AUS4133 Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Report No: AUS4133 Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Word Bank Office Jakarta Indonesia Stock Exchange Building, Tower II, 12th floor Contents Jl. Jend. Sudirman Kav. 52-53 Jakarta 12190 Tel : (+6221) 5299-3000 Fax : (+6221) 5299-3111 http://www.worldbank.org/id Published in December 2014 Photo credit: cover photo courtesy of Nugroho Nurdikiawan Sunjoyo, chapter 1 photo courtesy of Ed Wray, chapter 2 photo courtesy of Rezza Istily, chapter 3 and 4 photos courtesy of the World Bank Acknowledgements Acknowledgements ii viii Abbreviations ix List of Abbreviations iii Assessing the Role of the School Operational Grant Program (BOS) in Improving Executive Summary 1 Education Outcomes in Indonesia is a product of the staff of the World Bank. The findings, Executive Summary v CHAPTER 1 interpretation, and conclusions expressed herein do not necessarily reflect the view of the World Bank, its board of Executive Directors, officers or any of its member countries. The Indonesian school grants program - objectives and evolution 13 1.1 Introduction: a decade of education reform 13 The World Bank does not guarantee the accuracy of the data included in this work. The Chapter 1: The Indonesian School Grants Program - Objectives and 1.2 School funding in indonesia 16 boundaries, colors, denominations, and other information shown on any map of this work Evolution 1 1.3 The objectives of the Bantuan Operasional Sekolah (BOS) program 22 do no imply any judgment on the part of the World Bank concerning the legal status of any 1.4 The 1.1. inner workings of the BOS program Introduction: a decade of education reform 1 30 territory or the endorsement or acceptance of such boundaries. 1.5 Summary 34 1.2. School funding in Indonesia 3 The work is funded by the Government of the Kingdom of the Netherlands under the CHAPTER 2 Dutch Education Support Program (DESP) Trust Fund. DESP provides support to the 1.3. The objectives of the Bantuan Operasional Sekolah (BOS) program 8 Assessing the effects of BOS on household education spending and participation 35 Government of Indonesia through the World Bank for the purpose of developing policies, 2.1 Introduction 1.4. The inner workings of the BOS program 14 35 studies, and programs that help the government achieve its education strategic plan. 2.2 Trends in education spending 38 1.5 Summary 17 For any questions regarding this report, please contact Samer Al-Samarrai 2.3 Trends in school participation 47 (salsamarrai@worldbank.org) 2.4 Isolating the effect of BOS on household education spending and participation 50 2.5 Effect of BOS on school level finances 56 Chapter 2.6 Summary 2: Assessing the effects of BOS on household education spending 57 and participation 18 CHAPTER 3 2.1 Introduction BOS 18 management and school based 59 3.1 Introduction 2.2 Trends in education spending 20 59 3.2 The contribution of school based management to improved education 2.3 Trends outcomesin school participation 28 in Indonesia 61 3.3 BOS and school based management 2.4 Isolating the effect of BOS on household education spending and participation 31 62 3.4 The state of school based management in Indonesia 67 3.5 Effect 2.5 of BOS on school level finances Summary 35 77 2.6 Summary CHAPTER 4 36 Strengthening the BOS program 79 4.1 Adjusting BOS to enhance its focus on improving education quality 81 Chapter 3: BOS 4.2 Strengthening and the school poverty focusbased of BOS management 38 82 4.3 Making better 3.1 Introduction 38use of BOS funding through better alignment with other school funding 87 3.2 The contribution 4.4 Revitalizing school ofof the role based BOS to management empower to improved schools and education outcomes 90 in local communities Indonesia 39 4.5 Looking further ahead 95 References 3.3 BOS and school based management 40 97 Appendix 101 Assessing the role of the School Operational Grant Program (BOS) in improving education outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia v vi vii Boxes 3.4 The state of school based management in Indonesia 44 Boxes 2.10 Crude transition rates of children aged 12 and 13 years old into junior 1.1 An overview of the main mechanisms used to fund the education sector 16 secondary school 49 3.5 Summary 53 1.2 The ‘pupil premium’ in England 25 2.11 Trends in household education spending per student controlling for 1.3 The forerunners of the BOS program in Indonesia 26 household characteristics 53 1.4 School based management reforms in Cambodia and the Philippines 28 2.12 Regression-adjusted trends in primary and junior secondary net Chapter 4: Strengthening the BOS program 54 enrolment rates 54 1.5 BOS program implementation down to the school level 31 4.1 Adjusting 2.1 Sources of data to BOS enhance for its focus household on improving education spending education quality 55 and participation 40 2.13 Regression-adjusted trends in junior and senior secondary transition rates 55 3.1 The role of school based management in improving school decision making and 2.14 School hired teachers by year of hire 56 4.2 Strengthening the poverty focus of BOS 56 student learning outcomes in Indonesia 62 3.1 Channels of BOS support to school based management 60 4.3 The 3.2 Making 2003better of BOS use law education funding and school through based better alignment with other school management 63 3.2 BOS as a proportion of total school-level funding, 2010 64 funding 61 3.3 Connections between BOS funds and school based management 65 3.3 Percentage of primary and junior secondary school principals reporting they had 3.4 A recent evaluation of BOS training activities final decision making authority 68 4.4 Revitalizing the role of BOS to empower schools and local communities 63 72 3.5 Improving educational quality through enhancing community participation 75 3.4 Average participation in final school decision making 70 4.5 Looking 4.1 The 2012 further ahead regulations on67 fees for primary and junior secondary schools 85 3.5 School committee participation in key decisions 70 4.2 Local school grant programs 88 3.6 Committee members’ response to identifying objectives of BOS 71 4.3 Calculating the appropriate value of non-salary operational support 3.7 Composition of discretionary spending at the school level, Appendix 69 In Jakarta schools 89 primary schools, 2010 76 4.4 Increasing References 72 the knowledge of parents on the BOS program and their participation 4.1 Provincial poverty rates and the real value of BOS, 2013 83 in school affairs through school meetings and SMS text messages 92 4.5 School report cards for community based monitoring in Uganda 94 Appendix Figures 1 Composition of per-student spending (IDR thousands), education module 102 Figures Figures 2 Trends in household education spending per student in government and non-government schools 103 1 Channels through which BOS improves education outcomes 2 2 The value of BOS assistance for each student has increased considerably 3 Tables Tables 3 The introduction of the BOS program led to an initial drop in education spending by households 4 1.1 Characteristics of school grant programs, selected countries 24 4 Enrolment in secondary schooling has been growing 6 1.2 Comparison of BOS per-student grant and estimates of operational needs 5 School committees play their strongest role in decisions over the allocation of the (2012 Prices) 30 school budget and participation in key decisions 8 1.3 BOS allowable and unallowable expenditures, 2014 33 1.1 Major education policy changes in Indonesia 1997-2012 14 4.1 Summary of policy suggestions 80 1.2 Transfers and fund flows in the education sector 18 1.3 Composition of public education spending, 2009 19 Appendix Tables 1.4 BOS program allocations per-student and as a share of government spending, 2005-2014 20 1 Comparing education spending in the education and consumption modules 101 1.5 Average household education spending per student, 2003-2012 21 1.6 The objectives of the BOS program and its channels of influence 23 1.7 Steps in BOS allocation process, 2014 31 2.1 Channels through which BOS might influence households and students 36 2.2 Two measures of spending on education, 2002-2013 38 2.3 Average share of total household consumption required to enroll all children in primary and junior secondary school 39 2.4 Household spending at the primary school level 41 2.5 Household spending at the junior secondary school level 42 2.6 Household spending at the senior secondary school level 43 2.7 Composition of per-student spending (IDR thousands), education module 45 2.8 Students reporting fee payments 47 2.9 Primary, junior and senior secondary net enrolment rates, 2000-2013 48 Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Acknowledgements Abbreviations The team is grateful to officials and staff in the Ministry of Education and Culture for Abbreviation Indonesian (Bahasa) English their overall support of this study. Special thanks go to Pak Taufik Hanafi (Expert Staff for ACDP Kemitraan di bidang Pengembangan Analytical and Capacity Social Economy, Ministry of Education and Culture) for his support in the overall study. Kapasitas dan Analisis Development Partnership We are also grateful to Pak Didik Suhardi (Director General Basic Education, Ministry of Education and Culture) for his helpful comments and suggestions in a seminar on the report ADB Bank Pembangunan Asia Asian Development Bank held at the Ministry. AusAID Badan Kerjasama Pembangunan Australian Agency for Internasional Australia International Development AusaAID-ERF Badan Kerjasama Pembangunan Australian Agency for The core team responsible for the content of the report included Samer Al-Samarrai (Task Internasional Australia – Fasilitas International Development Team Leader, Senior Economist, GEDDR), Tazeen Fasih (Chapter 3, Senior Economist, Sumberdaya Pendidikan – Education Resource GEDDR), Amer Hasan (Chapter 2, Economist, GEDDR) and Daim Syukriyah Facility (Consultant, GEDDR). Ratna Kesuma (Senior Operations Officer, GEDDR), Muhammad Farman Izhar (Operations Officer, GEDDR), Nur Hidayat (Consultant, GSPDR) and BKM Bantuan Khusus Murid Special Student Assistance Santoso (Consultant, GEDDR) provided useful insights, advice and information on the BOS Bantuan Operasional Sekolah School Operational BOS program. The graph design and layout work was done by Yvonne Armanto. Assistance (Provided by the Central Government) / National School Grants The peer reviewers for the report were Sachiko Kataoka (Senior Economist, GEDDR), Dina BOSDA Bantuan Operasional Sekolah Daerah BOS Supplementary Abu Ghaida (Senior Economist, GEDDR) and Budi Susetyo (ADB Education Advisor). The Funding / Local School team is also grateful for helpful comments and reviews of the report from Susiana Iskandar Grants (Senior Education Specialist, GEDDR) and Andy Ragatz (Senior Education Specialist, BPS Badan Pusat Statistik Central Bureau of Statistics GEDDR). BSM Beasiswa Siswa Miskin National scholarship for the Poor The analytical work was generously supported by the Dutch Education Support Trust Fund. BSNP Badan Standar Nasional Pendidikan National Education Standard Board ECED Pendidikan dan Pengembangan Anak Early Childhood Education Usia Dini and Development ICT Teknologi Informasi dan Komunikasi Information and (TIK) Communication Technologies IDR Rupiah Indonesian Rupiah IES Standar Internasional Pendidikan International Education Standards IT Teknologi Informasi Information Technology KIP Kartu Indonesia Pintar Indonesian Smart Card viii Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia ix x LKS Lembar Kerja Siswa Student Worksheet Executive Summary MoEC Kementerian Pendidikan dan Ministry of Education and Kebudayaan Culture MoF Kementerian Keuangan Ministry of Finance MSS Standar Pelayanan Minimum Minimum Service Standards Pendidikan NES Standar Nasional Pendidikan National Education Standards NUPTK Nomor Unik Pendidik dan Tenaga Unique Identification Kependidikan Number for Teachers and Teaching Personnel OECD Organisasi untuk Kerjasama Ekonomi Organization for dan Pembangunan Economic Cooperation and Development OLS Simulasi Regresi Linier Ordinary Least Squares PISA Program Penilaian Pelajar Antarbangsa/ Programme for International Internasional Student Assessment RCT Pengujian Acak Terkendali Randomized Controlled Trials The Indonesian school grants program (Bantuan Operasional Sekolah - BOS) is SBM Manajemen Berbasis Sekolah School Based Management coming to the end of its first decade of operation. Over that period, the program has been continually improved and channeled ever larger amounts of funding directly to primary and SC Komite Sekolah School Committee junior secondary schools. Its success in delivering operational funding to schools has been SD Sekolah Dasar Primary School replicated in other parts of the systems and by many local governments. By 2014, all levels of SMP Sekolah Menengah Pertama Junior Secondary School the education system from ECED to tertiary have a ‘BOS’ type program and around a third SPP Sumbangan Penyelenggaraan Pendidikan Educational Management of all local governments implement similar programs. Contribution; tuition fee The purpose of the report is to assess whether the program has succeeded in contributing SUSENAS Survei Sosial Ekonomi Nasional National Socio-economic to expanded education access and improved quality. This executive summary outlines Survey the main findings from the analysis undertaken in the report and outlines the main policy suggestions that arise from these findings. TIMSS Studi Internasional tentang Matematika Trends in International dan Ilmu Pengetahuan Alam Mathematics and Science Improving education through the use of school grants Study (TIMSS) Education policy makers have increasingly recognized the importance of empowering schools to make their own decisions in the quest to improve education outcomes. Many countries have recognized that schools themselves are often in a better position than central government agencies to make effective decisions on some aspects of teaching and learning. This recognition has led to many countries introducing school based management reforms that aim to provide schools, and the communities they serve, with greater autonomy over their own affairs. These reforms have also supported stronger school accountability through the establishment of more inclusive school governing arrangements that include parents and the wider local community. The focus on school based management has usually gone hand-in-hand with direct funding to schools to support improvement. Funding of this kind differs from regular public funding of schools as it gives the school a degree of discretion on how the funds are spent. It also provides schools with a predictable income stream which has greatly facilitated school improvement planning. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 1 Executive Summary Executive Summary 2 3 The extent of decision making and resources devolved to schools varies significantly The school grant is allocated based on an amount for each student and currently covers across countries. For example, in Australia and the United Kingdom, schools are provided approximately 43 million primary and junior secondary school students. The real value with and control decisions over all recurrent spending including teacher hiring. In other of the per-student allocation has more than doubled since the introduction of the program in countries, school decision making is much more circumscribed. In Malaysia and Thailand, 2005 (Figure 2). In 2014, the BOS program provided funds to the average primary (junior schools are provided with resources to cover only non-salary operating expenses. School secondary) school of approximately US$10,000 (US$ 20,000). The program is financed by funding of this kind also varies according to the level of discretion schools have over its use. the central government and allows schools to utilize funds according to lists of authorized In some cases, funding is provided with relatively few conditions while in other cases schools and unauthorized categories of expenditure. have to spend funds according to approved spending plans with limited scope to alter pre- agreed budgets. Figure 2: The value of BOS assistance for each student has increased School based management and school grant programs have shown some success in considerably improving education access and raising education outcomes. Recent reviews of research BOS program allocations per-student and as a share of government spending, 2005-2014 exploring the impact of school based management reforms have shown that they have the (constant 2012 prices) ability to improve education access and learning outcomes as well as address education 1,200,000 14% inequality. 1 However, education reforms of this kind can take time to yield results and their % of total education budget success depends critically on political support and effective implementation. 1,000,000 12% Indonesian Rupiah 10% The Bantuan Operasional Sekolah school grants program succeeded a smaller school 800,000 grant program introduced in the wake of the Asian Financial Crisis in 1997. The 8% 600,000 objectives of the program are to reduce the public’s financial burden of education in the 6% framework of providing 9-years of good quality compulsory education and to support 400,000 4% school based management reforms. These objectives are designed to raise overall education 200,000 outcomes through three main channels (see Figure 1). 2% 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0% Figure 1: Channels through which BOS improves education outcomes (25) (44) (43) (36) (53) (48) (44) (62) (60) (60) Increased BOS per student amount: Primary state funding of BOS per student amount: Junior Secondary schools and reduced burden BOS spending as % of total education budget (RHS) on households Increased Note: Until 2011, there was a slight difference between the BOS amount for rural and urban schools. education participation In this figure, information for these years refers only to the rural school amount. Figures in parentheses and improved learning particularly for are the US$ equivalent of the primary per-student value of the BOS grant. the poorest Source: MoEC and MoF Introduction of Direct financial The role of BOS in reducing the burden of education costs faced by School Grants support (BOS) to poor students households • Greater school autonomy Unpacking the role of BOS in reducing the education costs faced by households is • Greater parental participation in school difficult. The BOS program covered all primary and junior secondary schools and was • Stronger school introduced to all parts of Indonesia at the same time. This makes it difficult to use formal accountability • Improved transparency of methods to evaluate the effect of BOS and out of necessity a second-best approach is adopted. Strengthened SBM school decision making The first step of the approach looks to see whether the introduction of and subsequent through establishment of rules and responsibilities changes in the BOS grant amount had any discernible effect on overall household education of schools and spending. Second, simple regression analysis is used to see if a ‘BOS’ effect remains after a local communities set of other explanatory factors (e.g. household income) are controlled for. Finally, trends in managing BOS. in household education spending at primary and junior secondary schools, where the BOS program operated, are compared with trends in senior secondary to identify any differences. 1 See for example, AusAID ERF (2011).‘School grants and school-based management’ and Bruns, B., D. Filmer, It is recognized that this approach cannot provide definitive conclusions on the effect of the et al. (2011). Making schools work: new evidence on accountability reforms. Washington D.C., The World BOS program but can provide some insights into its overall effect. Bank. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Executive Summary Executive Summary 4 5 Looking at overall trends suggests that the BOS program was initially associated with in contrast to per-student spending at the primary and junior secondary which registered a drop in household education spending (Figure 3). Looking at survey data reveals that small declines. These findings provide some tentative support to the perception that BOS annual education spending for households with children in primary and junior secondary reduced household costs. fell by about 6 percent in the first year after BOS was introduced. However, the drop in However, the drop in education costs faced by households appears to have been education spending appeared to be a relatively temporary phenomenon; by 2009 household relatively small compared to the size of the per-student grants given to schools. While education spending began to increase steadily again. These findings, outlined in the report, the analysis is only indicative it suggests that where overall drops in household per-student support other more detailed results showing that the incidence and level of charges made spending occurred they were relatively small, particularly at primary school, when compared on parents fell with the introduction of BOS but then began to rise as schools became more to the per-student amount given to schools through BOS. Drops in household spending for familiar with the workings of the BOS program. the poorest households were equivalent to around 5 percent of the BOS grant at primary school and around 30 percent at the junior secondary level. Figure 3: The introduction of the BOS program led to an initial drop in education spending by households The limited use of BOS funding to reduce charges faced by households is further Annual household per-student education spending, 2002-2012 supported by the significant increase in discretionary resources schools appeared to have after the introduction of BOS. It is possible that BOS only had a limited effect on the costs Introduction Large increase in BOS of BOS amount facing households because other sources of school funding fell when BOS was introduced. For example, local governments may have reduced their funding to schools in response to 900,000 the BOS program. Unfortunately, no detailed time-series information on school funding is 800,000 available. However, information on the number of teachers hired directly by schools, before IDR (constant 2012 price) 700,000 and after BOS provides important evidence on how the overall school funding situation 600,000 changed. In 2012, there were approximately six hundred thousand school hired teachers in 500,000 the education system and approximately half of these were recruited after the introduction of the BOS program. This suggests that schools had more resources to spend after BOS was 400,000 launched and they devoted a share of these resources to hiring additional teachers. 300,000 200,000 Improving education participation through the BOS program 100,000 0 Simple time trends show that enrolment in junior secondary, particularly for the 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 poorest households, increased significantly after the introduction of the BOS program. Year The reduction in household education costs were intended to raise education participation, average spending per student for all groups particularly amongst the poorest groups. The same approach that was used to look at how average spending per student for the poorest group BOS affected household education spending is used to explore its effect on education participation. Enrolment rates in primary school have been very high for a considerable time Note: (1) Average education spending per-student for households with primary and junior secondary and it is therefore unlikely that BOS has had any discernable effect (Figure 4). However, education children (2) Year here refers to academic year enrolment rates in junior secondary have been on an upward trend which appears to have Source: Susenas household survey, 2003-2013 accelerated after the introduction of BOS particularly for the poorest. Between 2000 and A closer look also reveals that initial drops in household spending were concentrated 2005, junior secondary enrolment rates for the poorest 20 percent remained relatively stable amongst poorer households and for children attending government schools. The general but increased 26 percentage points between 2005 and 2013. pattern shown in Figure 3 remains even after other factors that are likely to have influenced household education spending (e.g. levels of household income) are controlled for. Moreover, drops in household education spending directly after the introduction of the BOS program were relatively larger for the poorest 20 percent of households in Indonesia. The analysis contained in the report also reveals that declines were largely confined to government schools where the program was more strongly associated with reducing the cost burden of sending children to school. Household education spending trends at senior secondary did not follow the same pattern which provides some evidence to suggest that BOS reduced costs for at least some households. In the immediate period after the introduction of BOS household education spending per senior secondary school student continued to increase. This stands Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Executive Summary Executive Summary 6 7 Figure 4: Enrolment in secondary schooling has been growing The role of BOS in supporting school based management Primary, junior and senior secondary net enrolment rates, 2000-2013 The BOS program has been a vital component of government efforts to implement Introduction of BOS Increase in BOS amount school based management reforms. In 2001, the responsibility for basic education service 100 100 delivery was largely devolved to local governments. Further reforms were introduced in 2003 that provided the legal basis for school based management and school committees in an 90 90 effort to encourage local community participation and strengthen accountability between 80 80 schools and parents. The BOS program supported these reforms by providing resources to Net enrolment rate (%) 70 70 fund school improvement plans and by making use of established school based management 60 60 structures and processes to govern the use of its funds. 50 50 Evidence from Indonesia shows that improvements in school based management can raise levels of learning achievement. For example, a recent study found that primary schools 40 40 with better parental and school committee participation had better learning outcomes. The 30 30 study showed that the effects of better school based management worked through improved 20 20 resource allocation decisions and higher teacher attendance rates.2 10 10 Most schools in Indonesia have the institutions and processes required for school 0 0 based management. A nationally representative survey conducted to explore school based management issues showed that all schools had established school committees. However, the 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 selection of committee members was not very transparent. For example, in primary schools Year members were commonly either appointed or selected by consensus; less than 15 percent Primary: national average Junior: national average Senior: national average of school committee chairs and less than 25 percent of committee members were elected. Primary: poorest 20% Junior: poorest 20% Senior: poorest 20% Principals reported that they had considerable autonomy over a number of important Source: Susenas household survey areas of their schools’ affairs. For example, almost all interviewed principals felt that they set the overall vision and goals of the school and were the final decision maker on school More detailed analysis tentatively suggests that the BOS program may have initially planning and budget decisions. In making decisions, school principals typically involved contributed to the increase in junior secondary school enrolment amongst poor school teachers in the process. However, it was less common for school committees to be households. The analysis undertaken in the report attempts to control for a host of other involved. A national survey of primary and junior secondary schools estimated that school factors that may have contributed to the trend in participation rates. After controlling for committees were involved in about 40 percent of the decisions made at the school level. these variables a large increase of around 5 percentage points is registered, for the poorest households, in junior secondary enrolment rates directly after the introduction of BOS. This While school committee participation was less commonplace, their strongest role effect seems to have also been temporary and enrolment rates settled back onto a long term centered around the use of BOS funds and overseeing financial matters more generally. trend that did not fluctuate with subsequent increases in the per-student amount of the In 2010, more than 60 percent of school principals reported that school committees were BOS grant. Further support for these findings comes from the different rate at which poor involved in final decision making in these areas (Figure 5). Perhaps unsurprisingly, school households closed the participation gap in junior and senior secondary schools (Figure 4). committees were rarely involved in decisions about pedagogy and instructional issues. These findings highlight the importance of the BOS program in opening up school decision- However, transition rates were also examined for a similar effect but this analysis did not making to the broader school community. find a similar jump in participation at the time BOS was introduced. Enrolment rates are only one measure of the potential effect of BOS on school participation. The program was expected to improve the chances of all children completing the full nine years of compulsory education by improving transition rates between primary and junior secondary education. Transition rates of this kind have indeed increased since the introduction of BOS and have followed a similar trend to the enrolment rates shown in Figure 4. However, further analysis shows that the introduction of the BOS program and subsequent increases in its level were not associated with jumps in transition rates. 2 See for example, Chen, D. (2011). ‘School-based management, school decision-making and education outcomes in Indonesian primary schools’. Policy Research Working Paper No. 5809. The World Bank and Heyward, M. O., R. A. Cannon, et al. (2011). ‘Implementing school-based management in Indonesia: impact and lessons learned.’ Journal of Development Effectiveness Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Executive Summary Executive Summary 8 9 Figure 5: School committees play their strongest role in decisions over Strengthening the BOS program and potential future directions the allocation of the school budget and participation in key decisions The report shows that the BOS program had a temporary and small effect on the costs %ge of primary and junior secondary schools where school committee is involved in final decision faced by households. These findings were strongest amongst the poorest households and in selected areas, 2010 for households who sent their children to government schools. While these findings are 0% 20% 40% 60% 80% promising they suggest the need to explore ways in which BOS can more effectively support the education outcomes of poor households in Indonesia. In particular, the report shows that Allocation of school budget the real value of BOS can be much smaller in poorer regions of Indonesia once geographical Allocation of BOS funds cost differences are accounted for. School work plan Results on student participation were less conclusive but suggest that the program may School vision, mission and/or goals have contributed, at least initially, to increases in junior secondary school enrolment rates. Notwithstanding the contribution of BOS, enrolment in primary and junior secondary Teacher recruitment, hiring and incentives school is high and continues to rise. This suggests that it may be timely to consider a greater Student admission criteria focus on quality in terms of the objectives of the BOS program. School curriculum The BOS program has also supported efforts to provide schools with more autonomy Textbook and to strengthen links with local communities. In other countries, school based management reforms have improved educational attainment and in some cases levels of Academic calendar learning achievement. However, it is not clear that these gains have so far been realized in Student promotion to next grade Indonesia. In particular, these weaknesses manifest themselves in the relatively inefficient use of funds evident in the large amount devoted to hiring additional teachers. These findings Source: School Based Management Survey, 2010 suggest that greater efforts are needed to establish and deepen the implementation of school based management if outcomes are to be improved through this route. Despite the role that BOS has provided for school committees, there are significant weaknesses in how effective they have been. Focus group discussions, conducted as part From the findings of the report summarized here in the executive summary a number of of the survey, with the BOS team and school committee members generally agreed that policy directions are suggested and include: committee members were rarely, if ever, actively involved or consulted in making BOS fund 1. Adjusting BOS to enhance its focus on improving education quality allocation decisions. In practice, it was more common for the school principal and teachers to agree on the allocation of BOS funds and then to communicate their decision to the • Link BOS funding more directly to education standards. Efforts have been made school committee chair for approval. in the past to set the value of per-student financing provided under BOS to minimum service standards. Establishing a more formal link between BOS The role of the school committee in managing BOS funds is further weakened by the funding and education standards has the potential to signal the importance of requirement for schools to establish a separate BOS management team. While BOS using BOS resources to fulfil these standards. BOS funding could also be tied management has generally aligned with existing school governance arrangements it also to quality assurance systems by providing an incentive for schools to obtain introduced an additional team to manage BOS funds. Rules on the formation of this and maintain accreditation status. team explicitly prohibit membership for parents from the school committee. The school based management survey found that about 70 percent of primary schools had established • Review list of eligible items under BOS to provide schools with the flexibility to these teams. Given that all schools already have school committees, a separate team for the invest in quality enhancing inputs. For example, allow BOS funds to be used management of BOS funds dilutes the potential role of the school committee and introduces to purchase teaching and learning materials such as audio-visual equipment. a lack of clarity into the management of the BOS program. 