGURE 1.21: CURRENT EXPENDITURE COMPOSITION IN EDUCATION 6 New 10 6 Source: World Bank calculations based on data from the National Statistics Service (NSS), 0% Data Collection 5 Research Reports (Donors, academic research papers) COULD UNDERMINE EFFICIENCY Data Collection & EQUITY The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF -20% 2016 2017 2018 Total Education Sector Budget, 2014/15 FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 Republic of America FIGURE 1.21: CURRENT EXPENDITURE COMPOSITION IN EDUCATION (a) Average PCF Allocation per Student (b) Average School Size COULD UNDERMINE6 New EFFICIENCY Data Collection & EQUITY Source: World Bank calculations based on data from the National Statistics Service (NSS), (000s AMD) (Number of Students) Internal Resources 4 The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF Total Education Sector Budget, 2014/15 Development Partner 300 500 FIGURE 4: FIGURE 4: 10% (a) Average PCF Allocation per Student 250 (b) Average School Size Republic of America Shortage of Planned B 400 FOR PRIMARY EDUCATION RELATIVE TO SCHOOL SIZE 4% FOR PRIMARY EDUCATION RELATIVE TO SCHOOL SIZE 5 (000s AMD) 200 (Number of Students) 300 26% FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD 200 150 FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD 2 300 100 500 30 FIGURE 4: 10% 30 Primary Education 250 NEED TO BE TRANSFERRED TO 400 COMPY WITH JOINT DECREE 100 50 NEED TO BE TRANSFERRED TO COMPY WITH JOINT DECREE TO SCHOOL 4% 25 SIZE FOR PRIMARY 25 SIZE EDUCATION RELATIVE TO SCHOOL 200 0 0 Secondary Education 300 ra l ou s l us an nta us s 0 m y en ity 150 FIGURE Ru 70: PERCENTAGE ra OF PRIMARY SCHOOL TEACHERS THAT WOULD an FIGURE 70: PERCENTAGE OF nta us PRIMARY SCHOOL TEACHERS THAT WOULD s 0 m y o en ity ou 26% 40 om nit ou in 40 om nit ou ino Ru Urb ud n ts in ou ino Urb ud n ts in ta u St u 200 in ta u St u un # of existing teachers to move within districts m n-m nta 100 University Education # of existing teachers to move within districtsoun m n-m nta 20 20 30k w ol in om o w ol in om NEED TO BE M TRANSFERRED TO COMPY WITH JOINT DECREE No ou TO COMPY WITH JOINT DECREE NEED TO BE TRANSFERRED 50 No ou 30 M # of existing teachers to move between districts in same province < aC C 100 Primary Education M < aC ho in C hly # of existing teachers to move between districts in same province M ou Sc ol in 0 hly TIVET 0 15 15 30k Hig ol 0 Hig o 0 NUMBER OF PRIMARY SCHOOL TEACHERS ho nta On us Sch ith s ho 25 25k ith Secondary Education l s Sc l s ou Sc ra ou an nta us ou inonly s 0 m y en ity ra an ly y en ity Other ou 40 om nit in ly in 40 om nit ou ino 10 10 Ru Urb ly On ud n ts in Sc TEACHERS n-m nta t O Ru Urb ud n ts in nta u St u On 25k ta u St u # of existing teachers to move within districts m n-m nta University Education No ou No t # of existing teachers to move within districtsoun m 0 m 20 ou No 30k w ol in om 20k w ol in om No ou M # of existing teachers to move between districts in same province < aC ho in C M 5 5 39% M < aC ho in C hly # of existing teachers to move between districts in same province M 21% hly TIVET l 30k Hig oo 15 ol 20k Hig NUMBER OF PRIMARY SCHOOL TEACHERS HUN HUN ho NUMBER OF PRIMARY SCHOOL On Sch ith MDA MDA 25k UZB UZB ROU ith RQU CVE AZE LVA 15k CZE KAZ KAZ KGZ SVN KGZ POL POL TJK TJK MKD LTA MKD 0 0 Sc ARM BLR SVK SVK RUS UKR UKR Sc BGR nly BGR ly Other ly G THE LOWEST IN THE REGION 10 ly On NUMBER OF PRIMARY SCHOOL TEACHERS tO On 25k 15k No t No 0 100 200 300 400 0 100 200 300 400 20k 10k FIGURE 41 5 39% 21% 20k (c) Average Class Size (d) Student-Teacher Ratio SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS 10k 15k BY SOURC 0 12 5k Source: Education 15k 5k 25 400 0 100 200 300 400 10 10k 0 (c) Average Class Size 20 (d) Student-Teacher Ratio 8 10k 0 6 30% Kalimantan Selatan Jawa Barat MalukuJawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Kalimantan Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung SCHOOL SIZE BY NUMBER OF STUDENTS 15 5k 10 12 4 Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Barat Riau Barat Timur o Jakarta ou Timur s Tengah Darussalam 40 om nit Tengah ud n Selatan ts Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Barat Bangka Belitung 5k 25 5 10 2 0 0 20 08 l s 20% Sumatera Kalimantan Papua NoTenggara ly us Kalimantan 0 l s ra ou an nta us s 0 m y en ity 6 n-m Barat Timur lTengah w ol in Utara < Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung 15 ou ou 40 Com nit Secondary school student-teacher ratios in the ECA region (1999-2009) ra an nta us Kalimantan 0 m y Sulawesi ou Dki in 0 m y en ity ou ino en Kalimantan Ru Urb ud n ts in nta u St u Ru in Urb 2009 m nta in 4 in nta u St u 10 ou m n-m nta om 2004 Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Barat Riau Barat Timur o Jakarta ou Timur s Tengah ithAceh Darussalam 40 om nit Tengah ud n Selatan ts Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Barat ou Bangka Belitung ou 2009 ol Com M ou ho in C 2 m wit Aceh M M hly On Jawa 5 a 40 om nit < a C Sumatera On Jawa M hly Sc ol in Hig ly hoo No 0 u h in Jawa Sumatera FIGURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION SPENDING, CIRCA 2013 0 ig 10% ith Nusa l H o Secondary school student-teacher ratios in the ECA region 1999-2009 Sc s Onino Sch Sumatera Kalimantan Papua ho CONCENTRATION CURVE ON GOVERNMENT SPENDING Sc No NusaoTenggara ly Kalimantan ou Nanggroe l s ra an nta nlyus ly ity ra ou an nta us Kalimantan Sulawesi ou Dki in 0 m y en ity ou t Oino Kalimantan Ru Urb ud n ts St u Ru in nta Urb ot FIGURE 2: INTERNATIONAL COMPARISO n-m N nta in in ON EDUCATION AND VARIOUS BENCHMARKS nta u St u ou m N n-m ta w l in m o ou No ou n w ol in om Co M ou < aC M 2004 M hly < aC Sc ol in C M hly 0% Sc ol in Hig Source: IMF, PISA, and World Bank calculations o FIGURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION SPENDING, CIRCA 2013 o Hig On Sch ith ho N o On Sch ho NT SPENDING 100 Nanggroe 100% nly ly nly 18 tO tO RKS FIGURE 2: INTERNATIONAL COMPARISON: PISA SCORE AND EXPENDITURE PER SE No 90 88 87 82 81 (f) School Capacity Utilization Source: IMF, PISA, and World Bank calculations 600 100 80 80 79 79 (Student Enrollment as % of Educa % PUBLIC EDUCATION SPENDING 80% Total Builing Capacity) PISA 2012 AVERAGE SCORE 73 Capita 88 70 68 33% 550 16 90 87 32% (f) School Capacity Utilization 600 Educa 82 81 70% 80 80 6079 79 57 31% (Student Enrollment as % of 60% % PUBLIC EDUCATION SPENDING Total Builing Capacity) PISA 2012 AVERAGE SCORE 60% 73 30% 50% 500 50 68 49 49 33% 550 70 32% 29% 28% 70% 40% 30% Russia 14 Propoor Spending Estonia Turkey Poland 60 40 57 31% 30% 27% 60% 26% 20% 450 Slovenia 40% 49 49 50% 10% 500 Czech Republic 30 s 50 Hungary 29% l 40% ra ou an 0% nta us Progressive s 0 m y en ity Russia ou 40 om nit 40: Ru FIGURE30% in ou ino Urb ud n ts l s in 28% nta u St u ou m n-m nta ra an nta us s 0 m y en ity Propoor Spending ou Turkey ou 400 40 om nit w ol in om in Lorenz Curve of Income or Consumption ou ino Ru Urb 20 ud n ts No ou 40 27% BY SOURCES AND hlyUNIT COST PROJECTION in nta u St u M < aC ho in C 20% m n-m nta 450 M ou w ol in om No ou s Sc ol 26% ig Slovak Republic Regressive (Poorly Targeted) M Albania 12 < aC 10% Sector Strategy; ho in C H ou nly ho M Source: Education 2016-2025, January 2015 hly ith 719 10 Sc 709 20% 30 l us ol ra Hig an 0% nta us nta us nly Progressive s 0 m y o en ity ho 676 ou 40 om nit ith FIGURE 40: Ru in ou ino Urb ud n ts 350 FIGURE 6.1: ALBANI Sc ou ino t O l s in nta u St u in O Sc ou m n-m nta ra an nly 0 m y o en ity ou 400 40 om nit w ol in om n-m nta N ly in Lorenz Curve of Income or Consumption 0 Ru Urb 20 ud n ts No ou tO On BY SOURCES AND hlyUNIT COST PROJECTION nta u St u M < aC ho in C m M No ou 562 w ol in om No ou ol Regressive (Poorly Targeted) ig M Albania 1k 3k 529 < aC ho in C ho 522 H M Source: Education Sector Strategy; 2016-2025, January 2015 hly ith 719 120% Honduras Costa Rica Guatemala Panama Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep 10 Sc 709 2016 ol Sc 0% Hig nly ho 427 676 10 ly ith 455 350 Sc tO On 100% Sc 428 2017 420 nly No 411 ly 394 0% 0 20% 40% 60% 80% 100% 75 tO On No 80% 2018 562 1k 3k 5k 7k 9k 529 314 522 120% 2012 Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Finland Korea, rep 60% 2016 279 CUMULATIVE SHARE OF POPULATION 244 455 241 230 100% 2025 427 428 2017 420 40% 204 411 394 185 % 80% 100% 70 8 165 80% 20% 2018 TOTAL EXPENDITURES PER PUPIL IN SECON 314 93 96 60% 74 2012 279 71 62 72 ON 0% 52 244 241 37 27 33 230 40% 15 13 11 11 9 2025 204 2016 2017 2018 2021 2025 -20% 65 185 165 Preschool Primary Non-Formal 1st Cycle 2nd Cycle Technical Higher FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 20% Secondary Secondary and Education FIGURE 78: CLASS AND SCHOOL 6 SIZES ARE SMALL 93 96 71 74 62 72 Source: World Bank calculations based on data from the National Statistics Service (NSS), 0% 33 37 52 School School Vocational The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF 15 13 11 11 Internal 27 Resources 60 9 -20% 2016 2017 2018 2021 2025 Development Preschool Partners Primary Non-Formal 1st Cycle 2nd Cycle Technical Higher FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 Republic of America Shortage of Planned Budget Secondary Secondary and Education Source: FIGURE (a) Average PCF Allocation per Student (b) Average School Size 8: SOURCES World Bank calculations OF based on data fromEDUCATION SECTOR the National Statistics Service (NSS), FINANCE, 2013 (000s AMD) (Number of Students) School School Vocational Internal Resources FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 4 The National Center Source: Authors’ for Education estimate from CBN, OECD,Technology (NaCET), Nigeria, State Budget, the Assessment Federal Government andHousehold Testing Budget, and General Center Survey Panel(ATC) 2012/13 and MOF 55 Development Partners 300 500 Source: Kosovo BOOST Source: Kosovo BOOST Republic of America Number of Students per Classroom Number of Students per Primary School (a) Average PCF Allocation per Student (b) Average School Size Shortage of Planned Budget ATION SECTOR FINANCE, 2013 (000s AMD) 9 | 1% 250 (Number of Students) 400 get, and General Household Survey Panel 2012/13 200 300 FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 50 150 200 4.5% 4.5% 2 300 100 Source: Kosovo BOOST Source: Kosovo BOOST 500 250 426 | 18% 50 100 936 | 40% 400 LGA 0 4.0% 4.0% 200 300 l s 0 l s 45 ra ou ou an nta us s 0 m y en ity 150 ra an nta us Household s 0 m y en ity ou 40 om nit 4.5% 4.5% ou in 40 om nit ou ino Ru Urb ud n ts in ou ino Ru Urb ud n ts in nta u St u 200 in nta u St u 3.5% 3.5% m n-m nta 100 m n-m nta ou w ol in om ou w ol in om 295 | 13% No ou UBEC No ou 500 M < aC C | 18% 50 100 M M < aC ho in C hly M 40 ou Sc ol in hly 4.0% 4.0% 0 LGA 0 Hig ol State 0 Hig 40 o ho 3.0% 3.0% nta On us Sch ith s ho Capital Outlays Capital Outlays ith l s Sc l s ou Sc ra ou an nta us ou inonly s 0 m y en ity ra an nly Household 0 m y en ity ou 40 om nit in ly in Federal Gov’t 2008 2009 2010 2 40 om nit ou ino Ru Urb ly ud n ts in n-m nta t O Ru Urb ud n ts in tO nta u St u On nta u St u 3.5% 3.5% m Subsides & Transfers Transfers41: Subsides &FIGURE n-m nta No ou No m ou No w ol in om ou w ol in om 295 | 13% 2.5% 2.5% ou UBEC M 587 | 25% Donors FIGURE 41: < aC C M M < aC ho in C hly M Sc ol in hly Hig Utilities Utilities BY SOURCES AND UNIT COST PROJECTION; 2016, 2025 No ol State Hig BY SOURCES 3.0%AND UNIT COST PROJECTION; 2016, 2025 o HUN HUN ho 3.0% On Sch ith MDA ho BLR LTU UZB Capital Outlays Capital Outlays ith RQU AZE AZE LVA CZE LVA KAZ KGZ SVN POL POL TJK LTA MKD Sc BLR 2.0% 2.0% SVK SVK UKR UKR Combine Increas Sc ly BGR 76 | 3% Goods & Services Goods & Services ly ly On 35 Federal Gov’t ly On Source: Education Sector Strategy; 2016-2025, January 2015 Source: Education Sector Strategy; 2016-2025, January 2015 On t Subsides & Transfers Transfers41: Subsides &FIGURE No t No 2.5% 2.5% Wages & Salaries Wages & Salaries Health Spending 25% Donors (c) Average Class Size (d) Student-Teacher Ratio 1.5%41: FIGURE 1.5% BY SOURCES ANDUtilities 30% UNIT COST PROJECTION; 2016, 2025 Utilities BY SOURCES 4.0% AND UNIT COST PROJECTION; 2016, 2025 30% Education Spend 400 2.0% 2.0% 12 Goods 1.0% Sector Strategy; Source: Education & Services 2016-2025, January 2015 1.0% Goods & Services Source: Education Sector Strategy; 2016-2025, January 2015 Baseline FIGURE 12: THE STRUCTURE OF BASIC EDUCATION 25 FINANCING 10 Wages & Salaries Wages & Salaries 3.0% (c) Average Class Size 20 (d) Student-Teacher Ratio 8 1.5% 1.5% 30 6 30% 0.5% 20% 4.0% 0.5% 30% 20% 15 Internally Federal 4 Fund Generated Account & VAT Independent 10 12 Devp’t 1.0% 1.0% 2.0% Pool Revenue Partners 2 0.0% 0.0% BASIC EDUCATION Sources 25 FINANCING Revenue Derivatives 5 10 0 3.0% 20 08 l s 0.5% 20% 2007 2008 2009 2010 201110%2012 0.5% 2007 2008 2009 20% 2010 2011 2012 10% l s ra ou 1.0% an nta us s 0 m y en ity 15 6 ou ou 40 om nit ra Federal an nta us in s 0 m y 40 om nity en ity ou ino Ru Urb ud n ts ou in in nta u St u ou ino Ru Urb ud n ts m n-m nta 4 in Ministry ntaof 2.0% u St u VAT Independent 10 Devp’t ou m 0 m n-m nta w ol in om 25 ou No ou m wit l in om Pool Revenue Partners Finance M 0.0% 0.0% No ou < aC C 2 M 5 M hly 300 40 om nit < a C Sc ol in ho in C Source: EdStats database. M LGA SJGA State Federal 0 hly Hig 0% 0.0% 0% o l ig 201110% 10% ly us oo 0 Con. Rev. 2007 2008 2009 2010 2012 2007 2008 2009 2010 2011 2012 On Sch ith ho l H o s Onino Sch u h l s ra ou 2016 2017 1.0% 2018 2021 2025 2016 2017 2018 2021 an nta nlyus nly en ity Sc ou ly ra Federal an nta us in s 0 m y en ity ou t Oino Ru Urb ud n ts tO ou 40 om nit Federal St u in nta ou ino Ru Urb ud n ts Responsibility n-m N nta in Ministry ntaof u St u No ou m n-m nta w ol in om o ou Budget No ou w ol in om Finance M No ou < aC ho in C M M hly < aC C M Federal hly 0% 0.0% 0% Sc ol in ol UBFC (2%) Hig ho Con. Rev. Hig ith o 20 Sc On Sch ith ho Sc 2016 Education spending as share 2021 2017 2025 of total public spending 2018 2016 Education spending as share of total public spending 2021 2017 2018 ly ly On nly On ly Federal tO t FMF No Capital spending as share of total education spending Capital spending as share of total education spending No Budget UBFC (2%) (f) School Capacity Utilization Education spending as share of GDP Education spending as share of GDP Salary Non-Salary Salary Non-Salary Salary Non-Salary (Student Enrollment as % of Education spending as share of total public spending Education spending as share of total public spending 200 Salary Non-Salary Education FMF Lower Total Builing Capacity) Finances Primary Lower Unity Capital spending as share of total education spending Capital spending as share of total education spending 15 Secondary 33% School SUBEB Secondary College 2025 32% (f) School Capacity Utilization 70% Education spending as share of GDP Education spending as share of GDP Salary Non-Salary Salary Non-Salary 31% (Student Enrollment as % of 60% Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) 30% Total Builing Capacity) 50% Lower Unity 33% to the local government councils of the state from the Federation 29% Secondary “into which shall be paid all allocations College 40% 32%of the State” (Section 162 [6], 1999 Constitution of Nigeria). Account and from the Government 28% 2025 70% 30% FIGURE 5.12: DISTRIBUTION OF TEACHERS BY 10 31% 27% 60% 20% 30% 26% 50% 10% JLGA) 29% l s 100 40% ra ou an n 0% nta us s 0 m y en ity ou 40 om nit in ou ino Ru Urb ud n ts l s in 28% nta u St u ou 30% FIGURE 5.12: DISTRIBUTION OF TEACHERS BY FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED m n-m nta ra an nta us s 0 m y en ity ou ou 40 om nit w ol in om in ou ino Ru Urb ud n ts No ou 27% in nta u St u M < aC ho in C 20% m n-m nta M hly ou 12k w ol in om No ou s Sc ol 26% Hig M < aC 10% ho in C ou nly ho M hly ith Sc l s 5 ol ra ou Hig an 0% nta us nta us nly s 0 m y en ity ho ou 40 om nit ith in ou ino Ru Urb ud n ts Sc ou ino t O l s in nta u St u in O Sc ou FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED m n-m nta ra an ly 0 m y o en ity ou 40 om nit w ol in om n-m nta N ly On Ru in Urb ud n ts No ou 75 On ta u St u M 10k < aC C t un m M No hly 12k Sc ol in w ol in om o No ou Hig M < aC C o M hly On Sch ith Sc ol in ho Hig ly o ly On Sch On ith ho t nly No ly 75 8k 70 tO 0 0 10k No 6k 65 8k 70 ROU LTU BGR LVA EST RUS EUROPE MDA SVN SVK 63.1 CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK DNK SVK 6k 4k 60 65 63.1 56.9 4k 2k 60 55 54.9 E 14: SOURCES OF EDUCATION SECTOR FINANCE, 2013 om CBN Annual report 2013 and per capital allocation from monthly shares of distribution from 0 56.9 HYPOTHESIS: SIMULATING THE IMPACT unt by State, Nigeria Economic Report, The World Bank Group (2013) 2k 50 FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 55 UNQUALIFIED 54.9 MASTERS AND 49.7 Federal NANCE, 2013Allocation From Source: All BOOST Kosovo Sources Source: Kosovo BOOST /BEGINNER SECONDARY HIGHER PH.D DEGREES INTEGRATION OF COMMUNITY TEACHE COULD UNDERMINE EFFICIENCY &N a EQUITY surveys data (interna 1 (ie: policies, laws, and regulations) onal 10 Hondura Paragua Costa Ric Guatemal Colombi Nicaragu Panam Panama Chil (census, household Jorda Finlan Korea, re 10 assessment, cross El Salvad surveys) na onal sta s cal Total Education Sector Budget, 2014/15 COMPOSITION a EDUCATION IN databases) Research 1 (ie: policies, laws, and regulations) 4 2 Cross-National Data (International assessment, System-Level Country statistical cross national Data (Budget (ie: BOOST), databases) 3 National Surveys (Census, household surveys) 0 a Cross-n surveys data (interna onal Country system- FIGURE 2: SIX TYPES OF(donors, reports DATA THAT CAN BE USED FOR PER ANALYSIS sectoral data (ie: EMIS, national assessment) Honduras Costa Rica Guatemala Chile Colombia Nicaragua Jordan El Salvador Paraguay Finland Korea, rep UITY 15 census, household academic research assessment, cross level data (budget 2 5 3 surveys) na onal sta s cal papers) Country System-Level Data (Budget (ie: BOOST), Research Reports (e.g. BOOST), sectoral 1 4 Cross-National data (ie: EMIS,assessment, Data (International 4 Cross-National Data (International assessment, 8 databases) data (e.g EMIS, sectoral national assessment) National (Donors, Surveys academic research papers) cross national statistical databases) Resea rch (ie: policies, laws, and regulations) cross national statistical databases) (Census, household surveys) na onal assessment)) 10% reports (donors, 4% academic research Government’s papers) 2 New data c (Budget (ie: BOOST), Country System-Level Data 5 National Surveys Research Reports 3 6 New Data Collection 5 Research Reports official 26% PER sectoral data (ie: EMIS, national assessment) (Census, household surveys) (Donors, academic research papers) 4 Cross-National Data (International assessment, cross national statistical databases) (Donors, academic research papers) FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD NEED TO BE TRANSFERRED TO COMPY WITH JOINT DECREE FIGURE NEED TO BE TRANSFERRED TO COMPY 4.5: JOINT WITH DECREEOF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 DISTRIBUTION Primary Education 6 10 6 Source: World Bank calculations based on data from the National Statistics Service (NSS), New data c 3 National Surveys (Census, household surveys) Secondary Education 6 New Data Collection 5 Research Reports (Donors, academic research papers) New Data Collection The National Center for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF 26% PER FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD FIGURE 70: PERCENTAGE OF PRIMARY SCHOOL TEACHERS THAT WOULD University Education # of existing teachers to move within districts FIGURE 4.5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY # of existing teachers to move within districts TYPE, 2009/10 Primary Education NEED TO BE TRANSFERRED TO COMPY WITH JOINT DECREE NEED # of existing teachers to move TO between BE TRANSFERRED districts in same province TO COMPY WITH JOINT 30k DECREE (a) Average#PCF of existing Allocation perto teachers move between districts in same province Student (b) Average School Size Republic of America TIVET 6 New Data Collection 30k Source: World Bank calculations based on data from the National Statistics Service (NSS), (000s AMD) (Number of Students) NUMBER OF PRIMARY SCHOOL TEACHERS 4 Secondary Education 25kTechnology (NaCET), the Assessment and Testing Center (ATC) and MOF The National Center for Education Other NUMBER OF PRIMARY SCHOOL TEACHERS 25k # of existing teachers to move within districts 300 University Education FIGURE 4: FIGURE 4: # of existing teachers to move within districts 30k 20k (a) Average PCF Allocation # of existing per teachers to Student move 250 between districts in same province (b) Average School Size 500 Republic of America # of existing teachers to move between districts in same province 400 39% 21% FOR PRIMARY EDUCATION RELATIVE TO SCHOOL SIZE FOR PRIMARY EDUCATION RELATIVE TO SCHOOL SIZE 5 TIVET 30k 20k (000s AMD) 200 (Number of Students) NUMBER OF PRIMARY SCHOOL TEACHERS 300 25k 15k 150 Other 200 THE REGION 2 100 NUMBER OF PRIMARY SCHOOL TEACHERS 25k 300 500 FIGURE 4: 30 FIGURE 4: 30 15k 20k 250 10k 50 400 100 21% 25 SIZE FOR PRIMARY EDUCATION RELATIVE TO SCHOOL 25 SIZE FOR PRIMARY EDUCATION RELATIVE TO SCHOOL 20k 200 0 300 l s 0 s 10k ra l an nta s s 0 m y ou en ity 150 ra an nta s ou s 0 m y ou en ity ou it 15k ou ou in 40 om it 5k Ru Urb un ud n ts Ru in Urb -m tain un ud n ts in nta St u 200 -m tain in nta St u Bali wit l in Comm 100 m 20 20 n om ou n w l in om Non ou ou ou Non ou 30 30 ou 15k M < C 100 M 5k 50 M < aC hoo in C hly M h a l in 0 hly 40 10k 0 0 Hig oo 15 15 l 0 Hig hoo s oo 0 nta On s Sch 25 25 ith h l s Sc l s ou Sc Sc ra an nta s ouly 10k Utara s 0 m y ou en ity ra an ly ou 0 0 m y ou en ity ou 40 om it 162 in ly -m tainOn Kalimantan Selatan Jawa Barat Jawa Timur Sc Tengah Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Selatan ly Maluku oo Papua Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung in 40 Com it 10 10 Ru Urb ly On Di. Yogyakarta ith < in a C mun ud n ts Ru in Urb -m tain un ud n ts in 5k nta St u On nta St u Non ou Not Not m 20 20 Education Public Expenditure Review Guidelines n Sulawesi Utara whool in Com ou n w l in om Non ou ou ou ou M Kalimantan Selatan Jawa Barat MalukuJawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Kalimantan Selatan Maluku Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung M 5 5 M hoo in C hly M 5k hly a Sumatera Sulawesi Balinly Sc hool Hig 15 15 l 0 Hig < HUN MDA Jawa Kalimantan UZB UZB ROU On Sch ith RQU Kalimantan CVE AZE LVA KAZ KGZ SVN KGZ POL TJK TJK MKD LTA MKD 0 0 ARM BLR SVK SVK RUS UKR Sc BGR nly BGR 10 10 ly 0 On O Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Maluku Papua Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung O Secondary school student-teacher ratios in the ECA region (1999-2009) Not Not 0 400 0 100 200 300 400 100 200 300 5 5 Jawa Barat Jawa Timur Jawa Tengah Sumatera Utara Sumatera Selatan Lampung Sulawesi Selatan Nusa Tenggara Barat Banten Sulawesi Tenggara Jambi Sumatera Barat Riau Kalimantan Barat Nusa Tenggara Timur Dki Jakarta Kalimantan Timur Kalimantan Tengah Nanggroe Aceh Darussalam Sulawesi Tengah Papua Bali Sulawesi Utara Di. Yogyakarta Gorontalo Kepulauan Riau Malukuutara Bengkuu Sulawesi Barat Papua Barat Bangka Belitung (c) Average Class Size (d) Student-Teacher Ratio SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS 0 SPENDING, CIRCA 2013 GURE 16: WAGE BILL AS A PERCENTAGE OF PUBLIC EDUCATION 0 12 25 0 100 200 300 400 0 100 200 300 400 10 (c) Average Class Size 20 (d) Student-Teacher Ratio 8 FIGURE 2: INTERNATIONAL COMPARISON: PISA SCORE AND EXPENDIT SCHOOL SIZE BY NUMBER OF STUDENTS SCHOOL SIZE