th Autnoou egoni Publc Epenitr I InsituinlRve A*S fAR ic Educatio npenAing A JOINT REPORT BY DFAT AUSTRALIA AND THE WORLD BANK AustralianmGovemnment THE WORLD BANK X, Department of I oreign Allrairs and Trade IERRO * IDA Cover photo courtesy of GOV.ph http://www.gov.ph/2014/08/22/australian-govt-funded-program-turns-over-24-newly-rehabilitated-classrooms-in-sulu/ PHILIPPINES: Making Education Spending Count for the Children of Autonomous Muslim Region of Mindanao Public Expenditure & Institutional Review for ARMM Basic Education A JOINT REPORT BY AUSTRALIA DFAT AND THE WORLD BANK Preface T his ARIM Public Expenditure and Institutional Review (PEIR) resulted from a request by the Governor Mujiv Sabbihi Hataman for assistance in improving public resource use and basic educational outcomes in the region. We would like to thank the Governor and his office for their openness to carry forward a frank and forward looking effort to improve governance and performance in the region's basic education sector. In particular, Executive Secretaries Attys. Anwar Malang and Laisa Alamia provided support and guidance during the process. ARIM Secretary of Education Jamar Kulayan, Assistant Secretaries Noor Mohammad Saada and Maisara Latiph, Alfadar Pajiji (Chief of Staff), Mosbicak Ibrahim (Budget Chief), Analee Biruar (Accountant), Rosana Hassan (Cashier), Virginia Besana (Planning), and Marjuni Maddi (SPCO Managing Director) all spent a great deal of time assisting the team. In addition, many other officials from DepEd ARMM, school division and district administrators and individual staff and teachers gave their time and enthusiasm during both research and workshop activities. In DepEd National Under-Secretary Francisco Varela, Assistant SecretaryJesse Mateo, and Assistant Secretary Armando Ruiz participated in the project workshop on strategic budgeting and offered their time and expertise on DepEd National policies and procedures. In DBM, special thanks are due to Director Tina Rose Marie Canda (BMB - D), Grace de los Santos (Division Chief in charge of ARIM), and Assistant Director Dante de Chavez (BMB-B in charge of DepEd National) for their participation and provision of key data. The Consortium of Bangsamoro Civil Society (CBCS) ably implemented the schools survey under challenging circumstances. Without the efforts of Guiamel Alim, Wahab Guialal, Tarhata Sambutuan, Amier Dodo and over 90 dedicated enumerators from all parts of the ARIM, the study would have been impossible. We also thank all the respondents in individual schools who gave their time during the survey. The BE-PEIR project is overseen by Teresita Felipe (Education Specialist, Australian Embassy Manila), Kai Kaiser, Lynnette Perez, and Matthew Stephens (WB).'The report was produced by Hamish Nixon, Alma Mariano, Futoshi Yamauchi, Roy Tordecilla, Pinky Esguerra, Miro Capili, Carole Belisario, and Lea Neri. The following reviewers gave helpful comments: Samer Al-Samarrai, Elizabeth King (Sector Director, HDNED), Yongmei Zhou (Manager, Fragile and Conflict-affected Countries Unit (OPCS)), Daan Pattinasarany (Senior Economist, EASPI), Francisco Varela (Undersecretary for Finance, Department of Education), Bruce Stacey (Advisor, Performance Informed Budget (PIB), DBM), Prof. Paul Hutchcroft (Adviser, Australian Embassy Manila), Sam Chittick (Adviser, Australian Embassy Manila), Evelyn Daplas (Portfolio Manager, Australian Embassy Manila), and Assistant Secretary Noor Mohammad Saada (DepEd-ARMM). Financial support for the exercise came from a joint Australia WB Philippines Development Trust Fund and the World Bank State and Peacebuilding Fund. Table of Contents Acronyms i EXECUTIVE SUMMARY iii KEY FINDINGS IV SUMMARIZED RECOMMENDATIONS V TOWARDS BANGSAMORO IX CHAPTER ONE: INTRODUCTION 1 STRUCTURE OF THE BE-PEIR PROJECT 3 CHAPTERTWO: ARMM DEVELOPMENT & BASIC EDUCATION CONTEXT 5 HISTORICAL DEVELOPMENT OF THE ARMM 5 STRUCTURE OF THE AUTONOMOUS REGION IN MUSLIM MINDANAO 7 ARMM DEVELOPMENT CONTEXT 9 WIDESPREAD CONFLICT 9 WEAK GOVERNANCE 9 ARMM EDUCATION CONTEXT 11 EDUCATION OUTCOMES 12 CHAPTERTHREE: PUBLIC EXPENDITURE MANAGEMENT IN ARMM 16 FRAGMENTED PUBLIC EXPENDITURE MANAGEMENT OF ARMM RESOURCES 17 POOR INTEGRATION OF ARMM REGIONAL AND NATIONAL BUDGET PROCESSES 22 INCONSISTENCY BETWEEN ARMM REGIONAL "DEVOLUTION"AND NATIONAL DECENTRALIZATION 25 WEAK STEWARDSHIP AND ACCOUNTABILITY IN HANDLING PUBLIC FUNDS 27 CHAPTER FOUR: BUDGETING FOR BASIC EDUCATION IN ARMM 29 AGGREGRATE BASIC EDUCATION SPENDING LEVELS IN THE ARMM 29 NON-GOVERNMENT BASIC EDUCATION SPENDING 29 GOVERNMENT SPENDING ON BASIC EDUCATION 32 COMPOSITION OF GOVERNMENT SPENDING ON BASIC EDUCATION 33 PERSONAL SERVICES 35 MOOE AND THE "ROUTINE OPERATIONS AND MAINTENANCE GAP" 37 TEXTBOOKS 42 OTHER PROGRAMS 43 CAPITAL OUTLAY 47 LOCAL GOVERNMENT UNIT CONTRIBUTIONS 49 THE ARMM PEM SYNDROME AND DISTORTIONS IN BUDGET COMPOSITION 50 CHAPTER FIVE: ISSUES IN BUDGET EXECUTION AND SERVICE DELIVERY 55 BUDGET EXECUTION AND THE ARMM PEM SYNDROME 55 PERSONAL SERVICES AND PAYROLL MANAGEMENT 56 MOOE 60 EXECUTION PROCESSES OF CENTRALLY-MANAGED FUNDS 61 FRONTLINE SERVICE DELIVERY: THE SCHOOLS SURVEY 63 TEACHER ABSENCE AND GHOST TEACHERS 65 PUPIL ABSENCE 66 CHALLENGES IDENTIFIED BY SCHOOLS 71 INCONSISTENT AND INEQUITABLE LGU AND COMMUNITY CONTRIBUTIONS TO EDUCATION 72 CHAPTER SIX: CONCLUSIONS AND RECOMMENDATIONS 74 SUMMARY ASSESSMENT OF THE PEM SYSTEM FOR BASIC EDUCATION IN ARMM 74 INABILITY TO LINK POLICY PLANNING AND BUDGETING 74 LACK OF BUDGET COMPREHENSIVENESS, INTEGRATION AND PREDICTABILITY 76 SLOW AND INEFFECTIVE BUDGET EXECUTION 76 WEAK ACCOUNTABILITY AND REPORTING 77 IMPROVING PUBLIC EXPENDITURE MANAGEMENT FOR BASIC EDUCATION IN ARMM 79 ENSURING QUALIFIED TEACHERS ARE BUDGETED, PAID, AND PRESENT 80 CORRECTING THE "ROUTINE MOOE GAP" 80 INTEGRATING CENTRALLY-MANAGED PROGRAMS WITH ARMM'S EDUCATION BUDGET 81 IMPROVING ARMM PUBLIC EXPENDITURE MANAGEMENT 83 STRENGTHENING THE INTRA-ARMM BUDGET PROCESS 83 INTEGRATING REGIONAL AND NATIONAL BUDGET SYSTEMS 85 DEVELOPING A FRAMEWORK OF HORIZONTAL AND DOWNWARD ACCOUNTABILITY 86 CONSIDERATIONS FOR BANGSAMORO 89 THE IMPORTANCE OF IMPLEMENTATION 89 FISCAL MANAGEMENT AND INTER-GOVERNMENTAL FRAMEWORK 89 FISCAL FRAMEWORK FOR LGUS AND REGIONAL AUTONOMY 90 COORDINATION BETWEEN BTC, ARMM AND NATIONAL GOVERNMENT 90 PRINCIPLES FOR FUTURE ENGAGEMENT 91 REFERENCES 93 Tables Table 1: Indicative Action Plan viii Table 2: LGUs in ARMM 8 Table 3: Principal sources of government funds to ARMM'Ihousands Pesos) 21 Table 4: PAP structures for Basic Education in National Government and ARMM 24 Table 5: Indicative key donor education sector support to ARMM (thousand Pesos) 30 Table 6: Aggregate government education expenditures in ARMM (thousand Pesos) 32 Table 7: Locus of control of government spending by economic category, FY 2013 34 Table 8: Centrally-managed fund allocations for ARMM, FY 2013 44 Table 9: Budgeting basis for education components 51 Table 10: Reasons for school closure during survey 64 Table 11: Provincial/division teacher absence rates 65 Table 12: Presence by teaching grade 66 Table 13: Absence rates (%) of students in classes in session by province, division and gender 67 Table 14: Student absence (%) by grade and gender 69 Table 15: Frequency of top challenges mentioned by school heads 71 Table 16: Summary of key weaknesses in the basic education PEM system 78 Table 17: Indicative ARMM regional budget preparation calendar (FY 2014) 84 Table 18: Indicative Action Plan 88 Figures Figure 1: A framework for improvement x Figure 2: ARMM Government Organization 7 Figure 3: Net enrolment rate (%), SY 2011-12 12 Figure 4: Educational Attainment of Females in ARMM and selected regions (2008) 13 Figure 5: Educational Attainment of Males in ARMM and selected regions (2008) 14 Figure 6: Funds flowing to ARMM by locus of budget control (PHP Millions) 18 Figure 7: Household Expenditure on education, 2009 (percent by region) 31 Figure 8: Nominal vs. Real government education expenditures in ARMM, 2009-12 33 Figure 9: PS per pupil by region (SY 2011-12 enrolment, FY 2013 budget) 36 Figure 10: Primary and secondary teacher-student ratios by region (SY 2011-12) 37 Figure 13: Per pupil MOOE in regular operations budgets by region (SY 2011-12 enrolment, FY 2013 budget) 40 Figure 14: Lack of operating and maintenance resources 41 Figure 15: Textbook budget comparison per pupil (SY 2011-12 enrolment, FY 2013 budget) 43 Figure 16: Total allotment for "Madrasah Education" by region, FY 2013 45 Figure 17: Planned ALS Allocation per Target Learner 2013 (PHP per Learner) 46 Figure 18: Basic Education Facilities Fund (BEFF) allotment per pupil by region, FY 2012 47 Figure 19: DepEd national and overall ARMM education expenditure composition, FY 2012 49 Figure 20: % LGU spending on education, culture & arts, and manpower development (2011) 49 Figure 21: The two main channels of ARMM education budget preparation 52 Figure 22: Payroll Cheque Collection in ARMM 58 Figure 23: Prevalent pupil absence 67 Figure 24: Teacher and student presence (%) by division 69 Figure 25: Improving ARMM education 79 Figure 26: Transition to ARMM responsibility for centrally-managed education funds 82 Figure 27: Relationship between national and regional budgeting guidelines 83 Figure 28: Decentralization and the service delivery framework 86 Figure 29: A framework for improvement 91 Boxes Box 1: Selected provisions of the Framework Agreement on Bangsamoro 6 Box 2: Selected legal provisions for ARMM's education responsibilities 10 Box 3: Selected provisions of RA 9054 17 Box 4: Expenditure norms for MOOE allocation in DepEd National 53 Box 5: Triangulation of HRIS, payroll and plantilla databases 59 Box 6: Areas for future research 92 Maps Map 1: Provinces of ARMM 1 Map 2: ARMM Education Divisions 12 Map 3: Student and Teacher Absence Rates by School Division 70 Acronyms ABM Agency Budget Matrix ADM Alternative Delivery Modes ALIVE Arabic Language and Islamic Values Education ALS Alternative Learning Systems ARG ARMM Regional Government ARMM Autonomous Region in Muslim Mindanao BAR Budget Accountability Report BE Basic Education BEIS Basic Education Information System BEFF Basic Education Facilities Fund BESF Budget of Expenditures and Sources of Financing BLGF Bureau of Local Government Finance BTC Bangsamoro Transition Commission CO Capital Outlay DBM Department of Budget and Management DILG Department of Interior and Local Government DPWH Departments of Public Works and Highways DSWD Department of Social Welfare and Development EQUALLS Education Quality and Access for Learning and Livelihood Skills FAB Framework Agreement on the Bangsamoro GAA General Appropriations Act (National Budget Law) GASTPE Government Assistance to Students and Teachers in Private Education GDP Gross Domestic Product HDI Human Development Index HRIS Human Resources Information System IRA Internal Revenue Allotment LGC Local Government Code (RA 7160) LGU Local Government Unit MFO Major Final Output MILF Moro Islamic Liberation Front MMA Muslim Mindanao Autonomy (Act) MNLF Moro National Liberation Front MOOE Maintenance and Other Operating Expenses NBM National Budget Memorandum NCA Notice of Cash Allocation NDHS National Demographic and Household Survey OPIF Organizational Performance Indicator Framework Acronyms (cont'd.) ORG Office of the Regional Governor PEIR Public Expenditure and Institutional Review PEM Public Expenditure Management PHP Philippine Peso PI Performance Indicators PS Personal Services QSDS Quantitative Service Delivery Survey RA Republic Act RBMO Regional Budget and Management Office REDPB Regional Economic Development Planning Board RG Regional Governor RLA Regional Legislative Assembly SAOB Statement of Allotments, Obligations and Balances SARO Special Allotment Release Order SBM School Based Management SBP School Building Program SDO School Division Office SDS School Division Superintendent SEF Special Education Fund SY School Year EXECUTIVE expenditure and education delivery and SUMMARY outcomes. It has three inter-related objectives: To improve the quality of basic edu- cation in ARMM by diagnosing public expenditure management issues and sug- gesting improvements; Children in the Autonomous Region in To improve public expenditure man- Muslim Mindanao (ARMM) have long agement processes and practices in the suffered from the effects of underdevelop- region; ment, neglect, poor governance, and conflict. To improve fiscal autonomy ar- The proportion of impoverished families in rangements and inform planning for the the ARMM is the highest in the country, proposed Bangsamoro political entity. amounting to over twice the Philippines average. The region is beset by multiple forms This review is one component of a joint of violence, has a history of weak governance Australian Aid and World Bank supported and corruption, and depends heavily on a project to enhance fiscal transparency and central government with whom it shares both budget effectiveness in the ARMM (particu- mutual distrust and ties of patronage politics. larly in the basic education sector), reinforce Since 2012, the new regional administration the reform agenda for ARMM, and con- has pursued a reform and anti-corruption tribute to preparations for establishing the agenda focused on strengthening services and Bangsamoro. Basic education (BE) is defined relations with the national government. Ne- broadly to include elementary and second- gotiations between the Moro Islamic Libera- ary schooling as well as other programs for tion Front (MILF) and government have also school aged children such as Alternative produced a "Framework Agreement" and road Learning Systems (ALS) and Islamic educa- map to establish the "new political entity" tion. The project builds on prior support to known as Bangsamoro. DepEd ARMM through the Australian Aid Basic Education Assistance for Mindanao This Basic Education Public Expenditure and (BEAM) activity. Institutional Review (BE-PEIR) examines the public expenditure management (PEM) The project was initiated at the request of system for basic education in the ARMM. It the ARMM regional government to enhance describes ARMM's Public Expenditure Man- dialogue and reduce the trust gap between agement "Syndrome," which is characterized ARMM government and national and by longstanding socioeconomic and politi- regional stakeholders, enhance efficiency and cal issues, and assesses its impacts on public effectiveness of spending in basic education, iv and support institutional reforms. It has three Grade 1 made it to grade 6 against a na- components: tional average of 73.5 percent. At secondary level, around 45 percent of 1st Year students * This Public Expenditure and Institu- continued studying until 4th year, whereas tional analysis and report; the national average was 78.8 percent. In * Technical assistance to ARMM Re- other terms, only 1 out of every 10 students gional Government and DepEd ARIM who began primary education in ARMM will on planning, budgeting, and dialogue with graduate from high school. national budget institutions; and * Technical assistance on issues includ- Such low completion rates decrease the ef- ing vulnerabilities in the education payroll: ficiency of a given level of education inputs, "ghost teachers," Human Resources Infor- amplifying the impact of problems in the mation and Payroll systems reform, and public expenditure management system. Due a facilities survey focused on teacher and to the attrition of students leaving the system, student absence. it takes 32 percent more spending and effort to produce a single high school graduate in Positive change is contingent on the peace ARIM than elsewhere in the Philippines, process, reform orientation of the regional and twice the effort to produce a Grade Six government, and improvements in the graduate in ARMM as elsewhere. relationship between ARMM and national authorities. These opportunities can be best In 2011-12, the average achievement test seized by enhancing ARMM and national score for Grade 6 students in ARMM was institutions in the short-term, and learning 54.1 percent, whereas the national average from the ARMM's history to inform the was 66.8 percent. The average achievement design of enhanced autonomy institutions for score of 4th Year high school students was the Bangsamoro. 37.1 percent, more than ten points below the national average of 48.9 percent. Literacy Key findings rates for people aged 10-64 were estimated ARMM suffers from extremely poor edu- to be 71.6 percent in 2010, well below the cation outcomes (Chapter 2). Enrolment, national average of 86.4 percent. completion, and attainment rates for elemen- tary education lag behind most regions in There have been longstanding issues on the the Philippines. These outcomes perpetuate effectiveness and quality of education delivery, "poverty traps" and decrease the prospects of including ghost teachers, poor infrastructure, children in the region to enjoy better living and lack of teaching and operating resources standards and attain secure employment. in schools. The project carried out a represen- From 2011-12, only 23 percent of students in tative surprise survey of schools in ARIM. The survey confirmed widespread school clo- with budgets, weak budget comprehensive- sures and absence rates for teachers of about ness, slow budget execution, as well as a lack 31 percent, and pupils of 29 percent. Areas of accountability and data availability. where teacher absence was high also generally experienced high student absence, and these Summarized recommendations problems appear particularly in Lanao del Sur For the 2014 and 2015 budget years, the (Chapter 5). ARG, DBM, and DepEd could take a num- ber of concrete steps to improve education After years of underfunding, the aggregate delivery and manage public expenditure: amount of regional funding for education inputs (sourced almost exclusively from Continue to strengthen payroll and national transfers) is coming in line with human resources.'The DepEd ARM national levels under the Aquino adminis- and ARG should build on substantive ac- tration. However, the effectiveness of public tions already initiated to clean the teacher expenditure for ARMM is hampered by an payroll and introduce systems to improve ARIM Public Expenditure Management its integrity. These include continuing to "Syndrome," discussed in detail in Chapter 3. integrate and cross-reference payroll and This is characterized by: Human Resources Information Systems (HRIS), develop internal structures and * Fragmentation of planning, budget- procedures for human resource manage- ing, and execution processes across a range ment, streamline merit-based recruitment of national, regional, and local actors and processes through appointment boards, and systems; communicate these achievements to both * Poor integration between regional ARMM citizens and the national govern- government and national budgeting sys- ment. An additional priority would be to tems including the OPIF framework; introduce incentives to improve teacher * Inconsistencies between autonomy attendance, which appears to influence arrangements for ARMM and national pupil attendance as well. A start would be decentralization under the local govern- utilizing attendance records during payroll ment code; processing through the URIS. * Weak stewardship and accountability Tackle a routine MOOE gap in the for public funds. region. As a result of budget fragmenta- tion, the level of operational "routine" These problems go beyond typical public MOOE available to schools is significant- expenditure challenges found throughout ly smaller in ARNM than elsewhere, and the Philippines, because of the exceptional elementary schools do not receive regular burdens they cause for poor linkage of plans MOE funds. DBM, ARG, and DepEds vi ARMM and National should negoti- tance to ARMM could be included in the ate an arrangement by which operating ARG budget appropriation for direct ex- MOOE levels can be based on a similar ecution in the region, taking into account norm to the national level, taking into accountability demands. Madrasah educa- account concerns about enrolment data tion is a good candidate, as almost half of quality. This will require increased budget its national appropriation is dedicated to allocations from the national government, ARMM at this time. In the medium-term as well as clearer policies on the alloca- (2016 budget), the ARMM or Bangsam- tion, delivery, and use of the MOOE and oro could begin making allocations among improved reporting by school divisions programs according to its priorities. and districts. Until a more permanent policy is in place, an interim measure such In addition to these short-term steps in as a capitation grant similar to School the education sector the ARMM Regional Based Management could be considered. Government and its partners can address Given ARMM's history, special empha- public expenditure management more broadly sis must be placed on governance and through the following initiatives: monitoring. Renewed MOOE payments could be linked to a grass-roots informa- Strengthen the intra-ARMM bud- tion campaign to encourage bottom-up get process. A more effective process of accountability through citizen report cards budgeting within the ARMM Regional or mobile reporting. Government is needed to shift the ORG * Plan and implement a transition of budget team's role away from preparing centrally managed education programs to the budget submission towards ensuring the region. DBM, ARG, and both DepEd the inputs of the ARMM sectoral agen- ARMM and National should commit to cies support regional plans and national the overall objective of eventually mov- requirements. To do so, the ORG should ing all or most funds for use in the region supplement the National Budget call with onto the ARG budget appropriation.'The guidelines setting out the responsibilities first step (already underway) is informa- of ARMM agencies, building in time for tion sharing between DepEd ARMM agencies to collate Budget Accountability and National about the basis for the al- Reports (BARs), and guiding preparation location of centrally-managed funds. For of ARMM agency spending proposals by the preparation of the 2015 budget the specifying data requirements and aiding DepEd National and DepEd ARMM integration with a regional performance should participate jointly in preparing the informed budgeting framework (see program budgets. At the same time, by below). Like a national budget call, these