2. Strengthen the poverty focus of BOS Weaknesses in the role of the school committee in the oversight of BOS funds are likely • Adjust the value of BOS periodically to account for regional price differences and to have contributed to the limited effectiveness of BOS spending on raising education inflation to ensure that all schools can meet operating standards. Indonesia is a large quality. Limited oversight and participation of local communities in deciding the use of and diverse country and providing the same amount of per-student funding to BOS funds has meant that around a half of all BOS funds have gone to hiring non civil- schools in areas with high costs is unfair. At the minimum, consideration should service teachers and paying teachers for additional activities. Given the very low student- be given to adjusting the BOS funding formula periodically for geographical cost teacher ratios seen in primary and junior secondary education it is questionable whether, in differences and inflation. many cases, hiring additional teachers is the best use of BOS funds. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Executive Summary Executive Summary 10 11 • Use the BOS formula to provide more funding to schools serving poor and vulnerable The central government could also consolidate more of its funding into the BOS children. Schools serving poor and disadvantaged students need additional support program and provide a greater share of funds directly to schools. Alternatively, formula to ensure that they are able to provide a quality of schooling similar to schools in funding mechanisms of this kind could be also used as part of the broader inter-governmental wealthier areas of Indonesia. transfer system to allocate central government education funds (including BOS) to local governments. The local governments could then add this to their own resources and allocate • Phase out the use of BOS resources to support the ‘out of pocket’ expenses of poor students. funding to schools on basis of a single formula. Given its initial success it is perhaps timely to Existing guidelines on BOS allow schools to cover the education costs of poor explore how the program and the mechanisms it has introduced for allocating and managing households. However, large cash transfer programs (e.g. Kartu Indonesia Pintar) resources can be adapted to make an even bigger contribution to improving education already exist and are perhaps more effective at supporting these costs. While these outcomes in Indonesia. programs require strengthening, they should be the principal way of reducing the direct costs of schooling. 3. Improve coordination of BOS with other school funding • Clarify school fee and contribution policy. Despite government efforts to provide clear guidelines, survey data reveal that fees and charges continue to represent a significant proportion of ‘out-of-pocket’ expenses. Efforts to clarify the rules governing voluntary contributions to schools should continue and consideration should be made to strengthening the role of school committees in managing the level of contributions. Regulations should also be clearly communicated to parents and other stakeholders. • Coordinate more closely with local governments. Many local governments also run school grant programs to support school operating expenses beyond basic BOS funding. It is important that these funds are used to raise overall school standards beyond the level provided by BOS. 4. Revitalize the role of the BOS program in empowering schools and local communities. • Strengthen school committees. Strengthening school-level management of the funds to improve their effectiveness is vital. Strengthening the role of the school committee by for example, transferring responsibilities of the BOS team to the committee and ensuring better representation in the committee have the potential to significantly improve the effectiveness of BOS funds. The introduction of BOS type programs in ECED, senior secondary and tertiary education is testament to the success of the BOS program. In the nearly 10 years it has been running it has established itself as a program that is able to deliver resources to schools on a regular and timely basis. It is popular amongst parents and is becoming increasingly well known. Other countries having successfully established school grant programs and their financing mechanisms have further developed them to address other education challenges. For example, programs of this kind have been used to consolidate the different channels of funding schools receive. In particular, other countries have included teacher salary costs in funding formula which has contributed to better spending efficiency. As the report shows, Indonesia suffers from significant budget fragmentation in its school financing system which contributes to the large and growing spending inefficiencies. Consolidating a larger share of their budgetary resources, and in particular teacher remuneration, into BOS type formulas could be useful for local governments striving to improve spending efficiency and sustainability. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 12 CHAPTER 1 The Indonesian School Grants Program - Objectives and Evolution 1.1 Introduction: a decade of education reform Over the last ten years, Indonesia has continued to make significant progress in improving educational opportunities for all its citizens. Participation rates have increased at all levels and by 2013 approximately 90 percent of 7-18 year olds were in school compared to 81 percent in 2003. Increases in access have been particularly large amongst the poorest and most marginalized groups. Between 2003 and 2013, an additional 1.5 million children from the poorest 20 percent of households enrolled in primary and junior secondary school, increasing participation amongst these children from 77 to 84 percent. Improvements in education outcomes have been driven by a comprehensive reform agenda that began with decentralization. In 2001, the responsibility for many aspects of basic education service delivery was devolved to local governments. Further reforms were introduced in 2003 that provided the legal basis for school based management and formalized school committees in an effort to encourage local community participation and strengthen accountability between schools and parents. In 2005 a new law on teachers was passed that addressed shortcomings in teacher pay and quality by introducing certification, improved remuneration and a strengthened program of continuous professional development. At the same time, the national school grants program (Bantuan Operasional Sekolah) was rolled-out and provided schools with the resources necessary to support the adoption of earlier school based management reforms. A national scholarships program was introduced a little later to complement the school grants program and support poor households with the ‘out-of- pocket’ school expenses (see Figure 1.1). Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 13 The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 14 15 Figure 1.1: Major education policy changes in indonesia 1997-2012 the program on household education costs and education participation. The review is used Education law - 9 to draw out policy options to improve the program further and ensure that it contributes years compulsory Abolition of SD effectively to emerging challenges in the sector. education and and SMP fees school based Introduction of management While the report aims to provide a detailed assessment of how successful the program has Introduction Free basic school grant and Decentralization of national education been in fulfilling its objectives it has a number of limitations. First, it is difficult to isolate scholarship projects of primary Teacher law 2005 scholarships regulations and secondary (BSM) the impact of the BOS program amongst the many other factors influencing education education services outcomes. For example, as the introduction has shown, there were significant policy changes occurring across the education sector at the same time that were also likely to affect 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 education attainment. With the data available it is not possible to utilize more experimental or quasi-experimental approaches to establishing program impact. While efforts have been made in the assessment to control for these other factors it is recognized that this limits the Asian introduction BOS value BOS value BOS value strength of the conclusions that can be drawn. Second, the focus of the report is on the financial of BOS increased increased increased crisis role the program has had in supporting improvements in education outcomes. It does not aim to explore or evaluate in detail program implementation because a series of previous BOS value reviews have already explored these issues (see for example, World Bank 2008; 2010; 2011; increased 2013b). While these reviews have shown on the whole that program implementation has been good it is possible that implementation weaknesses may lie behind some of the findings The ambitious reform agenda has been backed up with substantial increases in public reported. Third, time series information on the reaction of local government financing to education investments. In 2009, a constitutional obligation to devote a fifth of the national the introduction of the BOS program is not available and may be an important explanation budget to education was achieved for the first time. This resulted in a more than doubling, of the findings of the report. Finally, the report does not undertake new analysis on the link in real terms, of public education spending between 2001 and 2009, a rate seen in few other between the quality of school based management and student learning outcomes. Detailed countries. Since then, public investment in education has continued to grow rapidly. work on this has already been undertaken and the report summarizes rather than duplicates this work (see for example, Chen 2011; World Bank 2012e) Despite increased investment and significant policy change the education sector still faces a number of challenges. Indonesia’s rapid increases in educational access have not been Notwithstanding these limitations, the report shows that the introduction of the program accompanied by significant improvements in learning achievement. The most recent was associated with a reduction in the costs faced by households in sending their children international learning assessments show that between 2006 and 2012, mathematics to school. However, this drop appears to have been only temporary and while household proficiency among Indonesian 15 year old students has declined and reading and science education spending has grown more slowly in basic education it is now higher, in real terms, scores have stagnated. This is particularly worrying given that, in 2012, over half of all than it was 10 years ago. Moreover, the burden of enrolling all children in primary and junior students were judged to be below the lowest international benchmark in mathematics. secondary education on household income has not fallen despite significant improvements in household income and reductions in poverty. The report finds that the program had a Concerns have also been raised around the role of increased public investment in contributing limited impact on primary school participation but finds some evidence of a positive effect to growing inefficiency in the sector (World Bank 2013d). In particular, a large proportion on junior secondary enrolment. of additional education spending has been directed towards hiring new teachers and paying them more. A large share of the increased resources available for teachers has come through The upshot of these findings is that schools have had considerably more discretionary the national teacher certification program and the introduction and increased resources resources for school improvement activities since the introduction of BOS. These additional flowing directly to schools through the national BOS program. While this increased resources have provided important support to the implementation of earlier school based spending has absorbed significant resources it appears to have had little impact on learning management reforms. For example, almost all schools have functioning school committees achievement. Put another way, these findings suggest that raising the effectiveness of public and school improvement plans. However, evidence shows that these institutions remain spending and the mechanisms used to allocate resources to local governments and schools weak and have some way to go before they can ensure effective use of the increases in funding has the potential to improve the quality of education substantially. schools have experienced under the BOS program. The purpose of this report is to provide an assessment of the national school grants program, The remainder of this chapter provides a brief overview of school funding in Indonesia Bantuan Operasional Sekolah (BOS), and its contribution to progress in the education sector. before setting out the channels through which the BOS program was designed to improve The BOS program aimed to raise educational attainment through the provision of school education outcomes. The chapter also provides background information on the size and grants to all primary and junior secondary schools. The assessment is timely given that the mechanics of the BOS program as well as the major changes that have occurred over the last program has reached its tenth year of operation and efforts are already under way to expand ten years. Chapter 2 explores the effect of BOS on the education costs faced by households the program beyond basic education. The report draws together studies and data that have and whether these changes have been a key driver of the improvements in participation, been used to analyze the program and undertakes new analysis that looks at the effect of particularly of the poor, seen over the last decade. Chapter 3 looks at the role BOS has played in supporting school based management reforms and whether they are likely to have led to Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 16 17 improvements in education outcomes. The final chapter provides some suggestions on how to improve the BOS program and strengthen its role in improving education outcomes. Box 1.1: An overview of the main mechanisms used to fund the education 1.2 School funding in Indonesia sector (continued) In order to understand the potential role of the BOS program in supporting schools it is about 40 percent of DAK transfers allocated for education and used primarily for important to locate it amongst the other sources of funding primary and junior secondary school rehabilitation and quality improvement. The DAK allocation has a formula schools receive. In Indonesia, the mechanics of public education financing are extremely component that takes into account the fiscal gap and has a 10 percent matching complex and information on the size and timing of financial flows is frequently lacking. requirement. DAK is transferred in three tranches: the first is allocated after the Schools can receive funds from central, provincial and district governments and from at budget is submitted to the central government; the next two depend on the depletion least eight different sources (see Figure 1.2 and Box 1.1). This level of budget fragmentation of the previous tranche. Although DAK is earmarked to fund capital spending, the makes it extremely challenging to use resources effectively and avoid duplication of effort government allowed some routine maintenance expenditures. amongst the different actors financing education. Revenue Sharing Fund (Dana Bagi Hasil, DBH) Unlike DAU, which is a horizontal equalization grant, DBH is a vertical equalization Box 1.1: An overview of the main mechanisms used to fund the 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 education sector3 part of DBH on basic education. DBH represented approximately 20 percent of This box provides a brief description of the objectives and means by which the various total sub-national government revenues in 2009. transfer mechanisms from the central government to sub-national governments within Indonesia are determined. These transfers represent the major source of Special Autonomy and Adjustment Funds including BOS and DID financing for sub-national governments and thus, to a large extent, explain the level Special Autonomy Funds include specific grants for Papua, Papua Barat and Aceh and composition of their spending. (Dana Otsus) and Special Adjustment Funds (Dana Penyesuaian) which include additional allowances for teachers, such as professional benefits for certified teachers General Allocation Fund (Dana Alokasi Umum, DAU) and for uncertified civil service teachers, a School Operational Assistance program The DAU, according to Law No. 33/2004 Article 1 (21), is a discretionary block grant (Bantuan Operasional Sekolah, or BOS), and local incentive grants (Dana Insentif sourced from the Central Budget (APBN) and aims to equalize the fiscal capacities Daerah, or DID) for education. of sub-national governments. It is transferred monthly and directly from central to sub-national governments. The DAU is allocated based on a national formula and is ----------------------------------------------------------------------------------------------------- the sum of a basic allocation (a portion of the sub-national budget for public servant Central Government Spending at the Sub-national Level not Recorded in Sub- salaries) and the “fiscal gap” (the difference between the estimated fiscal needs and national Budgets (APBD) fiscal capacity) of the sub-national government. The basic allocation accounted for about 45.5 percent of the DAU in 2010. Fiscal needs are based on regional variables De-concentration (Dekon) and Co-Administered Tasks (Tugas Pembantuan, such as population, area, GDP per capita, Construction Price Index (IKK), and the TP) human development index. Fiscal capacity is measured by a region’s own-source Dekon and TP funds originate from the central government’s budget (APBN), revenue and a fraction of total revenue-sharing. Based on Government Regulation and are administered by the provincial Dinas. The funds cover a variety of projects No.55/2005, provinces only receive 10 percent of the total DAU, while districts and activities, including school and classroom reconstruction and school quality receive 90 percent. improvements, social assistance programs (which included BOS until 2011) and Specific Allocation Fund (Dana Alokasi Khusus, DAK) capacity building programs for civil servants. DAK is an earmarked grant allocated to finance specific investment expenditures that are aligned with national priorities and carried out under the jurisdiction of sub- Source: World Bank (2013d) national governments. The DAK cannot be used for research, training, administration, Recent evidence suggests that the necessary coordination between levels of government and or official travel. In 2011, 19 economic sectors received DAK allocations including across agencies (e.g. Bappeda and DINAS at the district level) to accommodate this level education, health, agriculture, forestry, trade and various infrastructure sectors (road, of fragmentation is largely absent (World Bank 2012d). Furthermore there are significant irrigation, water, sanitation, rural electricity, housing and local government and weaknesses in the capacity of local governments to plan and budget effectively for education. remote areas infrastructure). Education is a key priority for DAK spending, with The monitoring of such a complex system is also extremely challenging and the lack of good quality information on resource allocations and usage exacerbates this issue. A recent study 3 of local education governance highlighted the weaknesses in existing information systems. The box excludes local governments own source revenue which is also used to fund the education sector. In While 70 percent of surveyed districts had education information databases less than half 2011, it was estimated that these made up on average about 7% of district government budgets (World Bank 2012d). had systems in place to verify the information that was collected (World Bank 2013c). Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 18 19 Figure 1.2: Transfers and fund flows in the education sector Figure 1.3: Composition of public education spending, 2009 LEVEL BUDGET 100 Percentage of total spending 90 53 NATIONAL MINISTRY OF FINANCE MINISTRY OF EDUCATION & CULTURE 61 80 MoF MoEC 70 DAU/SDA DAK Adjustment Fund Tugas Dekon Central 60 71 Pembantuan Functions BOS 50 100 6 40 1 30 3 16 20 41 10 26 22 PROVINCIAL Own Provincial Budget: Provincial Central Government Source DAU/SDA Education Agencies in the 0 Revenue PAD Office regions ECD Basic Senior Sec. Universities Education Education Central BOS-school grant Province District District Budget: District DISTRICT Own DAU/SDA Education Source: World Bank Education Public Expenditure Review, 2013 Source PAD Office Revenue DAK However, the central government remains an important source of non-salary spending in primary and secondary education. Latest estimates show that the central government provided over three quarters of all non-salary spending for primary and junior secondary Private Schools Public Schools schools (World Bank 2013d). Approximately half of these funds were provided through the BOS program.4 The remaining central non-salary support came through a number of Fund originated from MoF Fund originated from Provincial Budget different programs including ICT provision, student scholarships, school rehabilitation and Fund originated from MoEC Sectoral Budget Fund originated from District Budget a range of quality improvement programs. The discretion schools have over public resources is generally limited. Teachers are the Source: World Bank (2013d) largest single resource that the government provides all public and some private schools but While education financing remains a shared responsibility between all levels of government, schools have relatively little control over these civil-service teachers. For example, they have the bulk of funds for pre-tertiary education are provided by district governments. In 2001, limited authority over decisions on teacher hiring and transfer decisions which are largely the the responsibility for many aspects of basic education was devolved to local governments. responsibility of local governments. Schools also receive other ‘in-kind’ resources to support Decentralization reforms were expected to lead to significant improvements in education their teaching activities (e.g. IT equipment) but again have little control over the resources outcomes by bringing decision-making closer to the parents and students that are directly that are offered. affected. In this way, decisions on the best way to deliver education services would be increasingly responsive to local needs and more aligned with the specific characteristics of each district. Decentralization was also expected to lead to greater innovation and experimentation in service delivery with the potential for successful reforms to be replicated across local governments. In 2009, approximately two-thirds of all public education spending on pre-tertiary education was provided by district governments (Figure 1.3). 4 In 2011 BOS funding appeared in district government budgets and since then has been directed through provincial budgets. While this shifts a significant part of non-salary spending to local government budgets the central government remains a key provider of non-salary resources in the pre-tertiary sector. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 20 21 Figure 1.4: BOS program allocations per-student and as a share of considerable flexibility to use these funds to support their own school improvement plans government spending, 2005-2014 (see next section). The amount of the grant has increased considerably since the program was introduced in 2005. In 2014, the average primary school would receive an annual grant Constant 2012 prices of approximately IDR 100 million (US$10,000) while a junior secondary school received 1,200,000 14% approximately IDR 200 million (US$20,000). This is equivalent to the annual salary of approximately 2-4 certified civil-service teachers. % of total education budget 1,000,000 12% Local government programs tend to be much smaller in scale than the national BOS program Indonesian Rupiah 10% 800,000 and are generally less discretionary. In 2012, approximately half of all districts in Indonesia 8% had their own grant programs although the size of the grants tended to be smaller. A survey 600,000 in 2011 found that the average grant was equivalent to about a third of the national BOS 6% 400,000 grant schools received (World Bank 2013a). 4% 200,000 Overall, the BOS program provides the lion’s share of discretionary school funding. The 2% school based management survey, used in this report, shows that in 2010 BOS accounted for 0 0% approximately 83 percent of all discretionary funding that primary schools received (World 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Bank 2012e). District and provincial grant programs made up another 14 percent. The (25) (44) (43) (36) (53) (48) (44) (62) (60) (59) remaining 3 percent of discretionary resources reported by schools were received from non- BOS per student amount: Primary government sources including parental contributions. BOS per student amount: Junior Secondary BOS spending as % of total education budget (RHS) Figure 1.5: Average household education spending per student, 2003- 2012 Nominal prices 1,600,000 1,200,000 1,400,000 1,000,000 1,200,000 Indonesian Rupiah 800,000 1,000,000 600,000 800,000 400,000 600,000 400,000 200,000 200,000 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0 BOS per student amount: Primary 2003 2006 2009 2012 2003 2006 2009 2012 BOS per student amount: Junior Secondary Primary Junior Secondary Note: Until 2011, there was a slight difference between the BOS amount for rural and urban schools. In this figure, information for these years refers only to the rural school amount. Figures in brackets are fees materials uniform transport the US$ equivalent of the primary per-student value of the BOS grant. Source: MoEC and MoF Note: Excludes pocket money National and local government school grant programs are the only publicly provided resources Source: Susenas that schools themselves have significant discretion over. The national BOS program provides While the BOS program is a very important source of funding, information from school a per-student amount to all public and private primary and junior secondary schools and reports is likely to overestimate the share of discretionary resources provided by the program. madrasahs in Indonesia (Figure 1.4). In 2012, the overall BOS budget was approximately Since 2009, government primary and junior secondary schools have been prohibited from IDR 24 trillion (US$ 2.5 billion) and covered approximately 43 million primary and junior charging monthly tuition fees but are still able to collect voluntary fees and contributions secondary students in 220 thousand schools and madrasahs across Indonesia. Schools have Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 22 23 from non-poor students (see Figure 1.5). However, there was significant confusion over Figure 1.6: The objectives of the BOS program and its channels of the rules on charges and this appears to have led to schools underreporting parental influence contributions. This underreporting is apparent when information from schools is compared to information on charges collected directly from parents through household surveys. Figure 1.5, shows that approximately 40 percent of all household education spending in 2012 went Increased on fees and charges. Looking at 2012, this suggests that households were providing primary state funding of schools approximately IDR 323,000 per student. This is equivalent to about 55 percent schools and reduced burden of the overall BOS grant per-student. In a similar way to BOS, schools can also use these on households resources at their discretion as there is no requirement to pass them onto district education Increased authorities. Taking parental contributions into account suggests that BOS makes up around education participation a half of the discretionary resources that schools have but it is important to recognize that the and improved learning particularly for ability to collect voluntary contributions will vary enormously across schools. the poorest Introduction of Direct financial 1.3 The objectives of the Bantuan Operasional Sekolah (BOS) School Grants support (BOS) to poor students program5 • Greater school autonomy • Greater parental The school grants program in Indonesia aims to improve access to and raise the quality of participation in school the 9-year basic education cycle. The program provides the same per-student amount on • Stronger school accountability a quarterly basis to all government and non-government schools. The grant program is • Improved transparency of expected to raise education outcomes through three main channels (Figure 1.6): Strengthened SBM school decision making through establishment • Direct support for school operating costs. This channel has the potential to of rules and responsibilities of schools and reduce fees charged to parents and increase enrolment and participation particularly local communities for poor households. in managing BOS. • Financial assistance for poor students. School grants can provide direct support to poor students to cover transportation, stationery, uniform and clothing expenses.6 • Strengthened school based management. Grants are intended to lead to greater school autonomy by providing resources to finance activities school’s themselves Characteristics of school grant programs feel are important in raising enrolment and education quality. The management Many countries have established school grant programs of this kind to provide vital resources of funds through the school BOS team and the school committee is expected for schools to support enrolment and learning activities. While financing mechanisms of this to increase transparency, strengthen school accountability and lead to improved kind are relatively common, the characteristics and objectives of programs differs widely education outcomes. (Table 1.1). In some recent cases, school grants have been introduced to offset the losses in income that schools were expected to face when policies aimed at reducing or abolishing fees were introduced. In other cases, school grants have been seen as part of efforts to improve the overall quality of education. The scope of school grant programs also differs considerably across countries. In England and Poland for example, schools receive an overall budget for all aspects of school operation including the management and remuneration of teachers (see Table 1.1). In other countries, school grants fund a more circumscribed set of functions. For example, in Korea and Malaysia school grants are provided for non-salary recurrent costs only and teachers are provided to schools on the basis of staffing norms. 5 The definition of school grants, used in this report are funds that are provided to schools and which are spent by authorities at the school level. Schools must have an element of discretion over the use of these funds to qualify as grants. Grants are usually from public sources and exclude school income from fees and contributions by parents and the local community. 6 Assistance under BOS was initially limited to transport but expanded to these other areas in 2012. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 24 25 Table 1.1: Characteristics of school grant programs, selected countries Some countries in East Asia also adjust levels of school financing for the socio-economic status of children and other markers of disadvantage. For example, funding to schools in Country Coverage of Basis of basic Factors that increase/decrease basic Malaysia include additional allocations based on the number of students with low language school grant/ funding allocation funding allocation proficiency. In other countries, financing provided to schools through grant programs can funding to schools also be used to support the education of poor and disadvantaged children. For example, England All personnel and Student numbers 1. Social deprivation (e.g. students eligible for Malaysia provides additional per-student allocations for poor students and Korea provides operating costs free school meals) additional resources in kindergarten for children living in poor households (World Bank 2. Students with special educational needs 2014b). 3. Students with English as an additional language 4. Site and school factors (e.g. size of school, Box 1.2: The ‘Pupil Premium’ in England business rates etc.) Introduced in 2011, the  pupil premium provides government funded schools in Korea Non-salary recurrent 1. Teacher numbers None England additional per-student funding to raise the attainment of disadvantaged costs 2. Classroom pupils and narrow inequalities between them and other students. In 2014/15 schools numbers received an additional £1,300 (US$2,031) for primary-aged students and £935 3. Student numbers (US$1,461) for secondary-aged students. Rough calculations suggest that an average Malaysia Non-salary recurrent Student numbers are 1. Language groups in schools sized secondary school would receive approximately £200,000 (US$ 312,500) in costs main determinant 2. Subject based adjustments but there are many additional funding through the pupil premium which is the equivalent of five full- 3. Students with low language proficiency other smaller grants time teachers. 4. Additional per-student grants for poor children (e.g. milk, supplementary food and The main criteria of deprivation used to calculate eligibility is the number of students uniforms) in the school that have received free school meals over the last six years. Head teachers 5. School size and school governing bodies are accountable for the use of these funds in two ways. 6. Resource center, guidance and counselling provision First, performance tables that outline the performance of disadvantaged students 7. Special programs compared to their peers are made available to the public. Second, schools are required to publish details online each year of how they have used the premium and what Australia All personnel and Standard for primary 1. Students with non-English language operating costs and secondary school background and low parental education impact it has had. per student 2. Students from minority groups Schools typically use the additional resources to hire more teachers and teaching 3. Students with special educational needs assistants in order to introduce special programs for disadvantaged students. In 4. School size addition, resources are frequently used to allow eligible students to participate fully 5. School location (e.g. remoteness) in after school activities. 