BY NUMBER OF STUDENTS 15 6 Source: IMF, PISA, and World Bank calculations F PUBLIC EDUCATION SPENDING, CIRCA 2013 10 12 4 00 25 5 10 2 18 0 20 08 l s FIGURE 2: INTERNATIONAL COMPARISON: PISA SCORE AND EXPENDITURE PER SECONDARY STUDENT 90 88 l s ra an 87 nta s s 0 m y ou en ity 6 ou 15 ou 40 om it ra an nta s in s 0 m y 40 om ity ou en ity Ru Urb un ud n ts 600 ou ou -m tain in nta St u Ru in Urb un ud n ts Source: IMF, PISA, and World Bank calculations 2009 m -m tain 4 in 82 nta St u 10 81 m 0 m n w l in om ou Non ou ou n m wit l in om 80 79 79 ou M Non ou < aC hoo in C 80 2 ou M 5 M hly 40 om it < C hoo in C M hly un h a l Hig hoo PISA 2012 AVERAGE SCORE 0 73 l ig ly s oo 0 ith l H Sc s OninouSch CONCENTRATION CURVE ON GOVERNMENT SPENDING Sc s ra 550 l an nta nly s ly 68 ou en ity 70 16 Sc ou O ou ly ra On an nta s in s 0 m y ou en ity Ru Urb ud n ts ou ou On 40 om it -m Nottain St u Ru in nta Urb 600 un ud n ts Not -m tain in ON EDUCATION AND VARIOUS BENCHMARKS nta St u m n w l in om ou Non ou n w l in om ou M Non ou < aC C Estonia ou 79 79 M Poland 2004 M 60 hly < aC l in 57 hoo in C M hly Hig oo l hoo Hig oo PISA 2012 AVERAGE SCORE 73 On Sch ith 500 Czech Republic On Sch ith GOVERNMENT CONCENTRATION CURVE ON 49 SPENDING Sc 49 100% ly Sc 50 550 ly Hungary On 68 ly ly On Russia Not ON EDUCATION AND VARIOUS BENCHMARKS Not 14 Estonia Turkey Poland 40 57 CUMULATIVE SHARE OF GOVERNMENT SPENDING ON EDUCATION (f) School Capacity Utilization (Student Enrollment as % of 450 Slovenia Slovak Republic 49 49 100% 80% 500 Czech Republic 30 FIGURE 40: 33% Total Builing Capacity) Russia Hungary Turkey 20 400 CUMULATIVE SHARE OF GOVERNMENT SPENDING ON EDUCATION 32% (f) School Capacity Utilization BY SOURCES AND UNIT COST PROJECTION 31% (Student Enrollment as % of 70% 450 60% Slovak Republic Albania 12 80% Source: Education Sector Strategy; 2016-2025, January 2015 719 Total Builing Capacity) 10 30% 709 60% 50% 676 33% FIGURE 40: 29% 40% 350 0 Propoor Spending BY SOURCES AND UNIT COST PROJECTION 32% 31% 28% 70% 30% 400 562 27% 60% 20% Albania 1k 3k 5k 7k 529 522 60% Source: Education Sector Strategy; 2016-2025, January 2015 30% 26% 719 120% Honduras Costa Rica Guatemala Panama Chile Colombia Nicaragua Jordan El Salvador Paraguay Panama Finland Korea, rep 50% 709 10% 40% 2016 427 676 29% l s 10 455 40% ra 0% an nta s Progressive 350 s 0 m y ou y ou ou 100% 40 om it den nit 428 in 2017 Ru Urb un ts 420 s -m tain l 411 in 28% nta St u 30% m 394 ra an nta s s 0 m y ou en ity ou Propoor Spending The average class size is 16 students, compared to 21 and 24 students in OECD primary and lower ou 40 om it n w l in om in Lorenz Curve of Income or Consumption ou Ru Urb u un ud n ts Non ou ou -m tain 27% in 2018 nta St u 80% M 562 < aC hoo in C 20% m M hly n w l in om ou 1k 3k 5k 7k 9k TOTAL11k 13k EXPENDITURES PER PUPI 529 Non ou 314 ou l 522 26% Hig ou nly hoo Regressive (Poorly Targeted) M < aC 10% hoo in C 120% M 2012 Nicaragua Jordan El Salvador Paraguay Finland Korea, rep 40% hly ith 60% 2016 279 Sc 20% l s l s Sc Hig hoo ra 0% an nta s nta s nly Progressive s 0 m y 244 ou en ity 455 241 ou ou 40 om it 230 ith Ru in Urb 2025 un ud n ts 100% Sc ou O s -m tain 427 428 l in 2017 nta St u 420 in O Sc 40% 204 -m tain Not 411 m ra an ly 0 m y ou en y 40 om it 394 it 185 n w l in om ly On in Lorenz Curve of Income or Consumption ou Ru Urb 8 un ud n ts Non ou ou On 165 nta St u M < aC hoo in C Not m M 80% 2018 hly n w l in om 20% ou Non ou ou l Hig TOTAL EXPENDITURES PER PUPIL IN SECONDARY EDUCATION oo Regressive (Poorly Targeted) M 314 < aC hoo in C M hly On Sch ith 93 96 20% 60% 74 2012 l 279 Sc Hig 71 62 72 0% secondary, respectively (Figure 80). Class size in rural areas is only 9.4, against 20.4 in urban areas. oo ly 0% 52 ly On Sch On ith 244 241 37 27 33 230 2025 Sc 15 13 11 11 Not ly 40% 9 204 ly On 0% 20% 40% 60% 80% 100% -20% 2016 2017 2018 2021 2025 185 165 Not Preschool Primary Non-Formal 1st Cycle 2nd Cycle Technical Higher 5: DISTRIBUTION OF EDUCATIONAL IMPACTS BY COMMUNITY TYPE, 2009/10 20% 93 96 Secondary Secondary and Education 0% 6 CUMULATIVE SHARE OF POPULATION 71 62 72 74 calculations based on data from the National Statistics Service (NSS), 0% 33 37 52 School School Vocational for Education Technology (NaCET), the Assessment and Testing Center (ATC) and MOF Internal 15 13 11 11 9 27 Resources 0% 20% 40% 60% 80% 100% -20% 2016 2017 2018 2021 2025 Development Preschool Partners Primary Non-Formal 1st Cycle 2nd Cycle Technical Higher TS BY COMMUNITY TYPE, 2009/10 Republic of America CUMULATIVE SHARE OF POPULATION Shortage of Planned Budget Secondary Secondary and Education (a) Average PCF Allocation per Student (b) Average School Size School School Vocational (000s AMD) (Number of Students) Internal Resources 4 nd MOF Development Partners FIGURE 78: CLASS AND SCHOOL SIZES ARE SMALL 00 500 50 (b) Average School Size Republic of America Shortage of Planned Budget 400 00 (Number of Students) 300 Figure 80: Class and school sizes are small 50 200 2 00 500 50 100 FIGURE 8: SOURCES OF EDUCATION SECTOR FINANCE, 2013 400 0 0 Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP 300 l s l s ra ou ou Source: Kosovo BOOST Source: Kosovo BOOST an nta us s 0 m y en ity ra an nta us s 0 m y en ity Ru ou 40 om nit Ru Number of Students per Classroom Number of Students per Primar ou in 40 om nit ou ino Urb ud n ts in ou ino Urb ud n ts in nta u St u 200 in nta FIGURE 8: SOURCES OF EDUCATION SECTOR FINANCE, 2013 u St u m n-m nta m n-m nta ou 9 | 1% w ol in om ou w ol in om No ou No ou M < aC C 100 M FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC M CATEGORY, % GDP CATEGORY, % GDP < aC C hly M ou Sc ol in Source: Authors’ estimate from CBN, OECD, Nigeria, State Budget, Federal Government Budget, and General Household Survey Panel 2012/13 hly Sc ol in 0 Hig Hig 4.5% 4.5% o 0 o nta On us Sch ith s ho On Sch ith ho l s Source: Kosovo BOOST Source: Kosovo BOOST ou ou inonly ra an nly y en ity in ly Ru 40 om nit 426 | 18% ly in n-m nta t O Urb ud n ts tO ta 936 | 40% u St u 9 | 1% 4.0% 4.0% No ou No un m 0 m LGA No FIGURE 41: w ol in om o M FIGURE 41: < aC C M hly Sc ol in Household 4.5% 4.5% Hig BY SOURCES AND UNIT COST PROJECTION; 2016, 2025 3.5% BY SOURCES AND UNIT COST PROJECTION; 2016, 2025 3.5% HUN HUN o MDA BLR LTU UZB On Sch ith RQU ho AZE LVA CZE LVA KAZ KGZ SVN POL POL TJK MKD 295 | 13% UBEC SVK UKR 500 BGR nly 426 | 18% 40 ly 936 | 40% Source: Education Sector Strategy; 2016-2025, January 2015 Source: Education Sector Strategy; 2016-2025, January 2015 tO LGA State 4.0% 4.0% No 3.0% 3.0% Household FIGURE 41: FIGURE 41: Capital Outlays Capital Outlays (c) Average Class Size (d) Student-Teacher Ratio Federal Gov’t 295 | 13% UBEC 30% BY SOURCES AND UNIT COST PROJECTION; 2016, 2025 3.5% 4.0% AND UNIT BY SOURCES 2.5% COST PROJECTION; 2016, 2025 30% 3.5% Subsides & Transfers 2.5% Subsides & Transfers 587 | 25% Donors 12 State Source: Education Sector Strategy; 2016-2025, January 2015 3.0% Source: Education Sector Strategy; 2016-2025, January 2015 3.0% Utilities Utilities 5 10 76 | 3% 3.0% 2.0% Capital Outlays 2.0% Capital Outlays Goods & Services Goods & Services 35 0 (d) Student-Teacher Ratio 8 Federal Gov’t 30% 20% 4.0% 2.5% 30% 20% Subsides & Transfers 2.5% Subsides & Transfers 5 6 587 | 25% Donors 1.5% Wages & Salaries 1.5% Wages & Salaries 0 12 4 2.0% Utilities Utilities 400 5 10 2 76 | 3% 2.0% 2.0% 0 3.0% 1.0% Goods & Services 1.0% Goods & Services 08 l s FIGURE 12: THE STRUCTURE OF BASIC EDUCATION FINANCING 20% 10% 20% 10% l s ra ou 1.0% Wages & Salaries Wages & Salaries an nta us s 0 m y en ity 6 ou Ru ou 1.5% 1.5% 40 om nit ra an nta us in s 0 m y 40 om nity en ity ou ino Urb ud n ts Ru 30 ou in nta u St u in ou ino Urb 0.5% 0.5% ud n ts m n-m nta 4 in nta 2.0% u St u ou m 0 m n-m nta w ol in om ou No ou m wit l in om M No ou < aC C 2 Internally Federal M M hly 40 om nit < a C Sc ol in ho in C VAT Independent Devp’t 1.0% 1.0% M 0 hly Hig Fund Generated Account & Pool Revenue Partners 0% 0.0% 0.0% 0% 0.0% o l Hig FIGURE 12: THE STRUCTURE OF BASIC EDUCATION FINANCING 10% 10% ly us oo Sources On Sch ith Revenue Derivatives ho o l s Onino Sch u h ra ou 2016 2017 1.0% 2018 2021 2025 2016 2012 2017 2018 2021 an nta nlyus nly en ity Sc Ru ly in 2007 2008 2009 2010 2011 2007 2008 2009 2010 2011 2012 ou t Oino Urb ud n ts 0.5% 0.5% tO ta St u un n-m N nta No Federal w ol in om o o No ou M Internally Federal < aC C Ministry of M hly Fund VAT Independent Devp’t 25 Sc ol in Generated Account & Hig Sources Pool Revenue Partners Finance 0% 0.0% 0.0% 0% 0.0% o Revenue Derivatives 300 On Sch ith ho LGA SJGA Federal State 2016 2017 2018 Education spending as share 2021 2025 of total public spending 2010 2016 2017 2018 Education spending as share 2021 of total 2008public spending nly Con. Rev. 2007 2008 2009 2011 2012 2007 2009 2010 2011 2012 ly tO Federal No Responsibility Ministry of Federal Budget Capital spending as share of total education spending Capital spending as share of total education spending Finance (f) School Capacity Utilization Federal Education spending as share of GDP Education spending as share of GDP LGA SJGA State UBFC (2%) (Student Enrollment as % of Con. Rev. Education spending as share of total public spending Education spending as share of total public spending % % (f) School Capacity Utilization 70% Capacity) Total Builing Responsibility UBFC (2%) Federal Budget 20 FMF Capital spending as share of total education spending Education spending as share of GDP Capital spending as share of total education spending Education spending as share of GDP 2025 % (Student Enrollment as % of 60% Salary Non-Salary Salary Non-Salary Salary Non-Salary 200 Salary Non-Salary % Total Builing Capacity) 50% Education FMF Lower Finances Primary Lower Unity 15 % 40% Secondary School SUBEB Secondary College 2025 % 70% 30% % 60% 20% Salary Non-Salary Salary Non-Salary Salary Non-Salary Salary Non-Salary % Education Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 50% 10% Finances Primary Lower Lower Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) Unity l s Secondary “into which shall be paid all allocations to the local government councils of the state from the Federation 40% ra ou School Secondary College an 0% nta us s 0 mm y en ity Ru ou SUBEB it Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). in ou ino Urb un ud n ts l s FIGURE 5.12: DISTRIBUTION OF TEACHERS BY in nta St u ou 30% FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED m n-m nta ra an nta us s 0 m y en ity ou Ru 10 ou 40 om nit w ol in om in ou ino Urb ud n ts No ou o in nta u St u M < aC ho in C 20% m n-m nta M hly ou Source: Authors’ sketch following funding allocation arrangements in UBE following UBE Act of 2004 w ol in om 40 No ou s Sc ol Hig M Note: Each State shall maintain a special account to be called “State Joint Local Government Account (SJLGA) < aC 10% C ou nly ho M hly ith 100 Sc ol in “into which shall be paid all allocations to the local government councils of the state from the Federation Sc 0% Hig nta us nly o Account and from the Government of the State” (Section 162 [6], 1999 Constitution of Nigeria). On Sch ith FIGURE 5.12: DISTRIBUTION OF TEACHERS BY ho ou ino t O l s in O ou ra FIGURE 6.1: ALBANIA PUBLIC DEBT TO GDP RATIO UNDER INCREASED an nly y o en ity Ru 40 om nit n-m nta N ly in Urb ud n ts 75 tO nta 12k u St u m 0 m No ou w ol in om No ou M 5 < aC C M hly Sc ol in Hig o On Sch ith ho 10k nly ly 12k 75 70 tO No 8k 0 0 10k 70 65 63.1 6k 8k ROU LTU WEST. EUROPE 65 60 CHN CHL JPN ISR IDN ARG TUR BRA GBR IRL AUS FRA DEU ESP HUN PRT BEL CZE MEX FIN CHE AUT ITA POL EST ISL GRC RUS BLR PRK DNK SVK 63.1 56.9 6k 4k 60 55 54.9 4k 2k 56.9 FIGURE 14: SOURCES OF EDUCATION SECTOR FINANCE, 2013 50 BY ECONOMIC CATEGORY, % GDP FIGURE 5.11: EDUCATION EXPENDITURES BY ECONOMIC CATEGORY, % GDP Source: Calculated from CBN Annual report 2013 and per capital allocation from monthly shares of distribution from 55 54.9 49.7 Source: Kosovo BOOST The Federation Account by State, Nigeria Economic Report, The World Bank Group (2013) 2k 0 FIGURE 5.11: EDUCATION FIGURE 14: SOURCES OF EDUCATION SECTOR EXPENDITURESSource: BY ECONOMIC Calculated CATEGORY, from CBN Annual report 2013 and % per capital allocation from monthly GDP shares of distribution from 450 Federal FINANCE, Allocation From All Sources 2013 Internal Generated Revenue (IGR) Number of Students per Primary School 100 50 45 49.7 UNQUALIFIED /BEGINNER TEACHERS SECONDARY TEACHER HIGHER PEDAGOGICAL DEGREES MASTERS AND PH.D DEGREES 4.5% The Federation Account by State, Nigeria Economic Report, The World Bank Group (2013) 0 SCHOOL SCHOOL Source: Kosovo BOOST 400 Share of IGR (right axis) 90 UNQUALIFIED Source: Kosovo EMIS MASTERS AND 4.0% Federal Allocation From All Sources Per Capita Revenue (right axis) 80 40 /BEGINNER SECONDARY HIGHER PH.D DEGREES 350 100 45 4.5% 450 Internal Generated Revenue (IGR) Share of IGR (right axis) 3.5% 90 Source: NSI except Poland 70 (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 TEACHERS TEACHER SCHOOL PEDAGOGICAL SCHOOL 2008 DEGREES 400 300 2013 500 Source: Kosovo EMIS 4.0% Capital Outlays 350 Per Capita Revenue (right axis) 3.0% 250 Capital Outlays 80 “Western Europe” includes 60 other members of the European Union not mentioned40 in the gure, less Malta and 2008 50 Combine 70 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Increase 2019 2020 Education in 2021 2022 and Health and Spending 3.5% Subsides & Transfers 300 2.5% 200 Subsides & Transfers 60 Cyprus, plus Norway, Iceland and Switzerland. OECD at a Glance, 2011. 40 Health Spending Increases from 2.6 to 4 percent of GDP 2013 3.0% Utilities 250 150 Utilities 30 Education Spending Increases from 3 to 4 percent of GDP Goods & Services 200 2.0% Capital Outlays 100 Goods & Services 50 Notes: Year of reference 20 2008. Public institutions only (including for Belarus). Combine Increase in Education and Health and Spending Baseline 2.5% Subsides & Transfers 40 Health Spending Increases from 2.6 to 4 percent of GDP Wages & Salaries 150 1.5% 50 Wages & Salaries Utilities 10 30 Education Spending Increases from 3 to 4 percent of GDP 400 2.0% 100 1.0% Goods & Services 0 20 0 Baseline Borno Taraba Anambra Adamawa Bauchi Yobe Gombe Jigawa Delta Zamfara Kaduna Katsina Kebbi Sokoto Kano Benue Niger Kogi Lagos Nasarawa Plateau FTC Abuja Kwara Enugu Imo Anambra Ebonyi Abia Bayelsa Akwa Ibom Delta Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun Lagos 1.5% 50 Wages & Salaries 10 0.5% 0 0 FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 1.0% 0.0% Expenditure on Education and Teachers’ Salaries Borno Taraba Adamawa Bauchi Yobe Gombe Jigawa Zamfara Kaduna Katsina Kebbi Sokoto Kano Benue Niger Kogi Nasarawa Plateau FTC Abuja Kwara Enugu Imo Ebonyi Abia Bayelsa Akwa Ibom Cross River Edo Rivers Ondo Ekiti Osun Oyo Ogun Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 011 2012 0.5% 2007 2008 2009 2010 2011 North 2012East North West North Central South East South South South West FIGURE 16: TEACHER SALARIES RELATIVE TO PER CAPITA GDP, 2012 0.0% FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES Expenditure on Education BY BROAD ECONOMIC and Teachers’ Salaries 300 COUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH 2007 2008 2009 2010 2011 North East 2012 North West North Central South East South South South West CATEGORY SOURCE: DATA FROM MFB/SIGFP. Source: OECD 2014. PISA 2012 Results: What Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 PER CAPITA GDP OVER USD 20 000 PER CAPITA GDP LESS USD 20 000 FIGURE 15: EVOLUTION OF REAL EDUCATION EXPENDITURES BY BROAD ECONOMIC 200 2.5 FIGURE 3: AN OVERVIEW OF THE COMPLEXITY OF TRANSFERS & FUND FLOWS, 2012 CATEGORY SOURCE: DATA FROM MFB/SIGFP. 700 COUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH PER CAPITA GDP OVER USD 20 000 180 PER CAPITA GDP LESS USD 20 000 160 2.0 FIGURE 13: EXPENDITURE PERFO FIGURE 3: AN OVERVIEW OF THE COMPLEXITY OF TRANSFERS & FUND FLOWS, 2012 LEVEL BUDGET 600 200 2.5 700 BILLION CONSTANT 2013 AR. NATIONAL MINISTRY OF FINANCE MINISTRY OF EDUCATION CULTURE 180 140 500 MOF MOEC 200 Source: World Bank Edstats based on UNESCO statistics for education variab 160 120 2.0 1.5 LEVEL BUDGET DAU DAK BOS Adjustment Fund Tugas Pembantuan Dekon Central Functions 600 100 140 400 BILLION CONSTANT 2013 AR. NATIONAL MINISTRY OF FINANCE MINISTRY OF EDUCATION CULTURE MOF MOEC Adjustment Fund 500 120 80 1.5 1.0 Based on Averages 2009-2012* Indicator: Primary Completion Rates 300 Tugas Own Central Dekon Source DAU DAK BOS PROVINCIAL Pembantuan Functions District Budget Provincial Education Central Government 100 60 400 Revenue DAU/SDA Agencies in the Regions RIBUTION OF TEACHERS BY MOST EFFICIENT: DEVIATIONS FROM EXPECTED EDUCATION OUTCOME PAD Primary completion rate higher than expected, 80 40 1.0 0.5 Gov. Exp./GDP lower than expected Own 200 40 PROVINCIAL Source District Budget 300 60 20 100 Revenue DAU/SDA Provincial Education Central Government 20 Own District Budget Agencies in the Regions 100 PAD DISTRICT Source DAU/SDA Revenue PAD District Education 40 0 0.5 0 200 DAK 0 Fund Originated from MoF Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania Fund Originated from MoEC Sectoral Budget 20 -20 Own District Budget 0 FIGURE 79: MOST STUDENTS ARE IN URBAN AREA WHILE MOST SCHOOLS ARE IN RURAL AREAS Source DAU/SDA Fund Originated from Provincial Budget 100 DISTRICT Revenue PAD District Education 0 0 -40 DAK Fund Originated from District Budget UNDERACHIEVERS: Fund SCHOOLS from MoF Originated PRIVATE PUBLIC SCHOOLS Primary completion rate lower than expected, 2006 2008 2009 2010 2011 2012 2013 Hong Kong-China Germany Korea Portugal Spain Netherlands New Zealand Ireland Canada Denmark Japan Qatar Belgium United Kingdom Singapore Slovenia Luxembourg Finland Australia Italy Greece Austria Macao-China France Poland United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania Gov. Exp./GDP lower than expected Fund Originated from MoEC Sectoral Budget Fund Originated from Provincial Budget Source: Elaboration based on Permendagri 13/2006 on Guidelines of Sub-National Financial Management, World Bank (2009) and Law 22/2011 on AP3N 2012 0 -4 -2 0 2 DEVIATIONS FROM EXPECTED PUBLIC E Fund Originated from District Budget 0 Note: Adjustment Fund also includes the local incentive grant (Dana Insentif Daerah, or DID) Source: Belstat, MOE. Capital Expenditures Cumulative Expenditure by Educational PRIVATE SCHOOLS PUBLIC SCHOOLS 2006 2008 2009 2010 2011 2012 2013 Institutions per Students Aged 6 – 15 Source: Elaboration based on Permendagri 13/2006 on Guidelines of Sub-National Financial Management, World Bank (2009) and Law 22/2011 on AP3N 2012 Other Recurrent Expenditure Lower Secondary Teachers’ Salaries (after 15 years of Note: Adjustment Fund also includes the local incentive grant (Dana Insentif Daerah, or DID) Regular Salaries experience/minimum training relative to per capita GDP Cumulative Expenditure by Educational Capital Expenditures ROU LTU BGR LVA EST RUS WEST. EUROPE MDA SVN BLR GEO HUN UKR HRV SVK Institutions per Students Aged 6 – 15 Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP Other Recurrent Expenditure POL EST ISL GRC RUS BLR SVK Lower Secondary Teachers’ Salaries (after 15 years of experience/minimum training relative to per capita GDP Regular Salaries Lower Secondary Teachers’ Salaries (after 15 years of FIGURE 12: experience/minimum training relative to per capita GDP 250 100 4 7 6 6 5 5 4 6 8 HYPOTHESIS: SIMULATING THE IMPACT OF THE 12 90 FIGURE 12: 200 INTEGRATION OF COMMUNITY TEACHERS 100 80 250 4 7 6 6 5 5 4 6 8 70 100 150 90 60 65 66 59 65 55 48 55 59 50 HYPOTHESIS: SIMULATING THE IMPACT OF THE 600 HYPOTHESIS OF PERCENT MASTERS AND 200 50 INTEGRATION OF COMMUNITY TEACHERS 600 600 SECONDARY HIGHER PH.D DEGREES 80 586 RESOURCE NEEDS 40 573 Source: NSI except Poland90 100 TEACHER PEDAGOGICAL DEGREES 70 561 (BOOST data), Slovakia (Ministry of Education), statistics authorities of CIS countries. “Western H1 22.6 SCHOOL SCHOOL 30 550 10 65 66 59 65 55 48 55 59 50 150 60 50 20 600 500 HYPOTHESIS OF PERCENT MASTERS AND 50 10 586 600 600 H2 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 PH.D DEGREES 100 40 0 31 28 35 29 40 47 41 34 41 573 RESOURCE NEEDS 80 35 0 30 561 2001 2002 2003 2004 2005 2006 2007 2008 2009 2001 2002 2003 2004 2005 2006 2007 2008 2009 550 H1 MOBILIZATION Source: own calculations using NUPTIK Data, 2010 Europe” includes other members of the European Union not mentioned in the figure, less Malta and Cyprus, plus 20 50 10 500 400 H2 POTENTIALSCENARIO FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 60.4 Central 31 28 35 29 40 47 41 34 41 0 District S1 80 8 0 70 (Sistem Informasi Keuangan Daerah, SIKD), Ministry of Finance 2003 2004 2005 2006 2007 2008 2009 Source: own calculations using NUPTIK Data, 2010 Province 2001 2002 511 % OF TEACHERS WITH SENIOR SECONDARY 2001 2002 2003 2004 2005 2006 2007 2008 2009 OR BELOW AS HIGHEST EDUCATION LEVEL MOBILIZATION 400 428 S2 Norway, Iceland and Switzerland. OECD at a Glance, 2011. Central 300 350 POTENTIALSCENARIO District 80 60 Province (Sistem Informasi Keuangan Daerah, SIKD), Ministry of Finance 229 S1 % OF TEACHERS WITH SENIOR SECONDARY 511 60 OR BELOW AS HIGHEST EDUCATION LEVEL FIGURE 13: 428 S2 300 200 40 60 50 Public institutions only (including for Belarus). Notes: Year of reference 2008. 350 6 2 100 100 229 10 11 FIGURE 13: 12 23 90 40 20 200 9 80 15 6 80 70 100 2 100 100 60 44 60 0 POL 10 11 20 40 12 90 PERCENT ALARIES RELATIVE TO PER CAPITA GDP, 2012 PERCENT 23 100 9 51 50 77.4 4 80 15 80 DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat hers’ Salaries 6 40 75 70 73 40 100 0 FIGURE 1: PRESENT VALUE OF at Makes Schools Successful? Resources, Policies and Practices (Volume IV), pg. 96 60 44 60 30 (B MGA) 2016 2017 2018 2019 2020 0 CORE BLOCK RELATIVE TO DISC PERCENT 20 PERCENT 47 20 12 30 100 65 51 50 Nuggoe Acoh 25 Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta 10 40 75 73 0 40 0 0 Graphic Derivation of Crossover Discount Rate, Cho OUNTRIES AND ECONOMICS WITH COUNTRIES AND ECONOMICS WITH EARLY 30 BASIC SECONDARY UNIVERSITIES OTHERS 96 2008 2009 2008 2009 Chika and Hollow Concrete Block Alternatives in Eth (B MGA) ER CAPITA GDP OVER USD 20 000 PER CAPITA GDP LESS USD 20 000 20 47 CHILDHOOD EDUCATION 20 EDUCATION 2016 2017 2018 2019 2020 2 EDUCATION 20 25 TOTAL 10 250k 39.6 0 2.5 0 EARLY BASIC SECONDARY UNIVERSITIES Central OTHERS Central COUNTRIES AND ECONOMICS WITH CHILDHOOD EDUCATION EDUCATION 2008 2009 2008 2009 Central-80S District PER CAPITA GDP LESS USD 20 000 EDUCATION TOTAL District Province 10 Province 200k 2.5 2.0 Central Central 0 Central-80S District CROSSOVE DISCOUNT District Province PROJECTED ENROLLMENT GAP AND OPERATING Overall, lower student-teacher 0 ratios and smaller classes imply high costs per student, which do not Province 150k 2.0 1.5 EXPENDITURE GAP DUE TO POPULATION GROWTH PROJECTED ENROLLMENT GAP AND OPERATING Students Teachers Schools STR U 100k 1.5 1.0 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 EXPENDITURE GAP DUE TO POPULATION GROWTH 300k 1.4M 10 15 20 necessarily lead to better learning outcomes. Source: own calculations using NUPTIK Data, 2010 Chika 200k 1.3M 1.0 0.5 FIGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 4.0% HCB ENROLLMENT GAP 300k 100k 1.4M 1.2M Source: own calculations using NUPTIK Data, 2010 80 % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL 3.0% FUNDING GAP 200k 1.3M 0.5 0 - 1.1M 4.0% 80 60 ENROLLMENT GAP 2.0% % OF TEACHERS WITH SENIOR SECONDARY 100k 1.2M Canada Denmark Japan Qatar Belgium United Kingdom Montenegro Singapore Slovenia Croatia Luxembourg Finland Australia Italy Greece Austria ArgentinaMacao-China France Poland Indonesia United States Israel Sweden Norway One Republic Iceland Hungary Estonia Slovak Republic Jordan Malaysia Tunisia Turkey Mexico Colombia Montenegro Chile Croatia Thailand Shanghai-China Lithuania Bulgaria Peru Argentina Uruguay Latvia Indonesia Romania OR BELOW AS HIGHEST EDUCATION LEVEL (100k) 1M 3.0% Rural City of Minsk Vitebsk Oblast FUNDING GAP 40 OST SCHOOLS ARE IN RURAL AREAS 0 - 1.1M 1.0% (200k) 900k 60 FIGURE 3.2: SOURCES OF FUNDS: 2010 2.0% Malaysia Tunisia Turkey Mexico Colombia Chile Thailand Shanghai-China Lithuania Bulgaria Peru Uruguay Latvia Romania (100k) 1M (300K) 800k 20 0.0% FIGURE 2.1: IMPACT OF Urban Grodno Oblast Mogiev Oblast 100 (200k) 2012 2013 2014 2015 2016 2017 2018 2019 2020 2012 2013 2014 2015 2016 2017 2018 40 2019 2020 1.0% 6.9 900k by Educational FIGURE 3.2: SOURCES OF FUNDS: 2010 16 Aged 6 – 15 80 10 (300K) 800k with Current Funding and GER Operating Costs 20 0 0.0% 100 15 2012 2013 2014 2015 2016 2017 2018 2019 2020 GDP Growth: 1% 2012 2013 2014 2015 2016 2017 2018 2019 2020 500% Brest Oblast Gomel Oblast DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat rs’ Salaries (after 15 years of 6.9 ining relative to per capita GDP 16 29 GDP Growth: 2% 450% rs’ Salaries (after 15 years of 80 60 10 20 Operating Costs with Current Funding and GER GDP Growth: 3% 0 Nuggoe Acoh ining relative to per capita GDP 15 GDP Growth: 1% 400% Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta GDP Growth: 4% Minsk Oblast Belarus 29 40 GDP Growth: 2% 350% 60 20 GDP Growth: 3% 300% 20 48 54 GDP Growth: 4% 40 250% 12 20 48 0 PTCA 54 Non SBM Grantee SBM Grantee 200% 150% 0 LGU FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY FIGURE 76: CONSOLIDATION OF SCHOOL NETWORK 100% Non SBM Grantee SBM Grantee Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). Others 50% 10 Community 1991=100 PTCA LGU FIGURE 75: NUMBER OF STUDENTS HAS DECLINED, ESPECIALLY All Ages FIGURE 76: CONSOLIDATION OF SCHOOL NETWORK 0% URE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 Rural Others Age 5-14 Source: CISSTAT. “Statistics of the Countries of the CIS”, MOE (2011). 