mutual agreement, a few funds of impor- guidelines can include: indication of prior- ity areas of government activity, programs accountability mechanisms to improve Vii and projects; general and specific guide- the efficiency and effectiveness of public lines for agency budget proposals; and services in ARVM.'These tools could in- calendar of budget preparation activities dude: increased budget transparency and within the region. publication of sector budgets or ARMM's Develop a regional performance overall budget (for example, increased budgeting framework that works for the MOOE funding could be publicized so region and is acceptable to national gov- that schools and communities know their ernment. The ARG budget currently has entitlement); feedback mechanisms such to conform to the national performance- as citizen report cards and community based budgeting framework which is score-cards, hotlines, and websites; pen- ill-adapted for the region. Performance odic survey and monitoring by civil society accountability needs to be balanced with organizations to identify key issues and the need for financial control, and adapted locations for attention. A second civil so- to the region's unique conditions - its ciety implemented school survey is already cross-sectoral responsibilities, and its par- being planned under this project, and will tial control over its resources. The DepEd be used to pilot CSO monitoring tools. In ARMM, ORG, and DBM could agree doing so, it is important to consider the to modify the national OPIF framework challenges for demand side interventions for ARIM to promote a simple and in the region. Teacher appointments have limited results-based program budget- been at the mercy of entrenched political ing framework that integrates activities patronage; parents and children have been budgeted through the ARG and national displaced; education stakeholders such agencies and relates them to an agreed set as parents, civil society, and government of objectives and results indicators. The reformers are subject to intimidation and current ARMNIM HELPs framework of the violence; and popular expectations of what ARIM regional government should form constitutes acceptable basic education the basis for elaborating this framework. The have been undermined by years of poor need for a regional performance manage- governance and performance. ment framework will only become greater as more fiscal control is shifted to ARMM, Table 1, explained in more detail in Chapter and subsequently the Bangsamoro. Six, summarizes these recommendations ac- * Include civil society and communi- cording to the Public Expenditure Manage- ties. Effective public services depend ment problems they are intended to confront, on both top-down and bottom-up ac- and potential indicators for improvement. countability.anhere is potential for social Vii Table 1: Indicative Action Plan Problem Actions Indicators/Milestones Improving ARMM Basic Education Public Expenditure Management Weak payroll integrity (ghost teachers, 9 Continue data clean-up on HRIS and e Completeness of HRIS & unqualified teachers, leakage of personal payroll HRIS-payroll congruence services) * Complete and institutionalize payroll and e Organizational restructure complete, HR manual establishment of personnel boards * Enhance HRIS-payroll integration * National integration of HRIS-payroll system Teacher & pupil absence * Incentivize teacher attendance and preven- * Absence rates tion of early dropouts * Include communities & civil society in monitoring Routine MOOE gap * Provide additional MOOE operational funds * Operational MOOE availability & agree a norm-based budgeting approach * Reporting on MOOE utilization with DBM/DepEd National * Introduce policy for distribution to ARMM schools * Establish reporting system to region & regional budget reporting to DBM * Include communities & civil society in monitoring CMF budgets unconnected to policy or * Increase ARMM involvement in budgeting * ARMM/DepEd National liaison activities regional conditions Centrally-Managed Programs, beginning with enhanced liaison capacity, participation in budget process CMF execution slow * Move key centrally-managed programs * Proportion of program budget appropriated to ARMM appropriation for direct release directly to ORG (Madrasah, ALS) * Reporting on program fund utilization * Agreed reporting benchmarks * Capacity support & staffing for program implementation (temporary & long-term) Low or unknown completion of classrooms * Clarify monitoring responsibilities for BEFF/ * School construction projects completed SBP * Involve communities & civil society in monitoring LGU education spending * Integrate LGUs into regional planning & * Level of LGU spending on education inconsistent & inequitable budgeting, agree regional LGU expenditure increased responsibilities for education * Increased information on LGU spending to region Improving ARMM Regional Public Expenditure Management Budgets unconnected to policy * Revise OPIF framework for ARMM * Develop framework of key indicators for & planning for region * Establish sectoral budget working groups HELPs, agree with national agencies and/or joint national/regional program-bud- * Pilot ARMM open data initiative get based on HELPs framework with results indicators * Improve budget transparency & include civil society & communities in monitoring ARG budget proposal weak * Strengthen intra-ARMM budget process * Issuance of ARMM budget circular with prior to ORG submission (calendar, guide- calendar & guidelines lines, hearings) * Timely submission of ARMM agencies * Increase RBMO role Towards Bangsamoro lishment of effective regional budget and ix management institutions. A key concern International experience suggests that auto- for the negotiators of the Bangsamoro nomy in itself is no guarantee of better gov- institutions should therefore be asking ernance and public services. The above short how should implementation be measured, term steps need to be buttressed by a strategy monitored and enforced, and by whom? for Bangsamoro to avoid the pitfalls of the Fiscal management capacity and ARd\M PEM syndrome by addressing each framework. As a subnational govern- of its dimensions: ment the ARMM has not had to produce macro-fiscal estimates of revenue or ex- 1. Planning and budgeting for the penditure over the medium-term, nor has fragmented sources of funds for the region it had any opportunity to make strategic need to be unified as far as possible, along- cross-sectoral allocation decisions. Un- side improved revenues; der Bangsamoro these two functions will 2. Regional budget processes need to likely become much more important due connect effectively with national systems; to an increase in the amount and flexibil- 3. LGUs need a clearer framework ity of revenue under regional control. So of responsibilities and reporting to the the institutionalization of regional macro- region; fiscal and cross-sectoral strategic planning 4. Monitoring and accounting of public capacity should be a key priority. funds need to be strengthened both within LGU fiscal framework. ARMM government and with communities, citi- LGUs exist in a kind of limbo, having zens, and children. revenues that derive from the national de- centralization framework, but few respon- Some lessons from ARMM's experience sibilities or accountability to the region. are relevant to the work of the Bangsamoro The relationship between the LGUs and Transition Commission (BTC) and others the region needs elaborating. In return for involved in negotiating the Basic Law and this increased accountability to the region, public finance provisions for the Bangsamoro: LGUs should be given a role in planning and budgeting. In short, there is a need The importance of implementation. for a fiscal decentralization framework The failure of fiscal autonomy in ARIM between the region and its LGUs. A key is not simply the result of insufficient au- question is whether reform of the IRA is tonomy or poor governance. The ARMM more feasible nationally or through the PEM syndrome is also the result of regional peace process. incomplete implementation of the Ex- BTC-ARMM-GPH coordination. panded Organic Act, such as the estab- The historical experience of AR1\4M x shows that autonomy in the context of tion should also be established between continued central support increases the ARMM authorities and the negotiating need for all actors to share information teams to consider the transition of existing and influence, and enable capacity build- institutions - particularly for large depart- ing. When these processes are not in place, ments such as DepEd ARMM. negative cycles of weak accountability, mistrust, and blame will harm citizens and Figure 1 summarizes these recommendations in the case of education, children. Empha- in relation to the objectives of improving sis should be given to establishing prob- basic education public expenditure man- lem-solving, coordination and monitoring agement, strengthening the region's PEM institutions between the Bangsamoro systems overall, and informing preparations and the national government. Coordina- for Bangsamoro institutions. Figure 1: A framework for improvement Improving ARMM Basic Education Improving ARMM Regional Public Expenditure Management Public Expenditure Management * Continue Payroll & HRIS reform & * Strengthen intra-ARMM budget process reduce absences F Revise OPIF to joint regional/national results " Tackle routine MOiE gap framework based on HELPs to connect policy and SPhasing in greater AReMM involvement budgeting in program buageting & execution * Involve civil society & communities sContext sensitive social accountabilyty * Address inconcistencies with LGUs Co the nations for Bangs i ro Mprvn ARM Bai EdctoJmrvn RMRgoa * (si igrae A involvement in prgrambudgting exeutio CHAPTER ONE: ity to learn, and the poor economic climate 01 INTRODUCTION keeps emerging graduates from employment, reducing returns to household investment and commitment to education.2 Historically low investment and institutionalized corruption further deepened these problems.3 Children in the Autonomous Region in The current institutional form of the Au- Muslim Mindanao (ARMM) have long tonomous Region in Muslim Mindanao suffered from the effects of underdevelop- (ARMM) dates back to the 1996 "Final ment, neglect, and conflict. The proportion of Peace Agreement" between the Government families in poverty in ARMM is the highest of the Philippines and the Moro National in the country, and at 47 percent is over twice Liberation Front (MNLF).4 Its creation was the Philippines average.' Disruption due to an attempt to satisfy longstanding aspirations insurgent and clan-based conflict has been of Moro political actors for greater autonomy, periodic but severe - in 2008 some 750,000 while correcting years of actual and perceived people were displaced during confrontations. neglect of the region's development needs. Such levels of poverty, violence, and disrup- Unfortunately, the record of the ARVM has tion affect childhood opportunities and abil- been poor on both counts. The Moro Islamic Map 1: Provinces of ARMM LMM E SURE Source: Wikipedia (pending official NAMRIA maps) NAMIA aps AASILAN SULU > TAWI-TAWL Government of Philippines 2012 National Poverty Statistics. (http://www.nscb.gov.ph: accessed 4 April 2013). The national average family incidence of poverty is 22.3%. ARMM's incidence is almost 10 percentage points higher than the next poorest region, neighboring Region XII. 2World Bank (2003), pp 19-21. 3 In 2003 per capita spending on education in the ARMM was about three quarters the national average: World Bank (2003), p 31. 4 In 1977 two new administrative regions (IX and XII) were created in Mindanao, and the ARMM was actually formed in 1989 via Republic Act 6734. HDN/UNDP (2013). 02 Liberation Front (MILF), a breakaway group, In response, the national government in- did not accept the terms of the agreement, stalled a caretaker administration in the and there was a return to armed conflict. ARMM in early 2012. Regional Governor Over the same period, indicators of human in Charge Mujiv Hataman (subsequently development in the region stagnated or wors- elected in May 2013) has pursued a reform ened until they were at or among the worst in and anti-corruption agenda, and sought the Philippines. In the most recent Human to strengthen services and relations with Development Index compiled for 2009, all National Government. At the same time, provinces of ARMM except Basilan were in negotiations between the MILF and the the bottom ten provinces. Tawi-Tawi, Magu- Government progressed from an April 2012 indanao, and Sulu were the three provinces statement of principles to the signing of a with the lowest human development index "Framework Agreement on the Bangsamoro" scores in the country.s (FAB) in October 2012.1The FAB lays out a road map for a "new political entity" with a The national government has increased fund- ministerial form of government, its own basic ing for ARMM in order to correct historic law, and an "asymmetric" relationship with imbalances and support the peace process the Philippine national government.' Should between the national government and the the FAB be implemented on schedule, the MILF. Both the regional budget and national new entity, to be known as the Bangsamoro, agency budgets intended for the region have will replace ARM\4M in 2015. grown dramatically: in one estimate, these combined allocations to ARMM grew by One point of view is that reform or insti- 37.9 percent in 2011 and 72.4 percent in tutional learning in ARMM are irrelevant 2012.6 However, while budgets for ARMM because the end goal of the peace process is to may be growing, service delivery and develop- replace the regional administration with the ment outcomes have continued to fall behind Bangsamoro. This study's premise is that new because of public expenditure management institutions of autonomy can be improved problems.1These include low regional rev- through the analysis of previous attempts, and enues and extreme dependence on national the capacity to provide public services in the government, limited control over the budget, future can be influenced by improved systems and limited program and financial manage- and processes now, particularly as there will ment capacity, combined with vulnerabilities inevitably need to be some migration of larger in budget execution and accountability.7 institutions such as education systems to These issues have contributed to trust deficits the new entity. In the words of one ARMM between regional and national government, public servant: citizens, and international donors. 6 Antonio and Habito (2012), para 22. INCITEGov (2007), pp 55-59 and World Bank (2003), pp 36-39. 8 GPH-MILF, Framework Agreement on the Bangsamoro (October 2012), Art. 1.2-4. DepEd ARMM staff, personal communication (January 2013). 0 World Bank (2004), pp. 34-39 and 181-185. Ifthe establishment ofthe Bangsamoro Structure of the BE-PEIR project 03 is like the act ofprayer, then what we do until The project was initiated at the request of then is equivalent to the ablutions. The ablutions the ARMM regional government to en- are as necessary as the act ofprayer itself. hance dialogue and reduce the trust gap between ARMM government and national In this sense, this review is also addressing the and regional stakeholders, enhance efficiency question "What factors have prevented the of spending in basic education, and support ARIM from creating a successful form of au- institutional reformsincluding through the tonomy?" Public expenditure management for peace process. It has three components: basic education is an important window into understanding both failing services and trun- 1. This Public Expenditure and cated autonomy. It is public expenditure man- Institutional Review; agement institutions that determine if budget 2. Technical assistance to ARIM Re- composition, execution, delivery, and incentives gional Government and DepEd ARMM allow services to be delivered effectively.10 At on planning, budgeting, and dialogue with the same time, the subnational fiscal system is national budget institutions; the backbone of any effective territorial decen- 3. Technical assistance on issues including tralization or autonomy relationship. vulnerabilities in the education payroll (e.g. "ghost teachers," Human Resources Infor- Improving education outcomes is among mation and Payroll systems reform, and a the top priorities of the regional govern- March 2013 representative facilities survey ment, forming a key component of its stra- focused on teacher and student absence). tegic ARMM HELPs (Health, Education, Livelihoods and Peace Support) framework. The project builds on support to DepEd Education is by far the largest single category ARNIM through the Australian Aid Basic of expenditure for the ARNIM, yet it has Education Assistance for Mindanao (BEAM) been beset with widely publicized problems activity, and further initiatives by the World with "ghost teachers", inadequate resourcing, Bank and partners. and poor education outcomes.'The Depart- ment of Education in ARIM accounts for Chapter Two of this report proceeds with 57 percent of the regional government budget an introduction to ARMM's development in 2013, and when regional, national, and lo- context, including the evolution of regional cal expenditures on education in ARIM are institutions and the basic education system. combined they amounted to about 8.5 billion Chapter Three describes the unique ARIM pesos in 2012 (See Chapter Four). Public Expenditure Management "syndrome" that contributes to the weaknesses analyzed in this report, and in turn informs its policy 04 recommendations. Chapter Four analyzes pub- vey. Chapter Six summarizes the findings on lic expenditures for basic education in terms public expenditure management for education of aggregate spending, expenditure composi- in ARMUM and derives recommendations both tion, and specific challenges that the ARMM for reforms to ARMM institutions and for PEM syndrome poses for budgeting. Chapter consideration by the Transition Commission Five assesses budget execution and frontline and other stakeholders in the preparations for delivery aspects of basic education resources, a new Bangsamoro political entity. incorporating findings from the facilities sur- CHAPTER TWO: Liberation Front (MNLF) and the Philip- 05 ARMM DEVELOPMENT pine Armed Forces. In 1975, the Western and &BASI EDUC TIONCentral Mindanao administrative regions were & BASIC EDUCATIONestablished, and in December 1976 in Tripoli, CONTEXT the Government and the MNLF agreed in principle to an autonomous region in South- ern Philippines. However when the Marcos government established this in March 1977 it was viewed by the MNLF as a unilateral step Historical development of the ARMVM and armed struggle continued. The Autonomous Region in Muslim Min- danao (ARMVM) traces its roots to the long After assuming power in 1986, Corazon history of the quest for self-determination by Aquino revived peace talks with the MNLF the Muslims of Mindanao. The first Muslim and enshrined the creation of an Autono- settlements in the Philippines were recorded mous Region in Muslim Mindanao in the in the 1200s. By the 14th Century, the Sulu 1987 Philippine Constitution. In August Sultanate was well established, long be- 1989 the Organic Act of the Autonomous fore the Spanish arrived in Manila in 1571. Region in Muslim Mindanao was passed Muslims in Mindanao repelled attempts at into law, and in November a plebiscite on colonization throughout Spanish rule, and membership in the autonomous region was this resistance continued through the Ameri- conducted. Four provinces - Maguindanao, can occupation. Lanao del Sur, Tawi-Tawi and Sulu - decided to join. In September 1996, the MNLF and After the establishment of the Republic of the Government signed a peace agreement the Philippines there was large-scale migra- that led to MNLF Chairman Nur Misuari tion of Christians to Mindanao, and the becoming ARMM Regional Governor. Presi- proportion of Muslims in Mindanao has dent Gloria Macapagal Arroyo supported a dropped from 76 percent in 1903 to less than September 2001 plebiscite ratifying Republic twenty percent today. Settlement was pre- Act 9054 (the "Expanded Organic Act"), ceded by the establishment of military camps which further defined the powers and respon- to clear the entry of the settlers, driving Moro sibilities of the regional administration, and and Lumad communities on to less produc- expanded the autonomous region to include tive land, and formal land ownership arrange- Basilan province and Marawi City.12 ments generally favoured settlers." A breakaway group known as the Moro Is- In February 1973, the revolt escalated into an lamic Liberation Front (MILF) surfaced after armed conflict between the Moro National 1984 and continued armed struggle through- Lumad is the common term used to describe indigenous peoples in Mindanao as distinct rom both Muslim and Christian settlers. 