6. Students from two lowest socio-economic quartiles A recent study of the implementation of the pupil premium found: Poland All personnel and Student numbers None • Since the introduction of the premium an increasing number of schools (Kwidzyn) operating costs with a per-student are targeting the funding more effectively at improving the attainment of allocation for disadvantaged students and narrowing learning disparities. teachers, non- teaching staff and • The best schools combine a series of targeted interventions with robust tracking non-staff operating expenses systems to evaluate effectiveness. Note: This is a summary table and is designed to give a broad outline of the major components of the elements • Governing bodies in these schools take strategic responsibility for ensuring that determine the allocation to schools. Frequently, there are funding formulas used by central governments to allocate resources to local government or administrative units which these units then use to allocate to schools. the pupil premium supports eligible pupils. They also hold school leaders These formulas are excluded from the table. Information in the table does not include capital allocations. Refer to accountable for the use of these additional resources and the results obtained. original sources for more detailed information. Source: Alonso and Sanchez (2011), Chowdry and Sibieta (2011), Fazekas (2012) and World Bank (2014b). • Challenges remain in some schools with leaders and governing bodies in the weakest schools failing to ensure the pupil premium is used effectively to narrow School grant programs also differ in the extent of support they provide to poor and attainment gaps. marginalized children. In some countries, formulas that determine the allocation of funding between schools explicitly include adjustments to account for the poverty and/or special needs of students. For example, in the United Kingdom a ‘pupil premium’ is provided to Sources: OFSTED (2014) and www.gov.uk/pupil-premium-information-for-schools-and-alternative- schools based on the number of children that receive free school lunches (see Box 1.2). provision-settings Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 26 27 The grants provided through the BOS program do not provide additional resources to poor schools but grants can be used to support poor and marginalized students. Schools in Box 1.3: The forerunners of the BOS program in Indonesia (continued) Indonesia are prohibited from charging fees or seeking voluntary contributions from poor households. However, the ‘out-of-pocket’ expenses of school attendance can be a constraint Evaluations suggest that the combined scholarships and school grant program had for some poor households. While Indonesia has a large national scholarships program, BOS some positive impact on education outcomes. Results include: funding can also be used to support the direct costs of school enrolment for poor households. • Scholarship program had some success at protecting school enrolment which At the outset, BOS could be used to pay for the transportation costs for poor students and in remained relatively stable during and after the crisis. However, these effects 2012 its role was expanded to also cover the costs of uniforms and shoes. were most significant at the junior secondary level. The program did not appear While differences in the background of students a school serves are not adjusted for, it to have much effect on primary school progression or the transition between is important to recognize that BOS funds are likely to be a vital source of regular and primary and junior secondary school. predictable funding for schools serving poor children. In these schools, contributions from • Block grants were partially successful in protecting school incomes. However, parents are likely to be much lower. For example, household survey data reveal that in 2012, there was no evidence on whether the grants improved school quality or the poorest fifth of households provided about a half of the funding to government primary increased/maintained enrolment. Concerns were raised about the equally sized schools as the wealthiest 20 percent of households. BOS resources are therefore likely to be a grants across school levels and the lack of adjustment for school-specific factors much greater proportion of overall funding in schools with a large number of poor students. (e.g. number of students, school condition etc.). This will enable schools of this type to at least cover basic operating costs that may have gone unmet without BOS funding. • Qualitative studies pointed to a positive effect of school improvement grants. School improvements reduced disruptions to classes and improved student and School grants as a complement to school based management reforms teacher concentration. Teacher motivation was also improved because grants were used partly to provide additional payments to teachers. Questions were It is increasingly common for grants to be included as part of an overall reform package raised by stakeholders on the adequacy of the grant in schools with large student with the twin objectives of providing schools with greater autonomy and at the same time enrolments and the one-off nature of the grants which did little to foster proper strengthening accountability mechanisms between schools and the communities they serve. school improvement planning. School based management reforms of this kind require schools to undergo an improvement planning process that is used to identify strategies to raise education outcomes. This process Source: Ridao-Cano and Filmer (2004) is able to capture local knowledge of the factors constraining education performance and a way of identifying the most effective strategies in each school to address them. School grants Indonesia has followed a similar path and earlier experience of grants and school based have been seen as an important way for governments to support these reforms by providing management provided the building blocks for the BOS program. School grant programs directly to schools the funding they need to implement their own improvement plans. were initially introduced in Indonesia on a wide scale in the wake of the Asian Financial Crisis in 1997 (Box 1.3). Anxious to avoid the large increases in student drop-out and school closure that accompanied the crisis of the mid-1980s the government rapidly developed Box 1.3: The forerunners of the BOS program in Indonesia and implemented a school grant and scholarship program in 1998 (Baines 2005). These programs enhanced and in most cases established school committees to manage the use of In 1998, the Government of Indonesia introduced a program to respond to the school grants in schools. Evaluations suggest that the combined scholarships and school financial crisis of the previous year. The Scholarships and Grant Program (SGP) grant program had a positive impact on education outcomes (see Box 1.3). combined a scholarships program for children at risk of dropping out and block grants for schools. A second grants program (School Improvement Grants Program, Building on these positive experiences, the 2003 education law formally established school SIGP) was also introduced at the same time to compliment the SGP but also as a committees and introduced principles of school based management into the governance of strategy to raise education quality in a smaller number of schools. the education sector (Government of Indonesia 2003). Guidelines on the standards for the implementation of school based management soon followed in 2005 (Ministry of National Approximately two thirds of primary and secondary schools were targeted to receive Education 2007). These guidelines instructed schools on the process of formulating school grants in these early programs. Schools most affected by the crisis were intended as improvement plans, elaborated on the role of the local community and provided some the main beneficiaries and a combination of district poverty indices and school age clarification on the roles and responsibilities of school committees and other school level population sizes were used to allocate resources to districts.7 The same grant amount boards.8 With the introduction of the BOS program schools were provided with a regular was provided to each school at a particular level and the use of the grant was the and predictable funding stream to support the process of formulating and implementing the responsibility of school committees that were the precursor of the committees set up new school improvement plans. more formally in 2002 and the 2003 Education Law. 8 For a more detailed discussion of the roles and responsibilities of school committees and their role in the BOS 7 Schools were selected at the district level program see Chapter 3. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 28 29 Evidence from other countries points to the potential of school based management programs and associated grants to improve access and student progression indicators. Recent reviews Box 1.4: School based management reforms in Cambodia and the Philippines of the evidence on school based management have concluded that programs of this type (continued) have generally been successful at improving access and student progression (AusAID ERF 2011; Bruns, Filmer et al. 2011). For example, a program in Mexico that provided grants To assess the impact of SBM on learning outcomes the study exploited differences to support school improvement plans was shown to have reduced drop-out and repetition in the roll-out of SBM and associated grants between 2006 and 2009.9 The study rates (Skoufias and Shapiro 2006). Evidence from Cambodia and the Philippines also shows found that, over three years, the introduction of SBM and the provision of grants that school grant programs can lead to significant improvements in education outcomes (see improved scores on the National Achievement Tests (NAT) by 4 to 5 percentage Box 1.4). Programs of this kind have also been shown to have contributed to increased and points (approximately 0.25 standard deviations). Simple comparisons with other more equitable access to educational opportunities after the abolition of fees in sub-Saharan common interventions to improve education quality tentatively suggest that the Africa (Deffous 2011). SBM reforms have also been cost effective. Sources: Benveniste and Marshall (2004); Shoraku (2008); Khattri, Ling et al. (2010); Yamauchi and Box 1.4: School based management reforms in Cambodia and the Liu (2012); World Bank and AusAID (2013); Yamauchi (2014). Philippines However, evidence on the impact of school based management reforms on student learning School based management principles were first introduced in Cambodian schools in achievement are more mixed. A recent study that explored the results from the randomized 1998 through a pilot project that provided grants to clusters of schools. The objective experiments literature found no overall impact of school grants on learning achievement of the program was to improve school quality by providing resources for school in primary schools when they were introduced on their own (McEwan 2013).10 In improvement but also by building management and planning capacity in schools. Indonesia, a randomized controlled trial that provided schools with a small grant additional Grants were given to schools on the basis of priorities determined by school cluster to the national BOS grant found no statistically significant impact on learning (Pradhan, committees. Suryadarma et al. 2011). However, the additional grant was relatively small compared to the An evaluation found significant differences across clusters in the use of the grants BOS grants so the findings are perhaps not that surprising. Positive impacts were found in a and that these variations were the result of the interplay between schools and their study conducted in India and here school grants only had a positive effect on learning when communities in determining grant usage. The evaluation found that the school grants were not anticipated by schools and parents (Das, Dercon et al. 2011). When grants grants were associated with modest improvements in drop-out and promotion rates were anticipated, parents adjusted their own education spending downwards which meant as well as in student learning achievement. that the overall funding situation for education remained relatively unchanged. Without additional resources, levels of learning achievement remained the same.11 School based management and grant funding to schools are now part of the regular government budget and a key strategy to support Cambodia’s education priorities. The same review highlights the difficulty in assessing the impact of school based management They have played an important role in supporting the abolition of school fees in reforms because of the wide range of different interventions that fall under this reform area the first nine years of education and have reduced the cost burden on households. (McEwan 2013). In some cases efforts to strengthen school committees have had positive However, challenges remain in implementation and more recent studies have impacts on learning. For example, a randomized controlled trial in Indonesia found that identified remaining weaknesses in school support committees and community linking school committees with village councils and ensuring that elections took place for participation. committee members improved learning achievement (Pradhan, Suryadarma et al. 2011). In other cases, experiments introducing elements of school based management have had The Philippines has also had a similar experience where, in 2003, a pilot project little impact. For example, an experiment in the Gambia that provided a grant and a introduced school based management into 6,000 elementary schools. A study of this comprehensive program of training on school based management to a range of stakeholders initial pilot showed that schools that had introduced school based management had had no impact on learning achievement despite registering improvements in student and been successful at raising levels of learning at a faster rate compared to control group teacher attendance (Blimpo and Evans 2011). These results echo similar findings from the schools. Improvements in education quality brought about through school based broader evaluation evidence on school based management and highlight the importance of management and associated interventions have also been shown to have positive the overall policy context, design and implementation in determining success (AusAID ERF long-term development impacts particularly for women. 2011; Bruns, Filmer et al. 2011). Since the pilot project, school based management has been introduced nationally and have been accompanied in some cases by school grants. A recent impact evaluation showed these reforms continue to have a significant and positive impact on education 9 Propensity score matching techniques are used to develop an appropriate control group for the study performance (World Bank and AusAID 2013). 10 The evidence is fairly limited with relatively small scale experiments in only four countries. 11 It is possible however that the reduced levels of household spending that resulted from the introduction of the school grant prevented early drop-out and allowed poorer children to stay in school longer. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 30 31 1.4 The inner workings of the BOS program are able to purchase with their BOS funds. However, adjustments to account for price differences have not been made as they were seen to negatively affect the transparency of the Previous sections have documented the level of BOS funding and have shown how program and increase the risk of fund leakage. important these discretionary funds are for schools. This section provides a brief overview of the operation of the BOS program in order to highlight implementation factors that may be Figure 1.7: Steps in BOS allocation process, 2014 important in explaining program outcomes in the following chapters of the report. Schools submit data on student numbers to BOS district management team The size of per-student allocations in the BOS program have been calculated on the basis of studies to determine the operational expenditure needs of schools as well as the availability of the budget. At the outset, per-student allocations to the BOS program were considerably District, provinces and central BOS teams reconcile data and larger than the grants provided under the previous Scholarships and Grants Program (SGP) finalize data on number of students in each school but lower than those provided under the School Grant Improvement Program (SIGP). In 2006, the value of per-student amounts were adjusted on the basis of a school survey that was designed to determine average school operating costs (Ghozali 2005; Arze del Granado, Basis of disbursment and distribution to each province: BOS central team provides: Fengler et al. 2007). After 2006, periodic changes to the grant amount have taken place largely to adjust for inflation (see Figure 1.3). However, in 2012 the grant amounts were Student numbers to Proposed allocation to each province which forms Directorate General of basic education again changed after an exercise to estimate the operational costs necessary to achieve national for each school’s allocation basis of Ministry of Finance regulation education standards was conducted by BSNP. In 2012, BOS grant levels were above these initial estimates of the operational expenditure requirements at primary schools and slightly below estimated needs at junior secondary school (Table 1.2). However, a recent study concluded that levels of BOS were adequate to cover all non-salary operating expenditures National treasury transfer funds on a quarterly basis to provinces which then send to school bank accounts in primary schools with at least 130 students and in junior secondary schools with at least 150 students (ACDP 2013). The BOS program has largely been administered and financed directly by the central government. In order to receive their allocations, schools are required to submit data on Table 1.2: Comparison of BOS per-student grant and estimates of their enrolment numbers to their district administrations (Figure 1.7). This information operational needs (2012 prices) passes up through the system to the central ministry where the allocations for each school are determined. This information forms the basis of an annual Ministry of Finance regulation Current per- BSNP estimate ADB estimate determining the allocation for each school. Schools receive their grants on a quarterly student amounts based on based on basis and are able to check online to ensure they have received their full entitlement.14 minimum service minimum service Implementation of the program has been good with schools, on the whole, receiving the standards standard appropriate BOS grant in full and generally on time (see Box 1.5). Primary school 580,000 412,139 468,000 Box 1.5: BOS program implementation down to the school level Junior secondary school 710,000 731,894 831,000 Recent assessments of BOS program implementation show that the management Note: ADB survey was conducted in 2010 and unit costs are adjusted for inflation to report in 2012 of the program has been largely successful at providing schools with the correct prices. amount of BOS funds. Schools receive grants on the basis of the number of students attending. A monitoring and evaluation exercise conducted in 2012 showed that 93 While there has been considerable debate about adjusting grant amounts for regional percent of schools received the correct BOS grant. Moreover, the difference between cost differences, the nominal value of the per-student grant has remained the same across the amount received and the correct amount in the remaining 7 percent of schools Indonesia.12 Given the huge geographical spread of Indonesia it is unsurprising that large was small. While the accuracy of allocation has generally been good since the outset regional cost differences exist. Regional price indices show that prices in the education, of the program it has continued to improve through small changes to program recreational and sports category can be twice as high in some regions compared to others.13 implementation. For example, better data gathering systems were introduced to This can result in significant differences in the goods and services schools in different regions collect more accurate and timely information on student numbers over the course of 12 implementation. Between 2009 and 2011, the per-student amount did differ between rural and urban schools. However, the difference was very small and provided an additional 1% to city schools. 13 2012 data from BPS (www.bps.go.id/eng/tab_sub/view.php?kat=2&tabel=1&daftar=1&id_ 14 subyek=03¬ab=11) Schools can check their allocations and when they are released on the dedicated BOS website: http://bos. kemdikbud.go.id/ Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution The Indonesian School Grants Program - Objectives and Evolution 32 33 Table 1.3: BOS allowable and unallowable expenditures, 2014 Box 1.5: BOS program implementation down to the school level (continued) Allowable expenses Unallowable expenses In addition, the central government set up a buffer fund to address inaccuracies in 1. Purchase/reproduction of textbooks Saving to earn interest student numbers which led to a much quicker resolution of these kinds of problems. Activities around new enrolment (e.g. 2. Lending Schools also generally received their BOS grants on time which meant they had a registration fees etc.) predictable source of financing to carry out their plans. In 2012, it was estimated Activities that are not priorities (e.g. that 80 percent of all primary and junior secondary schools received their grants in 3. Learning and extracurricular activities study tours) the first month of the disbursement quarter. Delays in disbursement were usually Financing activities run by government associated with the late approval of the overall government budget which meant that 4. Tests and examinations offices most delays were seen in the first quarter of the financial year or the second semester of the school year. Consumables (e.g. notebooks, snacks 5. Teacher bonuses and routine transport etc.) The decentralization of the program to district governments in 2011 caused significant Clothes for teachers and students (excl. disruption to the running of the program. The requirement that schools fulfilled 6. Utilities poor) local government planning, budgeting and reporting guidelines led to significant delays. It also had a negative effect on the predictability of school-level funding and 7. School maintenance and repairs Medium and major rehabilitation resulted in less transparency in overall school funding. Owing to the delays in receipt Payment of honoraria teacher and of BOS funds many schools borrowed from school committees, parents and staff 8. education teaching personnel (max 20 Construction and this lending went largely unrecorded. Faced with growing pressure from media percent in government schools) reports of schools not receiving their funds on time the government reverted back to Materials that do not support learning a provincial level funding model in 2012. 9. Teacher professional development process External audits of the BOS program have also been positive. The state audit Contributions for poor students (e.g. 10. Buying shares agency audited the use of BOS funds and, with the exception of 2011, these audits transport, uniforms, shoes) resulted in an unqualified opinion. The 2011 audit gave a qualified opinion largely Activities already funded by other due to discrepancies in reported amounts transferred to schools. However, further 11. BOS management agencies investigation found that these differences were due to weaknesses in district reporting Activities unrelated to school operations rather than to fund misuse. 12. Computers and printer/scanner (e.g. festivals, events) While program implementation has generally been good the effectiveness of the BOS Activities organized by non-government program rests on what schools decide to use their resources for. The management 13. Other expenses when 1-12 are covered institutions related to BOS program and governance of BOS resources at the school level is the subject of Chapter 3 of the report. Purchase of software for the purpose of 14. BOS financial reporting Source: Ministry of National Education (2009); World Bank (2013b) Source: BOS Manual, MoEC, 2014 In 2011, attempts were made to decentralize the program and transfer BOS funds through The BOS manual also provides detailed guidelines on the roles and responsibilities of all local government budgets. These changes met with little success and were reversed in the stakeholders involved in the program. BOS teams are established in all levels of government following year after delays in the allocation of BOS funds and complications in reporting and at the school level. These teams form the basis of the implementation, monitoring and requirements arose. Since then BOS funds have been allocated to provincial governments evaluation of the program. The national BOS management team’s main responsibilities who directly transfer funds to schools. include: The annual BOS manual provides the implementation guidelines for the program and • Coordinating the collection of student enrolment data and calculating the outlines eligible and ineligible uses of grants (Table 1.3). Schools are given significant allocations of BOS for all primary and junior secondary schools across Indonesia. flexibility on the use of BOS funds which can be used for a wide range of improvement • Preparing annual implementation guidelines and disseminating changes of the activities. With the exception of the payment of school hired contract teachers, schools program to the management teams responsible for BOS in local governments and and their communities are also left to decide the composition of their spending across the at the school level. different allowable expenses. Restrictions on the proportion of BOS funds that can be used for spending on contract teachers was introduced in 2009 after concerns were raised about • Handling grievances through the Public Complaint Service and Handling Unit. the over hiring of teachers and the very low student teacher ratios in some schools. • Monitoring and evaluation of the program. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia The Indonesian School Grants Program - Objectives and Evolution 34 A BOS team also exists at the school level and works with the school committee to manage BOS funds. The BOS team at the school level is made up of the school principal, a treasurer CHAPTER 2 and a parent representative. The team is the main focal point in the school and manages all administrative procedures associated with the BOS program. The school committee oversees the planning and use of BOS funds as well as participating in the more general school Assessing the Effects of BOS on improvement planning process. It is important to recognize that school improvement and annual work plans that are central processes of school based management are required for Household Education Spending the implementation of BOS. Chapter 3 looks in more detail at the effectiveness of the school BOS team and the school committee in managing BOS funds and implementing school and Participation based management more generally. 1.5 Summary This chapter has highlighted the significant reforms that have occurred in the Indonesian education sector over the last 10 years. The BOS program has been a central component of this overall agenda and has provided vital funding for school improvement and to support the necessary changes to their management that these reforms have demanded. The large share of the budget the program absorbs and its coverage of approximately 220,000 schools is testament to its importance to the reform process. The program is designed to contribute to the overall objectives of the sector by reducing the costs faced by households in sending their children to school and by providing schools with resources to implement better school based management. The extent to which the program has supported these ambitious objectives is 2.1 Introduction the subject of the next two chapters. The next chapter looks at the role of BOS in reducing the Schools grants such as the BOS program are meant to increase state funding of schools, costs faced by households and improving levels of participation and attainment. Chapter 3 thereby reducing the need for schools to raise resources directly from parents. Chapter then takes a closer look at the role that BOS has played in strengthening the implementation 1 noted that the BOS school grants program was expected to raise education outcomes of school based management and how effective the management of BOS resources has been through three main channels (Figure 2.1). This chapter focuses on two of the three channels: at the school level. • Increased state funding of schools and reduced burden on households. This channel has the potential to reduce fees and other charges made by schools on parents. BOS resources would reduce the need for schools to charge parents and shift the overall burden of education costs from households to the government. • Direct financial support to poor students. Schools are also allowed to use BOS resources to support the ‘out of pocket’ expenses of poor students. This extra support would reduce the overall burden on poor households and again shift it to the school and the government. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 35 Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 36 37 Figure 2.1: Channels through which BOS might influence households what might have happened at lower levels of schooling had there been no BOS program. In and students light of this, the chapter also attempts to answer a third question: 3. Is there evidence of a different trend in basic education when compared with the trend in post-basic education? Increased state funding of Given that the BOS program transferred different amounts of funding to primary and schools and junior secondary schools it is also likely that spending and participation trends will also vary reduced burden on households between the two levels. The chapter also looks at differences in trends between government Increased and non-government schools because the effect of BOS in lowering fees may have been education participation stronger in government schools than in non-government schools. and improved learning particularly for BOS was introduced as a national program in 2005 and this limits the analysis that can be the poorest undertaken to assess its impact. For example, there are no exogenous sources of variation Introduction of Direct financial that would allow the causal impact of the BOS program on household education spending School Grants support and student participation to be determined. Instead, an attempt is made to analyze the (BOS) to poor students trends in these indicators before and after the start of the BOS program. The chapter also • Greater school autonomy takes advantage of the fact that the per-student allocation changed over the life of the BOS • Greater parental participation in school program and explores whether subsequent increases, after the program’s introduction, were • Stronger school associated with changes in spending and participation. accountability • Improved transparency of A major drawback of relying on an analysis of trends is that this does not control for other Strengthened SBM school decision making through establishment factors that may have also influenced household spending and participation decisions but of rules and responsibilities which are not related to the BOS program. These include, but are not limited to: of schools and local communities a. The growth in the Indonesian economy since 2005 which has in turn resulted in in managing BOS. greater consumption overall and lower poverty rates. b. The introduction of scholarship programs which may have reduced household If BOS grants are successful in reducing the burden on households of sending their children spending or eliminated it altogether for some households. to school it would be expected that participation and attainment would improve. Studies Such limitations notwithstanding, the chapter finds that the BOS program affected levels of have highlighted the significant costs households face in sending their children to school. household education spending in a limited way. Reductions in spending were seen immediately Efforts that make education more affordable for households are likely to result in more after the introduction of the program in 2005 and were particularly pronounced for poor children starting and staying in school for longer (World Bank 2012b). Reductions in the children. These initial effects of the program were also larger at the junior secondary level. burden are likely to be particularly important for poor households that typically devote a However, the magnitude of household education spending reductions was small relative to greater share of their income to education compared to better-off households. the size of the BOS grants. These findings which suggest that schools did not pass on much of This chapter assesses whether the BOS program reduced the costs faced by households and the BOS grant to parents in lower charges are further supported by the large improvements if this led to more children enrolling and completing primary and junior secondary school. seen in overall school funding since the introduction of the BOS program. The chapter also While the available data to make this assessment are not ideal, the chapter attempts to provides some evidence to suggest that the introduction of BOS was associated with a jump answer the following questions: in junior secondary enrolment rates. However, a similar improvement in transition rates between primary and junior secondary was not found. 