130 own calculations using NUPTIK Data, 2010 Community 1991=100 Urban PRIMARY All Ages Age 5-24 EDUCATIO COMPLET ECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 Rural 12k Age 5-14 130 110 80 Source: UN, Beistat 8 Urban % OF TEACHERS WITH SENIOR SECONDARY Age 5-24 OR BELOW AS HIGHEST EDUCATION LEVEL FIGURE 67: APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES Source: own calculations using School Based Management 60 10k Share of total household expenditures, by region 12k 110 90 FIGURE 2.2: PRIVATE RAT Source: EMOP 2014 Source: UN, Beistat FIGURE 67: APPENDIX FIGURE 29: SHARE OF TOTAL HOUSEHOLD EXPENDITURES 50 40 Share of total household expenditures, by region 2% 10k 8k 90 70 Source: own calculations using School Based Management 6 45 Source: EMOP 2014 20 50 40 25% 2% 15% 6k STUDENT TEACHER RATIO Bamako 8k 70 50 Koulkoro 45 35 20% 0 15% 11% Koyos 4k 40 30 6k STUDENT TEACHER RATIO 50% Gilo Bamako 4 50 DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Banton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Awa TungahSunatrabarat Bali Dulosalami DI. Yoguaunta Jama Timur Jamaluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Koulkoro Segou 35 25 15% Mepti 2% Koyos 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 11% Gilo 10% Silcasco 4k 2k 30 20 50% Tombouctou Nuggoe Acoh Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali Sunatrabarat Nuggoe Acoh Dulosalami Jama Timur Dou Jauta Segou 10% 25 15 Mepti 2% 1991 1993 1995 1997 1999 Number 2001 of Schools 2003 2005 2007 2009 Silcasco 10% 2k 0 20 10 10% Number of Students 2 Tombouctou 5% 2% 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 Number of Teachers 15 5 10% Number of Schools 0 10 0 0% Number of Students 2% 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 Number of Teachers 5 0 1 2 3 4 5 6 7 PRIMARY EDUCATIO 0 COMPLETE 0 FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 BUDGET FOR STUDENT IN RUPIAH Source: EMOP 2014 0 1 2 3 4 5 6 7 Male Female FIGURE 78: Source: EdStats database. BUDGET FOR STUDENT IN RUPIAH FIGURE 52: SHARE OF HOUSEHOLD BUDGET SPENT ON EDUCATION, BY REGIONS, 2014 Source: EMOP 2014 Primary school student-teacher ratios in the ECA region (1999-2009) Secondary school student-teacher ratios in the ECA region (1999-2009) STR Urban Rural IGURE 67: SHARE OF TEACHERS WITH SENIOR SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 FIGURE 78: urce: own calculations using NUPTIK Data, 2010 Source: EdStats database. 30 18 R SECONDARY OR LESS AS THEIR HIGHEST EDUCATION BY PROVINCE, 2010 80 FIGURE 4.14: FISCAL IMPACT OF PARENTAL FEE ELIMINATION BY REGION Primary school student-teacher ratios in the ECA region (1999-2009) 30 25 Secondary school student-teacher ratios in the ECA region (1999-2009) 18 16 14 FIGURE 4.14: FISCAL IMPACT OF PARENTAL FEE ELIMIN % OF TEACHERS WITH SENIOR SECONDARY OR BELOW AS HIGHEST EDUCATION LEVEL 20 Additional cost as percent of GDP 12 Additional cost as percent of GDP 16 FIGURE 67: 60 25 14 10 Source: own calculations using School Based Management 15 TOMBOUCIOU .13% 8 FIGURE 67: City of Minsk Vitebsk Oblast 20 12 40 10 10 Source: own6 calculations using School Based Management 50 TOMBOUCIOU .13% GAO .41% 15 4 45 8 5 20 2 50 Grodno Oblast Mogiev Oblast KOULIKORO .51% 10 6 40 UDENT TEACHER RATIO MOPTI .11% 0 0 GAO .41% 4 45 35 0 HUN HUN MDA MDA BLR UZB ROU UZB LTU RQU CVE AZE AZE LVA CZE LVA KAZ KAZ KGZ SVN KGZ SVN POL POL TJK MKD LTA TJK MKD ARM 5 BLR SVK RUS UKR SVK UKR BGR BGR KAYES .50% SEGOU .63% BAMAKO 2 40 KOULIKORO .51% 1.75% 30 POLRATIO Brest Oblast Gomel Oblast aunta Jama Timur aluat Dou Jauta anton Awa Tungah Bali DI. Yoguaunta Jamaluat Banton ungahSunatrabarat Bali Dulosalami aunta Jama Timur aluat Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat e Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat Banton Sunatrabarat e Acoh Dulosalami Jama Timur Dou Jauta Awa Tungah Bali DI. Yoguaunta Jamaluat MOPTI .11% 0 0 35 25 2009 HUN HUN MDA MDA BLR UZB UZB LTU ROU RQU CVE AZE AZE LVA CZE LVA KAZ KAZ KGZ SVN KGZ SVN POL TJK MKD LTA TJK MKD ARM BLR SVK RUS UKR SVK UKR BGR BGR KAYES .50% SEGOU .63% EACHER BAMAKO 1.75% SIKASSO .79% 2004 30 20 e Acoh Timur Jauta ungah Bali aunta aluat anton abarat salami Timur Jauta ungah Bali abarat salami Timur Jauta 25 163 Example 22: Analysis of unit costs and outcomes Albania PER (2014) Overall comment: The Albania analysis provides an example where the efficiency discussion in terms of public spending and learning outcomes may not be straightforward. (Both inputs and outputs for the country are low, compared to those for other countries). The Albania case also includes an analysis of the relationship between financial resources and learning outcomes at the subnational level and reveals problems of both inefficiency and equity in spending. Compared to other countries in the region, Albania’s education system gets poor learning outcomes and has low public spending. Hence, it is difficult to determine whether Albania’s system is efficient given that both inputs and outputs are low. The figure shows Albania on the curve that best fits international data on outcomes and public spending per pupil. Albania appears among the lowest spending countries in the sample. While the efficiency discussion in terms of public spending and learning outcomes is not clear it is evident that the system has not been effective at providing students with the necessary basic competencies. Figure 2: International comparison: PISA score and expenditure per secondary student (Reading, Mathematics and PISA 2012 Average score Science) Total expenditures per pupil in secondary education (USD converted using PPPs for GDP) Source: IMF, PISA, and World Bank calculations 164 Education Public Expenditure Review Guidelines FIGURE 78: STUDENT TEACHER RATIOS ARE AM Primary school student-teacher ratios in the ECA region (1999-2009) subnational level Measuring efficiency at30 Results from Matura, the university 25entrance exam, suggest that financial resources alone do not explain learning outcomes. Figure 2 illustrates the relationship between financial resources and learning 20 outcomes (as measured by 2013 Matura mathematics scores) by district. While the flat blue line is an unweighted average score for 15 the county (average of the blue dots), the flat red line measures expenditures per pupil in each county (right-hand side axis). The figure shows that some counties 10 in using their resources to produce good learning outcomes. seem to be more efficient than others For instance, Qarku Tirane, Vlorë, Elbasan, and Fier spend approximately the country’s average and get relatively good results, while other 5districts spend as much and do not get good learning outcomes. 0 It is important to stress that this analysis is based on correlation, and thus causality between expenditures HUN MDA UZB ROU CVE AZE LVA KAZ KGZ SVN POL TJK LTA MKD ARM BLR SVK UKR RUS BGR and scores cannot be claimed.a The figure also suggests that the financing mechanism does not target poorer prefectures (as measured by the share of population in poverty). For instance, Gjirokastër has low poverty incidence and the highest expenditure per pupil, whereas Kukës is poorer and gets a low per capita FIGURE 3:allocation. UNIVERSITY EXAMINATION SCORES AND COUNTY FINANCIAL RESOURCES Source: INSTAT. National Examinations Agency, and Ministry of Education and Sports. Figure 3: University examination scores and county financial resources Note: Matura results are from 2013: poverty headcount comes from INSTAT’s analysis of the LSMS 2012 and measures the share of the population below Albania’s poverty line. 8.0 80k 7.5 75k 7.0 70k EXPENDITURE PER PUPIL RED MATURA MATHMATICS BLUE 6.5 65k 6.0 60k 5.5 55k 5.0 50k 4.5 45k 4.0 40k 3.5 35k 3.0 30k Shkodër Berat Kucove Shrapar Bulque Dibër Mut Durrës Kruje Elbasan Granish Libradad Peqin Fier Lushnje Malbraster Gjrobasier Pernet Tepelene Korçë Devoll Kolonje Pogrades Kukës Has Tropoje Lezhë Kurbin Mirdie Maleseradhe Pude Karuke Delvine Sarande Vlorë Tiranë Berat Dibër Durrës Elbasan Fier Gjrobasier Korçë Kukës Lezhë Shkodër Tiranë Vlorë POVERTY HEADCOUNT % 20 15 10 Berat Dibër Durrës Eibasan Fier Gjrokas Korçë Kukës Lezhë Shkodër Tiranë Vlorë Source: INSTAT, National Examinations Agency, and Ministry of Education and Sports. Note: Matura results are from 2013; poverty headcount comes from INSTAT’s analysis of the LSMS 2012 and measures the share of the population below Albania’s poverty line. a For instance, it may be that Gjirokastër’s results would have been much lower had it not spent this amount of resources on education. 165 Belarus PER (2013) Overall comment: The Belarus analysis explains why higher per student costs are not translating into better performance outcomes: While a low student-teacher ratio (STR) is driving up per student costs, the quality of teaching has not improved. In Belarus, demographic change is affecting the demand for education services. Declines in school- age population and urbanization have reduced the number of students, especially in rural areas. As a result, student-teacher ratios at both the primary and secondary level are among the lowest in the region. Teacher salaries remain relatively low, making it difficult to attract and retain skilled labor in the teaching force. While per student costs have risen, this has not necessarily led to better learning outcomes (Figure 12). FIGURE 12: LOW STUDENT TEACHER RATIO IS DRIVING COSTS BUT NOT LEADING TO BETTER OUTCOME Source: World Bank sta calculations based on BOOST, MOE and Belstat data. Each dot represents a rayon. Figure 12: Low student-teacher ratio is driving costs, but no leading to better outcomes FIGURE Per Student 12: BYR Cost, Thousand LOW STUDENT TEACHER RATIO IS DRIVING COSTS BUT NOT LEADING BETTER OUTCOMES TOScores Average Math Per student cost, thousand BYR Source: World Bank sta calculations based on BOOST, MOE and Belstat data. Each dot represents a rayon. 10k Per Student Cost, Thousand BYR Average Math Scores 45 10k 45 40 8k 40 35 8k 35 30 6k 30 6k 25 25 20 20 4k 4k 15 15 2k 10 2k 10 5 5 ATIO IS DRIVING COSTS BUT 0 NOT 0 LEADING TO BETTER OUTCOMES 0 esents a rayon. 2 4 6 8 10 12 0 0 2k 4k 6k 18k 10k 2 4 6 8 10 12 0 2k 4k 6k Average Math Scores STUDENT TEACHER RATIO PER STUDENT COSTS, THOUSAND BYR NT TEACHER RATIO IS DRIVING COSTS BUT NOT LEADING TO BETTER OUTCOMES OE and Belstat data. Each dot represents a rayon. 45 STUDENT TEACHER RATIO PER STUDENT COSTS, TH Average Math Scores 40 35 45 40 30 35 25 30 20 25 20 15 15 10 10 5 5 0 0 6 8 10 12 0 2k 4k 6k 18k 10k 10 12 0 2k 4k 6k 18k 10k UDENT TEACHER RATIO PER STUDENT COSTS, THOUSAND BYR RATIO PER STUDENT Source: World Bank staff calculations COSTS, based on BOOST, THOUSAND MOE BYREach dot represents a rayon. and Belstat data. 166 Education Public Expenditure Review Guidelines Indonesia PER (2013) Overall comment: The Indonesia review provides an example of subnational analysis by correlating “changes” in total spending on education and “changes” in enrollments or learning outcomes across districts. It takes into account that districts with different characteristics may need to spend different amounts to reach the same outcomes. By looking at the rates of change in spending and outcomes, this comparison controls for district characteristics that are constant over time. Correlating spending per student with outcomes may miss an important point: districts with different characteristics may need to spend different amounts to reach the same outcomes. So a lack of correlation in one year may not mean that district spending does not matter for outcomes. A cleaner comparison is obtained by regressing changes in resources with changes in outcomes. If district spending on education matters for education outcomes, districts that increase their spending faster than other districts should experience a faster improvement in outcomes. By looking at the rate of these changes, this comparison controls for district characteristics that are constant over time. The Indonesia 2013 PER analysis shows no correlation between changes in total spending on education and changes in enrollments or learning outcomes across districts (Figure 57). The fact that increasing spending is not correlated with improvements in outcomes means that further increases in spending, given all other factors unchanged, will be unlikely to be associated with improved outcomes, and that outcomes may vary across districts regardless of changes in spending. 167 FIGURE FIGURE 57: CHANGE 57: CHANGE IN PUBLIC IN PUBLIC EDUCATION SPENDING EDUCATION SPENDING ANDAND CHANGE CHANGE IN EDUCATION IN EDUCATION OUTCOMES OUTCOMES AT THE AT DISTRICT LEVEL THE DISTRICT LEVEL Source: Source: MoF MoF SIKD SIKD (budget), Susenas (budget), (NER) Susenas (NER) (UN and MoEC MoEC (UN exams) and exams) FIGURE FIGURE 57: CHANGE 57: CHANGE IN PUBLIC IN PUBLIC EDUCATION SPENDING EDUCATION SPENDING ANDAND CHANGE CHANGE IN EDUCATION IN EDUCATION OUTCOMES OUTCOMES AT THE AT DISTRICT LEVEL THE DISTRICT LEVEL Source: Source: MoF MoF SIKD SIKD (budget), Susenas (budget), (NER) Susenas (NER) (UN and MoEC MoEC (UN exams) and exams) FIGURE FIGURE 57: 57: CHANGE CHANGE FigureIN PUBLIC 57: EDUCATION IN PUBLIC Change SPENDING in EDUCATION SPENDING public education ANDAND CHANGE CHANGE spending and IN EDUCATION change EDUCATION IN in OUTCOMES educationOUTCOMES AT THE outcomes AT DISTRICT THE DISTRICT LEVEL LEVEL Source: MoF Source: MoF SIKD SIKD (budget), Susenas a)Susenas (budget), (NER) andNER, SMA a) SMA MoEC (NER)total andNER, (UN total spending exams) MoEC spending (2002-2009) (UN exams) (2002-2009) b) SMA b) SMA NER, non-salary NER, non-salary spending (2002-2009) spending (2002-2009) at the district level 150 150 a) SMA a) SMA NER, NER, total total spending spending (2002-2009) (2002-2009) 150 150 b) SMA NER, b) SMA NER, non-salary non-salary spending (2002-2009) spending (2002-2009) % % % % ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT 150 150 a) SMA a) SMA NER, NER, total total spending spending (2002-2009) (2002-2009) 150 150 b) SMA NER, b) SMA NER, non-salary non-salary spending (2002-2009) spending (2002-2009) 100 100 100 100 % % % % ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL 150 150 150 150 100 100 100 100 % % % % 50 50 50 50 ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL IN SENIOR IN SENIOR IN SENIOR IN SENIOR 100 100 100 100 50 50 50 50 0 0 0 0 SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL SENIOR HIGHSCHOOL INCHANGE INCHANGE INCHANGE INCHANGE 50 50 50 50 0 0 0 0 -50 -50 -50 -50 CHANGE CHANGE CHANGE CHANGE -100 -100 0 0 100 100 200 200 300 300 0 0 500 500 1000 1000 0 0 0 0 -50 -50 -50 -50 CHANGE IN CHANGE IN CHANGE IN CHANGE IN of SMA NER Change Change & Education of SMA NER & Education Expenditure Expenditure of SMA NER Change Change & Education of SMA NER & Education Expenditure Expenditure -100 -100 0 0 100 100 200 200 300 300 0 0 500 500 1000 1000 Fitted Line Fitted Line Fitted Line Fitted Line -50 -50 -50 -50 of SMA NER Change Change & Education of SMA NER & Education Expenditure Expenditure of SMA NER Change Change & Education of SMA NER & Education Expenditure Expenditure -100 -100 0 0 100 100 200 200 300 300 0 0 500 500 1000 1000 Fitted Line Fitted Line Fitted Line Fitted Line Change of SMA NER Change of SMA & Education NER & Education Expenditure Expenditure Change of SMA NER Change of SMA & Education NER & Education Expenditure Expenditure Fitted Line Fitted Line Fitted Line Fitted Line c) SMP c) SMP NER, NER, total total spending spending (2002-2009) (2002-2009) d) SMP NER, d) SMP NER, non-salary non-salary spending (2002-2009) spending (2002-2009) 100 100 100 100 c) SMP c) SMP NER, NER, total total spending spending (2002-2009) (2002-2009) d) SMP NER, d) SMP NER, non-salary non-salary spending (2002-2009) spending (2002-2009) % % % % ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT 100 100 100 100 c) SMP c) SMP NER, NER, total total spending spending (2002-2009) (2002-2009) d) SMP NER, d) SMP NER, non-salary non-salary spending (2002-2009) spending (2002-2009) % % % % ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT 50 50 50 50 HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL 100 100 100 100 % % % % ENROLLMENT ENROLLMENT ENROLLMENT ENROLLMENT 50 50 50 50 JUNIOR HIGHSCHOOL JUNIOR HIGHSCHOOL JUNIOR HIGHSCHOOL JUNIOR HIGHSCHOOL IN JUNIOR IN JUNIOR IN JUNIOR IN JUNIOR 0 0 0 0 50 50 50 50 HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL HIGHSCHOOL CHANGE CHANGE CHANGE CHANGE 0 0 0 0 JUNIOR IN JUNIOR IN JUNIOR IN JUNIOR IN -50 -50 -50 -50 CHANGE CHANGE CHANGE CHANGE 0 0 0 0 0 0 500 500 1000 1000 -100 -100 0 0 100 100 200 200 300 300 CHANGE IN CHANGE IN CHANGE IN CHANGE IN -50 -50 -50 -50 of SMP NER Change Change & Education of SMP Expenditure NER & Education Expenditure Change Change of SMP NER of SMP NER & Non-Salary & Non-Salary Education Education Expenditure Expenditure -100 -100 0 0 100 100 200 200 300 300 0 0 500 500 1000 1000 Fitted Line Fitted Line Fitted Line Fitted Line -50 -50 -50 -50 of SMP NER Change Change & Education of SMP NER & Education Expenditure Expenditure Change 0 of SMP0NER Change & Non-Salary of SMP NER & Non-Salary Education Education 500 Expenditure 1000 Expenditure 500 1000 -100 -100 0 0 100 100 200 200 300 300 Fitted Line Fitted Line Fitted Line Fitted Line Change of SMP NER Change of SMP & Education NER & Education Expenditure Expenditure Change ofChange SMP NER & Non-Salary of SMP Education NER & Non-Salary Expenditure Education Expenditure Fitted Line Fitted Line Fitted Line Fitted Line e) score, e) SMP UN SMP UN score, total total spending spending (2005-2009) (2005-2009) f) SMP UN SMP UN f) score, score, non-salary non-salary (2005-2009) (2005-2009) 60 60 60 60 e) score, e) SMP UN SMP UN score, total total spending spending (2005-2009) (2005-2009) f) score, f) SMP UN SMP UN score, non-salary non-salary (2005-2009) (2005-2009) 60 60 60 60 % IN SMP % % IN SMP % % IN SMP % % IN SMP % 40 40 40 40 e) score, e) SMP UN SMP UN score, total total spending spending (2005-2009) (2005-2009) f) score, f) SMP UN SMP UN score, non-salary non-salary (2005-2009) (2005-2009) SCORE SCORE SCORE SCORE 60 60 60 60 IN SMP IN SMP IN SMP IN SMP 40 40 40 40 20 20 20 20 EXAM EXAM EXAM EXAM SCORE SCORE SCORE SCORE % % % % NATIONAL NATIONAL NATIONAL NATIONAL IN SMP IN SMP IN SMP IN SMP 40 40 40 40 20 20 20 20 IN EXAM IN EXAM IN EXAM IN EXAM 0 0 0 0 EXAM SCORE EXAM SCORE EXAM SCORE EXAM SCORE NATIONAL NATIONAL NATIONAL NATIONAL INCHANGE INCHANGE INCHANGE INCHANGE 20 20 20 20 0 0 0 0 -20 -20 -20 -20 NATIONAL NATIONAL NATIONAL NATIONAL CHANGE CHANGE CHANGE CHANGE -100 -100 0 0 100 100 200 200 300 300 -100 -100 0 0 100 100 200 200 300 300 0 0 0 0 -20 -20 -20 -20 CHANGE IN CHANGE IN CHANGE IN CHANGE IN of SMP UN Change Change of SMP Score UN Score & Education & Education Expenditure Expenditure Change Change of SMP UN Score of SMP UN Score & Education & Education Expenditure Expenditure -100 -100 0 0 100 100 200 200 300 300 -100 -100 0 0 100 100 200 200 300 300 -20 Fitted Fitted Line Line -20 Fitted Line Fitted Line -20 -20 of SMP UN Change Change of SMP Score UN Score & Education & Education Expenditure Expenditure Change Change of SMP UN Score of SMP UN Score & Education & Education Expenditure Expenditure -100 -100 0 0 100 100 200 200 300 300 -100 -100 0 0 100 100 200 200 300 300 Fitted Line Fitted Line Fitted Line Fitted Line Source: ChangeMoF SIKD SMP of Change UN (budget), Score of SMP UN Susenas Score & Education & Education (NER) Expenditure and MoEC (UN exams) Expenditure Change of SMP UN Change of SMP Score UN Score & Education & Education Expenditure Expenditure Fitted Line Fitted Line Fitted Line Fitted Line 168 Education Public Expenditure Review Guidelines Madagascar PER (2015) Overall comment: This example shows a method for judging the efficiency of a country’s public-education expenditures relative to its outcomes. It compares the expenditure-outcome relationships for one country (Madagascar) with those for other countries. The methodology is that of regression analyses against other countries and time series, with the results displayed in four-quadrant graphs. The public expenditure review has an extensive methodological annex (Annex 3) that explicates the models and assumptions underlying the results presented in this example. Although the Annex is too long to be included here, interested users can activate the link to the review to read the full Annex. Overall efficiency analyses are particularly helpful to benchmark the structural performance of the system in comparison to other countries (Box 5). The objective of this analysis is to establish a link between the level of public education expenditures and general education outcomes using information from a large set of countries of all income groups and regions. The methodology and motivation as well as limitations of the analysis in terms of the outcome indicators chosen are presented in Box 5. The structural analysis, which evaluates a long-run “Madagascar effect” based on estimation from an unbalanced panel of 144- 176 countries over the period 1980-2012, is presented in Annex 3 along with details on the methodology and regression results for both structural and current efficiency analyses. However, various limitations to the analysis impose caution when interpreting the results. In order to grasp the overall picture and make policy recommendations regarding the size of public expenditure, one cannot look at results of the efficiency analysis in isolation of analyses of inequalities in terms of financial burdens and outcomes. In addition, because education outcomes are usually affected with substantial delays by changes in the learning and teaching environment, it