06 out the 1990s. In 2000, clashes between the parties signed the Framework Agreement on military and the MILF escalated to so-called Bangsamoro (FAB), which created a Transi- "all-out war," displacing hundreds of thou- tion Commission to draft a Basic Law for sands of people. President Arroyo initiated the Bangsamoro and established security peace talks with the MILF in January 2001. arrangements (called "normalization"), wealth, A major breakthrough seemed imminent in and power sharing. Announcing the FAB, August 2008 when the parties were set to President Aquino emphasized the need for sign a Memorandum of Agreement on An- structural reform and referred to the ARd\UM cestral Domain (MOA-AD).The MOA-AD as an experiment that had "tightened the was declared unconstitutional by the Supreme shackles of poverty."14 Court in response to cases raised by local and national politicians, and renewed fighting The Basic Law will provide for a MILF-led displaced some 600,000 people. Bangsamoro Transition Authority until a duly elected Bangsamoro government is installed Peace talks were revived after the election in 2016. The implication is that if the new of President Benigno Aquino III in 2010. transitional authority is installed prior to the In April 2012, a set of "Decision Points on 2016 elections, it will take over from the cur- Principles" were agreed upon in Kuala Lum- rent ARNIM administration, elected in May pur to guide the negotiations towards "the 2013. However, the degree to which ARM creation of a new political entity in place of institutions will continue in force during the the ARNIM."'3 On October 15,2012, the transitional period is still unclear. Box1: .1 he artes gre tht te satuqparetie snd the Frngamork Agvrement o Selected ~ ~ ~ ~ ~ ~ ~ ~~agsmr (FA- i ncepal n h Bnsmr hllb V3 Te B),saor wilhareate athrani-y sion ofthe stalised t relactheAutnomusRgioin tion rcomivgrnsiond tonaftiaoBsi Lam forti e B.angnsor and estblished surity Agremnton .2 Th gvermet f he ansaorshll disarrang eentale "ornmaliztio",welh Band pwer sha ng annoingve a th F 1.3 Theprvines,ciies muicpaltie, a- apresien baquinoempased threreentaeed f rangys ad gograhicarea wihin ts estrucytu alreform and efCerred tovethen RM shall~ ~ ~ ~ ~ ~ ~~a an tecnttetuisothBagar. odexpodemren e hataha ce tgtnd fth - Theautoriy o rgulte n ts wn espnsbilty isackleina ofiacl poverty."14enue 1TheBaic aw wi poide forangsmL-l mentwiththe angamorGovrnmntsalle Bansaoursof Travn,it Auheoitye untladly reserlecte Bangsamor govrnen isgamr installeddnor shtanitina authrit iss installed priors tod theewllb c26 p r s ec tos it will t o fo t Provinces ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~inttuin wilvtdaanticuininteANMete n19r01aeCtbt,Dvodl c,Lnotine Nore duriwng Sothe KornaalIign,Paadathert rM."13aOn Octbobera 15 2012,meno the trasitiona pthemriodu s stlAnclesta.omi BMO1:D 1.1ne Th20ndanled Parties agpree that he stts qncuson yfa sm rahthdtievtdaanticuin SelectheAsoed pv thse aceale ad te iBangsamoro shal e sionsIoF, hecso esitashe to rncplsace te Autonomous(Apil202) Regint n Framewrk Mus0m Minanao.. Agreemoe esdet n f I.2 hebi gov ternmipns ul eto hpecftePeient of the Bangsamoro shallt ih h ML Bangsaobroaeamnstra.om the ffars f te cnsttuet u0tsis.uarntedS. I..Te.tosi of 0h Central Govern ~ Prvincs tat vtedalains is excluse pow iter ind19 ohre 2001 are Ctbt,DvodlSr aa e ot,Plwnot Cotaato,Sulan Kdart, Zmbonga el orteandZambang delSur and the teofCaBaoDpansamoo, GernelSntos Kornadllign,Pagdia, Pero Pincsaad ambang. Ake elmen IVfth opsTintthe Meanamo wiAcetal Doatity (MOA~D) sgnedin 208 ad anulle by he Sprem Cout wa itsincluionvyefitrofsomearea thanhadtwicsvotdraginstimelticn Undr te FB sme f hes aras illhod apleiscte n mmbrshp i BagVa.r shulhe% fresident eques intoenetlfs 13GPH-MILF DecisionPoints onolinciplesad comAprle122012r(prils2012,tpointo2 ~ Ofie f te reidet f heRepblc f te hiippne, ul Tet f he pechofthe Preidntsaoh ramework Aea ement thMIF (rert Octsorvebeira).e adf cta Structure of the Autonomous Region in three members from each national congressio- 07 Muslim Mindanao nal district. The ARMM Government also has As the only nominally autonomous region a cabinet of regional secretaries and corre- among seventeen administrative regions in the sponding line ministries. While RA 9054 itself Philippines, ARVIM has a regional govern- authorizes "no more" than ten cabinet secretar- ment headed by a Regional Governor and its ies, in practice there is a regional counterpart own Regional Legislative Assembly (RLA). As office for nearly every national government depicted in Figure 2 below, the Governor and agency. Currently there are four bureaus, 20 Vice Regional Governor head the ARMM departments, and nine locally created offices Regional Government, supported by an Ex- beyond the Office of the Regional Governor ecutive Office and an Executive Secretary. The and the RLA. ARMM RLA is unicameral and comprises Figure 2: ARMM Government Organization REGIONAL GOVERNOR REGIONAL VICE GOVERNOR .................. EXECUTIVE COUNCIL Regional Economic and Dev't. Office of the Planning Board DEPUTY REGIONAL GOVERNOR Office of the Office of the EXECUTIVE SECRETARY ATTORNEY GENERAL Office of the THE CABINET CHIEF OF STAFF -Line Departments Cabinet ARMM Liaison Office in Metro Manila Devolved Agencies/Offices Secretariat ARMM Satellite Coordinating Office Attached Offices/Bureaus/ CommissionsI L Locally Created Offices AMS TMS FBMS ISS Provinces Cities Municipalities Barangays Source: Barangays www.ARMM.gov.ph 08 A Regional Economic and Development ARNM also has ambiguous relationships Planning Board (REDPB) is chaired by the with the other Local Government Units Regional Governor and includes the region's within the region. Provinces, cities, mu- provincial governors and mayors, the RLA nicipalities and barangays appear formally Speaker and two additional RLA members, subordinate to the Regional Government and five private sector representatives. In (Figure 2). In reality as pre-existing entities practice the REDPB deals with the limited with their own elected local governments, resources raised in the region and the re- and guaranteed resources under the national gional share of the Internal Revenue Allot- decentralization provisions of the Local Gov- ment (IRA). Of much greater significance, ement Code, the LGUs within the ARM the regional government receives an appro- are themselves effectively "autonomous" from priation from the national government, and ART\4M control. This situation has compli- national line agencies allot funds from their cated the efforts of the regional government own budgets for the ARMM, although until to assert its own autonomy, and is a key struc- recently these were often neglected or chan- tural issue for the proposed Bangsamoro. neled through neighboring regions. Table 2: LGUs in ARGiMt No. of LGUs Population city Municipalities Barangays ARMM 3,256,140 116 2439 Basilan 293,322 Lamitan City 11 209 (excluding Isabela City) Lanan del Sur 933,260 Marawi City 39 1155 Maguindanao 944,718 Shariff Agualk 36 462 Sulu 718,290 Jo10 19 410 Tawi-Tawi 366,550 Bongao 11 203 Source: 2010 Population Census, NSCBG ARMM development context highly vulnerable, losing long periods of time 09 ARMd\M is the poorest region in the country, in school and ending up twice as far from with poverty incidence among families of available schooling, even after returning to 47 percent, about double the national aver- their home areas". age. Although the Philippines has enjoyed a decade of economic growth, poverty levels In addition to insurgency-related violence, are higher in ARMM now than in 2003. rido conflicts may be as "pertinent to the The region suffers a relative lack of economic daily lives of the people - and cause them integration even with growth centers in more concern."9 Rido are localized but can Mindanao itself, and this effect is worsened continue for long periods and trigger wider by conflict and related displacement and violence.'Ihe Quantitative Service Delivery disruption." Weak governance is a persistent Survey carried out in March 2013 found two factor undermining the delivery of services, schools closed due to ongoing rido violence - dampening investment, and further driving in one case, killings had taken place inside the continued poverty. school. Banditry and violence associated with groups such as the Abu Sayyaf are also found Widespread conflict primarily in coastal areas. Multiple inter-related forms of violence affect ARMM and neighboring provinces. Insurgen- Weak governance cy and counter-insurgency, clan feuds locally The early years of ARMM were colored by known as rido, and pseudo-ideological criminal a number of high profile corruption scan- banditry affect different parts of the region.'" dals, and to this day much of the region's The consequent lack of security is a major governance is characterized by clientelistic disincentive to the private sector investment practices, elite capture, and weak technical necessary for job creation. capacity. Governance in ARMM has been further undermined by a vicious circle of ex- Mass displacement is a defining characteris- treme dependence on a national government, tic of violence in ARMM. Nearly a million resulting in a distorted and limited version of people have been displaced in each of the the intended autonomy. At the same time, the 2000 and 2008 conflicts, and a 2011 survey region has been linked to important dimen- determined that 41 percent of households sions of national politics through patronage in central Mindanao had been displaced in return for the provision of votes by local between 2000 and 2010, rising as high as 82 leaders in national elections. Many national percent in Maguindanao province.'7 Dis- government agencies implemented programs placed people lose crops and livelihood and for the region through their regional offices often sell assets, locking them into debt and surrounding ARMM, or simply did not make long-term poverty. Displaced children are allocations for the region. ARMM effectively a World Bank/Australian Aid (2010), pp 12-27. 1 Torres (2007). 1 World Bank/WFP (2011). a World Bank/WFP (2011), "World Bank/Australian Aid (2010), p 13. 10 was "less autonomous than local government entrenched political clans with private armies such as cities, towns, and provinces and more who operate with impunity from the law. subservient and dependent on the national government than any other political sub- An important implication of endemic poor division in the country."20 governance in AR M is poor data. Indica- tors such as population and enrolment rates In this situation, the regional government, have been problematic, making per capita soundly re-elected in May 2013, has had and per pupil analyses difficult. Population limited traction against poor LGU governance. numbers were manipulated22, as both voter Accounts of absent local officials, non-existent numbers and the population-based share development plans and programs, and personal of the IRA could be increased this way.'The use of LGU budgets abound, a record reflected 2010 census was more carefully verified, and in national good governance measures: two of indicated an apparent slowdown in popula- five provinces, no cities and 29 municipalities tion growth from a suspiciously high 5.4 per- out of 113 in ARIM received the Depart- cent in 2007 to a more realistic level of 1.51.21 ment of Interior and Local Government's The total population in 2010 was measured (DILG) Seal of Good Housekeeping in 2011, at 3,256,140 million. During 2011-12 there as against a national average of 78 percent of were estimated to be 914,991 children of el- LGUs.21 Many LGUs lack a rating altogether ementary school age (6-11 years) in ARNIM, due to a lack of information, and are run by and 478,522 of secondary (11-15) age .24 Selcte leal undng or oca eucaionproram shll e te GvAnmotnt impllicsatliohaan of andmicport fgvenac nIR po qdta. Idca- eductionMMA ct N. 27 onasicEductionThetiona rsuasewpopulation mandnf renrlmevnt raes respnsiiliies RegonalGovrnmnt hal esablih,ainain resha sve ee probematics, mdaki n peapitaoso and uppot asa to prorit a cmplee aandh perpl neAlysems dficut.opulaetion56 integrated snumber we rel mnut2 RS bot voter annmbr andaioa theeor population-basedgshar relevant, ~ ~ ~ ~ ~ ~ ~ o the IRAoniv totenesdasad MgaCfou lde inrae this wMAy.AIh 6@ 6 20100 census was00 m6orecaefll veriied an aspratonsofthepeolein heeginSecion2 iNdica2)Te neapartent ow duon in opla Art 1). Autnoou Reio in Mulm Mindana shal tinrth fro edctal pcously thaigh 5.4tuper GAD CoMA Ac N. 280) Th Reioa As sestv and reeatt heesadeitn sct sit n 200 omue a m e reistic levlu 1. alTh totalm pouato ine 2010oou wasio measured10 atcani, 3,256,140 m . D n 2 ementary**..0 .0 schoo ag (-1ar)iAR M 21~~u toL a21).h lack of gohuskeinfratiLGon anarency byonailt,prtpto and 47,22o efodaye (11-15)arl age.24 Bicor ox2Reulici n omsino Act 54d(Art.I,Sc2,pr(g)supr()ChpeII;TeAtomusRina 2Slecte lpie ial g und no RM loal e uation proig rm side. hllpbe Cter overnmatietoushalsm st blis htata/pi.and/suport, proveais louaio rsonslty of the Regona Gov1.0 ermenRVMste aslkeylwe to irty,e coplete andm integrat7esse edctinMA AtN.29oBaiEdcto:Teinl frmeor that ismainfl relvan an fres Piilities Rgional GsisOvernen shall C estls h, mou atinaidHsn g responw w.ss .gvepht onthenneeds,-ieal n spiats- o pouainadhuigrasuphiprtnseapopulpriority2a4cmle andtepolinheAoomuRgon(cin56 DepEdNational,E n teted s ystm Iofq ai y e a ) a o A2 Xli)e. a edcaioa fameor tha is meanigful, *- .--. . 21DIL(2012aTheisalio hoekepn ratpeL ontrginSencactontability 162)TheDeatmntopEuation anefrac,pricual thei dicoue oiisArt Comisio AutnomusRegoninusimaMinanogsal sharp~sundertakeaan educational programiothatbis-culture 2TentioalopuAtio grodae inMMA1 At No.920) ee.AM Reioate s- sie nsiterive an eatto the orcinfonflaeds 2007 eisusn figresPhlipinesNaionly Saisic Ofeact0 eains ofophusrngh- station anofsn: tp/wwcesshcontie n terennudin pouainadhuingvand dlielopentio the Mdaai duain idgnuscmuite Sctoo9nAt ) 24 DILGd (202)Tion al nhanced Baicukeeiong IfrmatLG on ytransparenc, accounabili, prticiainadnefraeeatclal.hi ARMM education context tion is divided into twelve divisions across 11 Education is a regional responsibility. The the five provinces and two cities: one each first ARMM Organic Act specified that the in Lamitan and Marawi cities; one each in regional government would operate a "sub- Basilan and Tawi-Tawi provinces; two in each system" of the national education system, and of Sulu and Maguindanao; and four divisions devolution of "control and supervision over in Lanao del Sur (Map 2). [DepEd] offices in the region" took place in September 1991 by Executive Order 459. The Lanao del Sur subdivided its original two expanded Organic Act also stipulates that the school divisions into four, though DepEd regional government is responsible for fund- National does not officially recognize these ing education (Box 2). additional divisions. In general, there appears to be a greater problem of additional staff and The ARMM public basic education system facilities being attributed to Lanao del Sur includes 2,173 primary schools and 298 than the other provinces. For example, Lanao secondary schools.25 These physical and del Sur accounts for 43 percent of the region's human resources are spread across a varied secondary schools, despite having only 29 landscape on the Mindanao mainland as well percent of its population.2' DepEd ARM as the island provinces of Sulu, Tawi Tawi, has attempted to reassert a more streamlined and Basilan at the southern tip of the Philip- administration in the province, but with lim- pine archipelago. Limited access, particularly ited results so far. to these "island divisions", makes supervision particularly challenging, a problem deepened Basic Education echoes the national format: by historic understaffing of supervisors for the six years of elementary schooling followed regional DepEd. Basic education administra- by four secondary, with the phased introduc- Lignmd Map 2: m7 ARMM Regork ARMM Education ARMM EdSuutian DiMiaiog d Divisions PhiippinesM.niipaity Source: DepEd, NARMIA Mindanao TAVVWAWI SDpEdNaionl,Enhncd BsicEdcatonInfrmtio Sstemo divEisioninto four,2thughDepE 2Nationalnalestaotstfciciallyerecognizenlhese than th ote prvne.PoWxmpe aa secondaryschoosmepte hvngoly2 2'DpEdNatona, nhacedBascEuctioInormti haSs atem tE toraSser a2moe1sreamine 11Bsi Education echoesic thce nationa format: 12 tion of a K-12 program underway. Due to the is also addressing these conditions through, remoteness and displacement already dis- for example, the construction of community cussed, there is a higher than normal demand learning centers (e.g. the ARMM Social for Alternative Learning Systems to provide Fund) or community-based schools (under education services to out of school youth and/ the BEAM-ARMM project). or displaced populations. Private alternatives, primarily madaris, have been an important Education outcomes part of the system, and in recent years efforts The triple challenge of limited autonomy, to register and regulate their curricula have violent conflict, and weak technical capac- been introduced. Development assistance ity and governance are starkly reflected in Figure 3: Net enrolment rate (%), SY 2011-12 100 90 80 70 60 50 40 30 20 10 0 a) EG W CL I 2 0 .20 .2 .0 0 Il ll ll lIIl CM .00 a " Source: Primary/Elementary * Secondary EBEIS online (SY 2011-12) 17 EBEIS Net Enrolment ratio nationwide was recalculated before publication, hence at time of publication national figures do not match regional figures reported on August 13, 2013 and reflected in Figure 2. the region's education outcomes. The net serve out of school youth is thus greater in 13 enrolment rate is 64.4 percent for primary ARMM and similar regions than elsewhere. education and 26.1 percent for secondary education (SY2011-2012). While these rates High early dropout rates contribute to this are improving, they remain the lowest in the low cohort survival.'Ihe dropout rate in country by a large margin, and well below Grade One is the highest in the country as of the average national net enrolment rates of 2011. Lack of interest among students, dif- 91.2 percent and 62.0 percent for primary ficulty of travel between home and school due and secondary education, respectively.27 to geographic isolation of some communities and the poor condition of infrastructure, and Even worse than the problems with low conflict and disruption hinder the quality of enrollment, staying in school, completing sec- teaching and attendance of students. During ondary education and having opportunities the school survey two schools were closed for further study and/or employment is an due to ongoing Rido or local conflict, and extreme challenge. In SY 2011-12 only 23.1 four more indicated peace and order or local percent of students in Grade 1 had made it conflict were the number one challenge they to grade 6, against a national average of 73.5 faced.21 The situation is the same for second- percent.'Ihe situation is similar at second- ary, where the drop-out rate of 5.6 percent ary level, with only about 45.1 percent of the is higher than the achievement score of high Year 1 students staying on to year 4 (national school Year 4 in ARIM is only 37.1 percent, average is 78.8 percent). When considered more than ten points below the national aver- together, this means that only one out of ev- age of 48.9 percent. 29 ery 10 students who begin primary education will graduate from secondary school. Given Those who graduate from high school find it net enrolment rates, fewer than six percent difficult to continue in tertiary education, due of children will graduate high school in to a combination of factors including poverty. ARMM.'Ihe need for Alternative Learning Only 11.5 percent of females and 9.2 percent Systems (ALS), incentives and scholarships, of males in ARNIM are likely to have some and Alternative Delivery Modes (ADM) to college education, compared with the national 100 Figure 4: 90 Educational Attainment of Females 80 in ARMM and selected regions (2008) 70 60 50 40 U College or higher 30 U Completed high school 20 U Some high school 10 U Completed elementary Source: 0 U Some elementary NDHS 2008 ARMM Region IX NCR Philippines No education 28 BE-PEIR School Survey Data. 29 DepEd NETRC Directorate Presentation (May 2012). 14 Figure 5: 100 Educational Attain- 90 ment of Males in ARMM and selected 80 regions (2008) 70 U College or higher 60 U Completed high school 50 N Some high school 40 U Completed elementary 30 N Some elementary 20 No education 10 Source: 0 NDHS 2008 ARMM Region IX NCR Philippines averages of 22.4 percent for females cally been able to meet the needs of the sector. and 19.0 percent for males. As recently as 2009 ARMM had the highest classroom and teachers shortages among all These poor education outcomes result in regions, yet the Department of Education was graduates with reduced opportunities for train- only able to allocate resources to 10 percent of ing or employment. Low investment means the schools that needed new classrooms and few jobs, but people in ARMM are also ill 21 percent of those with teacher shortages. equipped for the few employment opportu- While some have protrayed resource shortages nities that are available in the formal sector. in the past as symptomatic of central govern- Limited labor force participation rates and ment neglect, at the same time governance a reliance on agriculture reduces returns on concerns limited national support for addi- education, limiting its relevance to youth in tional budgets for the Region. Well-founded ARMM who are required to work to sustain concerns over "ghost teachers," "ghost schools," their families. Taken together these reinforcing and "ghost pupils" and the lack of regular stra- dynamics threaten to keep more than half of tegic communication between the national and the families in ARMd\M trapped in poverty for regional governments were central factors in another generation. the chronic under-resourcing of the sector. Past underinvestment by government has rein- More recently the national government has forced the poor performance of the education dramatically boosted education budgets as a sector in ARMIM.o he allocation of national whole, increasing them by 18 percent in 2011, government funds for school buildings, fuarni- 15 percent in 2012 and 23 percent in 2013. ture, textbooks, and teachers has not histori- Since 2011 the DepEd National has priori- tized ARMM divisions for program funding, modes to respond to the specific needs of the 15 and in many areas the region is now relatively learners in the formal school system. Efforts advantaged in terms of available program bud- are also underway to expand the coverage of gets, as discussed in Chapter Four. the Alternative Learning System Program to reach more learners in areas that are exposed As with the increased prioritization of to the problems identified above and sup- ARIM by national government, the re- port them with better accreditation and gional government has also actively introduced equivalency testing. Recently, through the measures to address these problems. The BEAM-ARMM project, an Alternative De- government has supported efforts to make the livery Model has been introduced using the standard curriculum culturally relevant to the BRAC education model developed initially in Muslim population through the integration of Bangladesh. At the time of writing, BRAC Arabic Language and Islamic Values (ALIVE) (an international NGO) has established over in public schools and the development of 1,000 Community Learning Centers (CLCs) madrasah curriculum for private schools. The in the five provinces of ARMM. These are DepEd ARMM has continued implemen- located in communities with no govern- tation of the Region-wide Assessment for ment school or difficulty in access, and have Maths, Science, English, and Arabic Language targeted out of school children as enrolees. (RAMSEA) introduced by the BEAM proj- The CLCs are operated by BRAC and 11 ect. The results of the RAMSEA can inform Local NGOs.'Ihe BRAC model is an alter- targeted interventions such as teacher deploy- native to government financed public schools ment, supervision and continuing training. and could be an option under the Education Service Contracting Scheme, subject to the The Department of Education in ARMM development of an appropriate policy. is also aiming to expand alternative delivery 16 CHAPTER THREE: region with the fiduciary requirements of PUBLIC EXPENDITURE national government and principles of sound MANAGEMENT budget accountability. IN ARMM At the same time, there has been a tension between the nationwide decentralization pro- cess under the Local Government Code (RA 7160 of 1991) and regional devolution under RA 9054 (Box 3). These two frameworks have Weak public expenditure management insti- not been resolved coherently, resulting in am- tutions and practices have been central to the biguity over the relationships and responsibil- failure of ARMM to perform effectively as an ities ofthe ARMM regional government and autonomous regional administration. These other Local Government Units. In this sense problems derive from poorly developed fiscal ARNM has been doubly trapped between a arrangements to manage the unique relation- truncated devolution process from above and ship and dependence between ARIM and a lack of integration with the system of Local national government, contradictions between Government Units below. these arrangements and the wider decentral- ization process in the country, and a lack of The resulting ARMM public expenditure sound internal governance institutions in the management "syndrome" has four interre- region itself. While some problems are similar lated dimensions, discussed in the rest of this to those found elsewhere in the Philippines, chapter: the unique nature of the relationship between ARMM and national government adds par- Fragmentation of public expenditure ticular challenges. management processes for budgeting and budget execution between local, regional Delays and partial implementation of the and national agencies; regional autonomy granted first by the 1989 Poor integration of regional budget- creation of the ARMM, and revised via the ing processes with national systems; "Final Peace Accord" and resulting "Expand- Ambiguity between regional and na- ed Organic Act" of 2001, have undermined tional decentralization processes resulting the coherence of public expenditure manage- in weak application of LGU resources to ment institutions. This incomplete devolution devolved services; stems from both a failure to establish neces- Weak accountability and reporting on sary legal codes and functional institutions public funds within the region, poor data to support the provisions of the Organic Act availability and mistrust between national and a lack of capacity and compliance by the and regional stakeholders. 