1. How has household education spending and participation in basic education changed since the introduction of BOS? This chapter is organized as follows: Section 2 looks at trends in the level and composition of household spending on education. Section 3 examines trends in school participation before 2. Controlling for other determining factors what happened to spending and and after the BOS program and assesses whether there are any changes in enrolment and participation in basic education when BOS was introduced and when it was transition rates for children across different levels of education. Section 4 tries to assess the increased in subsequent years? trends in household spending and participation controlling for other factors that may have Since the BOS program only operated in primary and junior secondary schools it is influenced these outcomes. Section 5 examines overall school level finances to see if BOS instructive to compare these trends with those seen in senior secondary education. It would had improved overall levels of school financing while Section 6 concludes. not be unreasonable to expect that given the additional support the BOS program provided to primary and junior secondary schools trends in spending and participation would be different to those seen in senior secondary. These differences in trends could be indicative of Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 38 39 2.2 Trends in education spending15 Direct household education spending was growing before the introduction of BOS - having gone from 19 percent of total spending in 2002 to 23 percent of total spending in 2005. Total education spending – spending by households and the government – in Indonesia After the introduction of BOS, household spending fell sharply and reached 15 percent of has risen rapidly over the last 10 years. In real terms total education spending has almost total spending by 2006. However, since that initial reduction, household spending has been trebled in size since 2002 (Figure 2.2). Growth in total education spending has been more steadily increasing and represented 20 percent of total spending as of 2013. The second panel rapid in recent years, partly as a result of rapid increases in government spending since 2009 of Figure 2.2, shows that on average there were initial declines in per student spending – when the constitutional obligation to spend 20 percent of the total government budget on particularly among the poorest quintile – upon the introduction of BOS (2005) as well as for education was achieved. poor households when BOS was increased by a significant amount in 2009. Figure 2.2: Two measures of spending on education, 2002-2013 A key objective of the BOS program was to reduce the overall burden on households of Panel a: Total education spending, 2002-2013 sending their children to school. While Figure 2.2 provides a picture of how the burden of 450 25% education spending shifted between households and the government it does not assess the 23% 21% change in the overall share of a household’s budget that is required to send children to school. 400 19% 20% 20% Figure 2.3 shows that this share dropped a little with the introduction of the BOS program household spending as % 20% (constant 2012 prices) 19% 18% 19% 350 but for the poorer quintiles has begun to rise again. Only the wealthier quintiles have seen a 17% of total spending 300 15% IDR trillions 15% consistent drop in the overall share of spending needed to send their children to school. This 15% 250 reflects in part the faster consumption growth these households have experienced and the 200 growing levels of income inequality seen over the period (World Bank 2014a). 10% 150 100 Figure 2.3: Average share of total household consumption required to 5% 50 enroll all children in primary and junior secondary school 0 0% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 7 % of total household consumption Government spending (LHS) Household spending Household spending as % of total spending 6 Panel b: Household education spending per student, 2002-2012 5 Introduction Large increase in of BOS BOS amount 4 900,000 800,000 3 IDR (constant 2012 price) 700,000 2 600,000 1 500,000 400,000 0 300,000 poorest Q2 Q3 Q4 wealthiest 200,000 2003 2006 2009 2012 100,000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Note: The cost of sending all children to school is calculated by multiplying the average number of primary and Year junior secondary school aged children in a household by the average spending per student for each quintile. This average spending per student for all groups cost is then divided by average household consumption to estimate the share of total consumption needed to send average spending per student for the poorest group all children to primary and junior secondary school. Note: Susenas data are from the consumption module collected annually. Since the structure of the household Source: Susenas Education Module Data spending data for 2007 is different from other years, for comparability, it is omitted from the graph above. See Table 2.1 for items included in household spending. All amounts are in constant 2012 prices with differences in urban and rural prices taken into account to compute the national average. Source: 2012 revised plan budget laws, MoF and BPS for Susenas. 15 In this section the focus in on looking at average trends in household education spending and trends for the poorest 20 percent of households. Detailed results for the other quintiles are provided in the appendix to the report. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 40 41 Household spending at the primary school level Box 2.1: Sources of data for household education spending and Average household primary education spending per student increased rather than declined participation16 after the introduction of BOS. Figure 2.4 (Panel a) plots household education spending for the average primary school student between 2003 and 2012. A comparison of spending The analysis undertaken in this chapter draws on two sources of information on before (2003) and directly after the introduction of BOS (2006) reveals that household household education spending routinely collected as part of the Susenas household spending increased. This is contrary to expectations given that the introduction of BOS was survey: designed to reduce the burden on households of sending their children to school. 1. Education module information collected every three years. This provides However, the effect of the BOS program seems to have differed between government and the most detailed and accurate source of information on household education non-government students. Household spending for students attending government schools spending. In particular, the education module provides information on did decline a little after the introduction of BOS (see Panel b Figure 2.4). This perhaps reflects education spending separately for each child currently attending school. It also the stronger control the overall program had on government schools and in particular the collects information on all of the items of spending including spending on a greater importance that was placed on government schools to reduce the costs of education. variety of fees, transportation costs and pocket money. In general, information is collected for the school semester that ends just before the survey is conducted. Figure 2.4: Household spending at the primary school level 2. Consumption module information collected annually. This module collects Panel a: Overall spending at the primary level Panel b: Spending by type of school less detailed education spending information at the household level. Information 900,000 839,388 2,500,000 on education spending is only collected for the month prior to the survey which 800,000 2,204,412 IDR (2012 constant prices) means that many one-off payments made by households are not captured. A IDR (2012 constant prices) 700,000 630,827 2,000,000 576,870 593,516 further drawback of this source is the inability to disaggregate information for 600,000 1,496,638 1,500,000 individual students or for individual levels of education. 500,000 1,190,071 400,000 450,973 502,574 1,000,000 852,977 716,579 Given the differences in data collection it is perhaps not surprising that there are 600,000 361,220 346,861 530,323 525,115 548,723 differences in overall levels and trends of education spending between the two sources. 500,000 500,000 400,000 While the analysis conducted for the report has used both sources, the chapter 0 0 predominantly reports results from the education module data as this provides the 2003 2006 2009 2012 2003 2006 2009 2012 most accurate picture of household education spending over time. Results from the all households non government school students annual consumption module data are provided in an accompanying annex to the poorest 20% of households government school students report. Note: All amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between urban and Sections 2.3 and 2.4 of the chapter explore trends in education participation. rural areas. Education spending is the sum of all components including transportation expenses as shown in Table Information from the annual March round of the Susenas household survey is used 1 under the heading education module. Only pocket money is excluded. Source: Education Module Susenas for this analysis. After the program was introduced, average household spending on primary education has increased constantly. Average household primary student spending was 41 percent higher in As noted in Box 2.1, annual data on household education spending (such as that reported in 2012 compared to 2006. This is in striking contrast to the size of the BOS grant which also Figure 2.2) is less accurate and cannot easily be disaggregated by different levels of schooling. grew 41 percent over the same period. This suggests that on average schools did not reduce For this reason the remainder of the chapter focuses on analysis that uses the three-yearly the amount they required households to pay but in fact increased it despite increases in the education module information. This source has the advantage of being very detailed in terms amount of BOS funds they were receiving. A closer look at the data (Panel b) reveals that of composition of spending but has the limitation of not being available for every year since spending for both public and private schools also increased over the period. On average, BOS was introduced. As a consequence, it is sometimes difficult to pick up the immediate households report spending 36 percent more since 2006 on public primary school while effect of changes in the levels of BOS funding on household education spending. These they report spending almost 85 percent more since 2006 on private primary school.18 difficulties do not arise for the analysis on education participation because data from the When spending is disaggregated by wealth quintiles, there is evidence to suggest that those annual household survey is used to explore the effect of BOS.17 in the poorest quintile benefitted from the introduction of the BOS program. The average spending by a household in the lowest quintile in 2006 was 3.9 percent lower than in 2003 16 (Panel a Figure 2.4). This is small compared to the BOS grant amount and the drop was Appendix Table 1 provides more information on differences in education spending information between the only temporary. education and consumption module described in this box. 17 18 Analysis was undertaken on the consumption module data as well. The findings were broadly similar and the These differences are driven in part by differences in the growth of fees. Between 2006 and 2012 fees in results are available on request. government schools increased by 28 percent and non-government schools by 108 percent. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 42 43 Between 2006 and 2009, household spending increased by around 30 percent and at a time In a similar way to the analysis of primary school spending, reductions in the overall costs when the BOS grant was growing.19 Spending continued to grow between 2009 and 2012 faced by households were limited to children attending government schools (Panel b Figure albeit at a slower pace (11 percent). 2.5). Household spending fell by 7 percent after the introduction of BOS. While this was still a relatively small amount compared to the size of the BOS per-student grant it was Household spending at the junior secondary school level much greater than the overall average suggesting that BOS had a more significant role in government junior secondary schools. The BOS per-student grant amount increased by 25 In contrast to primary, average levels of spending on junior secondary school students percent between 2006 and 2009 and household spending continued to fall for students declined after the introduction of the BOS program (Figure 2.5). In particular the average attending government schools. However, by 2012 much of the reduction in household household went from spending IDR 1.48 million to spending about IDR 1.44 million – a spending seemed to have been washed away even though spending in non-government reduction of approximately 3 percent. Household spending declined further between 2006 schools was still slightly below its level in 2003. and 2009 at a time when overall levels of BOS funding were also increasing (see Figure 1.4). However, spending began to rise after 2009 even though there were further increases in the level of BOS. These increases were large enough to bring spending back to pre-BOS levels. Household spending at the senior secondary school level As mentioned earlier, there are no clear means by which to identify a causal impact of the While overall levels of household education spending declined the reductions represented BOS program. However, spending by households not covered by the BOS program may only a small fraction of the per-student amount schools received under the BOS program. provide a reasonable proxy for what might have happen had there been no BOS program For example, between 2003 and 2006 average household spending declined by an amount (a counterfactual). This is presented for all senior secondary schools (both general and equivalent to less than 10 percent of the overall BOS per-student grant. vocational) in Figure 2.6. Figure 2.5: Household spending at the junior secondary school level Figure 2.6: Household spending at the senior secondary school level Panel a: Overall spending at the junior secondary level Panel b: Spending by type of school Panel a: Overall spending at the senior secondary level Panel b: Spending by type of school 1,600,000 1,479,773 1,482,058 2,500,000 1,437,703 3,268,573 2,083,833 3,500,000 3,500,000 IDR (2012 constant prices) IDR (2012 constant prices) 3,146,096 1,400,000 1,291,241 2,929,287 2,000,000 1,704,751 3,000,000 2,863,866 3,000,000 2,804,473 IDR (2012 constant prices) IDR (2012 constant prices) 1,200,000 1,590,260 1,577,689 2,649,621 2,588,040 2,683,555 1,000,000 1,500,000 2,500,000 2,500,000 2,678,081 961,742 963,123 980,899 2,486,297 2,484,119 2,453,077 800,000 858,113 1,440,409 1,337,808 1,355,800 2,000,000 2,000,000 1,000,000 1,222,169 600,000 1,935,616 1,853,501 1,500,000 1,678,158 1,500,000 400,000 1,504,997 500,000 1,000,000 1,000,000 200,000 500,000 500,000 0 0 2003 2006 2009 2012 2003 2006 2009 2012 0 0 all households non government school students 2003 2006 2009 2012 2003 2006 2009 2012 poorest 20% of households government school students all households non government school students Note: All amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between urban and rural poorest 20% of households government school students areas. Education spending is the sum of all components shown in Appendix Table 1 under the heading education Note: All amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between urban and rural module. Pocket money is excluded. areas. Education spending is the sum of all components shown in Table 1 under the heading education module. Source: Education Module Susenas Only pocket money is excluded. Source: Education Module Susenas Households in the poorest quintile with children in junior secondary school fared somewhat better than average households – particularly at the outset. Between 2003 and 2006, Spending by households with children in senior secondary schools increased marginally spending per student for this group fell from IDR 961,742 to IDR 858,113 – a decline of between 2003 and 2012 (Figure 2.6). For households in the poorest quintile there was a about 11 percent. However, this reduction did not persist. By 2009 household spending steady increase between 2003 and 2009 followed by a slight fall in 2012. Spending per among the lowest quintile was 12 percent higher than in 2006 and by 2012 it was 14 student at the senior secondary school level for households in the poorest quintile was 11 percent higher than the first year of the BOS program.20 These results are striking given that percent higher in 2006 than in 2003. By 2012, spending at this level was 23 percent higher household spending for poor households was increasing at the same time that BOS grants than it had been in 2003.21 were being adjusted upwards. Comparing spending trends between basic education and senior secondary education does 19 The increase in spending by the poorest households was largely driven by increased spending on uniform not reveal a significant effect of BOS at least over the long term. Recall that patterns of which declined subsequently. See Figure 2.7 for further details. household spending at the senior secondary level are analyzed under the assumption that 20 The increase in spending by the poorest households was largely driven by increased spending on uniform and 21 transport. While spending on uniforms in subsequent years declined, transport costs continued to rise. See As with other levels of education, households whose students are in public or private schools seem to follow Figure 2.7 for further details. similar patterns of increased spending in the period under analysis. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 44 45 this reflects what would likely have happened at all levels of education had there been no pattern. In both cases however, there was a substantial increase in the proportion of spending BOS program. Directly after the introduction of BOS household education spending rose devoted to transport. While largely unchanged from 2003 to 2006 – the amount of spending significantly in senior secondary schools while spending in primary and junior secondary reported on transport increased six-fold between 2006 and 2012 and became equal roughly generally fell. However, at the primary level, overall household education spending increased to the amount households reported spending on fees.24 Thus it appears that any reduction in by approximately 46 percent between 2003 and 2012 and was much higher than the overall fees was more than washed away by the increases in transport costs households report facing. increase in senior secondary education. In junior secondary school, spending followed a Trends in fee payments at junior secondary schools were similar to those faced by primary similar trend to senior secondary spending. This comparison suggests that overall household school students. In junior secondary schools, the fees charged to parents also fell with spending in primary and junior secondary education rose at a similar or even faster rate than the introduction of BOS but began to rise again after 2009. For junior secondary school at senior secondary between 2003 and 2012 which suggests that BOS had a limited effect fees also represented about 50 percent of household spending before the start of the BOS over the longer term. program. This fell to 45 percent by 2006 and then again to 35 percent of the total educational However, changes in household education spending immediately after the introduction spending of the typical household. Overall between 2003 and 2006, fees declined by 12 of the BOS program reveals a different pattern. In general the introduction of the BOS percent and between 2006 and 2009 they fell another 25 percent. Thus in the case of junior program was associated with a fall in household education spending in primary and junior secondary school students spending on fees remained below pre-BOS levels. Household secondary schools (see Figures 2.5 and 2.6). These drops stand in contrast to the increase spending among those in the poorest quintile followed a similar pattern. However, as in the in household senior secondary education spending in Figure 2.6. This suggests that in the case of primary, this reduction in spending on fees was accompanied by large increases in short-term the trend in spending in primary and junior secondary schools was different than transportation costs which resulted in overall higher levels of total spending by 2012. in senior secondary school and may be associated with a short-term effect of BOS. The trends in data shown in this section show some small reductions in spending Figure 2.7: Composition of per-student Spending (IDR thousands), immediately after the start of the BOS program. These reductions are most pronounced education module for junior secondary school children in the poorest quintile though there is some indication Primary School Students of reductions for primary school children in the poorest quintile as well. Overall, any All Students Students in poorest 20 percent of households reductions we observe are much smaller than the amount of the BOS grant suggesting that 100% 100% 53 58 15 18 49 schools are using most of the grant amount in ways other than to lower household spending 62 70 136 316 50 54 160 172 on education. 80% 80% 157 118 147 60% 181 237 57 60% 85 42 The composition of spending 40% 131 143 40% 91 179 158 Thus far, the analysis has looked at trends in overall spending. We turn next to an analysis of 20% 281 313 20% 128 197 229 how the composition of spending has changed over time. Figure 2.7 looks at the amounts 207 spent by households on four main categories of education expenses – fees, uniforms, 0% 0% materials and transport.22 2003 2006 2009 2012 2003 2006 2009 2012 In primary schools, the fees faced by parents in sending their children to school dropped Junior Secondary School Students when BOS was introduced and only began to rise again after 2009. The main channel All Students Students in poorest 20 percent of households through which BOS should have operated is through a reduction in the amount of fees 100% 100% households report needing to pay. In the case of primary school students, these declined 335 302 418 540 176 154 273 371 between 2003 and 2006 and again between 2006 and 2009 but increased substantially in 80% 80% 104 106 80 83 2012 – to levels above those seen in 2003.23 It is important to note that this reduction in fees 211 60% 320 397 207 81 60% 251 59 charged by schools only occurred in government schools (see Appendix Figure 1). 202 206 133 40% 192 40% 132 The composition of primary education spending has changed significantly with fees making 721 up a smaller share and transportation a larger share of overall costs. In 2003, fees accounted 20% 633 475 659 20% 500 370 418 347 for about 50 percent of education spending by households. This declined to 40 percent in 0% 0% 2006 and a little over 30 percent in 2009 but then rose back to almost 40 percent in 2012. 2003 2006 2009 2012 2003 2006 2009 2012 The changes in spending composition for the poorest quintile followed this same general 22 This section does not include the comparison of government and non-government schools but these are fees materials uniform transport provided in Appendix Figure 1. 24 Transport spending increased in part because of the rising price of fuel that resulted from periodic adjustments 23 The drop in material spending after 2009 is largely due to drops in textbook and teaching supply costs which in the fuel subsidy and general increases in world prices. Average spending on transport also increased because may in part be a reflection that these costs were covered by other school funding including BOS. the share of students walking to school declined. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 46 47 Senior Secondary School Students Figure 2.8: Students reporting fee payments All Students Students in poorest 20 percent of households 100% 100% 100% 90% 379 780 739 358 559 535 767 660 80% 80% 114 80% 86 93 97 55 130 60% 533 261 276 60% 340 279 177 70% 454 255 259 154 40% 40% 60% 1293 883 1086 1299 1557 1582 774 944 50% 20% 20% 40% 0% 0% 2003 2006 2009 2012 2003 2006 2009 2012 30% 20% fees materials uniform transport 10% Note: All amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between urban and rural areas. 0% primary junior secondary general SS vocational SS Source: Education Module. Susenas. While fee payments declined in 2006 in schools receiving the new BOS grants they continued 2003 2006 2009 2012 to rise in senior secondary schools that were not part of the program. Again, analysis of Source: Education Module, Susenas senior secondary school spending provides an opportunity to assess if the patterns in school levels covered by BOS differed at all from other levels of the education system. This is not In principle, the BOS program could have reduced household spending on education through an ideal comparison group against which to evaluate the BOS program but it is the only direct financial support to the poorest households. For example, in the beginning, schools group available that is not eligible for BOS grants. This type of comparison is less conclusive could have used the grant to pay for the transportation costs of poor students. However, when looking beyond 2006. Trends in fees after 2006 were similar across different school regional independent monitoring shows that only a limited amount of the funds were used levels which suggest that if there was an effect of the BOS program on fees and charges at the for this purpose (see Figure 3.7). This is corroborated by the slight initial reductions in primary and junior secondary school it was relatively weak. transport costs for junior secondary school students in Figure 2.7. Over the period 2003-2012, the incidence of fee payment has also been changing. Figure 2.8 shows the proportion of students reporting the payment of fees in each round for which 2.3 Trends in school participation26 we have data. It shows that the introduction of BOS in 2005 coincided with a big drop in the number of students reporting fee payments.25 These drops are largest for primary A key objective of the BOS program was to raise participation in basic education particularly school followed by junior secondary school. The incidence of fee payment increased again for the poorest households (see Figure 2.1). Figure 2.9 assesses whether net enrollment rates in 2009 before falling again in 2012. The fact that the percentage of students reporting fee changed when BOS was introduced.27 The figure indicates the school year in which BOS payments is substantially smaller in 2012 than in 2009 coupled with the fact that average was introduced as well as subsequent years when its value was raised. The expectation would amounts of fees reported are higher in 2012 would suggest that those who do pay fees are be that these years would result in an increase in primary and junior secondary enrollment paying substantially more than they did in 2009. It is possible that that the decline in the rates. In contrast, it would be expected that a different pattern would emerge for senior percentage of households that reported paying fees in 2012 is a result of regulations issued secondary schools given that there was no BOS for schools at this level. around that time which attempted to outline the contributions schools were allowed to seek from parents and contributions that were prohibited (see Box 4.1). This may have resulted in Enrollment rates some schools dropping fee charges altogether while other schools increased the contributions There does not appear to be a strong link between BOS and primary school enrollment they sought from allowable charges. rates (Figure 2.9). This is perhaps understandable since the previous section established that 26 Splits showing changes in participation between government and non-government schools are not included in this section. Annual data on enrolment rates disaggregated by government and non-government schools are not available from the annual household survey. While the education module data includes information on the type of school attended this is only for a sample of children and it is difficult to match up with the overall trends. 27 Early enrolment in Indonesia is common so primary net enrolment rates can provide a slightly distorted view 25 Between 2003 and 2006 average fees (for those paying and not paying- including zeroes) dropped but given on access. A version of Figure 2.9 using adjusted net enrollment rates (the percentage of 7-12 year olds that are that far fewer households were being charged fees it seems as though the fee a student was paying increased. in school) shows virtually identical patterns. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 48 49 household spending on schooling did not decline much as a result of BOS. Furthermore, or by as much in the case of senior secondary. Thus, it is possible that BOS played a role in primary school net enrollment rates had little room for improvement as they were very close raising junior secondary school enrollment rates for the poorest quintile of children. to 100 percent throughout the period. While school grants were originally introduced in targeted schools after the wake of the Transition rates 1997 Asian financial crisis the national roll-out of the BOS program did not coincide with a As an alternative to enrollment rates, crude transition rates between primary and junior negative shock of this kind. In fact, in 2005 when the BOS program was being introduced, secondary school were also examined. These transition rates measure the percentage of 12 economic growth was averaging 5.7 percent and the proportion of the population that fell and 13 year olds who complete the sixth grade of primary school and report enrolling in the below the poverty line stood at 16 percent down from 18 percent in 2002. Thus while it is seventh grade (i.e. starting junior secondary school).28 It is possible that these transition rates possible that BOS prevented school participation from declining at a time when household provide a better measure of the effect of BOS because they focus on a key decision point that consumption was rising we are unable to test this assertion. BOS aimed to affect. Put another way, if BOS improved education participation, changes in Figure 2.9: Primary, junior and senior secondary net enrolment rates, these transition rates would likely pick it up more clearly and quickly than overall measures 2000-2013 of participation such as net enrolment rates. Introduction of BOS Increase in BOS amount Improvements in transition rates between primary and junior secondary school do not 100 100 appear to be strongly associated with the introduction of BOS or changes in the program 90 90 thereafter. Figure 2.10 shows that between the 2002/03 and 2005/06 school years there was a steady improvement in transition rates. The introduction of BOS in 2005/06 does 80 80 not appear to have changed this upward trend significantly. Transition rates seem to have Net enrolment rate (%) 70 70 dropped after 2005/06 and only really began to improve after 2010/11. Transition rates for 60 60 poor households followed a similar trend although they tended to improve more consistently after the introduction of the BOS program than the overall rate. 50 50 40 40 Figure 2.10: Crude transition rates of children aged 12 and 13 years old 30 30 into junior secondary school 20 20 Panel a: Transition rates to junior secondary 10 10 80% 0 0 64% 66% 66% 68% 70% 62% 64% 63% 64% 63% 64% 64% 1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 59% 69% 68% 60% 65% 63% 64% 65% 60% 59% 62% Year 50% 54% 56% Primary: national average Junior: national average Senior: national average 51% 40% Primary: poorest 20% Junior: poorest 20% Senior: poorest 20% Note: The horizontal access refers to the school year. BOS was introduced in the second half of 2005. 30% Source: Susenas, 2000-2013 20% In contrast, there is evidence that the expansion of junior secondary school enrolment rates 10% for the poorest children accelerated after the introduction of the BOS program. This is made clear by the fact that there is a greater than 10 percentage point gap between the poorest 0% quintile and the average before the introduction of BOS – a gap that does not appear to 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 close between the 1999/00 and 2005/06 school years. After the introduction of BOS there is a steady decline in the gap and by 2012/13 the junior secondary school enrolment rate for children in the poorest quintile is within 5 percentage points of the national average. It All children children in poorest 20% of households is possible that these increases in enrolment rates are linked to the declining share of fees documented in the previous section. Trends in senior secondary school enrolment tend to reinforce the view that poorer children 28 in junior secondary schools may have benefitted from the BOS program. Enrollment rates The transition rates used in this section are not the same as those typically measured with administrative data which divide the number of new enrollees in the first grade of junior secondary by the number completing the increased across the income distribution but the poorest quintile did not catch up as rapidly final grade of primary in the previous year. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 50 51 Panel b: Transition rates to senior secondary Such issues notwithstanding, this section tries to isolate any effect of BOS by using a simple regression analysis to control for other factors that may have driven the changes in spending 80% and participation. For example, improvements in the income of households, changes in the 70% demographic structure and changes in the overall supply of education at the district level 59% could have influenced the observed trends. These observable factors are controlled for by 60% 54% 55% 55% 56% 56% 51% 50% 50% 53% 53% using simple OLS regression models of the following form: 49% Yij=n btI[ T=t] + nbKXij +Xij 50% 56% 2012 K 53% 50% 40% 46% 47% 47% (1) 44% 45% 42% 44% t=2003 k=1 30% 34% 20% 27% These regressions use data from the education module. Yij denotes household spending on education for child i in schooling level j measured in constant 2012 prices adjusting for 10% differences between urban and rural areas. The right hand side of the regression includes 0% binary variables which equal 1 if the year is 2003, 2006, 2009 or 2012. The time dummies capture the average amount of household spending in each year. We interpret the coefficient 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 on the 2003 dummy as the amount of household spending in a pre-program year and the coefficients on 2006 and subsequent years as the amount of household spending in post- All children children in poorest 20% of households program years. A series of control variables which include household size and composition, mother’s education, consumption per capita30, an indicator variable for whether or not the Note: Crude transition rates from primary to junior secondary are calculated from the number of child receives a scholarship as well as dummy variables for province allows us to control children enrolled in grade 7 of junior secondary education divided by the number of 12 and 13 year old children who completed grade 6 of primary education. Crude transition rates from junior to for the influence of observable characteristics on spending. Regressions are run for the full senior secondary are calculated from the number of children enrolled in grade 9 of junior secondary sample as well as for each quintile.31 education divided by the number of 15 and 16 year old children who completed grade 10 of junior The main text reports the predicted spending using the regression coefficients and sample secondary education. averages of characteristics between 2003 and 2012. The predicted spending provides Source: Susenas, 2002-2013 an indicator of how household spending would have changed over the period if other Transition rates from junior to senior secondary that would perhaps be less influenced by observable factors (e.g. household income and educational support) had remained constant. the BOS program show a similar upward trend.29 Between 2001/02 and 2004/05 transition This predicted spending indicator gives a clearer picture of the potential effect that BOS may rates to senior secondary went from 51 percent to 54 percent (an increase of 5.8 percent). have had on household education spending. For example, if predicted spending dropped Between 2004/05 and 2009/10 the transition rate did not change much but from then after the introduction of the BOS program this drop would not have been driven by changes increased to 59 percent by 2012/13 (an increase of 11.2 percent). A comparison of these in the characteristics controlled for in the regression analysis. This would provide a slightly trends with those seen for the primary-to-junior school transition rate does not reveal major more refined view of the effect of BOS than the trends reported in Section 2.2. differences. This suggests that if BOS was a factor in improving transition rates between Similarly, for enrollment and transition rates, a probit model of the following form is primary and junior secondary school it was relatively weak. estimated: Pr(Yij= 1) = n btI[T=t] + nbKXij + hij 2013 K 2.4 Isolating the effect of BOS on household education (2) spending and participation t=2002 k=1 The previous sections have documented trends in spending and participation before and These regressions use data from the consumption module (see Box 2.1) and the dependent after the BOS program. As noted earlier, because BOS was a national program there are variable in the case of enrollment is a binary variable that takes the value of 1 if the child is no obvious comparison groups available to allow for a causal assessment of the impact of enrolled at primary or junior secondary level and is zero otherwise. In the case of transition BOS. A variety of factors will have impacted both household spending and participation rates it is a binary variable taking the value of 1 if the child transitioned from one level to decisions. These include but are not limited to economic growth and sustained poverty the next and zero otherwise. The coefficients on a series of indicator variables for all the years reduction, economic and climatic shocks and other policy initiatives – most importantly the from 2002 to 2013 are the main items of interest. Other controls are as previously specified. introduction of scholarship programs and increased government spending on education as a whole. 30 Spending on education is excluded from consumption per capita since this is captured in the dependent 29 variable. Figure 2.10 also shows transition rates from junior secondary school to senior secondary school – specifically 31 the percentage of 15 and 16 year old children who report completing 9th grade and report enrolling in 10th Single year regression estimates were calculated to see if estimated coefficients were different to the pooled grade (i.e. starting senior secondary school). estimates but this did not affect the results. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 52 53 These regressions are also run for the full sample and for each quintile. In the interests of Figure 2.11: Trends in household education spending per student brevity – only the results of the average and lowest quintile are reported in the chapter.32 controlling for household characteristics35 Regular Susenas data are used from early 2002 to 201333 to investigate the changes in the Primary school students enrolment rates before and after the policy implementation. The estimation covers children Raw trends in spending per student Regression-adjusted trends in spending per student within the official age groups because according to 2005 BOS guideline, the program is given to individuals between 7-15 years old attending schools so they would receive educational 900,000 839,388 1,000,000 875,929 services and complete 9 years of basic education.34 800,000 900,000 IDR (2012 constant prices) 768,811 754,243 761,522 700,000 630,827 800,000 576,870 593,516 600,000 700,000 a. The effects of BOS on household education spending 500,000 600,000 502,574 500,000 591,557 400,000 535,624 The results of the regression analysis do not alter significantly the findings from the trend 450,973 400,000 300,000 361,220 446,147 428,675 analysis reported in Section 2.2. Figure 2.11 summarizes the results of regression (1) for 346,861 300,000 200,000 200,000 each level of education and two groups – the average household and the poorest quintile. 100,000 100,000 By and large, the trends in household spending observed earlier do not change much when 0 0 other characteristics that may have had an influence on household education spending are 2003 2006 2009 2012 2003 2006 2009 2012 controlled for. Junior secondary school students Controlling for other factors shows that household education spending in primary and Raw trends in spending per student Regression-adjusted trends in spending per student junior secondary schools dropped after the introduction of BOS. Figure 2.11 shows that 2,000,000 1,722,851 household education spending, both on average and for the poorest households, fell after the 1,800,000 1,638,667 1,586,812 introduction of the BOS program. Given that the predicted spending trends show this after 1,600,000 1,479,773 1,437,703 1,482,058 1,600,000 1,476,911 IDR (2012 constant prices) controlling for other factors that may have influenced spending (e.g. household consumption) 1,400,000 1,291,241 1,400,000 1,200,000 1,200,000 provides further evidence of an effect of BOS on household spending. However, these drops 1,000,000 1,000,000 1,123,461 are relatively short lived. In primary, household education spending begins to rise soon after 961,742 963,123 980,899 1,022,148 1,038,344 800,000 800,000 889,364 the introduction of BOS. A similar trend is also seen for households with children in junior 858,113 600,000 600,000 secondary although spending only begins to rise after 2009. 400,000 400,000 The effects of the BOS program on reducing household education spending were strongest 200,000 200,000 in government schools. Regression analysis confirms the trends seen in Figures 2.4 and 2.5 0 2003 2006 2009 2012 0 2003 2006 2009 2012 (see Appendix Figure 2). Senior secondary school students Raw trends in spending per student Regression-adjusted trends in spending per student 3,500,000 3,500,000 3,173,790 2,969,390 2,889,218 2,925,769 2,863,866 3,000,000 3,000,000 IDR (2012 constant prices) 2,649,621 2,588,040 2,683,555 2,500,000 2,500,000 2,000,000 2,000,000 2,080,840 2,089,666 1,935,616 1,853,501 1,500,000 1,678,158 1,500,000 1,788,317 1,612,819 1,504,997 1,000,000 1,000,000 500,000 500,000 0 0 2003 2006 2009 2012 2003 2006 2009 2012 all households poorest 20% of households Notes: The left hand side of this table shows raw data. The right hand side shows the results of a regression-based 32 analysis. The dependent variable in the regression model is per-student spending. Education spending is calculated The full results are provided in the appendix to the report. from information on household spending for each child attending school in at the corresponding school level using 33 fees, uniforms, materials and transport (as described in Figures 3-5). Regression controls for a set of household Since regular Susenas data used here are from March rounds, following the analysis on education spending from annual data later in the analysis we also adjusted the year where BOS was commenced. So, 2006 March characteristics including household size and composition, number of enrolled children, mother’s education, age, round would be marked as 2005 because it belongs to 2005/2006 school year. 34 35 However, those who are outside such a cohort might also attend schools given that early age enrolment is Appendix Figure 2 provides the same information but disaggregates by government and non-government common in Indonesia. In analyses not reported here, we broaden the cohort using different age ranges to schools. Regression adjusted trends in spending per-student for the poorest quintile in government schools confirm these findings are unchanged. were also estimated. These follow similar trends in primary and junior secondary to those shown in Figure 2.11 Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 54 55 household per capita consumption (excluding spending on education), whether or not a child in the household The regression-based enrollment trends at the junior secondary level are also in line with received a scholarship and dummy variables for province. Predicted spending using sample averages are shown. All the raw trends seen earlier (Figure 2.12). There was a fairly consistent gap between the amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between urban and rural areas. Source: Susenas Education Modules lowest quintile and the average household between 2002 and 2005. Immediately after the introduction of the BOS program, enrollment rates for the poorest quintile jump up In a similar way to the raw trends reported in Section 2.2, spending on senior secondary by 5 percentage points. Over time the gap between the bottom quintile and the average school increased at the time the BOS program was introduced. This stands in contrast to the diminished and stood at 8 percent points in 2013 – substantially smaller than the 17 results in primary and junior secondary education and provides further weight to the initial percentage point gap seen in 2005. positive effect that BOS had on the costs faced by households. Transition Rates b. The effect of BOS on enrollment and transition rates Figure 2.13: Regression-adjusted trends in junior and senior secondary Enrollment Rates transition rates Figure 2.12 corroborates the broad trends described in Section 2.3 – primary enrollment Regression-adjusted trends in transition rates to junior secondary rates are high and do not seem to have been visibly influenced by the introduction of the 100% BOS program or by the increases in the size of the BOS amount. Given that the impact of 90% BOS on spending at this level was small and that enrollment rates were above 90 percent 80% even before the BOS program started, these findings seem plausible. 69% 67% 67% 70% 66% 63% 64% 64% 64% 65% 65% 67% Figure 2.12: Regression-adjusted trends in primary and junior 60% 60% 68% 64% 63% 63% 63% 63% 61% 65% 62% secondary net enrolment rates 50% 60% 57% 59% Regression-adjusted trends in primary net enrolment rates 40% 100% 96% 93% 94% 93% 94% 94% 93% 93% 93% 93% 94% 94% 30% 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 90% 94% 93% 93% 95% 92% 91% 92% 92% 92% 91% 93% 92% 80% All children children in poorest 20% of households 70% 60% Regression-adjusted trends in transition rates to senior secondary 100% 50% 40% 90% 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 80% 70% All children children in poorest 20% of households 60% 56% 56% 56% 60% 53% 51% 51% 53% 54% 56% 55% 49% Regression-adjusted trends in junior secondary net enrolment rates 50% 56% 51% 51% 40% 48% 48% 45% 47% 49% 100% 42% 41% 32% 38% 90% 30% 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 80% 68% 70% 70% 70% 70% 70% 66% 66% 67% 68% 62% 64% 63% All children children in poorest 20% of households 60% 61% 62% 62% Source: Susenas, 2002-2013 57% 59% 50% 54% 56% 57% 50% 48% 49% 49% Controlling for household characteristics suggests that there was a slight increase in 40% transition rates between 2002/03 and 2004/05 (Figure 2.13). The introduction of the BOS 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 program led to a slight increase 2005/06 but then transition rates fluctuated and ended at a slightly higher level in 2012/13. For the poorest quintile, transition rates increased between All children children in poorest 20% of households 2002/03 and 2004/05 but when BOS was introduced rates stayed relatively flat. Overall this Source: Susenas, 2002-2013 suggests little link between the introduction of BOS and the transition rates for children from primary to junior secondary. Given the small decreases in household spending seen in Section 2.2 this seems understandable. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation Assessing the Effects of BOS on Household Education Spending and Participation 56 57 It is hard to draw firm conclusions from a comparison between predicted junior and senior These findings support evidence presented in earlier sections that suggests that at best only secondary transition rates. Figure 2.13 also paints a corresponding picture for the transition a small proportion of BOS funding went to reduce the household burden of education rates between junior secondary and senior secondary. In contrast to junior secondary rates, spending. In contrast, levels of school funding appear to have improved significantly with transition rates to senior secondary fell in the year that BOS was introduced but have the introduction of the program. How effectively these resources are spent will depend since followed a similar path. It is not clear what may have driven the large fluctuations in crucially on how well schools are managed and operated. This is the subject of the next transition rates at senior secondary for the poorest quintile between 2002/03 and 2005/06. chapter. However, it is sobering to note that despite increased school level funding, levels of learning achievement, measured by the OECD PISA assessment, have remained relatively Taken together, the evidence on enrollment and transition rates tentatively suggests that unchanged (OECD 2013). the BOS program has been associated with an initial increase in junior secondary school enrolment rates for the poorest children. However, the evidence is not as clear cut as the trends outlined for household education spending seen in Section 2.3. 2.6 Summary It is important to recognize the limitations of the approach used in this chapter to assess 2.5 Effect of BOS on school level finances the effect that the BOS program may have had on household education spending and participation. In particular, given that BOS was a national program that all students The evidence that has been presented in this chapter suggests that the BOS program benefitted from, there are no good counterfactuals with which to compare recipients with. lowered household spending when it was introduced but that these reductions were small The chapter has made efforts to control for other factors that may influence trends in relative to the overall size of BOS grants. In turn, this would imply that the overall level of education spending and to use trends at the senior secondary level as a check on trends resources that schools had would have increased substantially with the introduction of the in earlier levels of education. However, it is recognized that these efforts do not allow the BOS program. Unfortunately, no detailed time series information on school funding in impact of the BOS program to be identified in a rigorous manner. Indonesia is available to confirm this conclusion. However, it is possible to explore trends in the number of school hired teachers which gives a glimpse at the status of school finances Notwithstanding these limitations the analysis outlined in the chapter has shown tentatively before and after BOS. that the introduction of the BOS program led to an actual decline in household education spending. These reductions appeared to be concentrated in government schools, tended to Data from the national registry of teachers (NUPTK) shows that the introduction of BOS be larger for poorer households and for junior secondary school students. However, the led to a sharp increase in the number of school-hired teachers. Figure 2.14 shows that size of the reductions was small relative to both the level of the BOS per-student grant and approximately 325 thousand teachers were hired by schools after the introduction of BOS. levels of overall household education spending. This stands in contrast to another study that Put another way, over half of all school hired teachers in primary and junior secondary looked at the effect of school grants on household education spending in India and Zambia. schools were hired after BOS was introduced. Evidence, presented in the next chapter, also The study showed that anticipated increases in school grants resulted in a similar decline in shows that a significant share of BOS grants went on hiring these teachers. household education spending (Das, Dercon et al. 2011). Figure 2.14: School hired teachers by year of hire Subsequent increases in the overall level of BOS funding did not appear to have a consistent effect on household spending and education spending in 2012 was significantly higher than 350,000 it was directly after the introduction of the BOS program. There are a number of reasons why schools may not have passed on more of the BOS grant 300,000 in the form of lower charges to households. First, it is possible that the increases in school 250,000 funding that BOS provided were not sufficient to cover the overall operating costs of schools no. of teachers and schools still needed to rely on parental contributions to cover the full costs. It is also 200,000 possible that schools did not receive the full amount of BOS funds and remained dependent on parent contributions. However, evidence on the implementation of the program suggests 150,000 that, on the whole, most schools received BOS funds in an accurate and timely manner (see Box 1.4). Second, other sources of funding may have dropped off with the introduction 100,000 of BOS. For example, a study on the predecessor to BOS found that local governments 50,000 reduced their own school funding which reduced the overall funding schools had (Ridao- Cano and Filmer 2004). This may have meant that schools had to continue to rely on 0 student contributions and charges to offset the reductions from other sources. It seems clear 1975-1984 1985-1994 1995-2004 2005-2013 that it is important to ensure that the BOS grant is truly additional to other school funding year of hire and that the overall level of the BOS grant, in combination with other sources, is sufficient to fully cover school operating expenses. The report returns to these topics in the concluding Source: NUPTK (2012), MoEC chapter. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Effects of BOS on Household Education Spending and Participation 58 Finally, the BOS program does not appear to have materially affected participation rates at the primary level but is associated with a jump in participation rates for poor households CHAPTER 3 at the junior secondary level. However, caution is warranted in this interpretation given the limited effect seen on the transition rate. Moreover, Indonesia has some way to go before junior secondary school participation and completion are universal and it is possible that BOS and School Based BOS could do more to support further improvements. Management 3.1 Introduction Decentralization of decision making to the school level or School Based Management (SBM) has been widely adopted across the world in an attempt to improve the teaching and learning conditions for children. The idea behind SBM is that by shifting decision making authority closer to the end users, who are better informed of their needs, the teaching and learning environment in schools may be improved (Caldwell 2005) . School based management is not a uniform or standard operational system and SBM programs vary in the level of authority devolved to schools and the how key decision making agents are held accountable. Another important dimension of school based management reforms is the resources that are made available to schools to implement the activities they have deemed necessary for improvement. The objective behind the introduction of SBM also varies as much as some of these key dimensions. In some countries such as Australia, the United Kingdom and the Netherlands, SBM was introduced to improve the efficiency of the education system. In others, including Canada, New Zealand and Czech Republic reforms were introduced to increase parental and community participation in school management and to democratize education. And a number of other countries (e.g. Thailand, El Salvador and Mexico) have identified improvements in the quality of teaching and learning as the major objective (Malen, Ogawa et al. 1990; World Bank 2007). Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 59 BOS and School Based Management BOS and School Based Management 60 61 Figure 3.1: Channels of BOS support to school based management literature appear to suggest that it still represents a recognizable picture of the current status of school based management. Increased state funding of 3.2 The contribution of school based management to improved schools and reduced burden education outcomes in Indonesia on households International literature suggests that the involvement of parents and community in SBM Increased education participation can lead to improvements in education quality and student learning. For instance, Gertler, and improved learning Patrinos et al. (2008) find that in Mexico, the provision of grants amounting to US$500- particularly for 700 and training to parent associations reduced repetition and failure rates by five percent the poorest although there appeared to be no effect on dropout rates. However, a doubling of the grant Introduction of Direct financial amount did reduce drop-out rates and also improved test scores in Spanish and mathematics School Grants support (Gertler, Patrinos et al. 2010). In the Philippines, Khattri, Ling et al. (2010) show that a (BOS) to poor students combination of grants and training offered to principals and head teachers resulted in a 1.5 • Greater school autonomy percentage point increase in test scores. • Greater parental participation in school Evidence in Indonesia also points to the potential of school based management reforms • Stronger school accountability to improve the quality of education. A recent quantitative study showed that good quality • Improved transparency of school based management led to better decisions at the school level which in turn improved Strengthened SBM school decision making through establishment student learning outcomes (Box 3.1). A mixed-methods evaluation of a USAID project of rules and responsibilities that supported the implementation of school based management in 50 Indonesian districts of schools and local communities also highlighted a number of important benefits. The evaluation found that the support for in managing BOS. school based management had resulted in high quality and relevant school development plans. These plans supported additional professional development activities for teachers and better learning resources such as supplementary readers and teaching aids. The evaluation found that through the project’s support, schools exhibited more ‘transparent, participatory In Indonesia, the formal introduction of school based management occurred at around the and responsive management practices’ and this demonstrated that school based management same time as far broader decentralization reforms were being introduced. In 2003, a new could be successfully implemented in Indonesia (Heyward, Cannon et al. 2011). education law outlined the new roles and responsibilities for different levels of government and also gave schools significant autonomy over their own affairs. The same law mandated the establishment of school committees as a key mechanism to both hold schools accountable but to also foster greater participation of parents and the local community in supporting their local schools. It was not until the introduction of the BOS program that schools received the resource boost necessary to execute their new roles and responsibilities. The program not only provided needed funding but also stipulated that these funds had to be managed using the newly created school based management structures (e.g. school committees) and processes (e.g. school improvement planning etc.). In this way the BOS program provided an important incentive to schools to adopt school based management. It also underpinned key government objectives to raise education participation rates and improve quality (Figure 3.1). This chapter reviews the implementation and outcomes of the school based management approach in Indonesia and in particular how it has influenced and been influenced by the BOS program. The specific focus is on the utilization of funds by the school based agents to achieve the objectives of BOS. In order to do this the chapter reviews the existing literature and analyzes data from a survey of schools that included a comprehensive survey of 10 key stakeholders in the SBM process. While the information in the survey is from 2010, it is invaluable in providing critical and detailed information on the working of the SBM process. Moreover, anecdotal evidence from recent school visits and reviews of other recent Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 62 63 Box 3.1: The role of school based management in improving Box 3.2: The 2003 education law and school based management school decision making and student learning outcomes in School-based management was enacted in Indonesia with the introduction of an Indonesia education law in 2003. This was a part of the broader decentralization efforts in A recent study explored the factors that drove better school level decision making the country and came in effect with decentralization of certain other responsibilities in primary schools across Indonesia. The study focused on two key decisions that to the districts including public works and health. The 2003 Education Law also schools had the most control over; the hiring of non-civil service teachers and the established national content and competency standards and learning assessments. allocation of discretionary resources. The law states that “The management of pre-school, primary and secondary education units shall be based on a minimum service standard by applying the principles of The results of the study provide some evidence of the importance of school based school/madrasah-based management”. management on the quality of decision making and ultimately student learning outcomes. In particular, the study found that schools that provided better information The Education Law (Act 20/2003) enacts a “moderately strong” system of school on the use of BOS funds tended to spend more of their resources directly on activities based management as in essence, the law transfers a high level of authority and to support students rather than on other items such as infrastructure. The study also autonomy to schools and communities. The law gives communities the right to found that the characteristics of the school committee chair were important factors provide community-based education in accordance with the religious, social and in better decision making. For example, a greater share of discretionary resources was cultural norms of the community. It also devolved the design and implementation of allocated to student learning activities in schools where the school committee chair curriculum and management of education programs and funds to the school level. had senior secondary education or above. Moreover, higher spending on student The hiring and firing of civil service teachers however, remains within the central activities was associated with statistically significant improvements in student government’s authority. mathematics scores. School Committees are mandated by the education law as autonomous bodies to The study also highlighted the importance of an effective district in supporting better ensure community participation in school management (established under Decree school decision making. For example, the share of resources devoted directly to of the Minister of National Education No. 044/U/2002 on Educational Board student activities tended to be higher in districts that were judged to play a very and School Committee). The school committees are given an advisory role and a influential role in budget allocation decisions. monitoring role for school management as well as an oversight role for financial matters at the school level (see Table 3.1). The study’s authors conclude: In 2005, the general standards for the application of SBM in schools were issued. ‘..although the scope of school based management in Indonesia is limited today, it These standards directed schools to: has begun to help schools make the right decisions on allocation of resources and hiring additional teachers, and to create an enabling environment for learning.’ 1. formulate a school vision, mission and goals with input from all stakeholders and decided by a school board meeting; Source: Chen (2011). 2. develop a four year mid-term plan that focuses on quality of learning by students and the enhancements to be made in the educational programs to improve quality; 3.3 BOS and school based management 3. develop an annual plan covering managerial issues such as student affairs, The formal introduction of school based management in Indonesia occurred at the same teachers and their professional development plans, learning activities and finance time as the devolution of responsibility for the delivery of pre-tertiary education services to and infrastructure among other issues; local district governments. The education law of 2003 devolved the responsibility for school management to primary and junior secondary schools (see Box 3.2). 4. engage in self-evaluation on the quality of education. While the necessary policies and regulations formally introducing school based management were developed, only a small proportion of funding fell under the control of schools. The Source: World Bank (2012e) introduction of the BOS program significantly increased the resources schools had to BOS makes up the bulk of education resources that schools have control over. While many implement their own improvement plans. BOS funding came with relatively few conditions resources, such as teachers and equipment, are provided directly to schools by local and attached as long as school based management structures and processes were used in its central authorities, the BOS program gave money directly to schools to manage. In 2010, management. For example, schools must have a medium-term school improvement plan BOS made up the vast majority of the total resources that primary and junior secondary and a related annual activity plan and budget in order to fulfil BOS management conditions. schools had available to them (Figure 3.2). While there is evidence that schools may understate the overall level of support they receive from parents and their communities, Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 64 65 it is clear that BOS represents an extremely important source of funding for schools. This is likely to be particularly true for poorer communities where the level of financial support schools can expect from parents is likely to be much less. Box 3.3: Connections between BOS funds and school based It is also important to recognize that the BOS grant provides a predictable and reliable management source of funding that schools can use to plan improvements over the medium term. This BOS funds have been a key source of implementation funds for school based stands in contrast to the fluctuations in parental contributions and the unpredictability of management in Indonesia. BOS operational manuals describe an elaborate hierarchy the government budget cycle (World Bank 2013d). of management and oversight of BOS funds from national level, all the way to the individual school levels, including steering committees at national provincial and Figure 3.2: BOS as a proportion of total school-level funding, 2010 district levels as well as BOS management teams at all three levels consisting of an executing person at every level supported by an implementing team. Detailed roles 100% and responsibilities of the implementing teams are also spelled out in the operational 83% manual. Based on the description of responsibilities, the most critical functions that 80% 72% connect the BOS funding to SBM, as mandated by the Education Law, 2003 are assigned to the district/city level and to the school level BOS management team. 60% The district level BOS implementing teams are responsible to inter alia, carry out socialization and school level trainings, plan and carry out monitoring and evaluation, 40% provide public services and handle public complaints. These activities are critical to ensure that BOS funds are utilized by the school based agents as per the specific needs 20% of a school within the specified BOS allowable expenses. This also helps reduce the 9% 11% 10% 4% 3% 5% possibility of misallocation or misappropriation of funds. 0 At the school level, the BOS management teams are responsible for inter alia, BOS District Provincial Other BOS District Provincial Other grant grant grant grant grant grant announcing the amount of funds received and managed by the school and the BOS use plan on the school notice board, producing monthly reports on BOS fund Primary Junior Secondary expenditures and on goods purchased by schools. All reports have to be signed by the Principal, Treasurer and School Committee Chairperson. The schools are also responsible for cases of discrepancies in the use of funds at the school, in addition Note: Other include school fees, parents or community contributions, donations etc. to providing public services and handle public complaints. The executing person Source: SBM Survey, 2010 for the BOS funds is the school principal, as such BOS funds are withdrawn by While BOS provided the resources to support school based management reforms it is also Principals (or school BOS treasurers) with the knowledge of the School Committee important to recognize that the success of BOS itself depends critically on the effectiveness Chairman and this may be done from time to time as needed. The key aspect is of school based management systems. These systems are responsible for allocating school- that the actual use of BOS funds by schools should be based on mutual agreement level funding to activities that support the BOS objectives of improving school participation and decision among the School BOS Management Team, Board of Teachers and and levels of learning. It is expected that schools with transparent and accountable systems of School Committee. Therefore, there is a strong role played by school based actors in governance are more likely to make better decisions on the use of BOS resources. In contrast, ensuring appropriate use of BOS funds. the quality of decision making is likely to be weaker in schools where decisions are made by the school principal alone and information on the use of funds is unavailable. Similarly, the Source: BOS Manuals, various years. quality and education of key personnel within the system will also help achieve the end goals of SBM. For instance, Chen (2011) finds that schools whose school committee chairs are The school committee and the BOS team provide the direct links between BOS and better educated and more experienced have better teacher attendance rates. school based management reforms at the school level (Box 3.3 and Table 3.1). The BOS management team consists of the school principal, a treasurer (generally a teacher) and a parent who is not a member of the school committee. The recent school based management survey found that while about two-thirds of primary schools had a BOS team, parents were only represented in about a third of them (World Bank 2012e). This is in part due to the limited knowledge in local communities about the BOS team but also because the overall authority to determine the composition of this team lies with the school principal alone. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 66 67 School committees are intended to have an advisory as well as oversight role in supporting School the implementation of school educational policies and transparency and accountability of Roles and membership allocation of funds, budget plans and other school matters. In addition, they are intended body to motivate parents and community to pay attention to the quality of education. School School committee committees need to be composed of no less than 9 members including representation Role • General role in providing support and advice on school management (e.g. from parents, alumni and current students, community members, education professional policies, programs), ensure school is transparent and accountable and strengthen organizations and specialists. engagement with community While the members of school committees are supposed to be elected this happens relatively • Participate in producing and signing off school development plan (RPS) infrequently and in general member selection is opaque. A preparatory committee is tasked • Acknowledge they have seen annual school budget and specifically the BOS with organizing the initial selection of candidates to sit on the school committee and arranges budget local elections. However, in 2010, a survey of schools found that less than 15 percent of • Sign off on the list of poor students who are exempt from any charges and fees school committee members were directly elected. In practice, most selection was based on • Sign off on the list of books bought by schools and funded from the BOS book consensus or direct appointment by the principal (World Bank 2012e). fund • Sign off on a notification sheet and report of BOS fund utilization Table 3.1:The main roles and responsibilities for BOS at the school level • Sign off on withdrawal plan of BOS fund by principal • Sign off on quarterly reports on actual BOS fund use (BOS K-2) School • Independent institution consisting of members of: Roles and membership Members body and o Public (e.g. parents, community members, alumni, business, relevant BOS team selection figures in the educational field etc.) Role • Overall responsibility for BOS funds o Community representatives (e.g. teacher councils) and village • Provide information on student numbers to MoEC government (maximum of 3 members) • Check that school receives correct amount of BOS • School committees should have a minimum of 9 members and should be an odd number • Publicize the receipt, amount and use of BOS funds on a school notification board • Selection is arranged by a preparatory committee which organizes candidates and elects members • Develop school budget activity plans (RKAS) including all sources of school revenue (BOS-K1 and BOS-K2) • School committee elects chair, secretary and treasurer and the chair is not allowed from the head of education unit • Support the development of the annual school budget plan • Produce quarterly reports on BOS fund use and upload information into an online system 3.4 The state of school based management in Indonesia • Submit annual report to Regional Working Unit (SKPD) of Education at district/city This section explores the quality of school based management in Indonesia to assess its usefulness in allocating school level resources, such as BOS, to support the goals of improved • Handle complaints on BOS fund usage education participation and learning. In line with the existing literature it assesses school • Sign off on a letter stating that the BOS funds have been used according to based management across three main dimensions (Bruns, Filmer et al. 2011). It first explores guidelines the degree of autonomy that schools have to make their own decisions. When schools have Members • Principal limited authority over key decisions their ability to materially affect education outcomes is and seriously circumscribed. However, autonomy is a necessary condition but on its own is not • Treasurer - usually a teacher nominated by principal selection enough to realize effective school management. Greater participation of a wider group of • Parent who is not a member of the school committee – elected by principal and stakeholders in school-decision making is also an important feature. A central component school committee of reforms of this kind has been to encourage greater participation of parents and other members of the community in providing support to schools. This almost always entails the establishment or strengthening of representative school committees. A third dimension to assessing management practices at the school level is to determine the strength of accountability relationships between the school and its community. The transparency of school decisions and the quality of information provided to stakeholders outside of the school are essential qualities of strong accountability relationships. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 68 69 Autonomy teachers receiving performance evaluations from principals have a significant effect on test scores. Schools have a considerable degree of autonomy over many important decisions that have important effects on student outcomes. The recent school based management survey showed The ability to hire teachers, largely through the resources that BOS has provided, allows that school principals have significant autonomy over a number of important areas of their schools an element of autonomy in teacher management issues. While school principals only schools’ affairs (Figure 3.3). For example, almost all interviewed principals felt that they have limited authority over civil service teachers they report that that they have significant set the overall vision and goals of the school and were the final decision maker on school authority over teacher management issues (Figure 3.3). It is likely that this authority refers planning and budget decisions. to their role in the management of school hired rather than civil service teachers. At the primary and junior secondary level, school hired teachers make up approximately 30 percent While overall decision making responsibility lies with the school principal, they rarely take of the teachers in the surveyed schools and almost all schools employ teachers of this kind.36 decisions on their own. In many cases, including BOS, the BOS team, the teacher board Schools can use the ability to hire local teachers to adjust for any short fall in civil service and the school committee are important contributors to the decision making process. In teachers and provide additional teaching resources in areas that the school may feel are the SBM survey principals rarely reported making decisions on their own. For example, the necessary (e.g. extra-curricular activities, remedial coaching etc.). The administrative data most common areas where school principals acted alone were in setting the school’s vision from the 2010 SBM survey reports that approximately 22 percent of discretionary funds and developing the school’s work plan. However, even in these domains less than a third of available to schools were used for the hiring of non-PNS teachers. The limited literature principals said they worked alone. In terms of the allocation of the overall school budget in Indonesia linking the learning outcomes to the use of SBM report that the number of and BOS funds, less than a fifth of principals reported making these decisions on their own. non-civil service teachers, and the teacher attendance rate have significant positive effects on math scores, controlling for other school and parent characteristics (Chen 2011). Figure 3.3: Percentage of primary and junior secondary school principals reporting they had final decision making authority Participation in school decision-making School vision, mission and/or goals The current rules and regulations for school based management provide for broad based Allocation of school budget participation in school affairs. A critical aspect in many SBM programs, as in the Indonesian School work plan context, is the broad based participation in school decision making processes (Box 3.2 and 3.3). Though in the Indonesian context, the key decision making process of the use of BOS School facility planning funds lies with the school principal, approval of the overall BOS budget plan is required Student admission criteria from the teacher board and the school committee. Therefore, de jure, the SBM model in Allocation of BOS funds Indonesia ensures participation of parents and community as well as the lower levels of School curriculum government agencies, specifically for the utilization of school level funds. Student promotion to next grade However, in reality participation of school committees and the wider community is generally Teacher recruitment, hiring and incentives low. De facto, however, many systems fail in guaranteeing the participation of other key Textbook stakeholders. The recent school based management study asked school principals who was Academic calendar involved in the final decisions on ten school related matters (Figure 3.4).37 School principals were involved in all key decisions with the exception, in some schools, of setting the academic 0% 20% 40% 60% 80% 100% 120% calendar. The school principal involved teachers in the decision making process particularly %ge of schools when decisions involved student issues (e.g. promotion etc.). Teachers were less likely to be involved in teacher recruitment issues. School Committees (SC) appear to be involved in Note: Percentage of school principals reporting that they had final decision making authority. In only 40 percent of the ten domains on average and parents and wider community are the most cases final decision making authority rested with more than one individual (e.g. principal in least involved. conjunction with teachers etc.) Source: School based management national survey, 2010 It is important to recognize that the resources provided to schools through BOS grants provide an important source of flexible and predictable funding for schools to act on this autonomy. For example, it can provide the additional teaching and learning materials that schools feel are necessary to support their students and the resources to support teacher 36 88% of primary schools surveyed in the school based management survey had at least one school-hired professional development. Chen (2011) utilizes a curriculum based test administered to 5th teacher. grade students as part of the SBM survey to determine the important correlates of improved 37 The ten areas are; a) School vision, mission, and/or goals, b) school work plan, c) school curriculum, quality in schools. She finds that teacher quality and management stand out as important d) teacher recruitment, hiring and incentives, e) academic calendar, f ) textbooks, g) allocation of BOS funds, factors in improved test scores. The share of teachers receiving training, and the share of h) allocation of school budget, i) student admission criteria, j) student promotion to next grade. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 70 71 Figure 3.4: Average participation in final school decision making funds, and only then communicate their decisions to the school committee chair and ask for his or her signature on the appropriate forms (World Bank 2012e). 100% A key factor explaining the limited role that school committees have in school decision 80% making is their lack of representativeness. Of the 446 schools that had a school committee 60% in the 2010 SBM survey, almost 60 percent had between five or less parents in the school committee whereas only 31 percent had six or more parents.38 As mentioned in the previous 40% section, a large number of school committee members were selected by consensus rather 20% than being directly elected.39 This may, in part explain the limited role the school committee plays in decision making. A recent randomized controlled trial highlighted the importance 0 of committee elections in raising levels of learning achievement in primary schools in Principal Teachers School District Parent Community Committee Indonesia. The trial found learning outcomes were significantly better in schools where elections for school committee members were held and efforts were made to forge stronger Note: The bars show the average participation of each actor in ten school matters (e.g. school links with local village councils (Pradhan, Suryadarma et al. 2011). committees are involved in approximately 40 percent of decisions made at the school level) Source: School based management national survey, 2010 The limited knowledge of key school issues also limits the participation of school committees and parents more generally. Despite the role that BOS has provided for school committees, However, the SC’s strongest role centered around the use of BOS funds and overseeing parents are still not well informed about the use of funds. For example, the SBM survey financial matters more generally. In 2010, more than 60 percent of school principals reported revealed that about 94 percent of the parents knew about the BOS program but only 43 that school committees were involved in final decision making in these areas (Figure 3.5). percent say that they received any information on the use of the funds. Further, only about Perhaps unsurprisingly, school committees were rarely involved in decisions about pedagogy 18 percent of parents reported that they knew that their school had designated a staff and instructional issues. These findings highlight the importance of the BOS program in member to receive complaints and respond to inquiries from the community and about the opening up school decision-making to the broader school community. same share of parents indicated that they had filed a complaint in the previous year. Figure 3.5: School committee participation in key decisions Interestingly, the knowledge of school committee members of their overall role and their 0% 20% 40% 60% 80% specific role on the BOS program was also relatively weak (Figure 3.6). Committee members were given a series or statements and asked to identify which were objectives of the BOS Allocation of school budget program. Only 19 percent correctly identified the objectives of the program while the Allocation of BOS funds remainder either did not respond or included or excluded a number of the real objectives School work plan of the program. With limited information on the objectives of BOS, it is unlikely that the committee would be able to play a significant role in the use of BOS funds. School vision, mission and/or goals Teacher recruitment, hiring and incentives Figure 3.6: Committee members’ response to identifying objectives of BOS Student admission criteria School curriculum 35 %ge of school committee Textbook 30 Academic calendar 25 members Student promotion to next grade 20 15 Note: %age of primary and junior secondary schools where school committee is involved in final decision in selected areas, 2010 10 Source: School Based Management Survey, 2010 5 While BOS may have improved the participation of school committees in some areas of 0 Correct BOS One error Two errors More than school decision making, significant concerns remain over the quality of this participation. objectives identified two errors A number of qualitative studies have explored the inner workings of school committees and their relationships with the school principal and teachers (SMERU 2006; Chen 2011; World Source: SBM survey, 2010. School Committee module. Bank 2012e). A common finding of these studies is that school committees were often only 38 passive participants in school decision making. For example, the school based management There were outliers with no parent and more than 20 parent members as well. study found that the school principal and teachers would agree on the allocation of BOS 39 This has been identified as a problematic issue in terms of effective SBM by an RCT. See Pradhan et al (2011). Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 72 73 A further factor underlying the weak capacity of school committees was the limited Transparency and accountability availability of good quality training opportunities on their role in the BOS program. While the data is relatively old the 2010 SBM survey revealed that around three quarters of the An important aspect of SBM is the importance that it places on establishing stronger surveyed school committee members had not received any training on their role in the BOS mechanisms of accountability between the school and the local community but also between program. Moreover, where training opportunities were available these were typically only the school and in the case of Indonesia, the local district government. provided to the school committee chair. Much of the training on offer was also provided by In Indonesia, since BOS constitutes a large part of a school’s discretionary funds, monitoring the central government. Less than 60 percent of the districts where surveyed schools were effective use of these funds becomes an integral part of the accountability process for the located provided any kind of training on the BOS program. implementation of SBM. Socialization and training play an important role in familiarizing More recent efforts to improve school committee capacity to carry out their roles have also the community and parents on the role that can be played by them and the expectations they suffered from limitations on the number of school committee members trained. However, a should have from the schools. Similarly, the actual school based managers, including the recent evaluation highlights the potential that a well-organized training program with follow principal and the teaching board also need to be fully aware of the role they have in effective up activities at the school level could have on strengthening the role of school committees in management of funds and delivery of better quality education. However, as the previous improving education outcomes (see Box 3.4). section has shown, even in 2010 – the sixth year of implementation of BOS, the functions of school committees were not fully understood by principals, and they still lacked some information on the role of BOS (World Bank 2012e). The teachers also lacked information Box 3.4: A recent evaluation of BOS training activities on the role of BOS, as did school committees and members (Figure 3.6). In 2011, a comprehensive BOS training program was delivered to almost all junior Information on the overall functioning and performance of the school is a vital component secondary schools across Indonesia. Approximately 650,000 people attended the of efforts by the school committee and the local community in holding their schools to training. The program provided training on the core elements of school based account. Without information on the quality of education, the total resources the school has management including planning, budgeting and financial management. Each school and the key challenges the schools face, the task of holding school managers accountable is sent a team consisting of the principal, BOS treasurer and a member of the school impossible. However, school committee members reported low levels of information being committee to a three day training event held in each district. Conclusions to an available to them about their schools. Specifically, about 60 percent reported not being evaluation of the training program included: informed of academic programs in their schools and about 45 percent of not receiving information about school expenditures. Such limited information sharing could be a • A follow-up survey found that the materials covered under the training program consequence of the insufficient information of the roles of the school committee and its were relevant to the needs of schools in terms of school based management. members as well as the overall representativeness of the committee (see previous section). • The training program contributed to a number of changes at the school Information on overall school performance appears to be lacking in many primary and level particularly in terms of school governance. These positive findings were junior secondary schools across Indonesia. The 2010 SBM survey reported that only 30 confirmed by interviews with parents and community members that schools percent of primary schools had sent information on the school’s performance to parents over were using BOS funds in a transparent and accountable way. the course of the last year and only around one in five had sent any information to parents • Schools reported that they were able to more accurately follow BOS reporting on how they were using BOS funds. As part of the BOS guidelines, schools are required guidelines which improved the overall transparency and accountability of the to post information on planned and actual use of BOS funds on a quarterly basis on a program. school notification board. However, monitoring surveys revealed that only around a third of primary schools and 40 percent of junior secondary schools actually complied with this • Results of follow-up visits highlighted overall improvements in the knowledge requirement (World Bank 2011). These findings highlight the difficulties that parents might of the management of the BOS program particularly amongst school committee face in holding their schools more accountable. members This limited information on the overall performance of schools may also be a driver of • The quality of training varied greatly across regions. There was a lack of what appears to be the low emphasis on education quality. In terms of selection of schools, supervision at the district level which resulted in differences in training duration, parents reported accessibility to be of high priority in selecting schools for their children (76 class sizes, capacity of trainers and the quality of training facilities. percent of parents). Approximately 86 percent of the parents reported that quality was not • The majority of trainees interviewed in follow-up visits felt that more guidance an important factor when selecting schools for their children. Without meaningful measures and assistance was needed to help them implement training elements in their to understand how their schools are performing in relation to other schools it seems unlikely own schools. that a focus on quality will emerge and the potential for school committees to hold schools more accountable for performance will be lost. Source: Shaeffer (2013) The SBM survey also revealed that there was minimal parental and community pressure to improve education (Table 3.1). The highest proportion of teachers reported having no pressure from parents or the community to improve the quality of teaching. This could Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 74 75 be due to cultural norms in the country. As suggested in the literature, in some cultures, parents are reluctant to give critical feedback to teachers and principals either due to the high respect accorded to these professions, or in certain rural settings, simply by the virtue of Box 3.5: Improving educational quality through enhancing the teachers being the highest qualified individuals in the village (Patrinos and Fasih 2009). community participation In the 2010 SBM survey, the supervisors reported having more pressure both from parents A Randomized Control Trial (RCT) with four different interventions aimed at and community (Table 3.1). This seems to indicate that the understanding of how the actual strengthening the role of school committees and subsequently improving learning school based actors, including teachers and principals can be held accountable to provide outcomes was conducted in Indonesia. The study took place in 520 randomly better services for the students is not well-developed in the social norms of the country. In selected rural public schools in central Java and Yogyakarta between 2007 and 2009. such situations, additional efforts of socialization can improve the role played by the SCs Under the RCT design, school committees were randomly allocated to receive a and parents. Another approach that also proved successful in improving student learning grant and a combination of one of three other interventions: outcomes was highlighted in a randomized controlled trial in Indonesia that linked schools and their committees with village councils and influential individuals in the village (Box a. Training of the school committee, 3.5, Pradhan et al, 2011). Interestingly, this may also point to the political economy of the b. Election of the school committee society, where the accountability pressure is higher when the more influential members of the community are considered as a part of the process. c. Fostering ties between the school committee and a local governance body called linkage Table 3.2: Percentage of stakeholders reporting pressure from parents The impact of these different interventions was measured on a host of education and the community to improve student achievement, by type of outcome indicators including school committee activities, parent’s effort and stakeholder, 2010 engagement, teacher motivation and student learning. Stakeholder Parents Community The findings of the trial show that there are no effects on learning from interventions that strengthened the school committees internally, that is, by providing grants and District head 37 39 training, even though, it does improve parental knowledge on SCs, and cooperation Sub-district head 45 59 between school and non-educational community organizations. This is in contrast to similar grant and training interventions, for instance in the Philippines (Khattri et Supervisor 24 26 al, 2010) or in Mexico (Gertler et al, 2008 and 2010). These studies find an increase Education board chair 42 44 in test scores and a reduction in repetition and failure rates respectively. Other similar interventions around the world also show some impact on intermediate or final Principal 37 48 outcomes. For instance, in Madagascar, providing parents with school report cards Teacher 57 65 and information about their role in improving quality of education and providing Source: SBM Survey, 2010 teachers with pedagogical tools, improved test scores by 0.1 standard deviations and The monitoring of the overall use of BOS funds at the school level seemed to be equally improved attendance rates. shared between the school committee and district level officials. The 2010 SBM survey The RCT does find significant effects on learning for interventions that strengthened found that school committees undertook school monitoring exercises, including looking at ties between school committees and external bodies. This includes democratic election the use of BOS funds, in approximately 80 percent of surveyed schools. While only around and linkage which together result in a 0.22 standard deviation improvement in 60 percent of schools reported visits from district or sub-district staff almost all reported Indonesian test scores. Linkage alone led to an improvement of 0.17 SD in learning. receiving visits from the local school supervisor. Most frequently monitoring visits occurred on a quarterly basis and may be indicative of a pattern where these key monitoring agents The factors driving these positive results appeared to be a combination of broader are just carrying out routine tasks of signing off on BOS funds use, rather than a regular in community action on school related issues (e.g. enforcing study hours in the village) depth monitoring and interaction. and in the case of elections an effect on improved school committee effectiveness. This reiterates that in the Indonesian case, efforts to improve the representativeness As the previous section highlighted the role of the school committee appears relatively passive of school committees and finding ways to strengthen accountability mechanisms in its day-to-day interaction with the school. The focus group discussions undertaken as part through the involvement of broader community members, particularly those who of the SBM survey found that in no school did the school committee ever question the use of have clout and voice, can foster greater transparency and accountability and enhance BOS funds and ask for changes in its use. This passive role is also reinforced by the relatively learning outcomes. limited authority that the BOS guidelines give to school committees for this oversight role. Most of the documents and forms that form the foundation of the management of BOS Source: Pradhan, Suryadarma et al. (2011). funds are only required to ‘be seen’ by the school committee rather than being approved or accepted. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management BOS and School Based Management 76 77 The outcomes of school based management Recent studies have questioned whether the large resources that have been devoted to the increased hiring and remuneration of teachers have had any significant effect on education The introduction of school based management was aimed at increasing the role and outcomes. The recent education public expenditure review highlighted the significant participation of parents, teachers, principals and the community in improving education inefficiencies in Indonesia’s education system that have resulted from very low student teacher outcomes. Funds allocated through BOS play an important role in increasing the ratios. The analysis also shows that the increased hiring, brought about in part through discretionary funding available and give operational freedom to the schools (Figure 3.2). BOS (see Chapter 2), has not been associated with any improvements in learning outcomes However, given the weaknesses in school based management outlined in previous sections it (World Bank 2013d). Moreover, a large teacher certification program that provided a is interesting to explore how schools use BOS funds. professional allowance equivalent to basic pay for teachers who qualified also had limited A significant share of BOS resources go to supporting teachers and in particular in hiring effects on learning (Chang, Shaeffer et al. 2013). This evidence suggests that BOS spending school hired teachers. The SBM survey suggests that of the discretionary funds available to on teachers may be exacerbating these inefficiencies and could perhaps be more effective. the schools, about 60 percent are used for instructional support on average, a large proportion While overall trends in levels of learning achievement are not driven completely by the of which goes to hiring non-PNS (non civil service) teachers (Figure 3.7). However, it does BOS program it is important to note that these have not improved significantly in recent reveal substantial variation in the utilization of funds across the schools. times. In general, the 2010 SBM survey of various stakeholders reported a relatively positive perception of the BOS program. Stakeholders highlighted their views that BOS had Figure 3.7: Composition of discretionary spending at the school level, contributed to increased transition to junior high school, better availability of textbooks, and primary schools, 2010 better student performance. Unfortunately, the perception of improved performance does percentage of discretionary expenditures, % not really emerge when the trends in learning achievement are examined. For example, the latest OECD PISA results show that student learning in mathematics and reading have only 0 5 10 15 20 25 improved marginally and in science have fallen (OECD 2013). Though there are a number of other factors that contribute to this performance, it does not appear that the overall Non PNS salaries objective of improved learning outcomes are being achieved through the implementation of SBM and the allocation of BOS funds. institutional support Student tests Student activities 3.5 Summary Textbook and libraries The chapter focuses on the role of SBM in utilizing the BOS funds and tries to identify whether it has led to the improved outcomes expected of these types of reforms. The review Staff development of survey data and literature suggests that SBM has played an important role in bringing school committees, and to some extent, parents, into school decision making on BOS. Consumable goods Given that the BOS grants link into other aspects of school decision making, most notably school development planning and budgeting, it has opened up these aspects of decision School maintenance making to school committees. Despite these improvements, the role of school committees remains weak and is unlikely to be influencing significantly the decisions made by the school facility support Furniture principal and teachers. Utilities A number of factors appear to weaken the potential role that school committees and the broader community can play in supporting the use of BOS funds and providing greater Transportation oversight. First school committees do not appear to be particularly representative and this has the potential to hamper their role as an accountability mechanism. Second, the chapter BOS management has shown that school committees have relatively little knowledge on what their role should be and this is partly related to issues around the limited training opportunities available to Other most school committee members and the lack of follow-up activities from training. Third it appears that the flow of information to parents and SCs is limited. Open sharing of information is a cornerstone of the theory of school based management. Without transparent sharing of information, the stakeholders are unlikely to be able to hold school based agents Note: Other spending includes registration formalities, school uniform spending, computers etc. accountable for better services. Fourth, the monitoring role of school committees on BOS Source: Reproduced from World Bank (2012e), Figure 6.1 funds outlined in the guidelines provides a relatively weak role for school committees. And finally, parents do not appear to have the information necessary to be overly concerned, Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia BOS and School Based Management 78 in general, about seeking better quality education. This reduces the pressure on schools to improve quality. Socialization on the importance of seeking or demanding better services CHAPTER 4 from schools and teachers can improve this apparently socio-cultural inertia in terms of pressure from parents. Despite these weaknesses, the evidence in Indonesia points to the potential that these reforms Strengthening the BOS Program can have on education outcomes. Research shows that where school based management is working effectively it has the potential to improve education outcomes and by consequence make better uses of the resources available to schools through BOS. Evidence from other countries also points to the relatively long length of time that reforms of this kind take to deliver results (Bruns, Filmer et al. 2011). It seems clear that further efforts to reinvigorate and strengthen school based management reforms could have significant pay-offs. The BOS program continues to be a central component of government efforts to raise education outcomes particularly for the poorest households. Since its introduction the budget for the program has increased continually and has led to an ever increasing availability of resources in primary and junior secondary schools across Indonesia. Its effectiveness at delivering ever increasing amounts of resources to schools has led to its replication throughout the rest of the education system. The aim of the report has been to assess the broader contribution of the program to education objectives and to identify ways in which the program can be strengthened to maximize impact. The report has shown that the BOS program had a temporary and small effect on reducing the costs of schooling faced by households. These findings were strongest amongst the poorest households and for households who sent their children to government schools. Results on student participation were less conclusive but suggest that the program may have contributed, at least initially, to increases in junior secondary school enrolment rates amongst the poorest children. While these findings are promising they suggest the need to explore ways in which BOS can more effectively support the education outcomes of poor households in Indonesia. The program has also supported efforts to provide schools with more autonomy and to strengthen links with local communities. In other countries, school based management reforms have improved educational attainment and in some cases levels of learning achievement. However, it is not clear that these gains have so far been realized in Indonesia. The findings of Chapter 3 suggest that greater efforts are needed to establish and deepen the implementation if outcomes are to be improved through this route. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 79 Strengthening the BOS Program Strengthening the BOS Program 80 81 This chapter outlines four main areas where efforts could be made to strengthen the role of Area Issue Policy suggestion BOS in supporting overall education objectives: Improve Many local governments have Encourage local governments to 1. Shifting the focus of BOS towards quality and linking to education standards coordination their own school grant programs introduce school grant programs 2. Strengthening the poverty focus of BOS of BOS with and better coordination with the Support local governments to link other funding national BOS program has the 3. Making better use of BOS funding through improved coordination with other total school grant funding to defined sources potential to improve spending funding sources education standards and equity objectives effectiveness and raise education 4. Revitalizing the role of the BOS program in empowering schools and local outcomes Experiment further with the use of local communities school grants to incentivize performance Table 4.1 provides a summary of the main suggestions for strengthening the current BOS Revitalizing the BOS management team at the The parent representative of the BOS role of BOS school level does not have any team should also be a school committee program that hold out the potential for increasing the impact of school funding on improving to empower school committee representation member to ensure greater oversight from education outcomes. schools and the community in BOS decision making Table 4.1: Summary of policy suggestions communities School committee’s oversight Increase authority by requiring school Area Issue Policy suggestion function on use of BOS funds is committees to approve rather than Adjusting BOS Identified need to strengthen the Link levels of BOS funding more weak acknowledge school plans for the use of to enhance link between BOS program and explicitly with national standards of BOS its focus on improving education quality education improving Complement the oversight role of Link eligibility for BOS funding to the school committee by tying it education regular school accreditation more strongly to existing government quality structures Schools are prohibited from using Review lists of eligible and ineligible BOS funds to purchase some BOS expenditures to ensure that items Weak capacity of school committees Build the capacity of school committees items that are required to improve associated with better education quality to fulfil their role effectively by increasing the coverage of members education quality are allowed trained and providing in-school follow up support Strengthening The real value of the BOS grant Adjust the value of BOS periodically to the poverty differs across Indonesia and declines account for regional price differences and Develop methods (e.g. school report focus of BOS over time because of inflation inflation cards) to provide school committees with information required to utilize resources Schools serving predominantly poor Introduce an additional component into more effectively and disadvantaged children require the BOS formula to provide schools additional support that could be with additional resources for poor and Explore ways to strengthen the role provided through BOS vulnerable children of local governments to support and BOS funds can be used to provide Change BOS guidelines to phase out the monitor the BOS program direct support to the ‘out of pocket’ use of funds to cover poor children’s ‘out school expenses of poor students of pocket’ expenses. despite a number of large cash 4.1 Adjusting BOS to enhance its focus on improving transfer programs designed more education quality specifically for this purpose. Regulations on school fees and Clarify further the costs BOS is intended In recent years and in recognition of significant progress on improving educational access, contribution policy are not clear to cover and the fees and contributions the Ministry of Education and Culture has started to shift the focus of the BOS program and have adverse consequences schools are allowed to charge towards education quality. This shift also coincides with a growing concern about the on the role BOS plays in reducing quality of primary and secondary education and recognition that efforts to improve the Strengthen the role of school committees education costs faced by households. skills student’s leave school with need to be strengthened. The latest guidelines for the and parents in determining acceptable charges and contributions implementation of the program state that: Consider adjusting eligibility rules for “The School Operational Assistance (BOS) program that was launched in July 2005 has the BOS program to exclude non- made a significant contribution to accelerating the completion of the 9-year compulsory government schools charging high fees education program. Therefore, from 2009 onwards the government has changed the BOS’ objectives, approach and orientation from outreach to quality improvement” (Ministry of Education and Culture 2014) Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 82 83 A shift of the focus of BOS towards quality can provide a strong signal to schools about the of schooling are also important. For example, the proportion of teachers with a bachelor’s importance of education quality and align the program more closely with other efforts to degree and above teaching in primary school is approximately 61 percent in urban areas of raise learning achievement. The government has embarked on an ambitious reform agenda Indonesia compared to only 37 percent in remote areas. to improve the quality of primary and secondary education and the BOS program has the potential to support this agenda. Figure 4.1: Provincial poverty rates and the real value of BOS, 2013 Indonesia already has a set of defined education standards and the overall value of BOS 40 funding should be reassessed to ensure alignment with these existing criteria. Chapter 1 has 35 Papua Papua Barat already shown that Indonesia has developed a set of detailed education standards which can be used to estimate average school operating expenses (see Box 4.3). In turn, these estimates 30 Poverty rate, 2013 can form the basis of overall BOS funding levels which can be linked more explicitly to a Maluku 25 defined education standard. For example, existing levels of BOS are similar to levels of non- NTT salary operating expenses associated with current minimum service standards (See Table 20 1.2). BOS funds could be explicitly linked to this standard of education. 15 While the government has already used operating cost estimates for specific standards to set 10 the value of BOS it has not articulated this link to schools and their communities. Socializing this link could clarify the role of BOS in insuring a minimum standard of education provision Sulawesi Selatan 5 and provide a basis for local communities to hold their schools more accountable. DKI Jakarta 0 Linking BOS funds more closely with quality assurance mechanisms in the long run could 300,000 400,000 500,000 600,000 700,000 800,000 further strengthen the link between BOS funding and standards. A recent review of the Real value of BOS at primary level (constant 2012 prices) main quality assurance systems highlighted the lack of incentives and sanctions of key parts of the system. For example, incentives and sanctions associated with accreditation and Note: The horizontal line included in the figure shows the provincial average poverty rate. The vertical fulfilling minimum service standards were relatively limited. While it is too early to link the line shows the value of the BOS grant for primary schools achievement of minimum service standards to eligibility for the BOS program, funds could Source: Susenas, 2013 be made conditional on schools seeking regular accreditation. Differences in the real value of BOS also drive differences in the resources available to schools Adjusting the allowable expenses under the program would also facilitate a greater focus on to meet education standards and can reinforce education inequalities associated with poverty education quality. While there have been adjustments to the items that are allowed to be and marginalization. For example, the price of school supplies can vary significantly between purchased with BOS funds in the past it is useful to review them in light of a stronger focus accessible and more remote areas because of associated transportation costs. A school in a on quality. For example, the purchase of audiovisual aids and learning media should always remote part of Papua can purchase fewer textbooks and other supplies than a school in be allowable expenditures under BOS instead of the current situation where they are only Jakarta even with the same amount of BOS. Frequently, the schools facing the disadvantage allowed after all other items have been purchased. of higher costs also tend to serve populations that are poorer and more marginalized. For example, Maluku and Papua provinces have some of the highest rates of poverty but due to high costs have some of the lowest real values of BOS (Figure 4.1). Schools in these areas 4.2 Strengthening the poverty focus of BOS tend to need more rather than less support to achieve similar outcomes as schools serving A key objective of the BOS program from its outset has been to improve education wealthier children. outcomes particularly for the poorest children. The report has shown that the introduction Adjusting the per-student amount of BOS to account for regional cost differences could of BOS was associated with a small drop in the costs faced by poor households but this was address at least some of this imbalance and provide all schools with the non-salary operating relatively temporary and education costs have risen relatively rapidly since. It also shows resources needed to meet set standards. It could also lead to a more equitable allocation of that other factors apart from BOS were at play in explaining the narrowing of education school funding. An annual cost index across the major regions of Indonesia is published inequality seen in recent times. Notwithstanding these improvements, large disparities in on a monthly basis by the Bureau of Statistics and this could, in the first instance, be used education persist and strengthening the poverty focus of BOS could narrow these education to adjust for the differences in costs across districts or provinces. In the longer term, cost inequalities further. adjustments could be based on a more refined index that was driven more closely by the Large disparities in education outcomes exist between children born into families in different goods and services usually bought by schools with their BOS funds.40 parts of Indonesia and with different social and economic circumstances. The chances of a child entering primary school in Papua and completing the full 9 years of basic education 40 are approximately half those of a child going to school in Jakarta. There are many drivers Addressing BOS values across districts would not account for intra-district cost differences which can also be of these outcomes that fall well beyond the education sector but differences in the quality quite large. Ensuring that these differences are included in local school grants program is an important aspect of coordination discussed in the next section. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 84 85 The per-student allocations for BOS also need to be adjusted periodically to ensure that Clarifying school fee and contributions policy overall levels of school funding are not eroded by inflation. Previous chapters have shown that adjustments to the BOS amount have happened irregularly and between adjustments The alignment of the BOS program with other sources of school funding has the potential to the real value of BOS declined significantly. For example, between 2009 and 2011the real improve the overall effectiveness of the program and at the same time provide greater support value of BOS allocations declined by 12 percent before adjustments were made in 2012 to poor students. The results of Chapter 2 have shown that while BOS may have reduced which just covered these losses at the junior secondary level. Ensuring a more regular and costs in the short term, poor households continue to face large direct costs in sending their explicit updating of the BOS per-student amount to adjust for inflation could reduce the children to school. Moreover, a significant share of these direct costs consists of fees charged need for schools to ask households for additional contributions and in particular avoid directly by schools. placing additional burdens on poor households. A number of regulations have been issued governing the amounts and purposes of fees and Indonesia already has programs that provide support to help poor households with the costs contributions that schools are allowed to seek from parents. However, these regulations lack of education but programs to support the teaching and learning of poor children once in clarity and provide an opportunity for schools to ask parents for contributions in areas that school are less common. Cash assistance is provided to poor families through a variety of are intended to be covered by the BOS program (see Box 4.1). Moreover, evidence from programs to assist in the ‘out of pocket’ expenses associated with school attendance. The household surveys, described in the report, have shown that a large proportion of household largest of these programs, Bantuan Siswa Miskin, provided scholarships to approximately 15 education spending goes to paying fees at school even in government primary and junior million children in 2014 (around a quarter of all children of primary and secondary school secondary schools. Poor households are also confronted by these direct fees and may be a age).41 While evidence from other countries has shown that these programs can be successful factor in explaining why poor children continue to drop out of school before completing at getting children into school they are less successful in raising learning outcomes (Fiszbein, junior secondary schooling. Schady et al. 2009). These findings demonstrate the need to provide additional support to disadvantaged children or schools serving these groups to make up for earlier deficiencies Box 4.1: The 2012 regulations on fees for primary and junior in education investments. For example, schools serving poor children may need to provide secondary schools more reading materials to encourage literacy compared to schools in wealthier areas where children’s own homes might be expected to provide much of this additional literacy support. In 2012, the Ministry of Education and Culture issued a set of regulations on the fees and contributions schools were allowed to collect. This built on previous BOS could be used to provide this complementary ‘supply side’ assistance to programs regulations that tried to define an appropriate cost sharing policy for primary and designed to reduce the costs to poor families of sending their children to school. Many other junior secondary schools. countries have used school funding mechanisms like BOS to target supply side support to poor children. In Chapter 1, England’s pupil premium was shown to provide additional The regulation differs on the type of schools that are able to charge fees for regular support to poor children that school’s used to introduce special programs to support the school operating expenses. The regulation defines regular operating expenses to learning of disadvantaged students (see Box 1.2). BOS could follow a similar approach include investment and operating costs, tuition assistance and scholarships. Regular and the unified database of poor households in Indonesia could provide the information government schools may not charge students or parents for these operating expenses necessary to adjust BOS funding in this way. whereas community, private and all international standard schools can. While BOS has a strong role to play in supporting poor children through the provision All schools can charge or seek contributions from students and parents for activities of good quality and appropriate education services, its role in supporting the direct costs and expenses that lie outside of these regular operating expenses. For example, schools of schooling for poor children is less clear. Indonesia has a large and growing national can charge parents for extra-curricular activities, additional instructional materials scholarships program that covered over 15 million students in 2014. Moreover local and uniforms. However, charging fees to cover expenses in the following areas are governments have similar programs which cover additional children. In total it is estimated prohibited: that in 2012, around a quarter of all children in primary and junior secondary school had • Academic requirements for student admission access to some form of scholarship. In addition, extremely poor households also participate in other programs that provide these households with conditional cash transfers to support • Student learning assessment the ‘out of pocket’ expenses to send children to school. Despite some weaknesses in these • Costs associated with student graduation programs they provide the best means to channel government support for the education of poor and vulnerable children (World Bank 2012a). As these programs continue to be • Welfare of teachers, school committee members or local and central government strengthened it would be possible to phase out BOS funding for poor children’s direct school staff costs. This would avoid confusion and remove any duplication of effort in supporting poor While schools are not able to ask for contributions in these areas or for the regular children. operating expenses to schools they are allowed to accept voluntary contributions 41 The newly elected government introduced at the end of 2014 the Indonesia Smart Card (Kartu Indonesia and donations from parents and the wider community. However, there is room for Pintar) as part of a commitment to expand universal education from 9 to 12 years. The new program is parents or the community to refuse paying charges that the school imposes. expected to take over from the Bantuan Siswa Miskin program but with an increase in coverage and changes to targeting and benefit delivery. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 86 87 poor households in Indonesia and the associated scholarship scheme could provide the basis of a definition that could be used in revised regulations to clarify exemptions. Box 4.1: The 2012 regulations on fees for primary and junior secondary schools (continued) The role of the school committee and parents more generally could be strengthened to ensure a fair school contributions policy. First, the school planning processes could be used to The revenue from these additional charges must be used in accordance with the identify the financing gaps that the school is seeking to fill with contributions from parents. school’s improvement plan and its annual work plan. Charges must be discussed In this way, a school’s contribution policy can be determined as part of broader school in a school committee meeting and a letter must be sent to all parents and students planning and trade-offs between the ambition of the school’s plan and the community’s outlining the amounts and the use of charges that are sought. Moreover, schools ability to pay discussed. Second, school committees or parents more generally should be must allocate at least 20 percent of the funds collected to support education quality given voting rights on a school’s contribution policy. For example, in Chile two-thirds of improvements. parents are required to approve the school’s contribution policy before it can be introduced While the regulations reflect efforts by the Ministry to outline a coherent policy on (Cuadra and Moreno 2005). fees it has a number of weaknesses: Existing provisions of the BOS program allow all non-government schools to receive support • The regulation provides no clear definition of the regular operating expenses of regardless of the level of fees they charge. Non-government schools are allowed to charge schools which can be interpreted in different ways and lead to schools charging fees to parents to cover differences in the value of BOS support and their own operating for a greater range of items than intended by the regulation. costs. No limits are set on the fees that schools can charge and while evidence on school fee levels is difficult to compile, anecdotal evidence suggests that some non-government • There is no clear mechanism to evaluate the use of these funds. schools receiving BOS funds charge high fees (see for example, Chapter 2). The rationale • School committees and parents are provided with limited opportunities to for providing government funding through the BOS program to schools catering for influence school decisions on fees. households capable of paying high levels of fees is unclear. Setting an upper limit on fees schools are allowed to charge if they are to be eligible for the BOS program has the potential • The regulation exempts poor households from additional charges but provides of increasing the programs overall effectiveness and reduce overall costs. no definition of the kind of households that would be exempted. A socialization program to outline changes to government policy on the role of BOS and • Private and community schools that receive government assistance, through for parental contributions in covering school operating costs is vital. Overall government policy example the BOS program, are allowed to set their own charges and fees. on school contributions has two potentially contradictory objectives. On the one hand, it is important that schools do not seek contributions that prevent children from attending or Source: Ministry of Education and Culture (2012) continuing in school. On the other hand, it is important that where parents and communities can contribute and these contributions raise the quality of education they do. In order to The BOS program is designed to remove the need for schools to charge households for basic ensure that communities are aware of their entitlements and obligations it is important operating expenses. It is vital therefore that schools do not charge households for these costs that a comprehensive socialization program accompanies any changes to regulations in this and that parents are aware that contributions can only be sought for activities that go beyond area. Indonesia has had some success in delivering messages on the BOS program and its these operating expenses. To the extent that the BOS program is tied to a defined standard relationship with school fees using innovative dissemination techniques (see Box 4.4). The of education, contributions could be permitted only where schools and their communities design of any new socialization program could draw lessons from these earlier programs to agree to raise standards above this minimum. ensure clear messages are communicated to the greatest number of communities. Regulations governing school contributions policy need to provide greater clarity for schools and parents. While the regulations provide some definitions of the items that schools are prohibited from charging households for, these are very general and open to different 4.3 Making better use of BOS funding through better interpretations. Providing a more detailed list of the eligible and ineligible items schools alignment with other school funding are allowed to seek contributions for would be an important first step. In addition, it is The report has also highlighted the need for central and local governments to look at the important that regulations ensure that in the first instance school operating expenses are overall funding schools have. While there is no reliable time series information on the covered by existing school resources (e.g. BOS, local school grants etc.). For example, overall funding situation of schools in Indonesia, there is some evidence to suggest that local government schools should not, on the whole, seek contributions from households to pay governments, in particular, adjust their funding depending on the support schools receive for items that can be funded through BOS (e.g. consumables or the costs of organizing from the central government (Amin, Das et al. 2008). This may be one reason for the results examinations). in Chapter 2 that showed that schools passed on a relatively small share of BOS resources to It is also important that the regulations clarify which children are exempted from school households in terms of lower fees and charges. charges. Existing regulations prohibit schools from asking for contributions from poor It is also important to encourage schools to use all of the funding available to support households. However, no definition of poverty is provided and this runs the risk that poor improvements in education quality and participation. Greater coordination between central children will not be considered poor by schools. It is possible that the unified database on Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 88 89 and local government funding of schools and linking this funding to agreed standards holds Local government school grant programs can be a natural complement to the national BOS the potential of raising the effectiveness of education spending generally as well as within program and together can drive better education standards. BOS funding can be seen to the BOS program. provide a minimum level of non-salary operating expenses to schools. Additional resources from local governments hold out the potential to raise standards above this minimum Collaborating with local governments in the provision of school level (see Box 4.2). However, in order to fulfil these additional objectives it is important that schools are provided with clear guidelines on how local funds can complement BOS funding resources and what the expectations are for the use of these funds in terms of education Many local governments have introduced schemes similar to BOS that provide additional standards and performance. A recent exercise conducted by the DKI Jakarta government funding to schools. These programs represent a very important potential source of additional demonstrates how levels of local school funding can be calculated in line with different non-salary support to schools and the available evidence suggests that they are associated with education standards (see Box 4.3). better education outcomes (see Box 4.2). While every effort should be made to encourage local governments to introduce school funding mechanisms like this there has so far been little effort to coordinate these programs with the national BOS program and this runs the Box 4.3: Calculating the appropriate value of non-salary risk that schools will not utilize these additional resources effectively. operational support in Jakarta schools Indonesia has a set of education standards for primary and junior secondary schools that have been used as the basis of calculating the non-salary operating needs of Box 4.2: Local school grant programs schools. There are three main standards: Using resources of their own, many regional governments supplement school grants 1. Minimum service standards (MSS). These are a set of minimum service standards provided by the national BOS program. The recent education public expenditure for schools that can be seen as a first step to the more comprehensive national review highlighted a positive association between learning outcomes and the receipt education standards of local school grants. 