is possible that the efficiency analysis does not capture accurately the capacity of the system to transmit competencies and knowledge to its pupils. Box 5. Efficiency Analysis of Public Education Expenditure Overall/system efficiency analysis is often presented comparing efficiency score estimated by Data Envelopment Analysis methods (DEA) to situate countries relative to an Efficiency Frontier (see Herrera and Pang (2005) for a review of the literature). It is, however, difficult to interpret DEA efficiency scores for countries with low levels of spending. In particular, a country can obtain a very high efficiency score despite very poor education outcomes. Keeping sight of where the country stands in terms of both levels of input and output while assessing efficiency is thus very useful. The methodology presented here uses the same data as for DEA but efficiency is assessed visually in two dimensions rather than using a summary indicator. One dimension is obtained using deviations from expected public expenditure and the other using deviations from expected education outcome. Expected outcomes 169 and expected expenditures take account of the country’s income level (in PPP), its size, and effects of regional and income groupings for LDCs. The unexplained variation is used to determine relative efficiency. Evaluating “structural” efficiency. Structural efficiency relates to the long-term standing of the country relative to others in terms of public expenditure performance. It is important to measure to establish some benchmark that can be used to interpret current efficiency, especially because these results depend on characteristics of the Madagascar’s education system (such as the relative importance of the private sector). A measure of structural efficiency can be obtained using all time periods and countries for which data are available and exploiting the panel structure of the data. A random-effects regression separates the residual variance in two parts: one that is common to all countries and one that is country specific. The country specific residual variance or “random effect” can be used to measure the structural advantage/disadvantage of the country (the country’s typical deviation from expected outcomes). Evaluating changes in “current/conjectural” efficiency. Beyond looking at the general position of the country relative to others, it is important to gage how it has changed across two time periods. In this case, expectations are estimated for each time period averaging values over 2 or more years (dampening measurement errors) and running an Ordinary Least Square regression on the cross- section of countries for which data is available. Residuals are then calculated as the difference between observed and predicted values. The choice of outcomes indicators. Different types of outcomes can be considered depending on whether efficiency is assessed over a long time period (structural efficiency), or using averages calculated over short time periods (current efficiency). In general, outcomes that are most contemporaneously related to current expenditures, sufficiently dependent on the overall situation in the country, and sufficiently available and comparable across countries are best suited to evaluate changes over time. Herrera and Pang (2005) use primary and secondary enrollment (gross and net), completion rates (first and second level), years in school, and learning scores. Based on data availability, the gross primary and secondary combined enrollment ratios, primary completion rates, and the youth literacy rate (only available 2000 and 2009) are used here. The youth literacy rate is most relevant to the structural analysis as it is less contemporaneously related to expenditure than the other two but is only sparsely available. Enrollment rates are somewhat problematic for Madagascar given the inaccuracy of population estimates. Combining primary and secondary rates dampens (but does not solve) the problem. Learning outcomes would have been a good indicator in the structural analysis to capture the quality dimension but are not sufficiently available (number of years) or comparable across countries. Unfortunately, this was the case for all other internationally comparable indicators of quality. The choice of expenditure indicators. Government education expenditure data in percentage of GDP (or GNI) are widely available in the UNESCO Institute of Statistics’ database for the period 1980-2012 (covering 177-211 countries, depending on the year). The indicator is comparable across countries and goes through country clearance procedures. The indicator includes all public expenditures externally or internally financed. 170 Education Public Expenditure Review Guidelines The four-quadrants graphical illustrations. Results can be plotted into a four quadrant graph where (0,0) is the point where both education outcome and expenditures are at levels predicted by the model. Deviations from expected education outcomes are represented vertically (y-axis) and deviations from expected expenditure horizontally (x-axis). For most education outcome indicators, higher is better, so a position in the NW quadrant is most efficient and a position to the SE is least efficient. The SW quadrant includes countries which underperform in terms of education outcomes but also have relatively low expenditures. They are called underachievers. Those in the NE quadrants are overachievers. Why focus on government expenditure? Although outcomes are the combined result of public and private education, and all education expenditures are expected to have some effect on outcome, the present analysis only considers public expenditures. First, as discussed above, there are no reliable data on total expenditures in education and no data that are comparable across countries. Second, according to the evidence provided above, public education represents the bulk of education expenditures, and it is also closely connected to the private education system; third, the outcomes considered focus on primary and primary/secondary education, accounting for more than 80 percent of children under 14 years. Finally, and importantly, even if comparable data on private expenditure are available, the point here is to evaluate public expenditure needs on efficiency grounds. If a country can obtain better education outcome overall with less public expenditure because the private sector takes on a larger part of the burden, it is indeed an efficiency improvement in terms of public expenditure. Structurally, Madagascar places in the group of the most efficient countries in terms of enrollment but as an underachiever in terms of completion (Figure 13). In other words, considering the full variation of expenditures and outcomes over 1980-2012, Madagascar tends to spend slightly less on education (in percent of GDP) than its level of income, size and geographical location would predict while it achieves average or better than expected outcomes in terms of enrollment (primary and secondary combined) and youth literacy. However, the result is different when looking at completion rates: Madagascar achieves slightly lower completion rates than the levels of expenditure would predict. These results are not to be interpreted as reflecting the current situation but need to be taken into consideration when interpreting the results presented below. These results are not to be interpreted as reflecting the current situation but need to be taken into consideration when interpreting the results presented below. Overall, the system demonstrated some resilience to the severe cuts in public funding during the crisis and remains broadly efficient compared to other countries. The analysis shows a slight increase in efficiency during the period 2009-2012 compared to 2000-2008 in terms of completion rates (Figure 14). In other words, the decrease in completion was less than what could have been predicted given the simultaneous decreased in spending. However, it is likely that the delayed effect on completion explains part of this result. On the other hand, the analysis shows that although the country’s advantage in terms of GERs was reduced, it remains well above expectation. This may indicate a certain degree of resilience of the system, likely to be related to the rapid increased in private spending to compensate the cuts in public spending. In addition, it is important to note that this analysis does not fully take into account the potential loss in terms of education quality which would be measured through learning outcomes.a Overall, the above analysis shows the relatively good structural performance of the system, but hides important weaknesses, including in terms of learning outcomes. Madagascar has a CHANGE IN NATIONAL EXAM SC CHANGE IN JUNIOR HIGHSCH CHANGE IN JUNIOR HIGHSCH 20 Change of SMP NER & Education Expenditure Ch 0 Fitted Line 0 Fitt 0 171 -50 -50 -20 -100 0 100 200 300 0 500 1000 e) SMP UN score, total spending (2005-2009) -100 0 100 200 Change of SMP NER & Education Expenditure 60 Change of SMP NER & Non-Salary Education Expenditure 60 Fitted Line Fitted Line Change of SMP UN Score & Education Expenditure Fitted Line CHANGE IN NATIONAL EXAM SCORE IN SMP % CHANGE IN NATIONAL EXAM SCORE IN SMP % high potential in education as it 40has managed to achieve better or similar outcomes than comparable 40 countries with lower expenditures. So far, the crisis does not appear to have had a significant impact on efficiency if 2009-2012 averages are considered. This would seem to indicate that there some room e) SMP UN score, total spending (2005-2009) 20 f) SMP UN score, non-salary (2005-2009) to increase public expenditure on education without moving to a situation of overspending relative 20 60 60 to other countries. Important caveats include however (a) it is likely that the full impact of the recent years is not fully captured, especially on indicators which react with a delay (such as completion rates), FIGURE 13: ofEXPENDITURE PERFORMANCE OF PUBLIC and EDUCATION: CHANGE IN NATIONAL EXAM SCORE IN SMP % CHANGE IN NATIONAL EXAM SCORE IN SMP % 0 0 40 (b) the efficiency the system is so far analyzed through the lenses 40 of enrollments completion, PRE/POST ignoring therefore CRISIS changes in EFFICIENCY the equality of education. BASED Given ON theCOMPLETION dramatic decrease in education RATES learning *Predicted outcomes values over based on OLS the past regressions decade, on GDP/c, GDP/c^2it is probably with the group e ects forcase incomethat although and region (LDCs), Madagascar is GDP relatively population and size in PPP per capita for all regressions. 20 -20 20 -20 Source: World Bank Edstats based on UNESCO statistics for education variables, including expenditures, more efficient than others at getting children to school, it is performing particularly poorly in terms of IMF data for income variables -100 0 100 200 300 -100 ensuring that resources are translated into improved learning outcomes. 0 0 Figure Overall E 13: Expenditure ciency performance of Public Spending inChange of public of SMP Education UN Scoreeducation: & Educationpre/post-crisis Expenditure efficiency Changein Overall E ciency of Public Spending based Based onon completion Averages 2009-2012*rates Fitted Line Based on Averages 2000-2008*Fitted Lin -20 Indicator: Primary Overall Efficiency ofCompletion Ratesin Education based on averages 2009-2012* Public Spending -20 Indicator: Primary Completion Rates Indicator: Primary Completion Rates -100 0 100 200 300 -100 0 100 200 300 MOST EFFICIENT: OVERACHIEVERS: DEVIATIONS FROM EXPECTED EDUCATION OUTCOME DEVIATIONS FROM EXPECTED EDUCATION OUTCOME Primary completion rate higher than expected, Primary completion rate higher than expected, Change of SMP UN Score & Education Expenditure Gov. Exp./GDP lower than expected Change of SMP UN Score & Education Expenditure Gov. Exp./GDP higher than expected 60 MOST EFFICIENT: 40 Primary completion rate higher than expected, Fitted Line Fitted Line Gov. Exp./GDP lower than expected 40 20 20 3: EXPENDITURE PERFORMANCE OF PUBLIC EDUCATION: LDCs 0 SSA T CRISIS EFFICIENCY BASED ON COMPLETION RATES 0 -20 LICs -20 ed on OLS regressions on GDP/c, GDP/c^2 with group e ects for income and region (LDCs), population and size GDP in PPP per capita for all regressions. Madagascar dstats based on UNESCO statistics for education variables, including expenditures, IMF data for income variables -40 NCE OF PUBLIC EDUCATION: -40 UNDERACHIEVERS: LDCs LEAST EFFICIENT: UNDERACHIEVERS: SSA Primary completion rate lower than expected, ON COMPLETION RATES Primary completion rate lower than expected, Gov. Exp./GDP lower than expected LICs Primary completion rate lower than expected, Gov. Exp./GDP higher than expected Gov. Exp./GDP lower than expected -5 y of Public come Spending and region in Education (LDCs), population and size GDP in PPP per capita for all regressions. -4 -2 Overall 4 E ciency of Public Spending in Education Madagascar 0 2 g expenditures, es 2009-2012* IMF data for income variables Based on Averages 2000-2008* DEVIA DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES y Completion Rates * Predicted values based on OLS regressions on GDP/c, GDP/c^2 with group Indicator: Primary Completion effects Rates for income and region (LDCs), population and size GDP in PPP per capita for all regressions. Overall E ciency Data Sources: Bank Spending of Public World Edstats based in Education on UNESCO statistics for education variables, including expenditures, IMF CIENT: OVERACHIEVERS: DEVIATIONS FROM EXPECTED EDUCATION OUTCOME mpletion rate higher than expected, Based on income data for Averages 2000-2008* variables. Primary completion rate higher than expected, GDP lower than expected Gov. Exp./GDP higher than expected 60 Indicator: Primary Completion Rates MOST EFFICIENT: OVERACHIEVERS: Overall Efficiency of Public Spending in Education based on averages 2000-2008* Primary completion rate higher than expected, Gov. Exp./GDP lower than expected Primary completion rate h Gov. Exp./GDP higher than Indicator: Primary Completion Rates 40 VERACHIEVERS: DEVIATIONS FROM EXPECTED EDUCATION OUTCOME mary completion rate higher than expected, ov. Exp./GDP higher than expected 60 MOST EFFICIENT: OVERACHIEVERS: 20 FIGURE 14: EXPENDITURE PERFORMANCE 0 OF PUBLIC EDUCATION:PRE VS. POST 40 Primary completion rate higher than expected, Gov. Exp./GDP lower than expected Primary completion rate higher than expected, Gov. Exp./GDP higher than expected EFFICIENCY BASED ON GROSS ENROLLMENT -20 RATIOS PRIMARY/SECONDARY COM 20 *Predicted values based on OLS regressions on GDP/c, GDP/c^2 with group e ects for income and region (LDCs), population and size GDP in PPP per capita for all regressions. World Bank Edstats based on UNESCO statistics for education variables, including expenditures, IMF data for income variables Source: 0 -40 -20 UNDERACHIEVERS: LEAST EFFICIENT: HIEVERS: LEAST EFFICIENT: Primary completion rate lower than expected, Primary completion rate lower t ompletion rate lower than expected, Primary completion rate lower than expected, Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expe GDP lower than expected Gov. Exp./GDP higher than expected -40 Overall E ciency of Public Spending in Education Overall E ciency of Public Spending in -5 0 5 FFICIENT:-4 -2 0 2 Based on 4 Averages UNDERACHIEVERS: 2000 Primary completion –than rate lower 2008* expected, LEAST EFFICIENT: Primary completion rate lower than expected, Based on Averages 2009 – 2012* completion rate lower than expected, ./GDP higher than expected Indicator: Gross Primary & Secondary Enrollment Combined Rate** DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected DEVIATIONS FROMIndicator: GrossEXPENDITURES EXPECTED PUBLIC Primary & Secondary -5 0 5 10 DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES EXPECTED EDUCATION OUTCOME EXPECTED EDUCATION OUTCOME 60 MOST EFFICIENT: Primary completion rate higher than expected, MOST EFFICIENT: OVERACHIEVERS: Gov. Exp./GDP lower than expected Primary completion rate higher than expected, Primary completion rate higher than expected, 40 Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected 20 20 0 0 LICs ncome and region (LDCs), population and size GDP in PPP per capita for all regressions. Change of SMP NER & Education Expenditure Change Madagascar ng expenditures, IMF data for income variables Fitted Line Fitted L Overall E ciency of Public Spending in Education 172 Based Education on Averages Public 2000-2008* Expenditure Review Guidelines FIGURE 14: EXPENDITUREe) PERFORMANCE Indicator: Primary Completion Rates PUBLIC EDUCATION:PRE VS. POST CRI OF (2005-2009) SMP UN score, total spending OVERACHIEVERS: 60 ON GROSS ENROLLMENT RATIOS PRIMARY/SECONDARY 60 EFFICIENCY BASED COMBIN DEVIATIONS FROM EXPECTED EDUCATION OUTCOME Primary completion rate higher than expected, Gov. Exp./GDP higher than expected *Predicted 60 values based on OLS regressions on GDP/c, GDP/c^2 with group e ects for income and region (LDCs), population and size GDP in PPP per capita for all regressions. MOST EFFICIENT: OVERACHIEVERS: Bank Source: WorldPrimary Edstats completion based rate on UNESCO higher than expected, Primary expenditures, statistics for education variables, including IMF completion rate higher thandata for income variables expected, Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected CHANGE IN NATIONAL EXAM SCORE IN SMP % CHANGE IN NATIONAL EXAM SCORE IN SMP % 40 40 40 Figure 20 14: Expenditure performance of public education: pre- vs post-crisis Overall E ciency efficiency of Public based on Spending grossin Education enrollment Overall E ratios (primary/secondary combined) ciency of Public Spending in Educa Based on Averages 2000 – 2008* 0 Based on Averages 2009 – 2012* 20 20 Indicator: Overall Gross Primary & Secondary Enrollment Combined Rate** -20 Efficiency of Public Spending in Education based on averages 2000-2008* Indicator: Gross Primary & Secondary Enrollm Indicator: Gross Primary & Secondary Enrollment Combined Rate** -40 DEVIATIONS FROM EXPECTED EDUCATION OUTCOME DEVIATIONS FROM EXPECTED EDUCATION OUTCOME UNDERACHIEVERS: 0 LEAST EFFICIENT: 0 T EFFICIENT: Primary completion rate lower than expected, Primary completion rate lower than expected, 60 ary completion rate lower than expected, MOST EFFICIENT: Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected Exp./GDP higher than expected Primary completion rate higher than expected, MOST EFFICIENT: OVERACHIEVERS: Gov. Exp./GDP lower than expected Primary completion rate higher than expected, -5 0 5 Primary completion rate higher than expected, 10 40 Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected 20 -20 DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES -20 20 0 -100 0 100 200 3000 -100 -20 -20 Change of SMP UN Score & Education Expenditure Change of SM -40 Fitted Line -40 Fitted Line UNDERACHIEVERS: LEAST EFFICIENT: UNDERACHIEVERS: Primary completion rate lower than expected, Primary completion rate lower than expected, Primary completion rate lower than expected, Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected Gov. Exp./GDP lower than expected -5 0 -5 -10 -4 -2 0 ANCE OF PUBLIC EDUCATION:PRE VS. POST CRISIS LDCs DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES DEVIATIONS FROM EXPECTED LLMENT RATIOS PRIMARY/SECONDARY COMBINED * Predicted values based on OLS regressions on GDP/c, GDP/c^2 with group effects for SSA income and region (LDCs), population and size GDP in PPP per capita for all regressions. ncome and region (LDCs), population and size GDP in PPP per capita for all regressions. ** Gross enrollment rates for Madagascar: 2005-2008 average LICs ng expenditures, IMF data for income variables Madagascar EXPENDITURE PERFORMANCE OF PUBLIC EDUCATION: Data sources: World Bank Edstats based on UNESCO statistics for education variables, including expenditures, IMF data for income variables. LDCs SSA CRISIS EFFICIENCY BASED ON COMPLETION RATES Overall E ciency of Public Spending in Education LICs OLS regressions on GDP/c, GDP/c^2 with group e ects Based onfor income and Averages region 2009 (LDCs), population and size GDP in PPP per capita for all regressions. – 2012* Madagascar based on UNESCO statistics for education variables, including Overall Indicator: Efficiency Gross expenditures, of Public Primary IMF data for Spending & Secondary income variables in Education Enrollment based on Combined averages 2009-2012* Rate** Indicator: Gross Primary & Secondary Enrollment Combined Rate** DEVIATIONS FROM EXPECTED EDUCATION OUTCOME 60 Public Spending in Education MOST EFFICIENT: Primary completion rate higher than expected, ciency of Public Spending in Education Overall EOVERACHIEVERS: Primary completion rate higher than expected, 009-2012* FIGURE 1: PRESENT VALUE OF CHIKA VERSUS Based on Averages 2000-2008* HOLLOW FIGURE 1: PRE OVERACHIEVERS: Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected Primary completion rate higher than expected, 40 mpletion Rates Indicator: Primary Completion Rates Gov. Exp./GDP higher than expected CORE BLOCK RELATIVE TO DISCOUNT RATE 20 CORE BLOCK R Source: Jimenez and Patrinos, gure 4, p. 34 Source: Jimenez and Patrinos, g 0 OVERACHIEVERS: DEVIATIONS FROM EXPECTED EDUCATION OUTCOME n rate higher than expected, Primary completion rate higher than expected, er than expected Graphic Derivation of Crossover Discount Rate, Choice Between -20 60 Gov. Exp./GDP higher than expected MOST EFFICIENT: Graphic Derivation of Cr OVERACHIEVERS: Chika and Hollow Concrete Block Alternatives in Ethiopia Primary completion rate higher than expected, Gov. Exp./GDP lower than expected Chika and Primary Hollow completion rateConcr higher th Gov. Exp./GDP higher than expect -40 40 T EFFICIENT: UNDERACHIEVERS: LEAST EFFICIENT: ary completion rate lower than expected, Primary completion rate lower than expected, Primary completion rate lower than expected, Exp./GDP higher than expected Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected 250k 20 250k -10 -4 -2 0 2 4 0 DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES * Predicted values based on OLS regressions on GDP/c, GDP/c^2 -20 with group effects for 200k 200k PRESENT VALUE BVR PRESENT VALUE BVR income and region (LDCs), population and size GDP in PPP per capita for all regressions. Data sources: World Bank Edstats based on UNESCO statistics for-40 education variables, : including expenditures, LEAST EFFICIENT: IMF data for income variables. CROSSOVER UNDERACHIEVERS: LEAST EFFICIENT: DISCOUNT RATE Primary completion rate lower than expected, Primary completion rate lower than exp n rate lower than expected, Primary completion rate lower than expected, Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected er than expected Gov. Exp./GDP higher than expected 150k 150k -5 0 5 4 -2 0 2 4 DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES a The full set of results is presented in Annex 3. 