17 Box 3: Art. Ill, Sec. 3: "The regional government shall Art. VII, Sec. 21: "The form, content, and manner Selected adopt policy on local autonomy whereby regional of preparation of the budget shall be prescribed provisions powers shall be devolved to local government by law enacted by the Regional Assembly. Pend- of RA9054 units particularly in the areas of education, health, ing the enactment of such law, the budgeting pro- human resource, science and technology and cess shall be governed by existing laws and rules peoples empowerment. Until a law implementing and regulations prescribed by the Department of this provision is enacted by the Regional Assem- Budget and Management." bly, Republic Act No. 7160, the Local Government Code of 1991, shall continue to apply..." Art. IX, Sec. 2: "The Regional Government shall enjoy fiscal autonomy in generating and budget- Art. Ill1, Sec. 11: "The Regional Government shall ing its own sources of revenue, its share of the provide, maintain, and ensure the delivery of, internal revenue taxes and block grants and among other things, basic and responsive health subsidies remitted to it by the central government programs, quality education..." or national government or any donor." Art. IV, Sec. 1: "The powers and functions already vested upon and the shares of the national taxes ciyfonaoalntrlrenutxs,esad provided by...the Local Government Code ... to provinces, cities, municipalities, and barangay... shlbed tr ueda foow :()T iy-vep - shall not be reduced."cet(5)tthprvneociy(bTit-fe Art. VI, Sec. 20: "The annual budget of the Re- gional Government shall be enacted by Regional ntoa oenet 000e 0. e d T0 0 IXiSec 9 0leto of aoprovinceror Acharges, and taxe sesou 0moe 0n na00 resources, cen (35%)y to the4 province ore ciy;()Nhit-fv prt5 ttregional government; wt a ams 0 1 A Re an (c nThirty percent (30%) 00 thecenra govenmen oro clea natona powrs uch s freig afairs2. ongrssinatioaly govement. Fragmented public expenditureTe are si main t ypes ogernment rev- management of ARMVM resourcesenetaflwioARVM In theory, RA 9054 provides the ARMM regional government with an almost "pie- 1 RVMRgoa nary authority"over the region's develop- ernm ens a ment and welfare, excepting a negative list of wihnteaiolbug; clear national powers such as foreign affairs 2. Conresional Priort Development and defense.32 T regional government islAssistane Fndsa granted fiscal autonomy including controlunespcaproefndschste over the budgeting of the revenues that come scolbidnfu; to it, whether locally raised or remitted by the 3 ud amre rsbalte o central government (though these are subject uei h R Mb te ainlgv to audit).3 However, in reality the major-emntagcis ity of funds for the region are not under the4.tergosowsuceevnefm cotrlfhereioalgoermet,loa aes man tyeso oenetrv 32 RA 9054, Art. IV, Sec. 3. See also LGSPA/InciteGOV (2007), p 25. 33 RA 9054, Art. IX, Sec. 2. 18 5. the Internal Revenue Allotment level, and in Local Government Units below (IRA) share reserved for ARIM as a the region level. region; 6. and the IRA allotted to the other First, consider resources that can be budgeted various LGU levels. autonomously by the regional government as opposed to those budgeted by other entities. This "regional fund" comprises locally raised While these funds are all resources intended taxes and fees and the regional administra- for use in the ARMM, they are each bud- tion's share of the IRA, but excludes the geted, appropriated, and executed in different shares for other LGU tiers. The regional ways and by different actors. This fragmenta- government has powers to raise revenue from tion is partly explained by parts of the Organic an open list of taxes and fees and can enact Act that remain unimplemented, as well as a regional tax code. The regional share of the fiduciary requirements for funds originating IRA is set at 35 percent, with 35 percent from national government. The Organic Act distributed among the other LGUs, and 30 calls for the creation of a regional Depart- percent remaining with national govern- ment of Budget and Management and the ment.36 This regional fund is under the full passage of a regional budget law, neither of budgetary control of the ARMM Regional which has occurred.34 Without these, central Government within audit limits set by the government funds remain under the "rules and Organic Act, and in practice forms a de- regulations of the Department of Budget and volved regional budget. This regional budget Management of the central government".3s is programmed by the Regional Economic This means they must be budgeted through and Development Planning Board (REDPB) the standard national budget process, or be described in Chapter Two, approved by the appropriated by other national agencies under Regional Legislative Assembly and managed their own procedures. At the same time, the by a Regional Budget and Management Of- provisions of the Local Government Code fice (RBMO).37 regarding the use of IRA funds for local gov- ernment remain in force in ARMM, leaving A second group of resources could be termed these funds outside the public expenditure the "national fund" and includes support to management systems of the region as well. the ARMM Regional Government in the national budget, the Congressional Prior- These six resource flows can be notionally ity Development Assistance Funds (PDAF), divided into three broad "funds" based on and funds appropriated by other government where primary control over programming and agencies for use in the ARMM. In the case of budgeting them takes place. These correspond the Regional Government budget, the ARG to the national budgeting level, the regional negotiates with the national government via 3 There is a Regional Budget and Management Office but it deals only with locally raised finances. 3RA 9054, Art. VII, Sec. 21. 3 RA 9054, Art. IX, Sec. 9. From 2003 to 2007 the 30 percent national share of ARMM-derived IRA was directed to the ARG as well, pursuant to RA 9054 Art. IX, Sec. 15. 37Sevilla (2008), pp 2-3. the national budget process. In practice, this will be seen in Chapters Four and Five. 19 appropriation-at least for education-has A last component of the "national" funds are only covered the recurrent components of the the appropriations according to congressio- budget: Personal Services (PS) and Main- nal district for priority projects identified by tenance, Operations and Other Expenses national representatives, known as the Prior- (MOOE). It has been budgeted incremen- ity Development Assistance Funds (PDAF). tally, with periodic percentage increases set by These funds have grown substantially for the Department of Budget and Management. Mindanao as a whole in relation to other In practical terms, the ARG is treated as a island groups since 2010, and contribute to primary budget unit just like any other, not a a significant improvement in potential fiscal devolved political entity with the political and flows to Mindanao during that period. They fiscal autonomy set out by RA 9054. have been appropriated at PHP 70 million per district across the eight AR1\4IM congres- Beyond the ARG budget, many other national sional districts as in the rest of the country, agencies also budget or later sub-allot funds but this appropriation is not fiLly utilized as for use in the ARMM, but fiduciary, politi- payouts for PDAF projects vary among leg- cal or administrative barriers have prevented islators.'9 Over the two fiscal years 2010 and the devolution of these resources to regional 2011 only Tawi-Tawi used its fill appropria- government control. These funds include larger tion, and the region as a whole had a utiiza- Capital Outlays (CO) and various special tion rate of 57 percent. program or purpose funds. In addition, many funds are still centrally managed, and therefore Thirdly, one can consider the funds avail- not devolved to the agency offices in any of the able to and programmed by local govern- regions, ARMM included.38 A significant ex- ment units below the regional level. As noted ample from the education sector is the financ- above, while the regional government receives ing of textbooks, which are procured centrally an IRA allotment of 35 percent of national and distributed in kind to the different admin- revenue from the region, the other local gov- istrative regions of the country. Before 2012 ement levels share an additional 35 percent few national agencies identified their budget that is used at the discretion of the local allocations for the ARMM in the National government units in question.'This portion is Expenditure Plan or GAA, and others as- divided among the LGU tiers as follows: in sumed either that the ARG appropriation was provinces 45 percent goes to the province, 35 to cover their sector's activities within ARM percent to municipalities, and 20 percent to or implemented their activities via neighbor- barangays, while in cities it is evenly divided ing regions. It is therefore difficult to account between the city and barangays. completely for funds budgeted for the region, and harder still to measure find utilization, as A comparison of these sources of finds "9Antonio and Habito (2012), para. 3. Senators also receive a larger PDAF allocation of 200 million PHR. While PDAF is budgeted at a flat rate, its disbursement requires executive and DBI involvement and approval, which explains the varying utilization of the entitlement. 4t Antonio and Habito (2012), para. 26. According to the expanded Organic Act, separate Public Works funds were also appropriated for regional infrastructure projects approved by the Regional Legislative Assembly for the first six years of the Act being in force. In practice some funds were earmarked for regional projects, and some were a kind of regional PDAF for Regional Assembly representatives. 20 forcefully demonstrates the limited scope of The consequences of this fragmentation is il- ARMM fiscal autonomy (Table 3). Region- lustrated by analyzing the "locus of control" of ally raised revenue has been very limited and these resources by quantity and whom is able comprises only 25 percent of the regional to budget them (Figure 6). As discussed in fund; when these revenues are added to the the introduction, there is indeed a significant regional share of IRA, the resulting total re- upward trajectory in the aggregate resources gional fund is a mere 2.2 percent of the total available to the ARNIM. However, it is also resources intended for use in the ARMM, a clear that a small and even declining share of proportion that has actually been falling. By those resources are truly under the purview of contrast, the national government appropri- ARMM'S own public expenditure manage- ates 38 percent of the funds for ARIM di- ment systems.'The resultant scenario is of a rectly to the ARG budget, while an additional regional government squeezed between much nine percent is earmarked to the ARNIM larger flows of resources controlled from in other agencies'budgets.41 Surprisingly, above and below. Another way to look at this the "local budget" is the largest of the three domination of local resources by national components. In fact, the local level IRA al- and LGU flows is through the lens of weak locations comprise the largest single source of revenue generation in the region, a central revenue flowing into the ARMM, accounting concern for a future Bangsamoro. for about half. In millions 350 300 250 200 Total "Local Fund" 150 b Total "ARMM Regional Fund" 100 E e u Total "National Fund" 2009 2010 2011 2012 Figure 6: Source: Funds flowing to ARMM by locus GAA, NEP of budget control (PHP Millions) (various years) This earmarking understates the true quantity at resources due to the implementation ot some AMM programming via neighboring regions. 21 Table 3: Principal sources of government funds to ARMM (Thousands Pesos)* Source of Funds 2009 % of total 2010 % of total 2011 % of total 2012 % of total Total ARG 9,448,957 38.6% 9,285,089 35.7% 11,179,638 37.8% 11,717,707 40.0% Appropriation PDAF 121,000 0.5% 133,500 0.5% 538,800 1.8% 576,700 2.0% NGA sub-Allotments 1,327,784 5.4% 1,806,160 6.9% 2,691,332 9.1% 2,138,129 7.3% for ARMM Total "National Fund" 10,897,741 44.5% 11,224,749 43.2% 14,409,770 48.8% 14,432,536 49.3% Regional Own Source 216,920 0.9% 300,320 1.2% 139,950 0.5% 190,050 0.6% Revenue ARG IRA Share 660,000 2.7% 660,000 2.5% 462,000 1.6% 462,000 1.6% Total "ARMM Regional 876,920 3.6% 960,320 3.7% 601,950 2.0% 652,050 2.2% Fund" Provinces IRA 3,609,289 14.7% 4,014,094 15.4% 4,352,476 14.7% 4,141,669 14.2% Cities IRA 470,336 1.9% 432,584 1.7% 465,022 1.6% 605,008 2.1% Municipalities IRA 5,942,938 24.3% 6,311,312 24.3% 6,840,396 23.1% 6,630,169 22.7% Barangays IRA 2,705,485 11.0% 3,069,806 11.8% 2,886,879 9.8% 2,802,080 9.6% Total "Local Fund" 12,728,048 51.9% 13,827,796 53.2% 14,544,773 49.2% 14,178,926 48.5% GRANDTOTAL 24,502,709 100% 26,012,865 100% 29,556,493 100% 29,263,512 100% *Totals may not be 100% due to rounding error. Figures are nominal. Sources: GAA, NEP (various years) Fragmentation of the public expenditure in deciding the level and allocation offunds management institutions for the ARMM intendedfor itspolitically distinct mandate has two important consequences. First, the than other non-autonomous administrative amount of ARMM resources whose budget- agencies ofthe National Government. 42 ary control is shared or rests entirely with na- tional institutions is vastly greater than those The ability to prioritize spending according to under regional control. This situation was local conditions - particularly across sectors summed up well by the World Bank's 2003 - is one of the key advantages of decentraliza- review of Human Development in ARMM, tion, and one of the key mechanisms by which which describes ARMM as it may help address socio-economic causes of intra-state conflicts.43 The lack of genuine fis- a legally autonomous regional government cal decentralization embodied in the ARMM's that has no more real orpractical autonomy fragmented public expenditure framework iiWorld Bank (2003), p 29. oSiegle and OoMahony (2006), pp 37-40. 22 therefore represents a fundamental flaw in the actual resource flows difficult, as illustrated strategy of using autonomy in the ARMUM to in the discussion of basic education budget promote either peace or socioeconomic devel- execution in Chapter Five. opment through self-determination. Poor integration ofARMM regional and A second, related consequence of this frag- national budget processes mentation is that no single budget document The second dimension of ARMM's public or process comes close to including the bulk expenditure management "syndrome" involves of resources to be used in the ARMM. It is challenges in budgeting the ARG appropria- therefore very difficult to promote a coherent tion itself. The national process led by the set of policy goals through budgetary alloca- DBM is essentially designed for national tions. This is both because no coordinated sectoral agencies, while the ARG budget budget process effectively captures all the submission represents all its constituent available resources, and because there is a lack sub-agency budgets across multiple sectors. of connection between key dimensions of the The Office of the Regional Governor (ORG), budget - particularly the link between recur- which leads the budget process for ARMM, rent and capital expenditures which helps to must both present and defend the budget ensure investments are properly sustained. requests of some 17 departments and agen- This lack of "comprehensiveness and integra- cies that have line items in the ARG budget, tion"is widely recognized as a major impedi- as well as the budget of the Office of the ment to effectively connecting policy with Regional Governor and Regional Legislative planning and budgeting. 44 Assembly. As an "autonomous" regional gov- ernment with unique cross-sectoral respon- A final point about fragmented public expen- sibilities its budgetary requirements do not diture management is that these weaknesses always fit neatly into the national process. apply equally to budget execution arrange- ments as to budget planning. When the flow This situation places considerable demands of funds depends on different release arrange- in terms of timing, preparation and presenta- ments the coordination of input purchase be- tion on budget management institutions in comes more difficult, as the timing and quan- the ARMM. The allotted time in the national tity of different resources becoming available budget calendar for an agency to report on depends on different conditions being met. In the previous year's budget and submit the new addition, information and reporting on these request is the same for a single national sector resources is more difficult to manage. The agency as it is for the entire ARMM Region- demands of fragmented budget execution on al Government. For the 2014 budget that was an already stretched regional administration two months to submit past budget reports make active monitoring and coordination of and three months to provide the new budget "World Bank (1998), p 1; DFID (2001), pp 27-28. submission.45 The ARMM ORG submis- as both a national agency and a regional 23 sion requires collecting and collating both government is most clear is adherence to the the budget accountability reporting from the OPIE The OPIF is a results-based logical past year from among the regional agencies, framework linking agency budgets to perfor- and their submissions for the coming year. mance as defined by a clear set of Major Final This process is made more difficult by the Outputs (MFO) each with a suggested maxi- challenge of receiving accurate reports from mum of two Performance Indicators (PI). An LGUs and other sub-regional offices that MFO is defined as a "good or service that a are not clearly subordinated to the ARMM. department/agency is mandated to deliver As will be seen in Chapter Five, in practice to external clients through the implementa- ,'47 these reports are not collected for education, tion of programs, activities and projects". meaning the budget accountability reporting DBM's guidelines for the preparation of the provided to DBM is not credible and fails to budget submission in these terms are rigidly support new budget requests. structured and require submission via Budget Preparation Forms A and B which respec- While this period may seem adequate, the tively detail the MFOs and PIs - failure to demands of budget preparation have been submit according to these matrices requires intensified in recent years, with each addi- that appropriations will be limited to the tional requirement having to be duplicated previous year's actual levels.48 among the constituent ARMM agencies. The Budget Call for preparation of the 2014 There is much to recommend moving towards budget includes expanded or new provisions a performance-based budgeting system such on consultations with civil society and the as this one. However, international experience private sector, submission according to Major suggests that such a transition should "respect Final Outputs (MFO) of the relevant Orga- institutional differences among agencies and nizational Performance Indicator Framework help them to develop approaches suitable for (OPIF), periodic Zero-Based Budgeting, their own situations and contexts", and be Program Budgeting, and Bottom-up Budget- accompanied by consultation, support and ing.46 While not all of these immediately updating of accounting systems.49 The ap- affect every ARMM agencies'preparation plication of a rigid results-based framework process, they complicate the budget prepara- onto a cross-sectoral regional agency such tion requirements in the region and place as ARG, rather than a sector agency with a demands on the Office of the Regional Gov- clear mission and mandate, results in program ernor's constrained budget preparation team. budgeting that is both complex and vague. One area where the mismatch between the The OPIF framework for the ARG as a requirements of the "dual" faces of the ARG whole has to encompass the regions overall "NBM 115 (December 28, 2012), Annex D. NBM 115 (December 28, 2012), Sec. 3.3.2-5, NBM 155 (December 28, 2012), Sec. 3.3.2. 