2. National education standards (NES). These standards are the key reference In 2012, a survey found that approximately half of all districts provided additional point for the education system and cover 8 areas including teacher standards, support in the form of BOS Daerah (BOSDA). In a similar way to the national equipment and infrastructure, management and financing. BOS program, the majority of districts allocated BOSDA to schools based on the number of students they enroll. While the amount provided for each student differed 3. International education standards (IES). Building on the NES and define a across districts it tended to be lower than amounts provided under the national BOS slightly higher standard of education. program. A survey in 2011 found that the average grant was equivalent to about a In order to better understand appropriate levels of school funding the DKI Jakarta third of the national BOS grant schools received. government carried out a study of the resources required by different schools to cover A recently completed pilot in 12 districts and 2 provinces supported efforts of local their non-salary operating resource needs. The regulations that lay out the three governments to include additional components in funding formula to improve the different education standards coupled with earlier estimates carried out by national equity of allocations. Pilot provinces and districts worked on developing a formula- standards agency (BSNP) provided the starting point to calculate the non-salary based BOSDA with three main components; a basic allocation, an allocation to operating expenses for each education level. To supplement this information a small address school inequalities (e.g. additional allocations for small and remote schools) school survey was conducted to identify a comprehensive list of the items required and a component that provided additional resources to schools that demonstrated to fulfil different standards and also to collect information on prices. On this basis, outstanding performance over the previous school year. the average per-student needs for non-salary operating expenses were calculated for schools across Jakarta (see table). The pilot also demonstrated that these kinds of grant programs can be developed even in the most difficult circumstances. Kaimana, a remote district in Papua province Estimated per-student Current per-student with challenging terrain and communication difficulties, participated in the pilot needs for different allocations and introduced a program that combined all three elements of the formula. standards (IDR ‘000s) A recent assessment showed that the pilot program had resulted in a more equitable (IDR ‘000s) allocation of BOSDA resources in participating regions. And while it was too early to MSS NES IES BOS BOP Total assess the impact of these changes on education participation and learning, feedback Primary schools 1,084 1,783 4,909 580 720 1,300 from participating local governments was positive. Junior secondary schools 1,261 2,142 4,911 710 1,320 2,030 Source: World Bank (2012c); World Bank (2013a); (World Bank 2013d) Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 90 91 management and the main decisions on the use of BOS funds are carried out by the school Box 4.3: Calculating the appropriate value of non-salary operational support BOS team in the absence of the school committee. While a parent is supposed to be part in Jakarta schools (continued) of the BOS team, existing regulations stipulate that this parent should not be a member of the school committee. This represents a missed opportunity to provide a stronger oversight These estimates have been used to assess the standard of education possible given role for the school committee on BOS management. Altering the make-up of the BOS team existing allocations provided to schools through the national BOS program as well to ensure that the parent representative is also a member of the school committee would be as the Jakarta school grant program (BOP). The Jakarta administration is using a relatively easy adjustment that may strengthen the link with the school committee and these estimates as the basis of proposals to adjust their own school grants program increase transparency around the use of BOS funds. to provide schools with sufficient funds to achieve national standards. In this way the funding that schools receive from both central and local government will be The report has also shown that a number of other issues weaken the oversight function better coordinated to ensure set standards are met. This basic allocation is also being of school committees. In many cases, the role of school committees in BOS is limited supplemented by a component to address the large cost differences between mainland to reviewing or acknowledging that key planning documents such as the school activity Jakarta and Kepulauan Seribu and a performance incentive based on school rankings plan have been prepared. For example, the school BOS team is only required to obtain the on the national examinations. signature of the school committee chair to acknowledge that they have seen the appropriate plan. This severely limits the role of the school committee in ensuring the best use of BOS resources to improve the school and ultimately levels of learning achievement. It is possible to increase the authority of the school committee by requiring schools to obtain the approval Local government school grant programs can also address intra-district funding inequalities of the school committee for the use of BOS funds. In particular, the school committee more effectively than the national BOS program. The previous section outlined ways in should have approval authority over the school activity and budget plan associated with the which the national BOS program could address some of the cost differences between BOS funds. districts in the real value of BOS. In a similar way, local grant programs can complement BOS by addressing some of the within district inequalities associated with the costs of school The report has also shown that in most cases school committees are selected rather than provision as well as to account for differences across schools in terms of the populations that elected. The school based management survey found that less than a quarter of school they serve. These local government programs are better placed to address these within district committee members were elected. It was much more common for schools to agree on inequalities given that the specific drivers of cost differences and markers of disadvantage are candidates by consensus or for the school principal to directly appoint members. likely to vary considerably from one district to the next. In order to function effectively it is crucial for school committees to maintain a degree of Some local governments have begun to utilize school grant programs to introduce independence from the school management. Increasing the proportion of school committee performance incentives for schools. For example, DKI Jakarta has introduced a component members that are directly elected can be a key strategy to achieve this objective. Regulations to their overall school grant program that rewards schools for improvements in their ranking on the school committee election process are already in place but it is not clear whether in national examination scores. Performance based components of this kind can improve schools and local communities are really aware of these provisions. It would be essential for education outcomes further by providing schools with further incentives to align spending socialization and capacity building activities to reaffirm the importance of these provisions. with education objectives. While it is too early to tell how incentive schemes of this kind will Consideration should also be given to introducing a more formal channel for schools to affect behavior they do provide a useful example of ways to link resource allocations more discuss their activities with village governments. This has the potential to provide broader closely to performance in the education sector. support for the school particularly with the introduction of the village law. In addition, fostering greater links with the village government through, for example, holding an annual 4.4 Revitalizing the role of BOS to empower schools and meeting on the school improvement plan has also been shown to be effective (Pradhan, Suryadarma et al. 2011). local communities While it is important for school committees to exercise some authority in overseeing the The effectiveness of the BOS program rests on how well schools are managed and governed. use of BOS funds and be broadly representative of the local community it is also essential Evidence from Indonesia as well as other countries demonstrates the potential of school based that these committees have the relevant capacity to run effectively. Chapter 3 highlighted management to improve education outcomes. However, as Chapter 1 outlined the success of many of the weaknesses of school committees in terms of participating in school affairs school based management depends on how well it is implemented. The report has identified and holding schools more accountable. Recent efforts to train school committees on their a number of areas where efforts to strengthen school based management implementation role in the management of BOS funds need to continue (see Box 3.4). However, it also could realize improvements in the effectiveness of BOS resources. In particular, the report seems necessary to extend training to a large number of school committee members and has highlighted the need to strengthen the role of the school committee in the management to the parents of students more generally. Moreover, as the evaluation on the recent BOS and governance of BOS resources. training revealed there is a need to provide some follow up support to help committee In order to ensure greater transparency and better BOS decision making non-school members members implement the approaches they learnt in training (Shaeffer 2013). Local district of the school committee should be represented on the BOS team. Currently the day to day education offices and their network of school supervisors are well placed to provide this Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 92 93 ongoing support. Efforts to build the capacity of local education offices to undertake this role and the provision of funding could play a pivotal role in ensuring the sustainability of Box 4.4: Increasing the knowledge of parents on the BOS program and their efforts to strengthen school committees. participation in school affairs through school meetings and SMS text messages A recent study in Indonesia has also shown that well-designed information campaigns can (continued) improve overall knowledge of the BOS program and also increase parental participation However, neither approach increased the participation of parents in the formal BOS (Cerdan-Infantes and Filmer 2014). An impact evaluation, carried out in 2011, explored planning process. The study acknowledges that even though schools are required to different methods of providing information to parents on the BOS program (Box 4.4). The involve the school committee in the planning process, opening this up further to all evaluation found that parent’s knowledge of the program and participation in school affairs parents would require additions to the guidelines associated with the BOS program. could be enhanced through the organization of school meetings to discuss the role of BOS and the use of simple SMS text messages. Beyond improving the knowledge and participation of parents, the information interventions also appear to have improved program transparency. In particular, school meetings increased the number of parents who felt that the use of BOS funds Box 4.4: Increasing the knowledge of parents on the BOS at the school was transparent. This may have been as a direct result of the increased program and their participation in school affairs through school knowledge and efforts by the school to make information on the use of BOS funds meetings and SMS text messages available on school notice boards. A recent randomized controlled trial evaluated the impact of providing information on the BOS program to parents through four different channels: Source: Cerdan-Infantes and Filmer (2014) • A school meeting for parents where A successful pilot in Pakistan also used a similar information strategy to strengthen the use of school councils (Cambridge Education 2014). The pilot hired a call center and • An SMS text to parents from the school principal used inbound and outbound calls, robot calls and SMS text messages to deliver important • A letter from the school principal to parents information on the role of school councils to their members. While no impact evaluation has been carried out of the program, an assessment of the pilot showed that school council • Giving parents a colorful notebook containing information on the program members’ knowledge of their roles and responsibilities had increased and school principals The study explored how these different approaches affected parental knowledge reported greater participation of the council in school affairs. The methods experimented about the BOS program and the extent to which improvements in knowledge with in the Indonesian study and the ongoing work in Pakistan could serve as important affected parental participation in school affairs. In each approach, information on strategies to deliver the information school committees need to take on the role envisaged what the program was about, levels of per-student funding, allowable fund use and for them under the BOS program. how parents were expected to participate were provided to parents. The report and recent surveys have shown that parents and school committees have relatively The impact evaluation found that school meetings and SMS interventions had a little information about the schools their children attend (World Bank 2012e). In general, significant impact on increasing parent’s knowledge of the program. In terms of parents are provided with information on their own child’s progress but little information raising parental knowledge, the school meetings had the largest impact. The study on the school he/she attends or its performance relative to other schools in the district found that school meetings increased a parental knowledge index by 0.3 standard or province. The provision of information of this kind can expand the role of parents in deviations. The SMS intervention had a smaller overall effect because very few supporting the school as well as a tool to hold schools more accountable for the quality of parents received or remembered receiving the message. When the impact is measured education they provide. In some countries, school report cards have been introduced to only amongst parents who received the text message, the magnitude of the impact is provide this kind of information and as a vehicle to increase local community participation similar to that of the school meeting. No impact on parental knowledge of the BOS in school affairs. Evaluations of these programs have not been altogether positive but in some program was found for the letter and pocketbook interventions. countries have been shown to have a significant impact on school outcomes (see Box 4.5) The evaluation found similar results of the different interventions on parental participation in school affairs. The study found that meetings and text messages increased the overall participation of parents in schools but in different ways. The effect of school meetings seemed to be largely limited to increases in parental visits to schools - a result in part of the school meeting itself. The text messages appeared to increase direct communication between parents and the schools through, for example, direct interactions between parents, the principal and the school committee. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program Strengthening the BOS Program 94 95 use of BOS funds at the local level. However, given that these funds are not part of local government budgets concerns have been raised about the perceived reluctance of some local Box 4.5: School report cards for community based monitoring in governments to fund the monitoring activities required of them under the program. Uganda An obvious way of strengthening the role of local governments would be to transfer BOS A randomized controlled trial in Uganda explored the impact of different interventions resources directly into their own budgets. This would result in local governments being that used school report cards as a tool for school managing committees to monitor directly accountable for BOS funds and ensure that financial control and management performance. The two interventions tested were: procedures associated with district budgets were followed. However, this approach was tried in 2011 and was not wholly successful. In particular, local government rules and regulations • Standard scorecard. School committee members were trained and supported seriously undermined the ability of schools to use the BOS resources in a flexible way and in the use of a school report card that was developed by education officials and delays in disbursement meant that the predictability of funds was also compromised. While NGOs revisiting the experience of BOS decentralization may be warranted, shorter term measures • Participatory scorecard. School committee members were trained and supported to strengthen the role of local governments in the BOS program could be explored. For in the use of a school report card that they themselves had developed and example, where the costs to the local government are prohibitive funding from the central included agreed indicators of school progress government to cover the costs of monitoring BOS could be provided. Information interventions like these are expected to improve education outcomes by providing good information for local communities to hold their schools accountable. 4.5 Looking further ahead Moreover, the provision of good quality information of this kind also aims to encourage better school performance either through social pressure or through better Many other countries have used formula funding mechanisms similar to BOS to address collaboration between the school and community. broader challenges in the education sector. The improvements outlined in the previous sections of this chapter have largely focused on changes that are consistent with the BOS The experiment found that the participatory scorecard approach had statistically program’s current role of providing national level funds to cover school operating expenses. significant impacts on education outcomes. In terms of student learning, schools However, many other countries have used similar funding mechanisms to improve the long where the participatory scorecard was introduced saw a statistically significant term efficiency and equity of education sector financing at a time when the demands for advantage in primary school test scores of approximately 0.2 standard deviations education services are changing. compared to control schools. Teacher attendance also improved: teachers working in participatory scorecard schools were 13 percentage points more likely to be present Recent studies in Indonesia have highlighted large inefficiencies in public education than teachers in control schools. spending. For example, the latest education public expenditure review found that despite large increases in government spending improvements in overall education quality had been Results for the standard scorecard approach were less promising. The experiment disappointing (World Bank 2013d). The report demonstrated growing levels of inefficiency found no statistically significant effects on student learning although it did appear to in public education spending driven in part by the combination of a large number of small have an effect on the attendance of some teachers. primary schools and staffing standards that did not take this into account. The authors argue that the better impact of the participatory scorecard was driven These inefficiencies are likely to be exacerbated by current demographic shifts in Indonesia. primarily by fostering greater cooperation between the school and the local Over the next 20 years the size of the primary and secondary school age population will community rather than because of differences in the information contained on the continue to fall and with it the need for primary and secondary school places. Moreover these different scorecards. These findings suggest that participatory methods to develop trends will be uneven in part due the rapid rates of urbanization that are being projected school report cards may improve education outcomes and strengthen the supporting for Indonesia. For example, it is estimated that over two-thirds of the population will live role of school committees. in urban areas by 2025 up from about a half in 2012 (World Bank 2014a). These shifts coupled with existing ways of providing schools and school places are having a profound Source: Barr, Mugisha et al. (2012) impact on the efficiency of the education system. For example, between 2010 and 2013, the proportion of schools in East Java with fewer than 120 students increased form 39 percent to Given the current weaknesses in school based management it is important to also make the 47 percent. These reductions have not been accompanied by significant declines in teacher best use of existing systems to monitor the use of BOS funds. Recent studies have shown numbers and will have resulted in large increases in the education costs per student and the relatively strong link that schools have with local education offices and the potential increased inefficiency. role they can play in both providing support for the use of funds and holding schools accountable (World Bank 2012e). Local education offices have BOS teams that are tasked Countries facing similar challenges to these have reformed financing mechanisms to shift with monitoring the program, providing support to schools for fund use, ensuring funds are the bulk of funding allocations from inputs (e.g. teachers, textbooks etc.) to outputs (e.g. used appropriately and providing a focal point for BOS grievance mechanisms. Beyond the number of students taught in a year). For example, during the 1990s and early 2000s school committees local government BOS teams are really the only external check on school many countries in East and Central Europe faced significant challenges brought about Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Strengthening the BOS Program 96 through declining population growth and increased rates of out-migration. As school-aged populations decreased, the number of schools and teachers remained unchanged because References decisions on the allocation of public resources were input-based. This reduced the efficiency of public education spending and led to significant imbalances in funding across regions. In response, a number of countries introduced school funding formula that covered the bulk of education funding (e.g. including teacher salaries) and were allocated mainly on the basis of the number of students or the size of the school age population. This had the effect of increasing allocations to urban areas that were experiencing rapid population growth and less to rural areas with declining populations. (Alonso and Sanchez 2011). The introduction of formulas of this kind also meant that efficiency gains were maintained without the need for one-off adjustments to allocations that were frequently difficult to achieve and politically costly. ACDP (2013). Analysis of School Operational Funds. Innovations of this kind could also be useful in Indonesia as a way of allocating a greater Alonso, J. D. and A. Sanchez, Eds. (2011). Reforming education finance in transition countries: proportion of overall education funds. As Chapter 1 showed, local governments provide the six case studies in per captita financing systems, The World Bank. bulk of education financing to schools and they could use formulas of this kind as the basis of allocating all funds, including teacher salaries, to schools. This would not only reduce Amin, S., J. Das, et al., Eds. (2008). Are you being served. the fragmentation of school financing (see Figure 1.2) but would also build in an automatic check to ensure more efficient input use because schools would receive funding based on the Arze del Granado, F., W. Fengler, et al. (2007). Investing in Indonesia’s education: Allocation, number of children actually taught. Equity, and Efficiency of Public Expenditures. Policy Research Working Paper The central government could also expand the BOS program to incorporate other central AusAID ERF (2011). School grants and school-based management. Current Issues in government funding for primary and secondary education and provide this greater level Education. Canberra. of resources to schools directly. Alternatively, formula funding mechanisms of this kind Baines, S. (2005). “Towards more transparent financial management: scholarships and could be used as part of the broader inter-governmental transfer system to allocate central grants in Indonesia (Ethics and corruption in education series).” Paris: IIEP-UNESCO. government education funds (including BOS) and transfers to local governments. Local governments could add this to local sources of resources and allocate to schools on the basis Barr, A., F. Mugisha, et al. (2012). Information and collective action in community-based of a single formula. This would also eliminate coordination issues between central and local monitoring of schools: Field and lab experimental evidence from Uganda. government grant programs and reduce the fragmentation of the overall education budget. Benveniste, L. and J. Marshall (2004). “School grants and student performance: Evidence The formula funding approach pioneered by the BOS program has spread to all levels of From the EQIP project in Cambodia.” Unpublished manuscript, World Bank, Washington, education and provides a vital source of funding for most education institutions. In the DC. 10 years that BOS has been operating it has established itself as a program that is able to deliver resources to schools on a regular and timely basis. It is popular among parents Blimpo, M. P. and D. Evans (2011). School-based management and educational outcomes: and is becoming increasingly well known. Other countries having successfully established lessons from a randomized filed experiment. Unpublished manuscript. programs of this kind have further developed them to address other education challenges that now confront Indonesia. Given the initial success of the BOS program it is perhaps Bruns, B., D. Filmer, et al. (2011). Making schools work: new evidence on accountability timely to explore how the BOS program and the mechanisms it has introduced for allocating reforms. Washington D.C., The World Bank. and managing resources can be adapted to make an even bigger contribution to improving education outcomes in Indonesia. Caldwell, B. J. (2005). School-based management, International Institute for Educational Planning Paris. Cambridge Education (2014). Review of Implementation of School Council Policy 2013. Cerdan-Infantes, P. and D. Filmer (2014). Information, knowledge and behavior: Evaluating alternative methods of delivering school information to parents. Chang, M.-C., S. Shaeffer, et al. (2013). Teacher Reform in Indonesia: The Role of Politics and Evidence in Policy Making. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 97 98 99 Chen, D. (2011). School-based management, school decision-making and education Ministry of National Education (2007). Standards for Education Management by Primary outcomes in Indonesian primary schools. Policy Research Working Paper No. 5809. and Secondary Education Units. Jakarta, Indonesia. Washington D.C., The World Bank Ministry of National Education (2009). Evaluation on the Effectiveness of the BOS (School Chowdry, H. and L. Sibieta (2011). School funding reform: an empirical analysis of options Operational Aid), Center for Educational Policy and Innovation Research. for a national funding formula OECD (2013). Pisa 2012 Results: What Students Know and Can Do: Student performance Cuadra, E. and J. M. Moreno (2005). Expanding Opportunities and Building Competencies in reading, mathematics and science. Paris, Programme for International Student for Young People: A New Agenda for Secondary Education, ERIC. Assessment. Organisation for Economic Co-operation and Development. 1. Das, J., S. Dercon, et al. (2011). School Inputs, Household Substitution and Test Scores. OFSTED (2014). The pupil premium: an update. Policy Research Working Paper No. 5629, World Bank. Patrinos, H. A. and T. Fasih (2009). Decentralized decision-making in schools: The theory and Deffous, E. (2011). Can school grants lead to school improvement? An overivew of evidence on school-based management, World Bank Publications. experiences in five countries, International Institute for Educational Planning. Pradhan, M., D. Suryadarma, et al. (2011). Improving Educational Quality through Fazekas, M. (2012). School funding formulas: Review of main characteristics and impacts. Enhancing Community Participation: Results from a Randomized Field Experiment in OECD Education Working Papers No. 74, OECD. Indonesia. Policy Research Working Paper 5795. . Washington, DC, World Bank. Fiszbein, A., N. Schady, et al. (2009). Conditional cash transfers: reducing present and future Ridao-Cano, C. and D. Filmer (2004). Evaluating the performance of SGP and SIGP: A poverty. Washington D.C., The World Bank. review of the existing literature and beyond. Human Development Sector Working Paper Series T. W. B. Human Development Sector Unit. Gertler, P., H. Patrinos, et al. (2010). “Parental Empowerment in Mexico: Randomized Experiment of the Apoyo a la Gestión Escolar (AGE) in Rural Primary Schools in Shaeffer, S. (2013). BOS Training: Its Implementation, Impact, and Implications for the Mexico: Preliminary Findings.” Unpublished manuscript, World Bank, Washington, DC. Development of Indonesia’s Education System. An Independent Review -- Prepared for AusAID Indonesia. Gertler, P., H. Patrinos, et al. (2008). Empowering parents to improve education: evidence from rural Mexico. The World Bank, Policy Research Working Paper Series: 3935. Shoraku, A. (2008). Educational Movement toward School-based Management in East Asia: Cambodia, Indonesia and Thailand. Background report for the UNESCO Global Ghozali, A. (2005). Educational Cost and Finance in Indonesia. Monitoring Report. Government of Indonesia (2003). Act of the Republic of Indonesia Number 20, Year 2003 Skoufias, E. and J. Shapiro (2006). The pitfalls of evaluating a schools grants program using on National Education System. non-experimental data. Policy Research Working Paper Series 4036, World Bank. Heyward, M. O., R. A. Cannon, et al. (2011). “Implementing school-based management SMERU (2006). “A Rapid Appraisal of The PKPS-BBM Education Sector: School in Indonesia: impact and lessons learned.” Journal of Development Effectiveness 3(3): 371- Operational Assistance (BOS).” 388. World Bank (2007). What Do We Know About School-Based Management? Khattri, N., C. Ling, et al. (2010). The Effects of School-based Management in the Philippines: An Initial Assessment Using Administrative Data. Policy research working World Bank (2008). Assessment In Preparation Of BOS KITA. Jakarta. Indonesia paper series. No. 5248, The World Bank. World Bank (2010). Making BOS effective under decentralisation. Policy Brief. Malen, B., R. T. Ogawa, et al. (1990). “What do we know about school-based management? A case study of the literature—A call for research.” Choice and control in American World Bank (2011). The BOS Review. education 2: 289-342. World Bank (2012a). Bantuan Siswa Miskin cash transfers for poor students. McEwan, P. (2013). Improving Learning in Primary Schools of Developing Countries: A World Bank (2012b). Bantuan Siswa Miskin: Cash Transfers for Poor Students. S. a. p. a. Meta-Analysis of Randomized Experiments. p. e. r. b. p. n. 5. Ministry of Education and Culture (2012). Regulation on the Contributions and Costs of World Bank (2012c). The BOSDA Improvement Program: Enhancing equity and Basic Education. Ministry of Education and Culture. Jakarta, Indonesia. No. 44. performance through local school grants Jakarta, World Bank. Ministry of Education and Culture (2014). BOS manual. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia 100 Growth. Effectiveness. for Negeri Sembilan. Empirical Investigation. Philippines. Policy Research Working Paper Series Number 6247. World Bank (2012e). School based management in Indonesia. Jakarta. World Bank (2013a). BOSDA review, The World Bank. Unpublished report. of local education governance in 50 Indonesian districts The World Bank. Indonesia. Education Public Expenditure Review. Jakarta, The World Bank. World Bank (2014a). 2014 Indonesia Development Policy Review: Avoiding the Trap. Assisatance - Knowledge Improvement for Transparency and Accountability Project. World Bank (2013c). Local governance and education performance: a survey of the quality Yamauchi, F. and Y. Liu (2012). School Quality, Labor Markets and Human Capital World Bank (2012d). Optimizing Subnational Performance for Better Services and Faster on students’ achievements: evidence from the Philippines.” Journal of Development World Bank (2013b). Implementation Completion and Results Report: School Operational Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia World Bank and AusAID (2013). School Based Management in the Philippines: An World Bank (2014b). Malaysia: Policy note on school funding formula and pilot options World Bank (2013d). Spending more or spending better: Improving education financing in Investment: Long-term Impacts of an Early Stage Education Intervention in the Yamauchi, F. (2014). “An alternative estimate of school-based management impacts Appendix Table 1: Comparing education spending in the education and consumption modules Education Module Period of Recall Period of Recall Consumption Module Period of Recall (Collected every 3 years) (‘03) (‘06, ’09, ‘12) (Collected annually) (Feb 02-07 and Mar 08-13) School Year 2005/2006, 2008/2009 Registration cost July - Dec Registration Cost One month prior to the survey and 2011/2012 Appendix School Fee (SPP) July - Dec Jan - June School Fee (SPP) and school Combined into one answer and committee conducted one month prior to School Committee July - Dec Jan - June the survey Laboratory Fees July - Dec Jan - June - Student Organization and Consumption Modules July - Dec Jan - June - Contribution Test/Evaluation Fee July - Dec Jan - June - Course fees (including July - Dec Jan - June Course fees One month prior to the survey enrichment) Other school fees (proficiency, Other fees July - Dec Jan - June One month prior to the survey course, test etc.) Student Worksheet (LKS)* N/A Jan - June - Teaching Materials Expenditure July - Dec Jan - June - Uniform and Sport Expenditure July - Dec Jan - June - Textbooks, photocopies of Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Textbooks/Manual/Lecture notes July - Dec Jan - June One month prior to the survey learning materials Stationary and other equipment July - Dec Jan - June Stationary One month prior to the survey Jan-June (2006) and 2 trips (2009 Transportation cost July-Dec - and 2012) Pocket Money N/A N/A (2006), daily (2009 and 2012) - Appendix Table 1: Comparing education spending in the Education Note: *only available in 2012. Spending from the consumption module is calculated based on information one month prior to the survey in line with the BPS approach. Due to changes in the survey methodology, spending between 2011and 2013 from the consumption module is computed from total household expenditures on education in the three months before the survey. - indicates that these items were not explicitly asked about in the consumption module. 101 102 103 Appendix Figure 1: Composition of per-student spending (IDR Appendix Figure 2: Trends in household education spending per thousands), education module student in government and non-government schools Primary School Students Primary School Students Government Non-government Raw trends in spending per student Regression-adjusted trends in spending per student 2,500,000 2,000,000 1,806,087 100% 100% 108 149 2,204,412 44 47 345 585 1,800,000 116 292 IDR (2012 constant prices) 61 68 67 85 2,000,000 1,600,000 1,390,107 80% 80% 1,283,909 178 70 1,496,638 1,400,000 221 335 1,145,161 155 257 1,500,000 1,200,000 60% 175 226 60% 224 1,190,071 1,000,000 56 1,000,000 852,977 800,000 133 716,579 40% 122 40% 548,723 600,000 782,193 530,323 525,115 683,065 251 458 749 1,293 500,000 657,238 665,551 621 400,000 20% 184 236 20% 200,000 155 0 0 0% 0% 2003 2006 2009 2012 2003 2006 2009 2012 2003 2006 2009 2012 2003 2006 2009 2012 Junior Secondary School Students Junior Secondary School Students Government Non-government Raw trends in spending per student Regression-adjusted trends in spending per student 2,000,000 1,846,414 1,887,062 2,500,000 1,765,140 1,800,000 1,594,891 IDR (2012 constant prices) 100% 100% 2,083,833 1,600,000 333 308 342 289 2,000,000 1,712,670 423 547 399 505 1,704,751 1,400,000 1,567,231 1,590,260 1,577,689 1,515,713 80% 107 80% 100 105 1,200,000 1,444,923 106 73 1,500,000 203 233 1,000,000 60% 387 82 60% 321 423 1,440,409 1,337,808 1,355,800 319 208 224 1,000,000 1,222,169 800,000 196 600,000 40% 184 40% 400,000 500,000 1,273 200,000 682 827 888 752 20% 537 408 531 20% 0 0 2003 2006 2009 2012 2003 2006 2009 2012 0% 0% 2003 2006 2009 2012 2003 2006 2009 2012 Senior Secondary School Students Senior Secondary School Students Raw trends in spending per student Regression-adjusted trends in spending per student Government Non-government 3,500,000 3,268,573 3,146,096 3,169,299 2,929,287 3,200,000 2,804,473 3,000,000 IDR (2012 constant prices) 100% 100% 3,100,000 3,151,143 812 2,500,000 3,096,225 782 743 686 773 728 2,678,081 2,974,718 2,951,488 740 645 2,484,119 3,000,000 80% 80% 78 2,000,000 2,486,297 2,453,077 117 115 109 2,900,000 2,962,380 2,866,409 138 273 90 236 305 60% 526 265 60% 471 545 280 1,500,000 2,800,000 444 243 1,000,000 2,700,000 2,760,472 40% 40% 2,158 1,355 1,516 500,000 2,600,000 1,164 1,390 1,185 1,530 1,806 20% 20% 0 2,500,000 2003 2006 2009 2012 2003 2006 2009 2012 0% 0% 2003 2006 2009 2012 2003 2006 2009 2012 non government students government students fees materials uniform transport Notes: The left hand side of this table shows raw data. The right hand side shows the results of a regression-based analysis. The dependent variable in the regression model is per-student spending. Education spending is calculated from information on household spending for each child attending Note: All amounts are in constant 2012 Indonesian Rupiah and adjust for price differences between school in at the corresponding school level using fees, uniforms, materials and transport (as described urban and rural areas. in Figures 3-5). Source: Education Module. Susenas. Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia Assessing the Role of the School Operational Grant Program (BOS) in Improving Education Outcomes in Indonesia