100k 100k 10 15 20 25 10 DISCOUNT RATE % KA VERSUS HOLLOW Chika FIGURE 1: PRESENT VALUE OF CHIKA VERSUS HOLLOW Chika NT RATE HCB CORE BLOCK RELATIVE TO DISCOUNT RATE HCB Source: Jimenez and Patrinos, gure 4, p. 34 ween Graphic Derivation of Crossover Discount Rate, Choice Between EXPENDITURE PERFORMANCE OF Chika andPUBLIC Block Alternatives in Ethiopia VS. POST CRISIS EDUCATION:PRE Hollow Concrete 173 Example 23: Cost-benefit analysis Ethiopia: Cost-benefit analysis on a project component (2008) Overall comment: This is a good example of a cost-benefit analysis of alternative materials that can be used to construct classrooms. It includes an analysis of the rates of return to the savings associated with different options relative to the base case of chika, a building material made of mud and thatch, and a comparison of the present values for two options relative to their discount rates. Note that the example is not from a public expenditure review, but from a policy research working paper (Jimenez and Patrinos 2008) Tables 1-3 and figure 1 illustrate a cost-benefit analysis. They involve the choice of the construction material for building new classrooms in Ethiopia. Table 1 lays out the alternative materials considered, with their advantages and disadvantages. Table 2 calculates the relative costs per lifetime year for each option. Table 3 calculates the rates of return on the savings from alternative materials relative to the base case of chika (mud and thatch) under different assumptions. Figure 1 shows the present value of chika versus hollow core block (HCB), assuming half maintenance costs and plotted against the associated discount rate. On the basis of tables 2 and 3, HCB emerges as the cheapest material per year of its lifetime and with the highest rates of return on savings under different assumptions. However, figure 1 shows that there is a discount rate at which building chika schools is preferable to HCB. The crossover discount rate is 23 percent in the case highlighted. What is interesting here is how benefits are treated. The benefit is implicitly assumed to be school seats resulting from the construction. De facto, this benefit is assumed to be constant across the materials options. Since benefits are assumed to be constant regardless of the material selected, the analysis becomes one of relative costs of the materials under different assumptions. Only a few advantages or disadvantages displayed in table 1 are independent of the relative costs (initial and recurrent or maintenance costs), which are properly captured in table 2. These “independent” variables include the popularity of the different materials within the community, the comfort that the material offers students and teachers in the classroom and environmental degradation (de-forestation). These variables could have been converted to categorical variables, weighted or unweighted, to measure the relative benefits of the different materials. For example, chika and corrugated iron sheets were unpopular materials with parents and students. The analysis could have argued that using either of these materials to construct new classrooms would prompt parents to be less likely to enroll their children in schools that were built of these materials, thus reducing their school seat benefits. The analysis did not use this option. 174 Education Public Expenditure Review Guidelines Table 1: Relative advantages and disadvantages of alternative construction materials (Ethiopia) Material Advantages Disadvantages Maintenance Chika (mud and thatch) Cool in dry Does not resist termites High and hot weather Washes away in rain Cheap initial investment Depletes forest Not popular with community, students Concrete element Resists termites and rain Costly initial investment Low Stone Resist fire, termites, rain Costly initial investment Low Hollow concrete block Resist fire, termites, rain Need stable foundation, Low (HCB) Better insulation supervision Brick Comfortable and cool Many trees cut to burn High brick Weak in rainy season Corrugated iron sheets Cheap initial investment Susceptible to corrosion High Not heat resistant Not popular with community or students Source: Harry Patrinos Table 2: Cost comparison of school buildings, 1992 Ethiopia Birr* Capital Recurrent Lifetime* Cost/Year Material (Investment) (Maintenance) (Years) Lifetime Concrete element 213,000 2,130 40 7,455 Stone 189,284 1,893 40 6,625 Brick 170,400 5,000 30 10,680 Hollow Concrete Block 127,800 1,278 30 5,538 Chika 85,200 6,000 10 14,520 Corrugated Iron Sheet 31,950 5,000 10 8,195 * With full maintenance Source: Participatory Evaluation of EICMA (Educational Institutes Construction and Maintenance Agency). A report to Ministry of Education and SIDA. 1992. Source: Jimenez and Patrinos, table 4, p.31. Table 3 includes the results of sensitivity analyses. Thus, school buildings are not well maintained in Ethiopia, and the assumption of full maintenance is not realistic. For this reason, a sensitivity analysis considers the more realistic case that maintenance will be less than adequate. For simplicity’s sake, half of the required maintenance is assumed to take place. The choice of technology also depends on the availability of local materials. It may simply not be possible to transport certain materials to remote areas. For many rural areas there will be impossible transportation problems. In such cases 175 there is no alternative other than building with chika. In less severe cases there may be roads, but the cost of transporting materials may be prohibitive. To examine this problem, sensitivity analysis was carried assuming scenarios of cost escalation due to transportation difficulties using orders of magnitude of 10 to 30 percent. Table 3: Rates of return on savings due to selection of different school building materials versus chika (Ethiopia) Full Half Transportation Difficulties: Material Maintenance Maintenance Cost Escalation 10% 20% 30% Concrete element 6% 9% 7% 5% 2% Stone 8% 11% 9% 6% 3% Brick 5% 11% 9% 6% 2% Hollow Concrete Block 17% 23% 19% 13% 8% Chika (base camp) (base camp) (base) (base) (base) Source: Jimenez and Patrinos, table 5, p.32 Figure 1 shows that even in the case of hollow concrete block, there are instances when it is not a good investment. All of the alternatives to chika have a lower present value. However, the total undiscounted cost of using the different alternatives is different, and they have differing time profiles. Therefore, the alternative one chooses may depend on the opportunity cost of capital. Figure 1: Present value of chika versus In the cases examined here, the higher the hollow core block relative to discount rate opportunity cost of capital assumed the Graphic Derivation of Crossover Discout Rate, Choice Between greater the likelihood of choosing chika over Chika and Hollow Concrete Block Alternatives in Ethiopia the alternative. This is because with the chika model one is postponing investment. In other words, there is a discount rate at which chika becomes the preferred option, despite higher undiscounted total cost or lower present value of the alternatives. To illustrate this example, the case Present value (birr) of chika versus HCB is used. The present value of the two methods, assuming half maintenance costs, is plotted against the associated discount rate. The graph shows that there is a discount rate at which building chika schools is preferred to HCB. This is known as the crossover discount rate, which is 23 percent in the case highlighted. This is the same as the internal rate of return calculated Discount rate (percent) for the benefits (cost-savings) stream presented above for the case of half-maintenance, which is Source: Jimenez and Patrinos, figure 4, p.34 probably the more realistic scenario. 176 Education Public Expenditure Review Guidelines Example 24: Data envelopment analysis Overall comment: The Kenya and Democratic Republic of the Congo (DRC) reviews have examples of data envelopment analysis at the subnational level. The Kenya example is a simple analysis based on one input (per capita spending) and one output (net enrollment rate or percentage of exam passes), while the DRC analysis conducts an analysis based on composite input and output variables. The DRC analysis provides a brief methodological note and explains the difference between the input and output approach in such analysis. These two examples provide different ways of examining and presenting data envelopment analysis results, such as the efficiency frontier graph in the Kenya example and the efficiency score quadrants diagram in the DRC analysis. Kenya PER (2013) Data envelopment analysis (DEA) is a linear programming methodology (developed by Farrell’s (1957), which can be used to estimate efficiency as the distance from the observed input-output combinations to an efficiency frontier defined as the maximum attainable output for a given input level. Per capita spending in Kenya tends to be low for northern and coastal counties with higher pupil teacher ratio (PTR), implying a shortage of teachers in these regions. Counties in these areas also tend to have lower net enrolment rates. But several counties, such as Turkana and West Pokot, report remarkable performance in exams, placing them on the frontier of translating per capita spending into exam performance measured in terms of the percentage of children passing the Kenya Certificate of Primary Education (KCPE) (Figures 1 and 2). The per capita spending in secondary level shows the reverse: unit costs are higher in the more remote areas in the northern and coastal regions. This fact is explained by the much lower enrolment rates in these regions and lower pupil-teacher ratio. The counties in these regions also show the challenge of being efficient with the resources. These counties are much more likely to be well below the frontier in terms of exam performance. Counties such as Kiambu, Nairobi and Nyeri where enrolments and school performance are relatively high are the most cost effective at this level. The counties far from the frontier at secondary level such as Garissa, Kwale, Laikipia and Nyamira are not cost effective (Figures 3 and 4). 177 Figure 1: Per capita spending and NER-primary, 2009 Source: Ministry of Education Figure 2: Per capita spending and percentage of exam passes, primary, 2010 Source: Ministry of Education 178 Education Public Expenditure Review Guidelines Figure 3: Spending per capita and NER–secondary, 2009 Source: Ministry of Education Figure 4: Spending per capita and percentage of passers at secondary level, 2012 Source: Ministry of Education PER CAPITA SPENDING BASED ON ENROLLMENTS TSC 2011 179 Democratic Republic of Congo PER (2015) Efficiency measurement with DEA Data Envelopment Analysis (DEA) is based on the construction of an empirical non-parametric production frontier and the measurement of the efficiency through the distance between the observed data and the optimal value of these data given by the estimated frontier. In the current analysis, the production frontier approximates the maximum quality or access to education (the output) that could be achieved given different levels of educational resources (the inputs). The figure below illustrates the efficiency measurement with DEA in a hypothetical case of one input x that is used to produce one output y. Illustration of the efficiency measurement with DEA ILLUSTRATION OF THE EFFICIENCY MEASUREMENT WITH DEA y C r ntie Fro y0 B FRONTIER A A B C Θx0 x0 x The frontier gives maximum levels of the output that could be achieved given different quantities of the input used. In the DEA literature, observations are called Decision Making Units (DMUs). DMUs that are on the frontier are relatively efficient (for instance, DMU at the point C) while those below the frontier are relatively inefficient (for instance, DMU at the point A). The level of efficiency is given by the distance to the frontier. Let’s consider the DMU0 initially at the point A. This DMU uses x0 FIGURE 49: SUMMARY OF PROVINCIAL GROUPING units of the input in order to produce y0 units of the output. As already mentioned, DMU0 is not BASED relatively EFFICIENCY efficient. InSCORES BASED order to be efficient, thisONDMU DEA MODEL can reduce its input in the way that it projects on Source: Author’s estimations based on Ministry of Budget, EMIS, SECOPE, HBS 1-2-3, 2012 and PASEC the frontier at the point B. In other terms, in order to be efficient, this DMU can keep its output level unchanged but has to reduce its input to the optimal level. The optimal quantity of input is given by with. The higher is, the closer the DMU is to the frontier and the more efficient is the considered DMU. The value of is the efficiency measure. This approach is called input oriented DEA. There is an alternative to the input oriented DEA (the output oriented DEA) which is about how to get the frontier2.by increasing the output given the input used. While there are also several DEA models, the Kasaï Oriental 1. Katanga that we Sud-Kivu model Equateur use is the one developed by Charnes, Cooper, and Rhodes (1981). Bas Congo 180 Education Public Expenditure Review Guidelines Efficiency measurement of the DRC education system In this study, we use input oriented approach because we would like to focus on the use of resources in the DRC education system. One can notice the high heterogeneity in terms of access and quality of education across DRC provinces. Provinces that seem to perform well in terms of access do not necessarily do so in terms of the quality of education. For this reason, we perform two different efficiency analysis, one for the access to education and the other for the quality of education. Efficiency in the provision of education access Recall that the illustration done above is a hypothetical case with only one output and one inputa. In order to estimate a DEA model for the DRC, we need to choose inputs and outputs. In fact, DRC education system uses many inputs in order to provide the observed access to education. In addition there are several indicators of education access. More specifically, in the inputs side, we need to have proxies for educational infrastructure, equipment, human resources and public expendituresb. In the current efficiency analysis, we use the total number of schools and classrooms per student as proxies for infrastructure, the total number of equipment materials (chairs, tables and other types of equipment) as a proxy of school equipment, the number of teachers per student as a proxy for human resources and government total spending per student as a proxy for government expenditures. The outputs are: the gross enrollment rate, the gender parity ratio, the pass rate at grade 6 and the repetition rate. As already discussed, according to the administrative organization, the DRC includes eleven provinces and each province is divided into districts. Inside the administrative districts, primary and secondary schools are differently managed. For this reason, we distinguished between primary and secondary schools inside each district. We aggregated schools by levels of education and we considered districts as DMUs depending on the level of educationc. In other terms, a DMU represents either all primary schools or all secondary schools in a given district. For instance, in the district of Beni, we have two DMUs, one for primary schools and the other for secondary schools. This approach is advantageous because it allows comparing primary and secondary schools within the same district and across districts. We have one frontier for primary and secondary schools and this provides a proper comparison. Due to the lack of data in several districts, we only consider 82 DMUs and 42 districts. Efficiency in the provision of the quality of education A DEA model is estimated using PASEC 2013 survey data in order to measure the efficiency in the provision of quality education in the DRC primary education system. Some key variables necessary for this analysis are not available for secondary education. For this reason, we concentrate on primary education. Inputs include the number of teachers per student, the number of classrooms per student, teachers’ level of education, teachers’ monthly salary and schools’ equipment. Outputs are: the success rate, average score in the PASEC French and mathematics tests. We have a total 160 schools. The overall results of the DEA analysis at the provincial level shows that Katanga and Bas-Congo provinces are relatively efficient and Kinshasa, Bandundu and Nord-Kivu relatively inefficient in their resource use. Figure 49 presents the summary of DEA results in four quadrants (I, II, III and IV). The 181 first quadrant (I), refers to provinces with relatively high efficiency scores both in terms of access and quality related variables (learning outcomes). Quadrant (II), designates provinces with a relatively high efficiency score in quality. This group of provinces are basically in line with first quadrant (I) in terms of learning outcomes efficiency score but have lower efficiency scores in access outcomes compared to quadrant (I). Quadrant (III) shows provinces with relatively lower scores in both access and quality outcomes efficiency scores. The final quadrant (IV), shows provinces with relatively better efficiency scores in the access model (in line with the first quadrant (I) in this aspect) but lower efficiency scores in learning outcomes. Figure 49: Summary of provincial grouping based efficiency scores based on DEA model Source: Authors’ estimations based on Ministry of Budget, EMIS, SECOPE, HBS 1-2-3, 2012, and PASEC a DEA accommodates multi input and multi output technologies. The principle is the same when we have more than one input and more than one output but it is difficult to be graphically illustrated. In addition, it is important to highlight that the efficiency assessment is done in a relative terms and results could change when the sample changes. b The choice of inputs and outputs is supported by the literature. In fact, similar choice of inputs are done by authors in studies on the efficiency analysis in education. Other authors provide discussion about possible inputs and output for the education system. See for instance, Correa (1963), Burkhead (1967), Michaud (1981), Charnes, Cooper, and Rhodes (1981) and Worthington (2001). c It should have been more appropriate to consider educational provinces as DMUs instead of districts given the fact that districts are linked to the country’s administrative organization rather than the organization in the education system. However, we could not have performed this analysis with the only educational provinces because DEA requires a certain number of observations for the results to be reliable. 182 Education Public Expenditure Review Guidelines Example 25: Internal efficiency indicators Bangladesh PER (2010) Overall comment: An internal efficiency analysis based on survival, repetition, and completion rate indicators reveals inefficiency problems in the education system in Bangladesh. The education system in Bangladesh is characterized by persistent low quality and inefficiency. Only half of all children beginning primary and secondary education survive up to the final grade. At the secondary level only one in five actually passed the SSC examination (Table 3.2). Failure to complete secondary education is a growing phenomenon. Between 2000 and 2005, the proportion of the 16-25 age group with an incomplete secondary education as their highest educational attainment increased from 23 to 33 percent (Al-Samarrai 2007a). Table 3-2: Internal efficiency (%) of the education system (2005) Sector Repetition Rate Survival Rate Completion Rate M F T M F T M F T Primary 12 11 11 49 57 53 - - - Secondary (general) 8 8 8 43 40 41 23 17 20 Secondary (madrassa) - - - - 57 61 - 14 22 Higher secondary - - - - - - - 59 57 Source: (DPE 2006a; BANBEIS 2007). Notes: Primary statistics cover government, registered non-government and experimental schools. Statistics for madrassa education are for 2003. Figures for survival and completion rates are calculated using the reconstructed cohort method. UNDERACHIEVERS: LEAST EFFICIENT: UNDERACHIEVERS: LEAST EFFICIENT: DEVIA DEVIA Primary completion rate lower than expected, Primary completion rate lower than expected, Primary completion rate lower than expected, Primary completion rate lower than expected, Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected Gov. Exp./GDP lower than expected Gov. Exp./GDP higher than expected -5 0 -5 -10 -4 -2 0 2 4 DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES DEVIATIONS FROM EXPECTED PUBLIC EXPENDITURES 183 FIGURE 1: PRESENT VALUE OF CHIKA VERSUS HOLLOW FIGURE 1: PRESENT VALUE OF CHIKA VERSUS HOLLOW Example CORE BLOCK RELATIVE RATE of returns to education 26: Rate TO DISCOUNT CORE BLOCK RELATIVE TO DISCOUNT RATE Source: Jimenez and Patrinos, gure 4, p. 34 Source: Jimenez and Patrinos, gure 4, p. 34 Graphic Derivation of Crossover Discount Rate, Choice Between Graphic Derivation of Crossover Discount Rate, Choice Between Chika and Hollow Concrete Block Alternatives in Ethiopia Chika and Hollow Concrete Block Alternatives in Ethiopia Sri Lanka PER (2011) 250k 250k Overall comment: PRESENT VALUE BVR 200k PRESENT VALUE BVR 200k CROSSOVER CROSSOVER DISCOUNT RATE DISCOUNT RATE 150k This review presents considerable differences in the rates of return 150k to education by level of education. Interestingly, the impact of education on earnings is higher for females at all levels except for those who only completed primary education. 100k 100k 10 15 20 25 10 15 20 25 DISCOUNT RATE % DISCOUNT RATE % Chika Education attainment and earnings are positively related in Sri Lanka. Earnings rise continually as Chika HCB the education levels of individuals increase [Figure 2.1). A male worker HCB with primary education earns 16 percent more per month than a man with no schooling, and a female worker with primary education earns 8 percent per month more than a woman with no schooling. Among men and women who have completed basic education, a male worker earns 43 percent more than a man with no education, and a female worker earns 45 percent more than a woman who is not educated. The highest gains are recorded among men and women who have completed higher education. Overall, these findings clearly support the notion that men and women benefit from their investments in education. FIGURE 2.1: IMPACT OF EDUCATION ON EARNINGS, MALE AND FEMALE WORKERS, 2008 FIG Figure 2.1: Impact of education on earnings, male and female workers, 2008 500% 450% 459 400% FEMALE 372 350% 322 300% MALE 284 250% 200% 210 150% 151 118 100% 87 50% 43 45 0% 16 8 PRIMARY BASIC GCE O/L GCE A/L UNIVERSITY POSTGRADUATE EDUCATION EDUCATION QUALIFIED QUALIFIED GRADUATE COMPLETED COMPLETED Source: World Bank staff computations, derived from econometric estimates of earnings functions. FIGURE 2.2: PRIVATE RATES OF RETURN TO EDUCATION, MALE AND FEMALE WORKERS FIG Source: World Bank sta computations, derived from econometric estimates of earnings functions Sourc 25% 184 Education Public Expenditure Review Guidelines The rates of return to education are positive and substantial at the secondary and higher education levels. Among primary educated workers the returns to education are positive but relatively low, at 2 percent for men and 1 percent for women, respectively [Figure 2.2]. Workers who have completed basic education receive slightly better returns: 7 percent for men and 10 percent for women. However, among workers who are secondary educated or higher, returns to education are considerably greater. GCE O/L qualified men earn returns of 13 percent, while women enjoy even higher returns at 21 percent. Among GCE A/L qualified workers, men receive a return of 15 percent, while women receive a return of 18 percent. Among university graduates the returns to education for both men and women are 21 percent. At postgraduate level, the returns to education for men are 9 percent and for women 17 percent, respectively. Figure 2.2: Private rates of return to education, male and female workers FEMALE 21 21 21 18 17 15 MALE 13 10 9 7 2 1 Source: World Bank staff computations, derived from econometric estimates of earnings functions. This pattern of returns to education is consistent with the fact that the supply of primary and basic educated human capital is relatively high, so that returns to education at this level are small. At secondary education level and higher education level, however, the supply of educated labor is lower, and returns to education are high. In addition, workers may be using their educational certificates at secondary education and higher education levels to signal their quality, while employers may be using these certificates to screen potential employees for quality. Overall, the pattern of returns to education is consistent with economic theories of human capital and of signaling–screening in labor markets with asymmetric information. The higher returns enjoyed by women in comparison to men is likely to be due to selection effects, as fewer women participate in the labor market. Therefore, the women who do work are likely to be more capable than average, resulting in better productivity and greater returns to human capital. 