4 NBM 115 (December 28, 2012), Sec. 5.1.3. * Shah and Shen (2007), p. 173-4. 24 objectives, but also all those of its constitu- tral Office framework involves 12 programs ent agencies. In the ARG budget, the newly under the basic education MFO, itself one revised Program/Activity/Project (PAP) of three education MFOs.'This result reflects structure involves only one program for basic a common failing in program budgeting education among 20 operational programs reforms worldwide, in which the new budget across the other sectors. A comparison of the classification simply echoes former line item PAP structures for basic education proposed categories, usually in order to remain mapped in November 2012 for the DepEd National to an unchanged organizational structure. with that of the ARMM Regional Govern- The ARMM submission includes similar ment is illustrative of these very different structures for all its departments and agen- program budget structures (Table 4). cies, diluting the detailed objectives that the adoption of a results-based framework aims The ARMM PAP entails only one program to bring to the budget process, while retain- for "implementation of basic education ing all the additional demands of preparing services" subdivided by regional office and the budget to this format. schools divisions, while the DepEd Cen- Table 4: PAP structures for Basic Education in National Government and ARMM DepEd National ARMM Regional Government MFO P/A/P MFO P/A/P Curricular programs and learning management Curricular and special learning support Alternative learning and delivery Regional Office modes including distance edu- cation Learning systems for special needs IVIF1 Basic School health and nutrition MFO12 Implementation ti Education Services 16 Regional Operations at kin- of basic education services dergarten, elementary, second- ary, and learning center levels (excludes ARMM) Assistance to private education (GASTPE) 9 Division operations at Acquisition, procurement and elementary, secondary and equitable distribution of learning technicalvocational levels resources Facilities and teaching and learning-conducive environments lCT enabled solutions Assessment systems Sources DBM, Authors As this framework is increasingly used by Naturally, the impact of this mismatch 25 the Department of Budget and Management depends largely on how it is interpreted to justify new spending requests, the lack of in practice, particularly at the level of the elaboration in the ARMM OPIF may - if national budget institutions. However, even if due consideration is not applied - undermine the worst effects of poor integration between the ability of the ARIM Regional Gov- regional and national budget processes are ernment to articulate its needs effectively. mitigated through dialogue and compromise, ARMM is left in a Catch-22, in which it this remains uninstitutionalized and vulner- cannot meaningfully adopt the performance able. At the very least it undermines regional budgeting framework as is, yet failing to fiscal autonomy by increasing dependence adhere to the framework can affect its bud- and reducing the bargaining power of the getary outcomes, regional government. Results-based budgeting systems are a kind of Besides the mismatch in the ARvIM Re- program budgeting - an approach that aims gional Government linkage to central gov- to group together as a single "Program" all ex- ernent budgeting, there is a lack of sound penditure activities that serve a coherent aim, budgeting processes among the AR IM operation, output, or outcome.10 A meaning- institutions. In addition to the lack of a ful performance or program budget needs to regional budget law or regional DBM, few if include and relate all budget components or any ARMVM-specific budget procedures are activities contributing to this goal - in this in place to help AR aM's sectoral agencies case defined as an MFO. In the ARNVIM support the ORG preparation process in the Regional Government budget this is simply time available. These shortcomings involve not possible, because the fragmentation of both budgeting and reporting, and undermine public expenditure systems described above the ability of the region to represent its bud- means that significant resources contributing getary performance and future requirements to any given MFO lie outside that budget. effectively to the national government. In this sense, the idea of program or results- based budgeting in such a fragmented public Inconsistency between ARM\'M regional expenditure environment is nearly meaning- "devolutionamand national decentralization less. The paradox was astutely expressed by The third dimension of the ARMM PEM one ARMM civil servant: syndrome is that, in addition to the lack of budgetary control over many centrally derived How can we be expected to submit a program resources, the regional government is disad- budget, when our budget does not cover most vantaged with respect to the resources flowing governmentprograms? to other levels of local government nominally within its jurisdiction. Tcis situation is the re- FInscher (2007), p 119-20. ARMM BEcEIR interview, AMM Regional Government Staff (March 2013). 26 sult of contradictions or tensions between the Autonomy Act (MMA) 25, the regional two distinct decentralization processes in the version of the local government code adopted region: the nationwide process embodied by nationwide. This regional act did not specify the 1991 Local Government Code, and the that local government units are responsible to unique regional devolution under RA 9054. discharge devolved responsibilities, so these services, including education, remained a The absence of the regional policy and regional function. When RA 9054 superseded legislation on "local autonomy" called for in this act, it specified once more that the LGC the Expanded Organic Act means that IRA applies until a regional local government code funds are distributed to LGUs in the ARNIM is adopted.The situation remains the same according to the provisions of the 1991 Lo- until now: cal Government Code. This law provides that LGUs "discharge the functions and Asymmetry exists between,ARMM-LGUs responsibilities of national agencies and of- and non-ARAMLGUs in terms ofthe fices devolved to them".52 At the time of the penditure responsibilities that are assigned to Philippines-wide decentralization instituted them because the regionalgovernment ofthe under the LGC, these functions and respon- ARMM has not devolved any of theirfunc- sibilities included a wide range of public ser- tions to the LGUs within theirjurisdiction vices, though notably not education excepting even fRA 6734 allows them to do so. 7hus, some construction and maintenance of school ARMM-LGUs did not receive any of the buildings.53 IRA shares are meant to support devolved activities that have been transferred provision of those services. In terms of their to non-ARMMILGUs underR- 716. 5 expenditure assignments, ARMVM LGUs continue to receive the full non-ARMM LGUs are treated on an almost IRA allocation as per the stipulation of RA equalfooting as the regional government 9054 that the shares of national taxes due to ofARMM ... although thefunctions and LGUs may not be reduced, regardless of the responsibilities assigned to the regional gov- expenditure responsibilities of the region or ernment ofARIM are slightly broader than the LGUs.5' As noted above, this share of those given to non-ARMMLGUs.54 resources is about half of all public resources flowing to the region, more than the ARG But this devolution is not implemented in the budget itself. In short, LGUs have substan- ARIM, where the responsibility for services tial external resources but reduced expendi- lies primarily with the regional government tures due to the responsibilities given to the under the regional autonomy provisions. ARMM regional level through the Expanded In 1994 the ARMM Regional Legislative Organic Act. This has the effect of reducing Assembly (RLA) passed Muslim Mindanao the incentives for LGUs to raise own source R2 A 7160, Sec. 17. ve Asian Development Bank/World Bank (2005), p 4. SWorld Bank (2003), p. 22. Asian Development Bank/World Bank (2005), p. 5. HA 9054, Art. IV, Sec. 1 revenues. In addition, when LGU resources in will be seen in the next Chapter in relation to 27 other regions are directed to social spending budgeting for Alternative Learning Systems. even if not directly to education, it increases The result is an ongoing cycle of incomplete the fiscal space in those areas to allocate other budget proposals and centrally-determined funds to education. In the ARMM by con- incremental budgets. trast, the IRA available in ARMM are widely reported as used for personal slush funds by The lack of sound information on the use of local LGU heads, or diverted to settle rido- government funds in the ARMM plays into type conflicts.17 and increases the mistrust in the relationship between the ARMM Regional Government Weak stewardship and accountability and the national authorities, particularly the in handling public funds DBM. At the same time, the ambiguity has The fourth dimension of ARMM's public served the purposes of some national political expenditure management syndrome is the forces, as relatively unaccountable support to history of poor governance as expressed in ARMM has been part of a system of inter- weak budget accountability and reporting. dependence in patronage politics, in which A 2010 review of ARMM's education budget support and autonomy are given in return control institutions found that there are for the provision of electoral support.'Ihese "[w]eak internal control systems" and "no problems are exacerbated by some national clear guidelines for reporting on budget agencies dismissing the need for improve- utilization by the implementing units, and ments to reporting systems, audit, and ac- more importantly a report correlating budget counting as a regional responsibility due to spent with achievement of physical outputs or the supposed "autonomy" granted under the performance targets is non-existent"." Expanded Organic Act. There is therefore a lack of systematic capacity support in meeting At least until the 2013 budget, the typical national fiduciary and budget accountability education budget reporting practices within requirements compared to other regions. That ARMM have consisted of schools, districts, being said, as awareness and communication and divisions failing to provide proper report- of needs between ARMM and the national ing to the regional office on budgets they government improve, there is increased will- received or used, while the regional office ingness to provide training and support from has responded by submitted budget account- national agencies, particularly the DBM and ability reporting on those funds to the DBM DepEd National. More concerted support is that it completed without corresponding data. needed to offset the deficits in this area. Without data on actual spending, central in- stitutions cannot budget additional resources A lack of qualified or appropriate staffing for ARMM even when they are needed, as for the important functions of budget * Torres (2007), p. 109. " Tribal Helm (2010), p 32-34. 28 accountability has also developed in the manipulating existing flows of resources rather ARMM. This is in part due to a historical than accounting for those used and planning politicization of appointments resulting in for those needed in the future. inappropriate skills and the emergence of vested interests in the continuation of the Finally, there is little or no provision for pub- system. But it can also be traced to the lack of lic or civil society scrutiny and accountability attention to requirements of a fully-fledged for public expenditure or service quality. In planning and budgeting function some respects, the lack of accountability has at regional level, leaving the ARMM been deepened by the diversion of electoral somewhere between being a true local processes towards patronage politics, and few government with the necessary staffing for incentives yet exist for active "horizontal" or public expenditure management, and an "downward" accountability to citizens. While administrative tier with minimal capacities some of these accountability issues can be to implement budgets given from above. In found throughout the Philippines, they are general, the continuation of these practices more severe in the ARIM, and compounded and entrenchment of clientelistic hiring by the lack of clear responsibility for proper has resulted in various kinds of productivity reporting created by the ambiguous nature of damaging forms of corruption, such as the the autonomy arrangements. introduction of ghost workers, and a focus on R RA 7160, Sec. 17. 5 Asian Development Bank/World Bank (2005), p 4. * World Bank (2003), p. 22. Asian Development Bank/World Bank (2005), p. 5. RA 9054, Art. IV, Sec. 1 CHAPTER FOUR: Whether measured as a proportion of GDP 29 BUDGETING FOR or real expenditures per pupil, government BASICeducation spending fell significantly after the BASICEDUC TIONfinancial crisis of the late 1990s until about IN ARiniM 2006, and in 2008 real per pupil expenditures had still not regained 1998 levels. Spending also fell considerably short of that required by the DepEd's Education for All (EFA) re- source estimates and was only about half the the relationship between expenditure and predicted GDP shares for countries of similar good education outcomes is complex. Levels of income levels.60 spending do matter, but as or more important is the composition of that spending in relation A 2011 Philippines-wide public expenditure to strategic goals and the input mix required review concluded that "[in education, trends for quality education services.'The budget is in public spending and key sector outcomes are the concrete expression of a government's closely related" and a 2012 education public goals, and if budgeting is poorly connected to expenditure review showed "compelling evi- policy and plans these goals have little chance dence that deterioration in outcome indicators of being realized. Indeed, it is broadly accepted has coincided with the weakening of govern- that the "[f]ailure to link policy, planning and ment support at the level of public spending budgeting may be the single most important on the education sector".61 More recently, real factor contributing to poor budgeting out- education budgets have been on the increase, comes at the macro, strategic and operational and it can be expected that this renewed re- levels in developing countries."p9 sourcing of the sector will bring benefits. This chapter outlines the size and composi- Non-government basic education spending tion of public budgets for basic education Education spending in ARMM - as in most in ARMM. It then traces the impact of the developing contexts - involves a combination public expenditure management syndrome of government spending, donor contribu- outlined in the previous chapter on budget tions, and private expenditure by households. preparation and composition. However, compiling comparable aggregate measures of spending on basic education Aggregrate basic education spending levels in AR tM is complicated by three factors: in the ARhMM the fragmentation of government spending A low level of aggregate education spending among multiple budget channels, the lack of in recent years has been a key contributor accurate donor information broken down by to poor education outcomes countrywide. region or by yeart and incomplete reporting axWorpd Bank (1998), p 31. 6 World Bank/Australian Aid (2012), pp 22-27 61 World Bank (2011), p6; World Bank/Australian Aid (2012), p 21. 30 on Local Government Unit supplementary Donor spending on education in ARMM can spending on education. Due to the impos- be traced primarily to Australian Aid through sibility of generating strictly comparable its BEAM project, UNICEF, and USAID. spending figures that include all government BEAM-ARMM is the continuation of the and non-government sources for ARMM, Mindanao-wide BEAM project, providing a and the public expenditure management range of capacity development, community- focus of this study, this analysis focuses on based schooling and other services. USAID's government spending on basic education. EQUALS II program is aimed at enhancing Nevertheless, before focusing on government community participation, improving teaching spending it is worth reviewing the contribu- capacity, access for out of school youth, and tion of non-government sources. support grants to secondary schools. Table 5: Indicative key donor education sector support to ARMM (thousand Pesos) Source Program Budget Annualized Budget BEAM - ARMM (2012-2017) 3,927,000 785,400 USAID (EQUALS II) (2006-11) 1,239,000 246,000 ARMM Social Fund N/A 658,140 Source: BEAM-ARMM Program Design Document (2011); ARMM BE-PEIR Interview, USAID (February 2013). Note: The annualized EQUALS II budget is based on US $60 million over a five year period. UNICEF is also a major donor to educa- focuses on development activity in conflict- tion, but available data is nationwide. For affected areas and is administered by the 2012-15 US $21.2 million is budgeted for World Bank. the Philippines as a whole. In addition, two major multisector and multidonor projects, While these amounts are estimated, the cur- the Mindanao Trust Fund and the ARMM rent donor spending specifically earmarked Social Fund make contributions to the educa- for education in ARMM would amount to tion sector.'Ihe Social Fund is a government about 1.7 to 2 billion PHP annually. This project with World Bank and JICA lending represents about a quarter of the amount of that contributes to the sector primarily via government spending. It is worth considering construction of mixed-use community learn- the sustainability implications of this figure. ing centers and earlier through support to the In general, donor spending on education in madrasah sector.'The Mindanao Trust Fund ARtM is geared towards capital invest- ment in the form of infrastructure, with some On the other hand, as these facilities are in 31 resources for alternative delivery modes or the many cases planned for areas where regular Alternative Learning System Program. Since primary or secondary schools are not planned - these donor contributions are not considered as in mixed use or community learning centers in the annual budgeting process the long term - the distortionary impact of donor spending on recurrent implications of the new infrastruc- facilities planning may be less important. ture are not necessarily being considered. Figure 7: Source: Household Expenditure on education, 2009 (percent by region) NSO, 2009 Family Income and Expenditure Survey Final Results 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% 0.0% I e d n > > > . 2 0 .0 0 0 .2 rM 00 .0 0 0 CL Information on household expenditure on tures on education is even greater in absolute education in the Philippines as a whole is terms: ARVIM families spend about 1,567 available from household level expenditure Pesos on education, against a national average surveys, but is not disaggregated by the type of 8,071 Pesos. 61 Private expenditure is thus of schools (private or public) that children small, and donors tend to support needs of a attend, nor by grades. It is therefore difficult particular type, addressing education sys- to make robust comparisons between public tem issues and infrastructure in particularly and private expenditures on education.62 As remote or conflict-affected areas that may be a relatively impoverished region, the impact less integrated to the public education sector. of the low percentage of household expendi- 62See also World Bank/Australian Aid (2012), p 31. 63 2009 Family Expenditure and Income Survey Final Results. 32 Government spending on basic education DepEd ARMM budget in the ARG Basic education in ARMM is heavily de- appropriation of the national budget; pendent on public sector resources, with the DepEd National allotments for majority of these flowing through the ARG ARMM, including program funds such budget in the form of Personal Services as textbooks, School-based Management expenditures for salaries and other compen- grants, and others; sation. Capital outlay and various program Spending on education via other expenditures form a smaller portion of the Ministries such as education facilities built total funds, and are primarily budgeted via by the Department of Public Works and DepEd National's programs. These resources Highways (DPWH) through the School have been growing steadily, in both nominal Building Program, and the cost of the and real terms (Table 6 and Figure 8). Conditional Cash Transfer grants under the DSWD, which provides support to Government spending on education can be families on the basis of both health and divided into four main categories: education conditions; Local Government Unit spending in- cluding the earmarked Special Education Fund (SEF). Table 6: Aggregate government education expenditures in ARMM (thousand Pesos) Source 2009 % 2010 % 2011 % 2012 % DepEd 4,644,785 89% 4,846,119 96% 5,818,269 84% 6,370,646 75% ARMM DepEd Na- 454,122 9% 203,449 4% 1,096,447 16% 2,086,615 25% tional SPF-SBP 77,709 1% N/A N/A N/A N/A N/A N/A (DPWH) LGU 17,690 0% 16,797 0% 42,868 1% 20,675 0% Total 5,194,306 100% 5,066,365 100% 6,957,584 100% 8,477,936 100% Source: BESF for DepEd ARMM, BLGF for LGUs, SAOB for DepEd National 2008-11. *Note: DepEd National ARMM allotments for 2009-2011 are obligations and not budgeted amounts due to the absence of regional allotment information in the national budget prior to 2012. Figure 8: 33 Nominal vs. Real government education expenditures in ARMM, 2009-12 9,000,000 8,500,000 8,000,000 7,500,000 7,000,000 6,500,000 - Nominal 6,000,000 - Real (2009 Pesos) 5,500,000 5,000,000 4,500,000 Source: 4,000,000 BESF for DepEd ARMM, BLGF for LGUs, SAOB 2010 2011 2012 for DepEd National 2008-11; Philippines IPIN deflator Direct comparison of the ratio of education While underspending in social sectors has spending in ARMM to total spending are occurred in the ARMM historically, more complicated by the fragmentation of education recently the primary issue for basic education budgets between regional and national agen- delivery is no longer the level of aggregate re- cies described in Chapter Three.'The ARIM sources, at least in comparison to other regions DepEd budget has grown from 49 percent of in the Philippines. Instead, the problems cen- the ARG appropriation in 2009 to about 57 ter on issues of composition (what is bought percent in 2013. However, this represents only with the available funds), execution (how effi- spending that is appropriated to the ARG, and ciently funds are spent), and delivery (whether excludes national agency spending in ARNIM the purchased inputs are available at the point for both education and non-education purpos- of service delivery). The next section discusses es. However, the aggregate spending on educa- budget composition and its relationship to tion from all government sources also grew as fragmented public expenditure management a proportion of aggregate public expenditure in the ARIM, and the following Chapter ad- from all sources.'Ihis proportion went from dresses issues of execution and delivery. 