185 Example 27: Analysis of inequity Costa Rica PER (2015) Overall comment: A simple analysis of learning outcomes, enrollment rates, and rates of preschool attendance reveals large regional differences and differences associated with socioeconomic status, indicating an inequity problem exists in Costa Rica’s education sector. In Costa Rica, urban and rural areas of the country present large disparities in enrollment in secondary and upper-secondary. While almost 60% of the 19 year old living in urban areas living are enrolled in school, less than 50% of those living in rural areas do so (Figure 21). Equally worrying, the disparities in post-secondary enrollment associated with household income have been dramatically increasing over time (Figure 22). FIGURE 21: AGE SPECIFIC ENROLLMENT FIGURE 22: POST SECONDA FIGURE 22: POST SECONDARY BY AND GEOGRAPHIC AREA ENROLLMENT ENROLLMENT BY INCOME BY INCOME Q QUINTILE Figure Source: World Bank Age 21: specific SSEIR team’s enrollment analysis of household surveys, by Figure 22: Post-secondary Source: World Bank SSEIR team’s analysis of household enrollment Source: World Bank SSEIR team’s analysis of household surveys, surveys, authors’ calculations based on ADePT software authors’ calculations based on ADePT software geographic area by income authors’ quintile calculations based on ADePT software 80 100 80 90 70 70 80 FIGURE 22: POST SECONDARY FIGURE 22: 60 POST SECONDARY ENROLLMENT BY INCOME 60 QUINTILE GROSS ATTENDANCE % GROSS ATTENDANCE % 70 FIGURE 22: POST50 SECONDARY UR ENROLLMENT BY INCOME QUINTILE BA FIGURE 22: POST SECONDARY 50 Source: World Bank SSEIR team’s analysis of household surveys, N 60 ENROLLMENT BY INCOME QUINTILE authors’ calculations based on ADePT software Q1 RU Source: World Bank SSEIR team’s analysis of household surveys, FIGURE 22: POST SECONDARY % ENROLLMENT INCOME BY80 QUINTILE Q2 RA authors’ calculations based on ADePT software 50 Source: World Bank SSEIR team’s analysis of household surveys, L 40 40 40 ENROLLMENT BY80 INCOME QUINTILE authors’ calculations based on ADePT software Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations based on ADePT software Q3 Q4 Source: World Bank SSEIR team’s analysis of household surveys, 30 30 80 70 Q5 30 authors’ calculations based on ADePT software 80 70 20 20 20 70 60 80 GROSS ATTENDANCE % PRIMARY SCHOOL 70 LOWER 60 UPPER 10 GROSS ATTENDANCE % AGES 6 11 SECONDARY SECONDARY 10 60 50 10 70 Q1 GROSS ATTENDANCE % 0 60 50 0 Q1 Q2 GROSS ATTENDANCE % 5 7 609 11 13 15 50 17 19 40 0 2007 2008 2009 2010 2011 Q2 Q1 2012 2013 2014 Q3 GROSS ATTENDANCE % 50 40 2007 2008 2009 2010 2011 2012 2013 201 AGES Q1 Q2 Q3 Q4 50 40 30 Q1 Q1 Q2 Q2 Q3 Q3 Q4 Q4 Q5 Q5 40 30 20 Q2 Q3 Q4 Q5 40 30 Source: World Bank SSEIR team’s analysis of household 20 Q3 Bank Source: World Q4 Q5 analysis of SSEIR team’s 30 surveys, author’s calculations based on ADePT software household surveys, author’s calculations based 20 10 Q4 Q5 30 on ADePT software 20 10 Q5 20 10 0 10 0 2007 2008 2009 2010 2011 2012 2013 2014 0 2007 2008 2009 2010 2011 2012 2013 2014 10 0 2007 2008 2009 2010 2011 2012 2013 2014 0 2007 2008 2009 2010 2011 2012 2013 2014 2007 2008 2009 2010 2011 2012 2013 2014 186 Education Public Expenditure Review Guidelines Moreover, the latest results from the Third Regional Comparative and Explanatory Study (TERCE) 2013 examination of Latin American students by UNESCO show that Costa Rica is the only country in the region where students in both 3rd and 6th grades performed worse in TERCE 2013 than they did in Second Regional Comparative and Explanatory Study (SERCE) 2006 in reading and in mathematics. Consistent with the findings for enrollment, there are large differences in learning outcomes associated with socioeconomic status. Data from the SERCE 2006 show that one standard deviation (sd) increase in a socioeconomic conditions index are associated with a 0.3sd increase in test scores (Figure 24). Figure 24: SERCE standardized score by socioeconomic status index Source: SERCE (2006) The drivers behind the large inequalities include gaps in the access to education services that start early in life. There is well established evidence that high quality Early Child Development programs act as equalizers, since they can reduce the effect of household socioeconomic differences on the child cognitive and non-cognitive development and, therefore, the ability to perform well in school. In Costa Rica, there is a large differential in daycare (age 1-3) and preschool (age 3-5) attendance by household income quintile. Only 3% of the children in the first quintile attend daycare, as opposed to 16% among those in the fifth quintile (Figure 27). Similarly, for children in preschool age, 27% of those in the first income quintile attend preschool, while attendance goes up to 49% for those in the top income quintile (Figure 28). 187 FIGURE 27: DAYCARE ATTENDANCE FIGURE 28: PRESC HOUSEHOLD INCOME QUINTILE, 2014 HOUSEHOLD INC Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using Source: World Bank SSEIR team’s analysis of h 27: Daycare Figurestandardized attendance ADePT software by household income quintile, 2014 (Education Module) standardized ADePT software (Education Mo 18 60 16 16 50 14 %3 5 YEARS OLD %1 3 YEARS OLD 12 40 10 30 8 29 27 7 25 6 20 4 10 3 4 2 2 0 0 ENDANCE FIGURE 28: PRESCHOOL ATTENDANCE Q1 Q2 Q3 Q4 Q5 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software Q1 Q2 Q3 INTILE, 2014 HOUSEHOLD INCOME QUINTILE, 2014 (Education Module) ors’ calculations using Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using Figure 28: Preschool standardized attendance ADePT software by household income quintile, 2014 (Education Module) 60 50 49 %3 5 YEARS OLD 40 30 34 29 27 25 20 10 0 Q1 Q2 Q3 Q4 Q5 Source: World Bank SSEIR team’s analysis of household surveys, authors’ calculations using standardized ADePT software (Education Module) 188 Education Public Expenditure Review Guidelines Example 28: Analysis of per capita financing (PCF) Tajikistan PER (2013) Overall comment: The report provides historical background on the reform of the country's per capita financing arrangements and the indicators used to measure their implementation. It also addresses factors that could undermine the sustainability of the reform. History of Per Capita Financing (PCF) in Tajikistan Since 2010, all general secondary schools receive their budgets according to a per capita financing (PCF) formula. In Tajikistan, PCF for general education was first piloted in five cities and rayons in 2005, with budget allocation primarily based on the number of students.a After successful implementation during the pilot phase, the PCF reform was gradually expanded and adopted nationwide— to all schools in 68 rayons—by 2010. The norms (unit costs) per student and per school have increased considerably over years (Figure 8), and the formula has been revised to better reflect the different needs of schools such as geographic location, type of schools, and multi- language requirements. Figure 8: Minimum standards (norms) for per pupil and per school, 2008-2013 Primary General Secondary Source: Ministry of Education 189 The introduction of PCF has considerably increased the role of schools and the central government in general education financing but it reduced the role of rayons. Schools prepare their budgets according to the norms and formula set by the Ministries of Finance and Education of the republican government. They then submit them to rayons, which in turn submit an aggregated education budget to their respective oblasts and finally to the Ministry of Finance of the republican government.b There remains a room for budget negotiations between the different levels of the government, and rayons may also allocate a significant portion of their local budget to education. However, the introduction of PCF has considerably reduced the role rayons in general education financing because the larger part of it is determined by the centrally defined formula. On the other hand, because the formula- based budget allocation cannot fully accommodate various factors and needs of schools under different conditions, the PCF model gives rayons the right to reallocate up to five percent of the formula-based allocations from schools with a surplus to those with a deficit. At the school level, schools may flexibly determine their budgets as long as they meet educational requirements and norms for wages. Subnational Variation in Education Expenditure Overall, the introduction of PCF has led to a more equitable distribution and efficient use of resources, increased budgetary autonomy at the school level, and greater transparency and community involvement in school planning and budgeting. Between 2007/2008 and 2011/2012, the nationwide student-teacher ratio increased from 17.0 to 18.0 on average. Over the same period, the share of personnel costs in local government education expenditures fell from 86.8 percent to 75.9 percent, freeing resources to improve the quality of education. In terms of equity, in 2010 only 82 percent of all general secondary schools had an approved budget in line with the formula-based budget, but in 2011 95 percent of schools had a PCF-compliant budget.c The switch to PCF provides greater budgetary autonomy for schools and gives them responsibility to manage resources effectively and efficiently, and work closely with communities to plan school development, formulate budgets and monitor expenditures. There has been regular monitoring and evaluation of reform implementation by the Ministry of Education.d The success of PCF in general education has encouraged the MOE to extend it to other levels of education. A comparison of student-teacher ratios between oblasts and rayons illustrates wide regional variations in terms of efficiency outcomes. Between 2007/2008 and 2011/2012, the overall STR improved from 17.0 to 18.0 ranging from 5.7 in Roshtkalla rayon in GBAO to 25.0 in the city of Dushanbe in 2007/2008 and from 6.5 to 25.5, respectively, in 2011/2012. Between the maximum and minimum, there are wide variations within each oblast (though all data are not shown in the graphs), but not all oblasts and rayons observed the similar change (Figure 9). As a result, per student spending and the percentage distribution of recurrent spending widely vary between oblasts, but without a clear correlation (Figure 10). Part of these variations can be explained by conditions—elevation and population density—but there also are considerable variations between rayons that have similar conditions (Figure 11). The remaining variations may be explained by other factors such as multi-language requirements, type of school, or catchment areas, but also by the inefficiencies in school network, managing class size, and recruitment of teachers and non-teaching staff. The Ministry of Education has observed irregularities in school financing, for example, payment of full-year salaries for seasonal workers (e.g., heating staff in winter) and electricity bills in schools without electricity. In-depth analysis of school-level data for each rayon is necessary to identify potential areas for further efficiency gains. 190 Education Public Expenditure Review Guidelines Figure 10: Per student unit cost and wage bill vary among oblasts, but without a clear correlation NIT COST AND WAGE BILL VARY HOUT A CLEAR CORRELATION Capital Other Recurrent Maintenance/repairs S Percentage Distribution of Recurrent Spending 2011 Heat & Utilities Food 100 Personnel 90 80 Personnel 70 Food 60 50 Maintenance/repairs 40 Heat & Utilities 30 Other Recurrent 20 10 0 GBAO Dushanhe Khation Sughd Total RRS GBAO Source: Tajikistan BOOST v0.4 government expenditure database 191 Figure 11: STRs correlate with elevation and population density, but there are intra-oblast variations among rayons with similar conditions student-teacher ratio student-teacher ratio Source: MOE, EMIS for STRs; Wikipedia (http://en.wikipedia.org/wiki/Provinces_of_Tajikistan) for the population density; and Avaneya, Vahram, 2012, Improving the system of school financing in the Republic of Tajikistan on the basis of normative formula funding, final, for elevation. The successful implementation of the PCF reform is undermined by distorted incentives for schools reducing their wage bills. General education budget for each rayon is approved by the central government according to the PCF formula. However, if some rayons fail to raise revenues as estimated based on their fiscal capacity, education budget could be reduced. However, local governments continue to finance “protected” budget items (e.g.), whereas other (non-wage) expenditures are the first to be reduced. This means that schools, that have rationalized staffing and increased their non-wage budget for quality improvement, get penalized, whilst those that have kept many teachers are rewarded. The gap between the approved budget according to the formula and the actual execution has compromised the successful implementation of the PCF reform. a Yovon, Kulob, and Vahdat rayons, and the cities of Khorog and Khojand. b See World Bank, 2012, Tajikistan Public Expenditure Review Interim Report, Box 2 (p. 17) on the local government budget preparation process. c Being in compliance was defined as having a budget that was 95 percent or greater as calculated using the PCF formula. This cut-off is based on the regulation that districts are entitled to reallocate up to 5 percent of the district-level budget, while ensuring that no school receives less than 95 percent of the budget as calculated by the PCF formula. d The PCF reform has been supported under the Fast Track Initiative grants and the Ministry has engaged consultants to monitor and evaluate the implementation and produced a number of reports analyzing the outcomes and aiming to improve the mechanism. 192 Education Public Expenditure Review Guidelines Mauritania PER (2016) Overall comment: The Mauritania example reports disparities in per student resources and flags several critical regional- distribution issues related to equity. The resources allocated to the regional offices are regressive. Poor and rural regions such as Guidimagha, Brakna, Gorgol, and Assaba, show a per student expenditure considerably below the national average (MRO 810 – 3 US$). At the same time, the number of students per teacher is high in these regions, between 45 and 64 (Figure 3.13). These data underline several critical regional distributional issues: (i) resources are not distributed based on need, as determined by the number of students, the conditions of the schools, and the poverty rate; (ii) the government struggles to hire and retain teachers in remote areas; and (iii) learning opportunities are more limited for poor students FIGURE 3.13: BUDGET PER STUDENT, PUPIL/TEACHER RATIO, in remote areas due to lower spending on school inputs and fewer teachers in the classroom. FIGURE 3.13: BUDGET PER STUDENT, PUPIL/TEACHER RATIO, AND THE POVERTY RATE BY REGION, 2015 AND Figure THE 3.13: POVERTY Budget RATE per student, pupil/teacher Source: Ministry of Education 2015 BY REGION, 2015 ratio, and the poverty rate by region, 2015 Source: Ministry of Education 2015 2,500 80 2,500 80 70 2,000 70 2,005 60 2,005 2,000 1,830 60 1,500 1,830 50 1,500 50 40 1,197 40 1,197 1,000 30 1,000 901 901 904 904 905 905 30 600 655 685 685 753 753 20 500 499 602 20 500 10 10 0 0 0 Guidimagha Chargui Brakna Gorgol Assaba Gharbi Nouadhibou Nouakchott Trarza Tagant Adrar Zemmour 0 Guidimagha Chargui Brakna Gorgol Assaba Gharbi Nouadhibou Nouakchott Trarza Tagant Adrar Zemmour Hodh Hodh Hodh Tiris Hodh Tiris Per Pupil Spending Per Spending Pupil Ratio (no. of students, right axis) Pupil/Teacher Ratio Pupil/Teacher Rate (in Poverty Rate Poverty (in %, right axis) Source: Ministry of Education 2015 193 Example 29: Analysis of cash transfer programs Overall comment: The Costa Rica and Indonesia public expenditure reviews examine the effectiveness of demand- oriented interventions, such as conditional cash transfer programs. The Costa Rica analysis finds poor targeting in the country’s program. The Indonesia example assesses the impact of a cash transfer program that targets poor students and challenges related to the program’s design, targeting, monitoring, and evaluation. The detailed Indonesia report also includes policy recommendations for addressing these challenges. Costa Rica PER (2015) The diagnostic, targeting and design of subsidies in secondary education can be improved. There are large socioeconomic differences both in the transition probability from lower secondary to upper secondary and from secondary to tertiary. Liquidity constraints can partly explain this difference. Evidence from the 2014 ENAHO shows that subsidies directed to students are poorly targeted (Figure 30). For instance only 29% of the 15-19 year old belonging to the first quintile of the household income distribution receives Avancemos, a conditional cash transfer (CCT) aimed at reducing dropout in secondary education among the poorest. On the contrary, among those in the fourth quintile of the income distribution, 8% report receiving this monetary transfer. A recent evaluation (Hidalgo and Romero, 2013) shows that this CCT has a positive impact on dropouts and enrollments. In other words, the CCT helps students stay in school and helps those dropping out getting back to school. Although some students remain in education by other factors, between 10% and 16% of the students do so solely Figure 30: Takeup of transfer and subsidies by because of the CCT and would, otherwise household income quintile, 2014 abandon their studies. Likewise, re-entering the educational system do so for various Avancesmo reasons, but even higher percentage (77% School Meal or more) did so because of the transfer Transportation were given. Although the sample in the case of reintegration is much smaller than FONABE in the case of desertion, it cannot ignore the positive impact of the transfer is much higher in reintegration. Therefore, a better targeting of subsidies and transfers can significantly weaken the liquidity constraints that affect poorest households. Source: World Bank SSEIR team’s, authors’ calculations using Household Surveys: ENAHO 2014 194 Education Public Expenditure Review Guidelines Indonesia PER (2012) The Bantuan Siswa Miskin (Cash Transfer for Poor Students, BSM) program provides cash transfers from central education agencies directly to students or schools. Both Kementerian Pendidikan dan Kebudayaan (Ministry of Education and Culture, Kemdikbud) and Kementerian Agama (Ministry of Religious Affairs, Kemenag) have BSM transfers providing cash payments once enrollment, attendance and other criteria have been verified. The amount of the transfers provided rises with the level of education, from Rp 360,000 for primary school to approximately Rp 1.2 million (per year) for a university student. Table 2: BSM at a glance Official Program Name Bantuan Siswa Miskin (BSM) Program type Cash transfer for poor students Program Type and inaugural year (start/usage year) Permanent, tax-financed, 2008 Coverage National (100% provinces, 100% districts) Number of beneficiaries (2010) 5.9 million Official value of benefit Rp 360,00-1,200,000 depending on level of schooling Public expenditure (2010) Rp 3,607 billion (US$ 397 million) Administrative cost per recipient (2010) Rp 15,608 (US$ 1.56) Percent of poor 6-18 years old covered (year) 3.2% (2009) Key policy and execution agencies Kemdikbud, Kemenag Key implementation agencies (role) Kemenag, Kemdikbud (verification of beneficiary lists), Education service providers (targeting and eligibility, fund distribution) Support operations partners (role) PT Pos/appointed bank (fund distribution) Local Government participation Targeting, verification, socialization, monitoring and Evaluation Sources and Notes: Kemenkeu, Susenas 2009, program manuals and World Bank staff calculations. BSM spending has risen rapidly and it now ranks as the third-largest household-based transfer (by central government expenditure). BSM spending has increased in line with the rapid expansion in target beneficiaries. In 2010, Rp 3.6 trillion (around US$ 397 million) was spent on the program, equivalent to 4 percent of central government education expenditures. BSM accounts for 10 percent of all central government resources devoted to household-based social assistance (SA). Around half of BSM spending goes to primary and junior secondary school students, with the remaining going to senior high school students and university students at similar portion. 195 Table 3: BSM expenditure summary, 2008-2010 2008 2009 2010 Total BSM (Constant, 2009 prices, Rp bn) 1,238 2,562 3,607 Primary school level (SD) 274 875 1,077 Junior secondary school level (SMP) 318 786 890 Senior secondary school level (SMA) 600 718 778 University level 46 183 863 Analytical Series: Total BSM (Constant 2009 prices, Rp bn) 1,343 5,856 3,339 Total BSM (US$, Rp million) 127 247 397 Share of central education expenditures (%) 2.2 3.0 4.0 Share of combined Kemdikbud and Kemenag (%) 2.1 3.0 4.1 Share of central government SA spending (%) 3.9 10.0 13.7 Analytical Series: Target number of beneficiaries (Million) 3.0 4.6 5.9 Average benefit level per beneficiary (Rp) 597,709 555,338 606,912 Sources and Notes: Kemdikbud and Kemenag records from Directorates General, Kemenkeu BPS, and World Bank staff calculations. Nominal expenditures have been deflated using the GDP deflator. In total, there are 10 BSM initiatives, each with its own manual, fund flow structure, and implementing procedures. There is little coordination between initiatives, even those located in the same institution. The BSM program is national in scope but reaches very few students overall and does a poor job of identifying poor students. In 2009 program coverage (through the senior secondary level) was still small at 2.3 percent of all 6 to 18 year olds in Indonesia. Overall coverage of enrolled children is 3 percent. However, students from the poorest 40 percent of households account for approximately half of all BSM scholarships (and half of all rupiah distributed through the BSM program) while the middle- class and richer households in the top 60 percent capture an equal 50 percent of all BSM scholarships. In other words, a rupiah from the BSM program is equally likely to end up the hands of a poor student as in the hands of a non-poor student. BSM does not target those who are unfamiliar with the school system and its administrators. BSM initiatives typically identify potential scholarship recipients by soliciting nominations from schools and school committees. Students nominated must have already achieved consistent attendance and demonstrated “good behavior” confirmed by the principal. Recently enrolled students or prospective new entrants have very little chance of being selected; likewise, those who have not made themselves known to the principal are unlikely to be selected. BSM does not effectively address difficult and costly transition periods - between elementary and junior secondary, and again between junior and senior secondary - when the overwhelming majority of dropouts occur. Verification procedures plus slow rates of disbursement together mean that recipients typically receive a BSM transfer for their first schooling year only after their second schooling year has already started. 196 Education Public Expenditure Review Guidelines Support operations like monitoring and evaluation are partly delegated to schools, which are also the main agency in charge of delivering BSM funds to beneficiaries. This has led to weakness in non-benefit program operations. Within the 10 different BSM initiatives, very few budgeted funds are spent on socialization, monitoring and evaluation, and complaints, appeals, or grievances. BSM could be a valuable program for poor households and for Indonesia generally if it were better targeted, better socialized, and benefit packages were revised to correspond to the risks that poor students face. Poor households are not translating primary school enrolments into frequent success in higher levels of education. Education costs, especially for senior secondary, are rising in real terms. Poor households in particular are facing the biggest increases in real education expenditures (at all schooling levels). By providing benefits adequate for meeting the real costs of education precisely when those higher costs arrive, the BSM program could serve an important risk-mitigation function and inducement to higher education for poor households. Unfortunately the BSM program does not have any of these characteristics and its effectiveness is consequently lowered. 197 Example 30: Benefit incidence analysis Democratic Republic of Congo PER (2015) Overall comment: The DRC example conducts a classic average (or simple) benefit incidence analysis of the distribution of education public expenditures across different levels of education. It provides an example of a common situation where public spending in lower levels of education (in this case, primary and secondary) is progressive, while higher-education spending is significantly pro-rich and regressive. The example also provides a subnational- (provincial-) level benefit incidence analysis revealing variance in public-resource distribution across quintiles by province. A BIA (Benefit Incidence Analysis) using the concentration curve to evaluate the targeting of government subsidies shows that the distribution of public expenditure in primary and secondary education is relatively more biased towards the poor than the distribution of income. Figure 34 includes the consumption concentration curve, which is a proxy for the general wealth and income inequality across quintiles. Compared to the consumption concentration curve, the expenditures on primary, lower secondary and upper secondary education are relatively more equitable than the general wealth distribution. This can be observed from the concentration curves in primary, lower and upper secondary lying above that of the consumption curve, indicating that spending in these levels tend to be more equitable. Therefore, while public spending in primary and secondary education levels is not pro-poor per se, this is somewhat mitigated by the fact that the distribution of spending promotes greater equality than the general observed income inequality. In contrast, higher education is significantly not pro-poor and is regressive. Given that the richest quintile receives the most benefit from public spending- the distribution of public spending in higher education is in fact worse than the general wealth inequality. Even though public spending on education is less regressive than is income distribution, such spending nonetheless benefits the rich much more than then poor. Moreover, the inequality becomes higher at higher levels of education. Figure 34(a) presents the BIA without adjusting for demographic factors and Figure 34(b) presents the analysis taking into consideration demographic factors. In general, public spending on education is pro-poor if the concentration curve for the particular level of education is above the 45-degree line. Figure 34(a) shows that the concentration curve for primary education spending is just above the line of perfect equity, while that of post- primary education spending is entirely below the line of equity. However, after adjusting the spending data in each quintile for variations in number of children by quintile, spending in all levels of education fall below the perfect equity line, including at the primary level. This suggests that public spending in education in the DRC favors the richer households at all levels of education since the poorest quintile receives lower shares of public spending. 198 Education Public Expenditure Review Guidelines FIGURE FIGURE 34:34: LORENZ LORENZ CURVE CURVE FOR FOR HOUSEHOLD HOUSEHOLD CONSUMPTION CONSUMPTION Line Line of of Perfect Perfect Equality Equality EXPENDITURE EXPENDITURE AND Figure 34: ANDPUBLIC Lorenz PUBLIC curve SPENDING for SPENDING household ONONEDUCATION EDUCATION consumption BY BY LEVEL expenditure LEVEL and public Primary Primary Upper Secondary Upper Secondary Source: Source: Authors’ Authors’ spending on education by level estimations estimations basedbased on Ministry of Budget, of Budget, on Ministry EMIS, SECOPE, EMIS, SECOPE, HBS 1-2-3, and 1-2-3, and HBS 20122012 Consumption Lower Consumption Secondary Lower Secondary Higher Higher Education Education Concentration Concentration Curve Curve Concentration Concentration Curve Curve 1 1 1 1 .8 .8 .8 .8 .6 .6 .6 .6 .4 .4 .4 .4 .2 .2 .2 .2 LORENZ CURVE FOR HOUSEHOLD CONSUMPTION 0 0 0 0 Line of Perfect Equality RVE FOR E AND HOUSEHOLD PUBLIC CONSUMPTION SPENDING ON EDUCATION BY LEVEL 0 0 .2 .2 .4 .4 .6 .6 .8 .8 1 Line of Perfect 1 0 0Equality .2 Primary .2 .4 .4 Upper Secondary .6 .6 .8 .8 1 1 USEHOLD BLIC CONSUMPTION SPENDING ON EDUCATION BY LEVEL d on Ministry of Budget, EMIS, SECOPE, and HBS 1-2-3, 2012 Line of Perfect Equality Primary Upper Secondary Consumption Lower Secondary G ON EDUCATION BY LEVEL , SECOPE, and HBS 1-2-3, 2012 Primary Upper Secondary Consumption Lower Secondary Higher Education Concentration Curve Sources: Author’s estimationsConsumption Concentration Curve Higher Education based on Ministry of Budget, EMIS, SECOPE, and HBS 1-2-3, 2012 Lower Secondary Concentration Curve Higher Education 1 Concentration Curve 1 The provincial .8 level BIA analysis reveals that public resource distribution across quintiles varies to 1 some degree by province where Kasai Orientale appears to be the most equitable province while .8 Equateur is.6the least equitable. Figure 35 depicts the overall distribution of public funds by quintile .8 .6 and province. .4 For example, about 27 percent of total public funding benefits the richest quintile in Kasai-Orientale while the lowest quintile receives about 17 percent (3 percent below their population .6 .4 corresponding figures for Equateur are 39 percent and 10 percent, respectively. share). The .2 .4 .2 FIGURE 35: 0 PROVINCIAL LEVEL BENEFITS INCIDENCE ANALYSIS OF .4.2 .6 .8 0 1 0 .2 .4 .6 .8 1 PUBLIC Provincial levelON EXPENDITURE Figure 35: EDUCATION benefits ALL LEVELS incidence analysis OF of public EDUCATION expenditure on 0 .8 1 0 education - all levels of education Source: .2 Authors’ estimations based .4 .6 EMIS, SECOPE, on Ministry of Budget, .8and HBS 1-2-3, 2012 1 0 .2 .4 .6 .8 1 100% 90% 27% 27% 31% 33% 33% 33% 33% 34% 37% 38% 39% 39% 80% Q5 70% Q4 Q1 23% Q2 60% 30% 20% 26% 22% 23% 20% 27% 24% 19% 19% 18% Q3 50% Q3 40% 16% Q2 17% 18% 16% 14% Q4 30% 23% 20% 14% 17% 12% 19% 18% Q1 Q5 18% 20% 18% 15% 14% 18% 17% 14% 12% 12% 14% 10% 16% 11% 17% 14% 13% 10% 12% 14% 13% 11% 10% 11% 10% 0% 4% C du le ta le a go vu le DR vu a a r n as eu em ng ta a du en Ki Ki on nt sh i en at ta - Or d- i n rie rd an -C Kn u Ba Ka id Su Eq No s i-O M cc Ba i-O sa Ka sa Ka Source: Authors’ estimations based on Ministry of Budget, EMIS, SECOPE, and HBS 1-2-3, 2012 199 Example 31: Analysis of private spending by income quintile Madagascar PER (2015) Overall comment: After Madagascar’s 2009 military coup, the Government faced a severe budget crisis. Where they could, families tried to fill public expenditure gaps in the education sector to keep their schools operating. Given this context, this PER prioritized an analysis of household contributions to education. It assessed the impacts of the shift to more private financing of education on households in different income quintiles and on enrollment outcomes. The analysis relied on two household surveys (2005 and 2010) and new survey data collection Distribution of household expenditure by income levels and regions In the recent years, private spending on education has generally increased, with some direct impact on enrollment. Since the crisis, a greater number of households with a child in Grade 2 has had to pay enrollment fees, PTA contributions and monthly school fees (Table 20). The share of households paying enrollment fees has increased the most, by 7 percentage points, against 4 percentage points for the share paying PTA contributions and one percent for those paying school fees. In addition to be more frequent, the amount paid in school fees have generally increased. Average household per pupil spending increased from Ar 6,561 to Ar 8,277, representing a 26 percent increase over three years.a This has had direct repercussions in terms of enrollment. Indeed, financial problems are the first reason given by household to explain dropout, and this has only worsened since the beginning of the crisis in 2009. Table 20: Share of households in the southern districts that paid enrollment fees, PTA contributions and monthly school fees, 2009 and 2012 (sample selection) Percent 2009 2012 2009-12 Gap (% Points) Households having paid enrollment fees 34.6 41.5 6.9 Households having paid PTA contributions 36.5 40.9 4.4 Share of parents having paid late 71.4 85.2 13.8 Parents exempt from PTA contributions 20.0 14.5 -5.5 Households having paid monthly school fees 12.5 13.8 1.3 Source: South Survey in the Districts of Amboasary and Betioky, 2009 and 2012. 200 Education Public Expenditure Review Guidelines The share of household budget spent on education increased more for the richest. To interpret this data, it is important to note that, in Madagascar, the poverty level is situated around the average level of consumption in the fourth quintile (and even closer to the high end of the fourth quintile in 2010). The distribution in 2005 was U-shaped, with the poorest spending a higher share of their income than households in the second, third and fourth quintiles. The situation was more clearly progressive in 2010, with shares of income increasing for the richer households, but this came with an increase in household expenditures on education for all but the bottom two quintiles, with all quintiles paying a higher share of their budget in 2010.b Looking at average education expenses per child in school, there was a slightly higher cost in the third and fourth quintiles in the bottom two and richest quintiles (Figure 26). Figure 26: Household direct education expenditure, 2005 and 2012 FIGURE 26: HOUSEHOLD DIRECT EDUCATION EXPENDITURE, 2005 AND 2012 Source: EPM 2005 and ENSMOD 2012 5.0 300 4.5 250 THOUSAND CONSTANT 2013 AR. 4.0 3.5 200 3.0 2.5 150 2.0 100 1.5 1.0 50 0.5 0.0 0 Lowest Quintile Second Quintile Third Quintile Fourth Quintile Richest Budget,2005 of Budget % of 2005 Average Expenditure Per Child in School, 2005 of Budget, % of Budget, 2012 2005 Average Expenditure Per Child in School, 2012 Average Expenditure Per Child in School, 2005 Source: EPM 2005 and ENSMOD 2012. Average Expenditure Per Child in School, 2005 The cost per child in school differs significantly by region, and is inversely related to poverty rates. Correlation coefficients with contemporaneous poverty are negative and significant (-0.70 in 2005 and -0.90 in 2010), after adjusting for regional price variations. The strong correlation results hide some important variance across regions, however, especially when looking at changes in costs between 2005 and 2010 (Figure 27). Changes across time show that all regions experienced an increase in cost in real terms, except for Ihorombe. Regions with the highest poverty rate in 2005 also experienced some of the highest increases in expenditure per child in school (Sofia and Atsimo Andrefana, in particular).c 201 FIGURE 27: PRIMARY ATTENDANCE RATE AND THE SHARE OF FIGURE 27: PRIMARY HOUSEHOLD BUDGETATTENDANCE ON EDUCATION RATEBYAND SHARE OF THE 2012 REGION, ENSOMD 2012 BUDGET ON EDUCATION BY REGION, 2012 HOUSEHOLD Source: Figure 27: Primary attendance rate and the share of household budget on Source: ENSOMD 2012 education by region, 2012 ANALAMANGA 90 RATE SAVAANALAMANGA 80 90 RATE SAVA ATTENDANCE 70 80 60 ATTENDANCE 70 2012 50 60 IHOROMBE 2012 40 50 10 YEARS IHOROMBE ANOSY 30 40 PRIMARY 6YEARS ANOSY 20 30 PRIMARY 10 20 6 10 0 10 00 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3 EDUCATION EXPENSE IN PRIMARY AS SHARE OF BUDGET, 2012 Source: ENSOMD 2012. EDUCATION EXPENSE IN PRIMARY AS SHARE OF BUDGET, 2012 The share of primary education in the household budget is weakly related to primary attendance rates, indicating that the private cost of public education can be higher for families living in regions where attendance is low. According to 2012 Household Survey data, the share of education in household expenditure increases with primary attendance, but the correlation coefficient is weak (0.34). Figure 28 shows differences across regions. The regions of Analamanga, Diana and 27: FIGURE Sava PRIMARY ATTENDANCE stand the highest RATE attendance rate with AND greater THE share SHAREexpenditure OF of household on FIGURE HOUSEHOLD 27: PRIMARY BUDGET ATTENDANCE ON EDUCATION RATE BY AND REGION,THE SHARE 2012 OF education, whereas the region of Anosy has a high household expenditure with less attendance HOUSEHOLD primary Source: rate. 2012 BUDGET ON EDUCATION BY REGION, 2012 ENSOMD Source: ENSOMD 2012 ANALAMANGA 90 RATE SAVAANALAMANGA 90 Cost per child in school by region, ranked by poverty Figure 28: 80 level, 2005 and 2010 RATE SAVA ATTENDANCE 70 80 60 ATTENDANCE 70 2012 50 60 IHOROMBE 2012 40 50 10 YEARS IHOROMBE ANOSY 30 40 PRIMARY 6YEARS ANOSY 20 30 PRIMARY 10 20 6 10 0 10 00 0.5 1 1.5 2 2.5 3 0 0.5 1 1.5 2 2.5 3 EDUCATION EXPENSE IN PRIMARY AS SHARE OF BUDGET, 2012 EDUCATION EXPENSE IN PRIMARY AS SHARE OF BUDGET, 2012 Cost Per Child in School, 2005 Poverty Ratio, 2005 (Secondary Axis) Cost Per Child in School, 2010 Poverty Ratio, 2010 Note: Cost per child is adjusted to reflect prices in the capital. Data source: Household Surveys 2005 and 2010. 202 Education Public Expenditure Review Guidelines The share of education expenditure in the household budget is also negatively correlated with poverty levels.d Correlation coefficients are negative and significant in both years (-0.51 in 2005, and -0.59 in 2010), although confidence intervals do not allow for establishing whether the relationship is indeed stronger in 2010. Figure 29 also shows that the strong correlation result hides significant variance across regions with a similar pattern as noted above. Relationship between out-of-pocket and public expenditures on education Households are financing an increasing share of the total costs of education (Box 9). In order to refine the analysis on how education is financed in Madagascar, it is important to compare public and household expenditure. Over 2006-2008, current public expenditure contributed, on average, to 73 percent of expenses of one child enrolled at school, whereas household expenditure represented, on average, 27 percent (Figure 30). In some regions (Vatovavy-Fitovinany, Melaky, Atsimo Atsinanana, Androy and Atsimo Andrefana), the share of public expenditure per child in school reached more than 80 percent over this period. However, from 2009 to 2013, the share of current public expenditure was, on average, 59 percent of total spending per child enrolled. This reflects a substantial increase in the share of the costs per child enrolled financed by household. The increase was higher in Atsimo Andrefana, Menabe and Vatovavy Fitovinany, three regions that are particularly vulnerable. Figure 29: Burden of education expenditures by region, ranked by poverty levels, 2005 and 2010 Burden of educ exp 2005 Poverty ratio, 2005 (secondary sxis) Burden of educ exp. 2010 Poverty ratio, 2010 Note: Burden=share of education expenditure in household budget averaged by region. Data source: EPM. 203 Box 9. Parents’ Contribution to School Financing: Results From a Field Study The results of the survey carried out for this study (non-nationally representative) show that household expenditures averaged 18,000 MGA for a child in primary school, ranging from 27,000 MGA in Analamanga to 7,300 MGA in Melaky. Parental contributions to school expenditures averaged more than 21,000MGA in the regions of Analamanga and Atsinanana. A survey conducted by MEN in 2013-2014 in 30 schools of the Antananarivo Renivohitra CISCO showed comparable results, included an average contribution of parents to the school budget of 18,410 MGA, varying from 5,000 MGA to 28,000 MGA. In 30 surveyed schools, the collected resources reached a total of 128 million MGA, an average budget of 5.3 million MGA per school. Parental contribution for a child enrolled Parental contribution to school primary school inPARENTAL (EPP) CONTRIBUTION expenditures : distribution by area TO SCHOOL EXPENDITURES: DISTRIBUTION BY AREA IN MGA/PARENT/YEAR –(in MGA/parent/year) Source: Survey conducted during the study breakdown of the use of school budget realized through parental contributions Urban Urban 7 Rural Rural Total Total 10.2 25k 23,850 20,800 20k 18,375 16,400 18,395 18,620 17,245 15k 15,475 Source : Survey conducted during the study 10k 27,089 5k 15,938 17,810 19,510 21,235 30k Source : Survey conducted during the study 0 Parental MGA contributions 2011-2012 are used to pay community essentially 2013-2014 2012-2013 2014-2015 teachers and the school keeper, purchase supplies, pay for repairs, and finance report cards and school sports. With the contribution of parents, the surveyed schools provide a monthly subsidy to subsidized and non-subsidized community teachers. The amounts are almost equivalent, 63,500 MGA/ month for the non-subsidized community teachers and 61,150 MGA/month for subsidized community teachers. Nearly 72 percent of parents consider that the level of the subvention is more than they can afford (52.6 percent) or is rather high (19.2 percent). Moreover, parents’ satisfaction with the performance of teachers differs according to their status. The highest satisfaction rates go to the subsidized community teachers (92 percent) surpassing the satisfaction rates of the civil servants (less than 86 percent), and largely surpassing the satisfaction rates for the performance of non-subsidized community teachers (37 percent). 204 Education Public Expenditure Review Guidelines Breakdown of the use of school budget realized through parental contributions 1.6% 0.5% 9.3% Remuneration of Subsidized FRAM Teacher Remuneration of Subsidized FRAM Teacher 7.4% 38.0% Remuneration of Non-Subsidized FRAM Teacher Reumeration of Non-Subsidized FRAM Teacher Guardian Salary 10.2% Guardian School Salary Operations School Operations Maintenance, Rehabilitation Facilities, and Furniture Maintenance, Rehabilitation Buildings (Classroom, Latrines, Housing) Facilities, and Furniture Buildings (Classroom, Latrines, Housing) Other Other 32.9% Source: Source : Survey Survey conducted conducted during during the the study study FIGURE 30: AVERAGE HOUSEHOLD EXPENSE PER CHILD COMPARED TO Figure PER 30. PUPIL Average household GOVERNMENT expense EDUCATION per child compared EXPENDITURE, to per 2005 AND 2010pupil government education expenditure, 2005 and 2010 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% ALAOTRA MANGORO AMORON’ | MANIA ANALAMANGA ANALANJIROFO ANDROY ANOSY ATSIMO ANDREFANA ATSIMO ATSINANANA ATSINANANA BETSIBOKA BOENY BONGOLAVA DIANA HAUTE MATSIATRA IHOROMBE ITASY MELAKY MENABE SAVA SOFIA VAKINANKARATRA VATOVAVY FITOVINANY Average Average ofexpenditure per child 2009-2013 Current in school, 2005 Expenditure on Education per Child in School Average AverageExpenditure per of 2006-2008 Child in current School, 2005 expenditure on education per child in school 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 10% 0% ALAOTRA MANGORO AMORON’ | MANIA ANALAMANGA ANALANJIROFO ANDROY ANOSY ATSIMO ANDREFANA ATSIMO ATSINANANA ATSINANANA BETSIBOKA BOENY BONGOLAVA DIANA HAUTE MATSIATRA IHOROMBE ITASY MELAKY MENABE SAVA SOFIA VAKINANKARATRA VATOVAVY FITOVINANY 205 Average of 2009-2013 Current Expenditure on Education per Child in School Average Expenditure per Child in School, 2005 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% ALAOTRA MANGORO AMORON’ | MANIA ANALAMANGA ANALANJIROFO ANDROY ANOSY ATSIMO ANDREFANA ATSIMO ATSINANANA ATSINANANA BETSIBOKA BOENY BONGOLAVA DIANA HAUTE MATSIATRA IHOROMBE ITASY MELAKY MENABE SAVA SOFIA VAKINANKARATRA VATOVAVY FITOVINANY Average expenditure per child in school, 2010 Average of 2009-2013 Current Expenditure on Education per Child in School Average of 2009-2013 current expenditure on education per child in school Average Expenditure per Child in School, 2010 Note: The expenditure data is averaged over 2006-2007 for salaries by region and 2006 and 2008 for other current expenditure. Data sources: EPM 2005 and 2010, MFB. a Annex B explains the difference between the two groups: Sample Selection and Tracer Study. b A lower budget share in the poorest quintile would not have led to a positive interpretation, however, as it would have likely meant that the poorest students do not go to school altogether, which is not desirable. c The correlation coefficient between the change in cost and poverty rates are, however, not significant. d The burden was also measured as the average household expenditure per child in school relative to per capita expenditures in the region, giving similar results (corr -0.55 in 2005 and -0.59 in 2010) NOTES 207 1. The Systems Approach for Better Education Results (SABER) is a World Bank initiative that produces comparative data and knowledge on education policies and institutions, with the aim of helping countries systematically strengthen their education systems. SABER evaluates the quality of education policies against evidence-based global standards, using new diagnostic tools and detailed policy data collected for the initiative. These key questions are aligned with the SABER School Finance Conceptual Framework discussed in "What Matters Most for School Finance: A Framework Paper" (2013), in the SABER Working Paper Series. A more detailed explanation of the SABER School Finance instrument is available at http://saber.worldbank.org/index.cfm?indx=8&pd=3&sub=0. 2. See Freinkman and Skhirtladze (2015) for a detailed review of 75 public expenditure reviews that cover multiple sectors. 3. "Fiscal space" is defined as "room in a government´s budget that allows it to provide resources for a desired purpose without jeopardizing the sustainability of its financial position or the stability of the economy” (IMF 2005). The education sector can create fiscal space by cutting inputs to the sector (e.g., reducing the size of the teaching force) or by obtaining those inputs at a cheaper cost (e.g., introducing procurement reforms to produce large savings in textbook costs). 4. World Bank staff may consult budget information through the Education Global Practice website. 