21 percent of the aggregate public expenditure for ARMM in 2009 to 29 percent in 2012, Composition of government spending on indicating growing prioritization of education basic education by both regional and national government.s4 The fragmentation of the public expendi- 1t Authors calculation based on data in Tables 3 and 6. 34 ture system between national, regional, and within the national purview, though it is itself local institutions outlined in Chapter 'Three divided among different programs and special has an important influence on how budget purpose funds. Finally, some capital construc- composition evolves in ARMM generally, tion does come through the PDAF congres- and for education in particular.'The bulk of sional allocations and LGU allocations. the recurrent budget is held by the ARMM regional government, with existing staff This kind of "dual" or divided budgeting benefits under Personal Services (PS), and in which recurrent and capital budgets are routine Maintenance and Other Operat- prepared separately and remain unintegrated ing Expenditures (MOOE) falling with is common in many countries and minis- the ARG appropriation. However, there are tries. Recurrent budgets are often prepared significant exceptions. Many of the funds for by finance departments within line minis- specific education programs and textbooks are tries or finance ministries, while planning classed as MOOE but budgeted by DepEd departments or ministries handle capital National. In addition, some expenditures are expenditure." However, the ARMM edu- slightly misleadingly categorized: for example cation case is extreme in three respects: the budgets for new staff and for Arabic and recurrent budget and capital budget proposals Islamic Education (ALIVE) teachers are with for education are split between two entirely DepEd National and managed through a separate government line agencies (the ARG cash allowance classed as MOOE despite its and DepEd National); they are prepared at actual use for salaries and benefits. different levels of government (the regional and the national); and the MOOE budget is Capital outlay, particularly for school build- spread across the two agencies and two levels ings and furnishings, is virtually entirely (Table 7). Table 7: Locus of control of government spending by economic category, FY 2013 DepEd ARMM DepEd National a Total PS 7,114,906 100% 3,118 0.04% 7,118,024 MOOE 359,222 36% 647,636 64% 1,006,858 co - 58,475 100% 58,475 Source: TFY 2012 NEP Appropriations/GAA. Note: Excludes LGU spending. MOE for "New Teacher Items" in DepEd National budget has been reclassified here as PS. 65 Schiavo-Campo (2007), pp 246-8. An analysis of composition of expenditure Personal Services 35 therefore involves estimating from dispa- As Personal Services are included in the rate sources of funds. The main components ARG budget, comparison with equivalent of government education expenditures in spending elsewhere is relatively straightfor- ARMM are: ward. For the SY 2011-12 the Philippines average regional operations spending on * Personal services expenditures (wages Personal Services was just over 8,500 PHP and benefits) for existing staff, budgeted per pUpil.66 ARMM had the lowest per pupil in the GAA under the ARMM Regional spending at 7,104 PHP, but this was close Government appropriation, and managed to the level of several other regions. Given via the ARMM regional government; concerns about enrolment figures, with enrol- * "Regular" MOOE under the regional ment discounted by 10 percent, the figure in government appropriation and managed ARMM would be 7,900, slightly below the by DepEd ARMM; national average. An alternative measure of * Cash allowance for new staff posi- equity in PS allocations could be operational tions, budgeted under the DepEd Na- PS per head of current population (based on tional appropriation as MOOE; the much improved 2010 census and popu- * A range of centrally-managed lation growth figures). This measure shows program funds including key operating that the 2013 budget ARMM allocations expenditures such as textbooks, key capital are about equivalent to the national average, expenditures through the Basic Educa- actually coming in at 2,070 PHP per capita, tion Facilities fund, as well as a range very slightly higher than the national aver- of programmatic activities budgeted by age of 1,980 PHP.67 In sum, the level of PS DepEd National and categorized variously spending in ARMM may be slightly low on as MOOE and Capital Outlay (CO); a per pupil basis, but not far out of line with * Congressional allocations via the levels elsewhere in the country PDAF and contributions to educational facilities from the infrastructure alloca- tion, also programmed by legislators and classed as CO; and * LGU spending on education from the internal revenue allotment and the SEF budgeted and managed through the LGUs themselves. This is in order to compare regions: the PS for each region is used, and PS at DepEd Central Office is excluded. Includes all public school enrollees, both elementary and secondary. 67 Population is trom 2010 Census with reported regional growth rate to 2013; 2013 GANER 36 Figure 9: Source: PS per pupil by region (SY 2011-12 enrolment, FY 2013 budget) EBEIS online; 2013 GAA/NEP 12,000 10,000 8,000 6,000 4,000 2,000 C 0 C > - > ** m .2 0000 a) ~ .2 .2 ' m . Note: Note: This and the per pupil analyses that follow rely on SY 2011-12 enrolment figures, the latest full set available, against the FY 2013 General Appro- priations Act. They are not therefore a strictly accurate per capita figure for either year However, current allocations are not them- 9 of the 13 divisions in the region have severe selves an indicator of adequate provision, par- secondary teaching shortages, and every divi- ticularly if past spending has been deficient. sion registers some shortage." Even if dis- In both elementary and secondary grades for counted by 10 percent, these ratios would still the 2011-12 SY, the reported pupil-teacher be higher than anywhere else, What these ratios are the highest in the country by a figures show is that while efforts are generally significant margin. In Elementary education, being made to adequately resource teachers ARIM education divisions account for 7 of for ARMM, there are still barriers to this the 11 with the highest "severe" need clas- budget being converted into actual teachers sification (50 students per teacher as against a in classrooms - these barriers are discussed in standard of 40-45 depending on grade). the next chapter. soaEBEIS Online, SY 201s1-12. 37 Figure 10: Source: Primary and secondary teacher-student ratios by region (SY 2011-12) EBEIS online, SY 2011-12 60 50 40 30 20 10 0 M > X X R 2 M C 0 a _ _ a p c l tha .2 0 of.2 ful studen .2 .2 is.n t Elementary Pupil-Teacher Ratio Secondary Pupil-Teacher Ratio However, officials in ARMVM have recently sources available for teaching materials and publicly claimed that with the elimination of the operation and maintenance of facilities. fraudulent student enrolments, it is no longer These "non-salary recurrent" expenditures the number of teachers or the PS budget are squeezed in many budget settings due allocation that is the primary constraint to to the relative difficulty of reducing staffing education performance, at least in terms of expenses and the separation of the recurrent national policy guidelines on teacher-pupil and the capital budgeting processes alluded to ratios or comparison with other parts of the above. In the Philippines such a dynamic was Philippines.69 Such a rebalancing of figures dramatic during the 1990s, when aggregate onto a more accurate footing represents a MOOE spending per pupil fell by half to a significant and important achievement for paltry 422 PHP per pUpil.70 ARMM governance. MOOE allocations for ARMM education MOOE and the" "rutineo oprtin MOQ ad te routn operations consist of both an "ordinary" component and maintenance gap" budgeted in the ARG education budget for The quality of education services is not just operations, and several programmatic alloca- a function of teachers, but also of the re- tions in the DepEd National Budget. The ThSeease (2013). ehoWorld Bank/Australian Aid (2012), p 46. 38 former operating MOOE is for distribu- allocation per pupil, about 1170 PHP against tion among schools to meet their routine a national average ofjust over 1400 PHP. maintenance and operating needs. The latter, nationally programmed expenditures include Figure 11 examines the allocations of key inputs for basic education institutions MOOE per pupil by region, taking into such as textbook allocations and school-based account both ARG and DepEd National pro- management grants, as well as many signifi- gram budgets for all regions.71 ARNM was cant special program costs that are gener- again the region with the lowest allocation ally, though not entirely, classed as MOOE. per pupil, about 1170 PHP against a national ARNIM was again the region with the lowest average of just over 1400 PHP. Figure 11: Source: MOME (operations and programs combined) budget per pupil by region EBEIS Online; 2013 GAA. (SY 2011-12 enrolment, FY 2013 budget) 2,500 2,000 1,500 1,000 500 M X X W a .> > X o~~ ~ .2 ox 0 .2 .2 o .2. Given ~ ~ ~ ~ ~ ~ ~ ~ e pupil,ern about 1170men PHPer aihcnitnccosrgins ANnationa aRM w as againthepreion wiothe thmeastveae of justnll aover 1400 PHPnavraeof1. MOOEbuge oroperations andprgascmie)bdtpr pibrgions. BEis Onlie 013 aton GAAha ot Sn (s 2011-12 eolethe21 budget)rP n OEintentoa n 2,5000 2,000 Given the concerns about enrolment numbers hg ossec cosrgos RVMi as a basis for comparison, another measure of magnlyboeteainlavreof1. aggregate MOOE provision by region could pecnwih1.pretoftspraos be to express it as a proportion of the overall adpormbde ossigo OE budget for operations and programs.'IhisThimlctosaehtbthPan can be thought of as the amount of MOOE M OEmybloagisrertdno- available per unit of spending on personnel metanbysilrmot. and programs.aOnbthistmeasure,iwhich shows 7This measure excludes the budget for PS and MOOE in the national and regional education headquarters as these are not allocated to regions. However, using the aggregate MOOE alloca- in ARMM relative to other regons. Figure 39 tion for ARMM overestimates the MOOE 12 illustrates this "routine operations and available for ongoing or routine maintenance maintenance budget gap" by comparing the and operations activities. Most programs' regional MOOE budget as a proportion of MOOE expenditures are linked to specific the regional operating budgets including activities and are therefore not available for programs, with MOOE as a proportion of more routine activities. As such, including all the regular operating budget (the DepEd MOOE allocations for programs obscures ARMM appropriation available to fund rou- significant underfunding of regular opera- tine school operations and maintenance). tions and maintenance for education services Figure 12: Source: MOtE as % of operating & program budget 2013 GAAiNEP vs. % of operating budget, FY 2013 14% 12% 10% 8% 6% 4% 2% 0% cc 4 . >. X X .2 0 .2 o .2- 0 0 .2 0C S02 .20 .' 2 .2 a) 0 . .2 0 5 .2 z I MOOEIRegional Operations & Program Budget I MOOEIRegional Operations Budget 40 The difference in the two measures is striking. ters of the national average, and 81 percent of When program specific MOOE allocations that in the next lowest region, Region I. are included, regional MOOE allocations range from 10.5 percent to 13.6 percent of This "gap", which is invisible if MOOE for the combined operating and program bud- both programs and operations is lumped gets. As just noted, ARMM's proportion together, is even more striking if calculated as is slightly over the national average of 12.1 a per pupil amount (Figure 13). ARMM again percent with an MOOE allocation of 12.3 has the lowest allocation in the country, 417 percent. However, separate analysis of the PHP per pupil. The national average is 715 operations budget reveals a different picture: PHP. The next worst off region (the National ARMM's MOOE allocation is only 4.8 Capital Region, which also enjoys the highest percent of the regular operating budget, well availability of Special Education Funds) still below 6.0 percent in the next worst off region has over 600 PHP per pupil. In short, ARMM and the national average of 6.5 percent. In enjoys only 58 percent of the national average short, the MOOE available for routine oper- in per pupil MOOE allocations for routine ations and maintenance as a proportion of the operations, a significant disadvantage in terms operations budget is only about three-quar- of the provision of basic education services. Figure 13: Source: Per pupil MOOE in regular operations budgets by region EBEIS Online; 2013 GAA/NEP (SY 2011-12 enrolment, FY 2013 budget) 1,200 1,000 800 600 400 200 0 > > - U .2 .2 0 .02 z This low routine operating and maintenance compensation of ALIVE teachers. 41 budget allocation within the ARG budget This "routine operations and maintenance is hidden by larger MOOE allocations for gap" has very significant real-world implica- ARMM within program funds budgeted tions. In the ARMM the operating MOOE and executed by DepEd National. In par- budget for primary and elementary schools ticular, the ARMM allocation for just one is entirely utilized by Division and School centrally-managed program accounts for District offices, with none being distributed almost all the difference between the two to primary or elementary schools.'This situ- measures: Madrasah education was allocated ation has until recently been considered by 209 million pesos in 2013, about eight times ARMM local education staff as an accepted the national average of 26 million. This high matter of policy, despite no such policy in the allocation makes sense contextually and is remainder of the Philippines.'This is worth without doubt welcome for its contribution to emphasizing: in part due to a shortage of Arabic Language and Islamic Values Educa- regular MOOE budget in the ARG ap- tion (ALIVE) programming in the ARMM propriation, primary and elementary schools region. However, despite its size it cannot in the ARMM receive no regular operating compensate for the low MOOE availability budget from the DepEd ARMM and must for routine maintenance since it is used for fund any routine operations and maintenance different purposes: primarily the training and expenditures from other sources.72 Figure 14 Lack of operating and maintenance resources ARMM BE-PEIR Interviews with DepEd ARMM and district staff, confirmed by eSiDS survey results in which no primary or elementary schools reported receiving a regular MO.E allotment. 42 Within DepEd National's program budget one quarter of that found nationwide, and there is one item that arguably can contribute argued that lack of maintenance and opera- to routine operation and maintenance and tions spending was a "critical" supply factor might therefore provide a remedy for the gap: explaining low performance." The overall School-based Management (SBM) fund- picture of MOOE for education in ARMM ing. School-based Management grants are has therefore improved a great deal as ag- intended to assist schools to face identified gregate allocations have been brought in line challenges, and is to be merged with MOOE with national averages. But when examined funds for schools in future budgets. more closely these numbers obscure a con- tinuing and crucial "routine operating and Unfortunately, as yet budgets for SBM maintenance expenditure gap", and this gap are both small when compared to regular has an ongoing impact on frontline service MOOE allocations, and they are distributed delivery, especially at primary and elementary unevenly, with ARMM at a small disadvan- level. tage as a proportion of its operating budget. Including the SBM budget as part of the Textbooks MOOE budget available for routine main- In addition to the ability to regularly maintain tenance increases the proportion of MOOE and fund operations in schools, MOOE also in the operating budget nationwide from 6.5 is a key source of educational materials, chiefly percent to 7.0 percent, but the correspond- in the form of textbooks. A comparison of re- ing rise in ARMM is from 4.8 percent to 5.1 gional textbook allocations reveals no system- percent, which leaves the ARMM alloca- atic disadvantage to ARMM, though of course tion at effectively the same relative level: 74 the impact of any allocation also depends on percent of the national level. SBM allocations existing textbook stocks. ARMM is currently therefore provide some increase in the overall budgeted for more textbooks than the national funds available for school improvements average, and this likely is needed to overcome everywhere, but do not compensate for the historically lower stocks (Figure 15). disparity in routine operations and main- tenance funding between the ARMM and Beyond the question of budget allocations, other regions of the Philippines. the most important concerns regarding text- books relate to the execution of the budget: Distorted expenditure composition with re- how many books are bought and delivered, spect to MOOE is not a new phenomenon in when and on what basis are they distributed? the ARMM, and not limited to education. A Textbooks are budgeted by DepEd National, review of education, health and social protec- procured centrally and delivered directly to tion in ARMM published in 2003 found the school divisions and districts. Even though proportion of MOOE spending to be about the DepEd National maintains records of " World Bank (2003), pp 28, 47 textbook allocations, poor data provision about what books they might receive, how 43 from ARIM division and district offices to many, and when. In the next chapter the inform national textbook distribution, and budget execution data shows low and variable limited dissemination of the information rates of budget utilization, and the data from available at national level on the distribu- the BE-PEIR schools survey also discussed tion within ARIM means that schools in in Chapter Five suggests uneven delivery of ARMMI generally have little information books to school level. Figure 15: Source: Textbook budget comparison per pupil EBEIS Online; 2013 GAA/NEP (SY 2011-12 enrolment, FY 2013 budget) 120 100 80 60 40 20 .2T - - M- W- 2 E - buge 0 ~~~ ..2~ 00 .2 .2 Z Other programs A large part of the government education regions. These programs have different objec- budget consists of about 20 programs bud- tives, from the provision of key inputs such geted and managed by DepEd central, of as classroom construction to programmatic which most now (since 2012) have earmarked activities such as support to special schools, appropriations by region, ARMM included. Islamic education, or Alternative Learning As noted earlier, previously these regional al- Systems. The main programs are listed in locations were sub-allotted to regions during Table 8 with the corresponding allocation for the yearI or administered from neighbouring ARMir in 2013. 44 Table 8: Centrally-managed fund allocations for ARMM, FY 2013 Program Amount for ARMM (PHP thousands) Cash Allowance for Newly Created Teaching Positions 3,118 Purchase of Textbooks and Instructional Materials 73,284 Support to Private Education - GASTPE 184,587 Financial Assistance for Regional Science High Schools 1,634 Support to Special Programs for the Arts and Sports 1,000 Support to Special Education Centers/Schools 2,140 Support to ESEP High Schools - Support to Special Elementary Science Schools Support to tech-Voc High school Programs - Human Resource Training and Development 62,140 Purchase of Office and IT equipment 5,693 Basic Education Facilities N/A Support for MultiGrade Schools 5,775 Computerization Program 46,414 Basic Education Madrasah 209,700 SBM Installation and Support 25,924 Kindergarten Education for All 52,110 Mass Production of Math and Science Equipment 27,692 Alternative Learning Systems 7,699 Alternative Delivery Mode Programs 4,900 Source: 2013 GAA/NEP; ALS is the allocated amount planned by DepEd Bureau of Alternative Learning Systems, and may not reflect actual appropriation. An analysis of the equity in spending across ing, and execution rules of the programs have regions for these programs shows variations been a barrier to transparency between the whether on a per pupil or per capita basis, regional education authorities and the national as would be expected of programs meeting level making these decisions. A more regular varying needs and applying varying planning dialogue has been developing and regional criteria and processes. Support to special arts input can be integrated into these programs as and sports programs has been given as an equal recommended in Chapter Six. share (1 million PHP) to all regions. Support to multigrade schools varies widely per pupil, One of these programs, Madrasah education with ARMM below the national average of (which in fact does not primarily fund ma- about 9 PHP per student at 6.7 PHP per stu- drasahs but rather ALIVE subject teachers) dent. In the past the varying planning, budget- is a key component of the national education authorities'spending plans for ARMM. About ing and execution of this program, which is 45 46 percent of national spending on the pro- so heavily weighted to the ARMM, may be gram is allotted for the region, far more than a good potential candidate for more regional any other part of the Philippines. The budget- planning and expenditure management input. Figure 16: Source: Total allotment for "Madrasah Education" by region, FY 2013 2013 GAA/NEP 250,000 200,000 150,000 100,000 50,000 0 - - 02 0 z .0o .2 .20 .. .2.l 0 .o z A second program that is of particular impor- rated according to these proportions across tance to the ARMM is Alternative Learning regions.74 Systems (ALS), as it targets the out of school youth of which there are disproportion- The regions and school divisions are required ate numbers in the region. ALS budgeting to provide reporting on the number and provides a good illustration of the problems identity of the mobile teachers and coordi- created by the ARMM's public expenditure nators, and the learners attached to each, to management syndrome. ALS needs are cal- verify their inclusion in the budget calcula- culated by the Bureau of Alternative Learn- tion. Due to the weaknesses in accountability ing Services in DepEd National, based on the and the unclear responsibilities for reporting numbers of mobile teachers and coordinators described in Chapter Three, the reporting on in each region's divisions. The appropriation ALS needs is very problematic. Despite the approved in the national budget is then pro- purported needs, in 2012 ARoM registered 1tBE-PEIR interviews, BALS and DepEd National Budget Department. 46 the lowest number of ALS enrolees of any out of 58 mobile teachers were disbursed in region (10,073), and as of May 2013, only ARMM, though this latter issue appears to 230 enrolees are in the records held at DepEd be a problem common to many regions.71 National for that year, and only four of twelve ARMM divisions have provided any report- In this respect, planning for ALS spending is ing at all for the 2013 year. largely driven by the completeness of admin- istrative reporting and current provision levels, These weaknesses have a double impact. First, rather than needs. An analysis of the alloca- they influence the budget allocation given to tion of operational budget for mobile teachers ARMM by underestimating needs. The lack and coordinators against the potential learner of involvement of ARMM in the budget- population reveals that allocations vary widely ing process may weaken the regional office's between regions from just over 4 PHP per ability to gather the adequate reporting from learner to 25 PHP (Figure 17) - all seemingly its divisions. Second, and perhaps worse, low allocations.76 ARMM is not particularly even the money that is allocated cannot be disadvantaged, with 9.9 PHP per learner disbursed without verification that there against a national average ofjust over 10 PHP, are both learners on hand, and teachers to so it would appear budgeting issues for ALS teach them. In 2012, allowances for only 11 are shared in other parts of the Philippines. Figure 17: Source: Planned ALS Allocation per Target Learner 2013 (PHP per Learner) Learner estimates from Koyama and Yamauchi, 2013, based on data from the Functional Literacy, Education and Mass Media Survey 2008; Allocation data from DepEd BALS planning allocations CY 2013. 30 25 20 15 0 00 a DepEd National BALS documentation. 7Bse on estimated learner populations as determined from FLEMM 2008 household survey by Koyama and Yamauchi (2008), and BALS allocation planning data. Capital Outlay has not typically been presented in the na- 47 Capital expenditure on basic education consists tional budget, so allocations for the BEFF are primarily of money for construction and repair available for 2012 only from year-end report- of school facilities, as well as smaller amounts ing. The BEFF is to be allocated according for programs such as technical and vocational to color-coded lists of schools with facilities education or computerization. The provision of shortages assembled from the EBEIS.77 The funds for construction and repair are primar- BEFF for 2012 reveals a strong intention to ily driven by two sources: the Basic Education improve ARMM's educational infrastructure, Facilities Fund (BEFF) within the DepEd with the region being allotted 928 PHP per National appropriation, and the School Build- pupil, against a national average of 387 PHP ing Program, a national special purpose fund (Figure 18). ARMM is thus the most favored channeled to the Department of Public Works region in the country for the construction and Highways for implementation. and repair of classrooms, with the National Capital Region close behind. However, as The BEFF is by far the larger of the two discussed in the next Chapter, there are ques- sources, with about ten times the allocation tions regarding the efficiency of the execution nationally of the special purpose fund. The of these funds and little information on the regional sub-allotment of both these funds completion of projects. Figure 18: Source: Basic Education Facilities Fund (BEFF) allotment 2013 NEP (2012 Actual Expenditures) per pupil by region, FY 2012 1,000 900 800 700 600 500 400 300 200 100 M X X cc > > X C pupil, ai 0 o 38 PH 0 .0 Z 7 DepEd Order 94/2011 "Guidelines on the Implementation of the Basic Education Facilities Fund' (November 29, 2011). 48 BEFF regional allotments do appear to the program implemented from ARMM's respond effectively to differing levels of neighboring regions. 2009 data, the last avail- classroom stock and the high allotment for able, shows about 90 PHP per pupil allocated ARMM appears amply justified. Indeed, the to ARMM, about 10 percent below the records show both NCR and ARMM as national average of 102 PHP.'9 The School areas of high priority for classroom shortages. Building Program appears therefore to be Seven of ARMMs 13 divisions are in "severe" allocated based on criteria other than class- shortage of elementary classrooms and there room shortages, although the other region of are over 62 pupils per classroom in the region. extreme shortage - the NCR - did in 2009 This is compared to a national average of 40 receive the highest per pupil allotment of pupils, and a national minimum service stan- this fund. A key question for the linkage of dard of 45 pupils for single shift schooling. planning needs with composition of budgeted The NCR appears even worse situated, with expenditure is therefore how planning for over 77 students per classroom in 2012. these two flows can be better integrated to address classroom needs as a whole. Secondary classroom shortages are more evenly distributed around the country than Overall, government expenditure composi- those in elementary schools. However, they tion for ARMM education is broadly in line are extremely severe in ARMM. In 2012 with national proportions when taking into the region had 82 enrolled pupils on record account both DepEd National and DepEd per classroom against a national average of ARMM budgets, but excluding LGU and 52, and a policy standard of 45 (55 for high SPF data. About three quarters of the budget school). In eleven of twelve ARMM school is expended on PS, 11 percent on MOOE, divisions, there is no way to meet the national and the remainder on CO. As noted above, standard even under double shifting. NCR is for 2012, the last year for which the BEFF again the second worst provided region for allocation by region is available, the share secondary school facilities. At the same time of ARMM expenditure for capital outlay as efforts to close these gaps are ongoing, in- has grown to 18 percent, as compared to 10 vestment in ALS and ADM canhelp address percent across the country. Given increasing classroom shortages by continuing to provide emphasis on school building and the lack of learning opportunities outside the classroom. routine MOOE already discussed, the ratio suggests increased attention will be needed to Determining the regional allotments within the recurrent operations and maintenance im- the School Building Program Special Purpose plications of the new infrastructure, perhaps Fund is more difficult, both as allotments to through continued provision of school repair regions are not made in advance, and because and maintenance grants. there has been some school building through " All classroom gap data from EBEIS Online (SY 2011-12) "Source on SBP Figure 19: 49 DepEd national and overall ARMM education expenditure composition, FY 2012 DepEd National ARMM Education Source: 2013 GAA/NEP (2012 Actual expenditures) Localgovernmentunit contributions other regions in the Philippines (Figure 20). LGU expenditures on education, culture and ARMM LGUs barely register any education arts, and manpower development show that spending, while other regions devote between ARMMIV LGUs spend little of their own re- about one and a half to 15 percent to these sources on supporting education compared to activities. Figur 20: Source: % LGU spending on education, culture & arts, 2011 Statement of Revenue and Expenditures, Department of Interior and manpower development (2011) and Local Government (Updated 2013). 16% 14% 12% 10% 8% 6% 4% 2% 0%/e F, .2 0 ~ ~ . 0 . 0 .2 o) 0 c2 0 50 One key source of education-specific resourc- In summary, beyond the aggregate allocations es through LGUs is the Special Education for basic education in ARNIM, this closer Fund (SEF).'Ihe SEF is levied as one percent look at the composition of expenditures in of the real property tax collected by the LGU terms of their economic classification and intended for specific support to finance public distribution across different programs reveals education.so The problem is, as with other el- several important concerns with respect to ements of LGU financing, the collections and education budgets. Chief among these are therefore expenditures through the SEF vary low routine MOOE spending, large and un- widely across the country and are generally even variations in program spending between higher in more prosperous localities. Across the ARNM and other regions of the Philip- the country, "LGUs account for roughly half pines in part driven by poor reporting and of capital expenditures in basic education lack of budget participation by ARNIM, a through the SEF financed through real estate growing emphasis on capital outlay that may taxes"."' The lack of such funds in AR1\4M not be matched by adequate recurrent spend- can be assumed to stem from a combination ing, and the regressive effect of LGU spend- of very low rates of LGU collections, low ing on education. property values and the ambiguity over devo- lution of education responsibilities outlined The ARMM PEM syndrome and distortions in Chapter Three. in budget composition It is important to understand that these ex- This means that LGU financing of education, penditure composition issues arise out of the to the degree it exists, is highly unequal. These ARIM Public Expenditure Management variations do not compensate for variations in Syndrome described in Chapter1Tree. As with national spending, and national government public expenditure as a whole in the ARNIM, spending on education is not designed to the fundamental issue facing basic education offset LGU-level inequalities in fiscal capac- is that the government "budget" for educating ity: there is "no evidence of such a pattern children is made up of different components of vertical coordination".82 In fact, LGU each planned, budgeted and subsequently spending tends to reinforce existing resource managed by different institutions across three inequalities, accentuating regional differences levels of budgeting - national sector agencies in human capital accumulation with serious (the DepEd National), the regional govern- implications for regional income distribution: ment (in the ARG appropriation), and other "LGU spending is strongly regressive and has local government units (Table 9). become more so, especially in education".Ia 80 Republic Act 2447 (1968). fWorod Bank (2011), p. 74. drWorid Bank/Australian Aid (2012), p 31. s3Wored Bank (2011), p 10. Table 9: Budgeting basis for education components 51 Budget Component Budget Relationship Budget Basis PS for existing staff ORG-DBM Bureau D Incremental New teachers DepEd CO-DBM Bureau B Teacher-Student Gaps MOOE ORG-DBM Bureau D Incremental/Negotiated Centrally Managed Funds DepEd CO-DBM Bureau B Various Criteria, more or less transparent School Building Textbooks GASTPE Madrasah Alternative Learning Etc... Special Purpose Fund - School Building DepEd CO-DBM-DPWH Executive Political Priorities PDAF Congress-President Legislative Priorities LGU Education Spending LGU Local Priorities& Fiscal Capacity Donor/NGO spending Donor-Partner Various Source: Authors By far the largest of the budgeting "chan- For this ARG channel, it is crucial that the nels" for basic education in the ARM\4M is the planning and budgeting relationships be- education budget within the ARG appropria- tween the constituent agencies of the ARMM tion and the ARMM allocation within the regional government and the ORG function centrally managed funds budgeted by DepEd well in order for the ORG to aggregate their National. As noted above, the former covers submissions into a single budget proposal un- the vast majority of PS and MOOE expenses der national guidelines. In past years, a lack of including the budget for routine operations information on DepEd ARMM performance and maintenance.'Ihis "ARG channel" is and final budget for the previous year have budgeted and negotiated with the DBM as meant that the ORG began budget prepara- part of the aggregated ARMM Regional tion without filfilling basic Budget Account- Government Appropriation by the Office of ability Reporting (BAR) requirements, and the Regional Governor, not by the DepEd was therefore at a significant disadvantage in ARMM. It is therefore determined by criteria its relationship to DBM. At the same time, the negotiated between ORG and DBM, and not relationship between the DepEd ARMM and directly based on education planning guide- ORG did not support the development of a lines. In practice this means it has tended to strong evidence-base for spending proposals in be determined incrementally education, a problem that also affected other departments. In this sense, the "intra-ARMM 52 budgeting process is as important as the rela- The second "DepEd National" channel, tionship with national government. budgeted or later divided as sub-allotments to the region within the DepEd National Frequent leadership and policy turnover in budget, covers the majority of capital outlay, the ARMM have worsened this situation. as well as operating expenditures for many Between 2006 and 2010 the region had five programs that bear on the effectiveness of education secretaries and experienced two basic education spending.'These programs "management takeovers" in which accounting, each have their own planning and budgeting budgeting and cash functions were taken on criteria, set by DepEd National - in some by the ORG in response to irregularities. A cases with input from its subordinate of- review in 2010 concluded there would "not fices - but without any direct input from the be one member of staff that has institutional ARMM authorities due to their "autonomous memory with regard to budget preparation", a status". In this sense, the ARA4M education striking illustration of the potential need for authorities have little or no direct input into capacity support from national institutions." the two main budgeting processes that affect their resources. Figure 21: The two main channels of ARMM education budget preparation Congress ctaey p N o-National GovernmenttDBM SBF ARMM ORG DepEd National DW PS an MOOECentrally Managed Funds (CM F) ARMM DepEd Divisions, Districs t Budget Influence and Schools two mi Weakinfluence dFunds Flowelease 84 Tribal Helm (2010), p 22. The fragmentation of budget preparation in regular MOOE allocations (Box 4), the ARG 53 this way helps explain the skewed composi- has received incremental budget increases tion of expenditure described in the previous from DBM. Furthermore, while the educa- section. The operating MOOE allocations tion budget for the rest of the Philippines is for ARMM are budgeted as part of the ARG prepared in cooperation with Bureau B of the appropriation, while the DepEd National's DBM, that of the ARMM is with Bureau D. planning and budgeting process determine Even inside the DBM, ARMM education those for the rest of the Philippines. While budgeting is fragmented and isolated from its DepEd National applies a needs-based own education budgeting specialists. planning norm that has recently increased Box 4:The NtionalDeparment o Eduction budge forlteretos the hlines isdalitotrfr Expeditue inrodced sysem fneds-bsprepedin cooperatoned it Brau Ber cof teude norms ~ ~ ~ ~ ~ ~ DM that fO thee ARMMs isr within Burea D.oaino MOrcie aEven in r c 00 0 budge isd the M AM e beu e f r fo w dt i s fragenfte nated fm D i avaiabl, i lie wih god owe n uationtc 1 hs mrvmns budgeting specialists. planin nomta aecently increase cretpann om DpdAIIMwsecuermtepoesa Boxs 4: The NOEaloationa Deatethtiesu of Edcaio has rcrrn bde allocations are increasted grdalaitotrfr normslowng fort MOO ture normsl for panning theoritin aloainoAMOsrcie aloato in00 reouce intebde. Thoorla is basedt onMMMO otnustebdee 3-... - 20 0HI 0e 0.pi inreetal bewe ORG -a- d DBIM 00miish DepEd Ntn 00r w0ch0 data i l U rtniatly wle Nuationa intoded 3,06 .0HP pe clsro th plnnn prces In 2~014 impovddalgu avalabe, n lneithgooinerntioalracice bteen DmproEmeNtnts dgeting fdMOOEafor For elemeta schools the cr pn nor DepEd AR e lderfr th pros as calmslateandMOOE allocatinntha the s um the ther raeurnt budetca ibes ngotrid anefolwig parfcorhsned-aeapoc t tof the oeverllpen ARG budgetaprpiain. Aors *i 40,00 PH per school 0ao rlA M E o ci to b bud * 2cre000 PHPdern pupil inree l b e OG a D , d Tbis gap is therefore not the result of conscious routine maintenance and operations require- neglect, and national government has even ments as described earlier in the Chapter. compensated in a sense with greater alloca- tions in other portions of its own education just as the DepEd AR dM is separated from budget. Nevertheless, because those budgets the preparation of the recurrent budget and are for specific programs, this effort does not therefore MOOE, they are also isolated from balance the shortage of routine MOgE down- the preparation of the program allocations for loads to schools through the ARG and DepEd ARM within the DepEd National appro- ARNIc. As a result, ARIvm schools become priation. As seen earlie DepEd National has significantly disadvantaged when it comes to in fact made generous allocations for ARM i ARMM BE-PEIR interview, DepEd National Planning Department Staff (November 2012). 54 to overcome historical neglect in many areas, maintenance expenditure or to staff them such as the construction of new education fa- may be compromised by the lack of integra- cilities and the provision for ALIVE education tion between capital and recurrent budgets. services. However, in other lower profile areas that are important for the quality of ser- In sum, the ARMM PEM syndrome re- vice delivery, the needs of ARMd\M can easily sults in a basic education budgeting process remain unconsidered since ARMM authorities that is very fragmented, not comprehensive, are excluded from the process. Weak report- and with limited participation by the main ing by districts and divisions within ARMM education provider in ARMM.'Ihe result- as with the Alternative Learning Systems or ing expenditure allocations are distorted, and textbooks exacerbates this problem, undermin- lack transparency and predictability for the ing even sound planning processes. ARMM government and the region's citizens. This makes the formation of strategic linkages Similarly, the planning process of DepEd Na- between regional education policy and plans tional determines allocations for the capital and the budget very difficult. In this sense, budget. As seen above, national government adjusting budget composition cannot simply has prioritized increased capital expendi- be a matter of reallocating available resources ture for ARMM to close the classroom gap. to new priorities. Instead, problems in budget However, this spending remains unconnected composition are a function of deeper struc- to the recurrent budget, meaning linking tural features of the ARAM PEM system, MOOE spending or increases in staff to the and need to be addressed in that light. Before new facilities is difficult or impossible. As looking at how to do so in Chapter Six, the new facilities become available due to gener- next chapter introduces some issue of budget ous construction budgets, the ability of the execution and service delivery. DepEd ARMM to provide operating and CHAPTER FIVE: efficiency and effectiveness of budget ex- 55 ISSUES IN BUDGET ecution, because institutions and processes for releasing budgeted funds vary between the different components of the education SERVICE DELIVERY system. This divided system can reinforce the disconnect that already exists at the planning and budgeting stage by causing slow, uneven, or inconsistent release of funds. While the process of budgeting is central to Unfortunately, the ARMM PEM syndrome public services, equally important is the ques- also makes it more difficult to accurately tion how effectively does the budget reach assess budget execution due to the gaps the frontline? Analysis of budget allocations between local, regional and national systems, and composition such as that presented in and poor accountability and reporting in the Chapter Four is only part of the picture. region. As the 2012 Basic Education Pub- Equally significant in determining outcomes lic Expenditure Review for the Philippines is technical or operational efficiency. This notes, "[t]racking actual execution against the chapter looks at the institutions that execute original authorization is a complicated task" the ARMM education budgets already dis- that "dilutes accountability for expenditure cussed, and the transformation of those funds management" even in those areas of the Phil- at school level into essential education inputs ippines outside ARMM.16 including teachers, textbooks, operating expenditures and classrooms. It also presents Budget execution procedures in the Philip- selected results from a school level survey of pines are complex and changing, and this is 163 ARMM primary, elementary and sec- doubly so for the ARMM. Allotment by the ondary schools, which provides a window DBM of funds appropriated to a given agency into the reality that these problems create for takes place through two main means: Agency students in the region. Budget Matrix (ABMs) for "standing" recur- rent or regular payments such as Personal Ser- Budget execution and the vices, and Special Allotment Release Orders ARMM PEM syndrome (SAROs) for specific programs or items that Chapter Three described the fragmentation require clearance. Once expenditure is ap- of public expenditure management in the proved, DBM releases a Notice of Cash Allo- ARMM, and Chapter Four spelled out the cation (NCA) that informs the agency that the implications of these divided institutions for money can be liquidated. As in many systems education budgets. Fragmentation of public worldwide, these budget execution steps are expenditure management also threatens the intended to control expenditure by checking reWored Bank (2012), p 56. 