5. SABER’s education management information system assesses data quality as one of the policy areas. Its four policy levers include (i) methodological soundness, (ii) accuracy and reliability, (iii) integrity, and (iv) periodicity and timeliness. For details about the assessment methodology, see http://saber. worldbank.org/index.cfm?indx=8&pd=2&sub=0. 6. UNESCO's International Standard Classification of Education (ISCED) is used to categorize and report cross-nationally comparable education statistics. UNESCO's Institute of Statistics uses these categories to report data on education finance, such as a government's per capita expenditure by ISCED-based levels of education. ISCED 2011 is the latest version and is available at: http://www.uis.unesco.org/ Education/Documents/isced-2011-en.pdf. 7. Fiscal imbalance is a mismatch in the revenue powers and expenditure responsibilities of a government. Horizontal fiscal imbalance occurs when subnational governments are able to raise either more or less funds from their tax base than they need to cover the cost of providing services. Equalization transfers can help to mitigate horizontal imbalances. Vertical fiscal imbalance describes the variance between a central government’s revenue and expenditures against those of regional governments. It is a structural issue and thus needs to be corrected by reassignment of revenue and expenditure responsibilities among different levels of the government. 8. In contract theory and economics, information asymmetry deals with the study of decisions in transactions where one party has more or better information than the other. This creates an imbalance of power in transactions, which can sometimes cause the transactions to go awry. 9. An externality is a side effect or consequence of an activity that affects other parties without this being reflected in the cost of the goods or services involved, such as the pollination of surrounding crops by bees kept for honey, or a factory's pollution of the air of nearby communities. 208 Education Public Expenditure Review Guidelines 10. Rent-seeking is the use of the resources of a company, an organization, or an individual to obtain economic gain from others without reciprocating any benefits to society through wealth creation. An example of rent-seeking is when a company lobbies the government for loan subsidies, grants, or tariff protection. The process of lobbying government is a political process. Economists call such lobbying “rent seeking” because the objective is to secure economic rents that are higher than the normal profits obtainable by competing in the economic marketplace. 11. Elite capture is a process whereby resources transfers designated for the benefit of the larger population are usurped by a few individuals of superior status–be it economic, political, educational, ethnic, or otherwise. 12. For detailed definitions of functional and economic classifications of expense, see International Monetary Fund (2014). 13. For a discussion on textbook management, see Read (2015). 14. See Chapter 7 in World Bank (2004). 15. Education improvements in Rwanda in the 2000s, for example, have been linked to better funding arrangements as a result of public financial management PFM improvements, a sector-wide approach program (SWAP), and decentralization policies (ODI 2009). Similar improvements were seen in Cambodia, where a program that supported PFM improvements across a number of sectors, including education, found lower repeat rates for students in select provinces (Wescott 2008). 16. See PEFA (2016a) and PEFA (2016b). 17. These international targets on education financing are included as reference points in UNESCO (2015b). 18. The Department for International Development is held to value-for-money standards by its oversight body, the Independent Commission for Aid Impact (ICAI). The Commission operates independently of Government, and reports directly to the International Development Committee of the UK Parliament. 19. See Barnett, C. et al. (2010) and Department for International Development (2011) on value for money. 20. See, for example, Coelli, et al. (2005) and Herrera and Pang (2005). 21. See Psacharopoulos and Woodhall (1985) and UNESCO et al. (2014), Chapter 2. 22. See Psacharopoulos and Woodhall (1985), Chapter 5. 23. Other relevant policies and programs range from inclusive education programs, to special training programs for teachers in disadvantaged schools, to education programs targeting specific groups of students. 209 24. See Birdsall, Ibrahim, and Gupta (2004) for public policies that can increase demand for education. 25. As pointed out in Technical Note 4, in 2013 Transparency International published a comprehensive analysis of sources of corruption by level of education. See: https://www.transparency. org/whatwedo/publication/global_corruption_report_education. 26. Data collection and analytical instruments are available at: http://saber.worldbank.org/index. cfm?indx=8&pd=3&sub=4. 27. See Jensen and Wolde (2016) for details. 28. SDG 4 aims to “ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.” (The 2030 Agenda for Sustainable Development and the Sustainable Development Goals, https://sustainabledevelopment.un.org/sdg4). 29. For information on 2001 survey, see Kattan and Burnett (2004); for information on 2005 survey, see Kattan (2006). 30. An introductory primer on PETS/QSDS is Reinikka and Smith (2004). Other resources are Reinikka and Svensson (2002); and Dehn, Reinikka, and Svensson (2003). 31. In Sub-Saharan Africa, annual losses of textbooks have been found to be as high as 65 percent in some countries (Read 2015). 32. These committees consist of three government representatives and three democratically elected representatives of the program's beneficiaries. 33. The World Bank developed this tool in partnership with the African Economic Research Consortium and the African Development Bank, and launched it in 2013. 34. For more on MAMS, see www.worldbank.org/mams; and Lofgren, Cicowiez, and Diaz-Bonilla (2013). 35. The J-PAL is a network of 136 affiliated professors from more than 40 universities. The methodology and research results can be found at: https://www.povertyactionlab.org/policy-lessons/education/ increasing-test-score-performance. 36. For example, see Glewwe et al. (2011). 37. A cost-efficiency analysis is typically not conducted in public expenditure reviews, but is often conducted in project appraisal documents during a project preparation. 38. For instance, Alonso and Sanchez (2011) compare different funding models. REFERENCES 211 Allen, Richard, Richard Hemming, and Barry Potter. 2013. “Challenges of Reforming Budgetary Institutions in Developing Countries.” In Public Financial Management and Its Emerging Architecture, edited by Marco Cangiano, Michale Lazare, and Teresa R. Curristine, 411. Washington, DC: International Monetary Fund. Alonso, Juan Diego, and Alonso Sanchez. 2011. Reforming Education Finance in Transition Countries: Six Case Studies in Per Capita Financing System. A World Bank Study. Washington, DC: World Bank. https:// openknowledge.worldbank.org/handle/10986/2332. Barnett, Chris, Julian Barr, Angela Christie, Belinda Duff, and Shaun Hext, 2010. Measuring the Impact and Value for Money of Governance & Conflict Programmes, Final Report. Itad Ltd. https://assets.publishing. service.gov.uk/media/57a08b1eed915d3cfd000b44/60797_ITAD-VFM-Report-Dec10.pdf. Birdsall, Nancy, Amina J. Ibrahim, and Geeta Rao Gupta. 2004. “Achieving the Millennium Development Goal of Universal Primary Education.” United Nations Millennium Project Task Force on Education and Gender Equality. http://www.eldis.org/go/home&id=14619&type=Document#.WO-czDvyu01. Chiche, Mailan. 2009. Sector Budget Support in Practice: Case Study of the Education Sector in Rwanda. London: Overseas Development Institute; Oxford: Mokoro. https://www.odi.org/sites/odi.org.uk/ files/odi-assets/publications-opinion-files/5580.pdf. Coelli, Timothy. J., Dodla Sai Prasada Rao, Christopher J. O’Donnell, and George Edward Battese. 2005. An Introduction to Efficiency and Productivity Analysis, Springer Science & Business Media. 2nd ed. New York: Springer. http://facweb.knowlton.ohio-state.edu/pviton/courses/crp394/coelli_Intro_effic.pdf. Dehn, Jan, Ritva Reinikka, and Jakob Svensson. 2003. “Survey Tools for Assessing Performance in Service Delivery,” in The Impact of Economic Policies on Poverty and Income Distribution: Evaluation Techniques and Tools, edited by L. A. Pereira da Silva and F. Bourguignon, Washington, DC: World Bank; New York: Oxford University Press. Demery, Lionel. 2000. “Benefit Incidence: A Practitioner’s Guide”. Working paper 35117, World Bank, Washington, DC. http://documents.worldbank.org/curated/en/574221468135940764/Benefit- incidence-a-practitioners-guide. --------. 2003. “Analyzing the Incidence of Public Spending.” In The Impact of Economic Policies on Poverty and Income Distribution (Evaluation Techniques and Tools), edited by F. Bourguignon and L. Pereira da Silva, p. 41. Washington, DC: The World Bank; New York: Oxford University Press. http://documents. worldbank.org/curated/en/973661468739790876/pdf/multi0page.pdf. Department for International Development. 2011. DFID’s Approach to Value for Money (VfM). https://www. gov.uk/government/uploads/system/uploads/attachment_data/file/67479/DFID-approach-value- money.pdf. Freinkman, Lev M. and Sophiko Skhirtladze. 2015. Taking the Stock of PERs: Emerging Patterns and Examples of Good Practice, Washington, DC: World Bank. http://pubdocs.worldbank.org/ en/561951490728666169/Taking-the-Stock-of-PERs-Emerging-Patterns-and-Examples-of-Good- Practice-Final-June-2015. 212 Education Public Expenditure Review Guidelines Glewwe, Paul W., Eric A. Hanushek, Sarah D. Humpage, and Renato Ravina. 2011. “School Resources and Educational Outcomes in Developing Countries: A Review of the Literature From 1990 To 2010.” Working Paper 17554, National Bureau of Economic Research, Cambridge, MA. http://www.nber. org/papers/w17554.pdf. Glewwe, Paul W., Michael Kremer, and Sylvie Moulin. 2009. “Many Children Left Behind? Textbooks and Test Scores in Kenya.” American Economic Journal: Applied Economics 1 (1): 112–35. http://pubs. aeaweb.org/doi/pdfplus/10.1257/app.1.1.112. Herrera, Santiago and Gaobo Pang. 2005. “Efficiency of Public Spending in Developing Countries: An Efficiency Frontier Approach.” Policy Research Working Paper 3645, World Bank, Washington, DC. http://siteresources.worldbank.org/INTQFA/Resources/ EfficiencyofPublicSpendinginDevelopingCountries_MAY05.pdf. Inter-Agency Network for Education in Emergencies. 2010. Reference Guide on External Education Financing. New York: INEE. http://toolkit.ineesite.org/toolkit/INEEcms/uploads/1003/INEE_ Reference_Guide-Ext_Educ_Financing_EN.pdf. IMF (International Monetary Fund). 2005. “Back to Basics—Fiscal Space: What It Is and How to Get It.” Finance and Development, June Vol 42 (2). http://www.imf.org/external/pubs/ft/fandd/2005/06/basics.htm. --------. 2014. Government Finance Statistics Manual 2014. Washington, DC: International Monetary Fund. https://www.imf.org/external/Pubs/FT/GFS/Manual/2014/gfsfinal.pdf. Jensen, Leif, and Hirut Wolde. 2016. Strengthening Public Expenditure Analysis: Application of BOOST Standard Tables and Selected Tools and Techniques. BOOST Guidance Note Series. Washington, DC: World Bank. http://wbi.worldbank.org/boost/tools-resources/boost/guidance-notes. Jimenez, Emmanuel, and Harry Anthony Patrinos. 2008. “Can Cost-Benefit Analysis Guide Education Policy in Developing Countries?” Policy Research Working Paper 4568, Washington, DC: World Bank. https://openknowledge.worldbank.org/bitstream/handle/10986/6544/wps4568. pdf?sequence=1&isAllowed=y. Kattan, Raja Bentaouet. 2006. “Implementation of Free Basic Education Policy.” Education Working Paper Series 7, Human Development Network, Education, World Bank, Washington, DC. http:// siteresources.worldbank.org/EDUCATION/Resources/EDWP_User_Fees.pdf. Kattan, Raja Bentaouet, and Nicholas Burnett. 2004. User Fees in Primary Education. A World Bank Study. Washington, DC: World Bank. http://siteresources.worldbank.org/EDUCATION/ Resources/278200-1099079877269/547664-1099079993288/EFAcase_userfees.pdf. Lewis, Maureen A., and Gunilla Pettersson. 2009. “Governance in Education: Raising Performance in the Sector.” World Bank, Washington, DC. http://siteresources.worldbank.org/EXTHDOFFICE/ Resources/5485726-1239047988859/GAC_in_education_master_12_April_2009.pdf. 213 Lockheed, Marlaine E., and Eric Hanushek. 1988. “Improving Educational Efficiency in Developing Countries: What Do We Know?” Compare 18 (1): 21-38. http://www.tandfonline.com/doi/ pdf/10.1080/0305792880180103?needAccess=true. Lofgren, Hans, Martin Cicowiez, and Carolina Diaz-Bonilla. 2013. “MAMS–A Computable General Equilibrium Model for Developing Country Strategy Analysis.” in Handbook of Computable General Equilibrium Modeling, Vol. 1A, edited by Peter B. Dixon and Dale W. Jorgenson, 159–276. North Holland: Elsevier B.V. Mincer, Jacob. 1974. Schooling, Experience, and Earnings. Cambridge, MA: NBER. Mingat, Alain, Jee-Peng Tan, and Shobhana Sosale. 2003. Tools for Education: Policy Analysis. Washington, DC: World Bank. Montenegro, Claudio E., and Harry Anthony Patrinos. 2014. Comparable Estimates of Returns to Schooling Around the World. Policy Research Working Paper 7020, World Bank, Washington, DC. http:// documents.worldbank.org/curated/en/830831468147839247/pdf/WPS7020.pdf. Narayan, Amber. 2014. “Who Benefits from Education Sector Subsidies?” A PowerPoint presentation prepared for “Economics of Education”, Education Staff Development Program, Washington, DC, April 1-2, 2015. OECD (Organisation for Economic Co-operation and Development). Education at a Glance. 1998-2016. http://www.oecd-ilibrary.org/education/education-at-a-glance_19991487. PEFA. 2016a. PEFA Assessment Field Guide. Vol. 2 of PEFA 2016 Handbook., Washington DC: PEFA Secretariat. https://pefa.org/sites/default/files/16_08_30%20Fieldguide_8.pdf. --------. 2016b. PEFA: Framework for Assessing Public Financial Management. Washington DC: PEFA Secretariat. https://pefa.org/sites/default/files/PEFA%20Framework_English.pdf. Psacharopoulos, George. 1973. Returns to Education: An International Comparison. Amsterdam: Elsevier. --------. 1995. The Profitability of Investment in Education: Concepts and Methods. Washington, DC: World Bank. Psacharopoulos, George, and Harry Anthony Patrinos. 2004. “Returns to Investment in Education: A Further Update.” Education Economics 12 (2): 111–134. http://siteresources.worldbank.org/ INTDEBTDEPT/Resources/468980-1170954447788/3430000-1273248341332/20100426_16.pdf. Psacharopoulos, George, and Maureen Woodhall. 1985. Education for Development: An Analysis of Investment Choices. New York: Oxford University Press. http://documents.worldbank.org/curated/ en/477701468137718173/pdf/multi-page.pdf. Psacharopoulos, George, and Richard Layard. 1979. “Human Capital and Earnings: British Evidence and a Critique.” Reviews of Economic Studies 46 (3): 485–503. 214 Education Public Expenditure Review Guidelines Read, Tony. 2015. Where Have All the Textbooks Gone? Toward Sustainable Provision of Teaching and Learning Materials in Sub-Saharan Africa. Directions in Development—Human Development. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/22123. Reinikka, Ritva, and Jakob Svensson. 2002. Assessing Frontline Service Delivery. A World Bank Study. Washington, DC: World Bank. ----------. 2004. “Local Capture: Evidence from a Central Government Transfer Program in Uganda” Quarterly Journal of Economics 119 (2): 679–705. https://academic.oup.com/qje/article- abstract/119/2/679/1894528/Local-Capture-Evidence-from-a-Central-Government?redirectedFrom =fulltext. Reinikka, Ritva, and Nathanael Smith. 2004. Public Expenditure Tracking Surveys in Education: Peru, Uganda and Zambia. Ethics and Corruption in Education. Paris: UNESCO International Institute for Educational Planning. http://www.iiep.unesco.org/sites/default/files/public-expend-track-surv.pdf. http://etico. iiep.unesco.org/uploads/media/136267e.pdf. Salmi, Jamil, and Roberta M. Basset. 2012. “Education for All: The Equity Challenge in Tertiary Education.” Paper presented at Salzburg Global Seminars, Austria, October 2–7. http://www. salzburgglobal.org/fileadmin/user_upload/Documents/2010-2019/2012/495/Session_Document_ OpportunitiesforAll_495.pdf. Shand, David. 2000. “Promoting Allocative Efficiency (Strategic Prioritization),” prepared for the World Bank’s Core Course on Budgetary Processes and Public Expenditure Management, Washington, DC, May 22. https://view.officeapps.live.com/op/view.aspx?src=http://www1.worldbank.org/ publicsector/pe/Core%20Couse-%20Strategic%20Allocn.ppt. Schick, Allen. 1998. “Allocative Efficiency,” In A Contemporary Approach to Public Expenditure Management, 89–110. Washington, DC: World Bank. Snilstveit, Birte, Jennifer Stevenson, Radhika Menon, Daniel Phillips, Emma Gallagher, Maisie Geleen, Hannah Jobse, Tanja Schmidt, and Emmanuel Jimenez. 2016. The Impact of Education Programmes on Learning and School Participation in Low- and Middle-Income Countries: A Systematic Review Summary Report. 3ie Systematic Review Summary 7. London: International Initiative for Impact Evaluation (3ie). http://pubdocs.worldbank.org/en/429341481774674870/The-Impact-of-Education-Programmes- on-Learning-and-School-Participation.pdf. Theunynck, Serge. 2009.  School Construction Strategies for Universal Primary Education in Africa: Should Communities Be Empowered to Build Their Schools? Africa Human Development Series. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/2637. Transparency International. 2013. Global Corruption Report: Education. New York: Routledge. https://www. transparency.org/whatwedo/publication/global_corruption_report_education. 215 Underwood, Julie K. 1995. “School Finance Adequacy as Vertical Equity.” University of Michigan Journal of Law Reform 28 (3): 493–519. https://litigation-essentials.lexisnexis.com/webcd/ app?action=DocumentDisplay&crawlid=1&srctype=smi&srcid=3B15&doctype=cite&docid= 28+U.+Mich.+J.L.+Reform+493&key=e01524aba9318b07adeedc50a54b6582. UNESCO (United Nations Educational, Scientific and Cultural Organization). 2015a. “National Education Accounts: A Comprehensive Framework to Improve Education Finance Statistics.” Paper prepared for the World Education Forum 2015, Incheon, Republic of Korea, May, 19–22, 2015. http://www.uis. unesco.org/Education/Documents/wef-national-education-accounts.pdf. --------. 2015b. Education 2030 Framework for Action: Towards Inclusive and Equitable Quality Education and Lifelong Learning for All. http://www.unesco.org/new/fileadmin/MULTIMEDIA/HQ/ED/ED_new/pdf/ FFA-ENG-27Oct15.pdf. --------. 2015c. EFA Global Monitoring Report, Education for All 2000-2015: Achievements and Challenges. Paris: UNESCO. http://unesdoc.unesco.org/images/0023/002322/232205e.pdf. --------. 2015d. “Pricing the Right to Education: The Cost of Reaching New Targets by 2030”. Education for All Global Monitoring Report. Policy Paper 18, July 2015 update. http://unesdoc.unesco.org/ images/0023/002321/232197E.pdf. UNESCO Institute for Statistics (UIS), UNESCO International Institute for Educational Planning (IIEP), and IIEP Pôle de Dakar. 2016a. Methodology of National Education Accounts. Paris: UNESCO IIEP and Montreal: UNESCO-UIS. http://unesdoc.unesco.org/images/0024/002457/245781e.pdf. --------. 2016b. Who Pays for What in Education?: The Real Costs Revealed through National Education Accounts. Montreal: UNESCO-UIS. http://unesdoc.unesco.org/images/0024/002462/246277e.pdf. UNESCO, UNICEF, World Bank and GPE (Global Partnership for Education). 2014. Education Sector Analysis Methodological Guidelines, Vol. I. http://unesdoc.unesco.org/images/0023/002305/230532e.pdf. van de Walle, Dominique. 2003. “Behavioral Incidence Analysis of Public Spending and Social Programs.” In The Impact of Economic Policies on Poverty and Income Distribution (Evaluation Techniques and Tools), edited by F. Bourguignon and L. Pereira da Silva, 69–84. Washington, DC: World Bank; New York: Oxford University Press. http://documents.worldbank.org/curated/en/973661468739790876/pdf/ multi0page.pdf. Welham, Bryn, Philipp Krause, and Edward Hedger. 2013. “Linking PFM Dimensions to Development Priorities.” Working Paper 380, Overseas Development Institute, London. https://www.odi.org/sites/ odi.org.uk/files/odi-assets/publications-opinion-files/8392.pdf. Wescott, Clay G. 2008. World Bank Support for Public Financial Management: Conceptual Roots and Evidence of Impact–Background Paper to Public Sector Reform: What Works and Why? An IEG Evaluation of World Bank Support. IEG Working Paper, World Bank Independent Evaluation Group, Washington DC. http:// siteresources.worldbank.org/EXTPUBSECREF/Resources/pfm_wescott.pdf. 216 Education Public Expenditure Review Guidelines Wils, Annababette. 2015. “Reaching Education Targets in Low and Lower-middle Income Countries: Costs and Finance Gaps to 2030.” Background paper prepared for the Education for All Global Monitoring Report 2015 “Education for All 2000-2015: Achievements and Challengers”. http://unesdoc.unesco. org/images/0023/002325/232560e.pdf. Woodhall, Maureen. 2004. Cost-benefit Analysis in Educational Planning, 4th ed. Paris: UNESCO International Institute for Educational Planning. http://unesdoc.unesco.org/images/0013/001390/139042e.pdf. World Bank. 2002. Peru: Restoring Fiscal Discipline for Poverty Reduction. A Public Expenditure Review. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/239311468074329638/ Peru-Restoring-fiscal-discipline-for-poverty-reduction-a-public-expenditure-review. --------. 2004. World Development Report 2004: Making Services Work for Poor People. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/5986. --------. 2009a. Core Guidance: Preparing PERs for Human Development. Washington, DC: World Bank. http:// siteresources.worldbank.org/EXTPERGUIDE/Resources/PER-Core.pdf. --------. 2009b. Specific Guidance for Education: Preparing PERs for Human Development. Washington, DC: World Bank. http://siteresources.worldbank.org/EXTPERGUIDE/Resources/PER-Education.pdf. --------. 2011. World Bank Group Education Strategy 2020, Learning for All: Investing in People’s Knowledge and Skills to Promote Development. Washington, DC: World Bank. http://web.worldbank.org/WBSITE/ EXTERNAL/TOPICS/EXTEDUCATION/0,,contentMDK:22474207~menuPK:282402~page PK:210058~piPK:210062~theSitePK:282386,00.html. --------. 2012. Public Financial Management Reforms in Post-Conflict Countries: Synthesis Report. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/945231468340162289/ pdf/699640WP0P1206070023B0PFM0Web0Final.pdf. --------. 2013. What Matters Most for School Finance: A Framework Paper. SABER Working Paper Series 2, Washington, DC: World Bank. http://wbgfiles.worldbank.org/documents/hdn/ed/saber/supporting_ doc/Background/FIN/Framework_School_Finance.pdf. --------. 2014. What Matters Most for Education Management Information Systems: A Framework Paper. SABER Working Paper Series 7, Washington, DC: World Bank. http://wbgfiles.worldbank.org/documents/ hdn/ed/saber/supporting_doc/Background/EMIS/Framework_SABER-EMIS.pdf. --------. 2015. Education Sector Public Expenditure Tracking and Service Delivery Survey in Zambia. Washington, DC: World Bank. https://openknowledge.worldbank.org/bitstream/ handle/10986/23884/K8641.pdf?sequence=2&isAllowed=y. --------. 2017. World Development Report 2017: Governance and the Law. Washington, DC: World Bank. http://www.worldbank.org/en/publication/wdr2017. 218 Education Public Expenditure Review Guidelines Education Global Practice WORLDBANK.ORG