56 it is being used for the intended purposes and funds. The national DBM has in recent years that it conforms to agency and national cash taken the curious position of both requiring planning and availability. this reporting to be completed, while consider- ing it to be spurious and therefore not a valid Just as budgeting is divided, the channels basis for future new spending requests." and procedures for executing basic education funds are also divided between the com- The execution rates of the Personal Services ponents executed by the ARMM Regional budget to the level of the region may be Government - PS and regular or "routine" presumed to be quite high (though not 100 MOOE - and the programs managed by percent), but the execution of salaries and DepEd National. In the case of the majority other compensation, and their distribution to of the recurrent budget, managed by DepEd teachers presents a number of major issues. ARMM and the ORG, funds are down- First, the processing of the payroll by DepEd loaded to the regional office for distribution ARMM has not been institutionalized and as salaries or school and district level MOOE experiences delays. Second, the integrity of payments. For the other programs, a range of the data in the payroll has long been suspect, different policies and procedures determine and ongoing efforts are needed to cleanse it how these funds are obligated and disbursed of spurious entries and introduce processes depending on the program. to prevent future problems.'The loose control of payroll and the resulting inaccuracies has Personal services andpayroll management also played into the historical role of ARM The execution of the Personal services budget institutions in providing a patronage vehicle is relatively simple, in that the appropriation for local and national politicians, rather than is held with the ARG budget and therefore as providers of quality public services. downloaded directly through the same proce- dures governing PS for other departments. As Widespread suspicions of ghost teachers and these are ABM allotments, they are released the integrity of the ARM education payroll without additional clearance. The ORG tasks and human resource information systems have the DepEd ARMM management with the been key concerns of the caretaker administra- transfers or physical distribution of salaries. tion since it took office in early 2012, and these The actual rate of budget execution for the were motivating factors behind the request ARG budgeted education PS and MOOE is for assistance through this project. DepEd impossible to derive: interviews verified that ARMM themselves estimated that these SAOB reporting for DepEd ARMM was pre- problems were resulting in a leakage of about pared by the ORG budget office without ac- 30 million pesos per month." A diagnostic of tual expenditure reports for recent years, giving human resources management and payroll ser- a notional 100 percent execution rate for these vices at DepEd ARis was therefore under- oAuthor interviews: DBM and DepEd ARMM staff. bPresentation of DepEd ARMM Assistant Secretary Noor Mohammad Saada, June 2012. taken as a component of the BE-PEIR project, ary increments for seniority." These factors 57 and issues of teacher absence were specified as impact the payroll within the parameters of a focus for the school survey.89 the plantilla as different staff fill approved staff posts. The HRIS was both incompletely The diagnostic process identified three broad populated (with only about 80 percent of the weaknesses in the hiring, payroll and human required 22,000 entries) and was unable to resource information management system: link unique teacher identities to particular school assignments or generate a region-wide 1. Lack of integration or communica- list as it only processed data by division.92 tion between HRIS and Payroll processes; This also made it difficult to validate teacher 2. Weak accountability and safeguards presence as part of the surprise school survey. in hiring and payroll entry and processing; It is important to note that school principals 3. Lack of skilled staff to ensure accu- appeared to report teacher's assigned to their racy in data in both HRIS and Payroll.9o schools relatively accurately, but this had no real feedback to the HRIS system.93 The There are three databases which may be used HRIS, due to technical issues related to its to keep track of the notional teacher workforce design and the scale-up and transfer of serv- in ARIM: the approved DBM plantilla (the ers to national level, does not have an inte- approved staffing establishment against which grated payroll processing capability. While in budgets are made), the payroll system, and theory the approval system for data entry to the Human Resource Information System the HRIS is stronger than that of the payroll, (HRIS). At the beginning of the BE-PEIR in reality a lack of trained data entry staff project these three systems showed different meant that the data was entered by casual numbers of employees and had not been cross- labor using a single user-ID, and errors were checked. Out of the 22,493 plantilla positions detected during the diagnostic exercise. filled, only 20,996 were in the payroll and only 18,780 in the Human Resources Information In turn, the payroll must be up to date and System (HRIS). These systems were not inte- free of spurious entries to ensure that only the grated with each other, and each was compro- correct salaries are processed and distributed. mised in important ways. The regional payroll system runs on a sepa- rate platform (earlier devolved from DepEd The Human Resources Information System National) with each School Division having (HRIS) should maintain the most current separate databases - thus duplicate names, information about staffing, including reg- errors, and discrepancies across Divisions istering new incoming teachers (and their cannot be validated automatically. There is qualifications) as well as retirements and sal- no system to check payroll data with that 8Australian Aid engaged Carole Belisario, HR operations analyst and Jonathan Masalonga, IT expert: Belisario and Masalonga (2013). atBelisario and Masalonga (2013), p. 6. The HRIS runs on a web-based application using MS SQL Server hosted under DepEd national. The HRIS program was developed through the BEAM Project of Australian Aid tor regions XI, XII and ARMM. It was turned over to DepEd Central Office at the end ot the project in 2011. The nationwide scale up on the use of HRIS is under DepEd Central management. 92uBelisario and Masalonga (2013), p. 5. s While the HRIS platform was initially rolled out for ARMM at the regional level, the server was subsequently transferred to DepEd Central. Connectivity issues then made it de facto impossible to update information in a timely fashion. 58 of official plantilla position numbers, and civil society organization (CSO) represen- employee absences, transfers, termination and tatives in April and May 2013 to the vari- retirement were not regularly being input- ous divisions to distribute payroll cheques ted. Informality, discretion and a the lack of manually, with the exception of Maguin- a processing calendar were observed in the danao school-based personnel that picked up processing of new entries, transfers and other cheques at the ORG Complex (Figure 22). payroll-related requests.94 A significant number of cheques were left un- claimed, suggesting ghost teachers. However, These weaknesses led to delays in the release of this approach had a number of limitations. salaries and inaccurate salary payments, exces- Even if cheques were collected, this was no sive retirement terminal leave payments, and guarantee that teachers would actually teach fraudulent payments to "ghost" employees. The regularly. For many bonafide teachers, physi- discrepancies and delays also have an impact cally coming to central points also involved on budgeting as they affect the release of significant time and financial costs. funding for additional plantilla positions from DBM as reports are difficult to reconcile. At the same time, the project team carried In this situation, DepEd ARIM man- out a triangulation of data between the three agement sought to validate the payroll by database systems for a sample of school deploying RPSU personnel accompanied by divisions (Box 5). Subsequently, the DepEd Figure 22: Payroll Cheque Collection in ARMM 94 Belisario and Masalonga (2013), p. 9. ARMM has taken steps to cleanse the payroll with the payroll database. A new operations 59 database. Following the diagnosis of the sys- manual (see below) specifies the use of this tem flaws, the technical assistance to DepEd- report as a validation document required for ARMM helped clean up the payroll list by the processing of the disbursement voucher correcting or deleting names that are not that releases the payroll cheques. This system, eligible for payment of salaries. As a direct however is a temporary fix that should not result of these efforts by DepEd ARMM, 400 substitute for continued efforts by DepEd spurious data sets have been removed from National or DepEd ARMM to launch a the payroll, resulting in an estimated saving of functional HRIS with local server access and 800 million PHP per year.95 an integrated payroll module. A software solution was also developed A final important issue for Personal Service to manually link the cleaned HRIS to the execution has been a history of varied hiring payroll database. This tool generates a list procedures and criteria for selection, resulting of duplicate or inaccurate entries using the in allegations of subjectivity fraud and delays HRIS as reference, and cross validating them in the deployment of teachers in the schools. Bnta ases He a o ceus this sysem, duplic66 Tringuatin Pyrol nd lanill daa) e at sioal or De e AMM si tof aunchantlafrBslnai of HRIS paroll dupicaed etris iallt funcdtione,as HiS wi locaal sre cc eessand bewn18(i 800bse millions PHP pernce year.epayoIn itgrteayroLnmo dSule ahprvne nLaa e u Ao aesl tion wa ualson eveloped A,7 finals iportan issueresnale Sincpetorspiou,wh tomnulyiik bh leandlct H ries s the teeto sbeen a histor of0 vatried shiringonsprustxae payol daabse Thsto weeraeslite pof rocedurnestiatind cieifrsction, ubresul Tng eros ab u ofe duplicateio ornife inaccurate entries usin the in allegnation of subecivty frau andmpet delay HRSaveeec,aros aidtingt them plni nd the dpalym entr,teache in thae hools. daandmr Tragltodayrolland. Plarintila da) ldaabs revealed spurious orbeso hsoresoviac o pcate enes winhallftrentemaaaes dues in n andplatilla(ayllown prcessig, ad acfaiue to acue leave datbastie b atalnepeatnd absences Thnr the payoll In total,ntoin ndevl LocalizaionLa hetriangulhatson reveale 7,772 g error or pos-io sige oodc sionsofte hirelscle dihuliatessent across the theessitst iiinrscolovrf of~~~~~n 862fiaios weren seosn n -ectied fouth inesonne.itdo hpyolaepyi teacheroverL Thser trgton otentiieds dulcatey eti of shos.Hwvrtesho As a rsult,the aious kecinstf nve bothibe theo plantill and thentpayroll in ARMMhasdatabse. The reinlpyoldtscoe rfteeale ftahr asne n teDpdAM ad ovenr0tatsed uplctdGSSnm bers Analysiso ofd thelng plantill fo8Bsian1Twi eliaro ad Msaong (213, p 8-. he ocaiztio La (A 890)ais t awproit and Larnaof delcSurs reveoaledetee 18 (in aonsfullat660 er s one spurios t identification . . nube. These eror maybedu to eg* a .omiato ofporadnomlt dat entry.6failure6to update6personne data andmor Subjective and partial interpretation of theThrhabenoaciem itigadevl Localization Law has meant that school divi- uaigtmatheRioasgndocnut sions often hire locally without assessment flo pvst odvsoso col ovrf of qualifications, often choosing non-eligible thtpronliseonheayllrehs- teachers over LET passers from other areas. cayreotnatshl.Hwvr,hecol As a result, the quality of teaching staffsuvydcrbdelwhpdtoinifte in AMM hs dclind.scoee thd pbeoteersnitornc andvl uting Deam atMV thaeein aned mcnuc followaurvisitsdoMaiviionsaor2chool, o3verif 96 Belisario and Masalonga (2013), p.8-41. TeLclzto a 10 ist rn roiyi iigo ecest oa eiet,bti consideration of qualifications. 60 needed division supervisor posts in the 2014 further strengthen the integrity of the system. budget, in line with staffing already available This could be through the approval of additional in other regions of the Philippines. Availability supervisor, finance and data entry posts, and of biometric scanners should also improve the continued attention to developing the HRIS to situation, assuming they are used to inform its technical potential. The current addition of payroll processing and the management infor- new division supervisors in the 2014 budget is a mation system. welcome first step in this direction. An operations manual has been developed MOOE in consultation with regional and division MOOE expenditures, like PS, appear out- level staff with roles in personnel and payroll wardly to be effectively executed - informa- processes. The manual sets out roles, structures, tion on the monthly downloads of MOOE and defines agreed principles guiding HR and to schools or district offices are available and payroll operations. It lays out steps for all key suggest the MOOE budget is executed in a HR processes including hiring, new forms timely fashion. However, liquidation reports of and timetables, and establishes enhanced ac- this routine MOOE from schools and districts countability for concerned offices/staff posi- are absent, meaning there is no reliable way to tions. Institutionalizing the procedures in this know the actual use to which the funds have manual throughout ARMM is a vital task been put. As noted in Chapter Two, the re- going forward to sustain these reforms. sponse by DBM to the failure to demonstrate expenditure needs and history is to budget The final report of the TA to DepEd ARMM incrementally rather than based on needs, includes recommendations for sustaining the thereby perpetuating the relatively low levels of new business processes, staff capacity building routine operating and maintenance funding. and internal reorganization." The creation of an internal audit unit to concentrate on data The weaknesses of budget execution for the integrity within the DepEd ARMM is rec- recurrent budget - PS and MOOE - chan- ommended to complement the oversight of neled through the ARMM are therefore not so the Action Center for Education (ACE) over much in the process of obligations or releases, procedures, timelines and hiring processes. The but in the poor integrity of the payroll and long-term integration of the payroll and HRIS the inadequate reporting on funds use. These system is the key structural reform needed. Cru- deficiencies mean that salary is used for non- cial also is continued dialogue with the DBM productive or even unauthorized purposes, and and DepEd National both to ensure alignment MOOE needs are not well known.This situa- with national human resources policies, but tion constrains the budget possibilities for the most importantly to coordinate on the use of ARVV by reinforcing a cycle of mistrust that the savings gained from payroll cleansing to encourages DBM to continue budgeting for iaBelisario and Masalonga (2013), p. 14-15. staff and operations and maintenance incre- additional steps to the national budget execu- 61 mentally rather than on the basis of plans. tion process.00 In general, the allotment for regions requires an additional "sub-Allotment Execution processes Release Order" or "sub-ARO" in addition to of centrally-managedfunds the SARO used for the lump sum. For some The execution of the many important programs funds, notably ALS, the ARMM regional al- for ARMM basic education that are managed lotment is handed over directly to the regional by DepEd National present considerably more education office via a funding cheque, poten- complications.Ihe funds must be released to tially saving some time. the DepEd National first, and then to regions or divisions, according to a set of guidelines and This procedure is cumbersome even for the procedures prepared by the DepEd National, national agency, never mind for the ARMM. In as well as several different mechanisms for the most cases funds or the inputs purchased with sub-allotment process set out by the DBM. them are distributed to schools divisions offices, The policies surrounding release of funds have with the regional offices playing a managerial or been frequently changed and this is a key factor oversight role. For example, funds for the distri- influencing the effective execution of education bution of textbooks are sub-allotted directly to resources. The 2012 Philippines-wide educa- school divisions offices, based on the issuance of tion public expenditure review found that policy a Notice of Cash Allotment (NCA) from the instability surrounding the procedures for key corresponding region's DBM office.'Ihe region flows such as furniture, school-building, and education office oversees delivery.101 textbooks had major negative impacts on the pace of budget execution Philippines-wide: "As However, for ARMM the purported "au- desirable as they may be, policies, operational tonomy" of the regional DepEd (as an agency changes, or rule changes create a risk of imple- of the ARG) works against it. The DBM does mentation failure".99 not have an office in ARMM and so sub-al- lotments and NCAs cannot be handled within Many of the major programs of the DepEd - the region and thus do not reach the regional notably Basic Education Facilities and text- authorities. Consequently, textbooks and funds books - have been appropriated to the DepEd for other programs may be downloaded di- budget as a lump sum.1That is, at the time of rectly to school divisions, at times without the budgeting they were not broken down by re- knowledge of the ARMM region, and poten- gion, and therefore specific clearance for each tially without subsequent reporting back to the allotment to regions are needed from DBM region. In addition, in the national DepEd ac- during the budget year. The DepEd then must counts, the execution of the funds for ARMM "sub-allot" the funds to the regions or other is not broken out, rather being subsumed administrative levels as needed, introducing under the central office. Therefore, while it is 9 World Bank (2012), p 73. 100 DepEd ARMM Regional Sub-Allotments work through an Authority to Debit Account (ADA) which authorizes download to Region Offices and to Schools Divisions. For ARMM these procedures involve either a reversal of SARO or a specific funding cheque to the new recipient. 101 DepEd Order 49/2012, "Policies and Guidelines on the Utilization of Distribution Funds for Centrally Procured Textbooks (TXs), Teachers Manuals (TMs), and K to 12 Basic Instructional Materials (BIMs)" (June 15, 2012). 62 possible to calculate obligation rates for several to illustrate just how different monitoring key education programs by region, the same of the execution of some of the most crucial information is not available for ARMM. parts of the education budget is for ARMM. Bear in mind, that it is this component of the What is clear is that central office obligation budget that was increased so dramatically to rates of key funds such as the Basic Education tackle acute classroom shortfalls, meaning the Facilities Fund (BEFF), textbooks, and alterna- increased budget allocation is likely having less tive delivery modes, fall dramatically behind impact on the ground than the budgets alone those of the individual, non-ARNM , regions. would suggest.03 For example, Regional BEFF and textbook fund utilization rates for 2012 were both above 70 A similar challenge might be found in Ma- percent for all regions, while the central office drasah education, which as seen in the last allotments for the same fund were only 8 percent chapter has been heavily funded for ARMM. and 24 percent used respectively. Unfortunately, In terms of utilization, the madrasah education it is not possible to impute a value for ARMM program was 32 percent obligated on average, execution or obligation rates from this difference, but the central office rate, in which ARMM both because ARMM finds are co-mingled with is included, was 16 percent." While a direct the other aspects of the national program, and conclusion about the ARMM execution rate because some programs also implemented some cannot be drawn, it appears certain to be very of their components for ARMM via neighbor- low. ing regions. Nevertheless, it appears clear that problems experienced in efficiently disbursing Some of these complications may be improved resources to the other administrative regions are somewhat by recent changes to national public compounded for ARMM. financial management processes. In the 2013 budget the regional sub-allotment is being The sheer lack of information, and likely specified in the appropriation. For the 2014 underperformance, is brought into relief by FY there will be no SAROs to approve allot- the 2012 BEFF project status report. While ments, only NCAs. In this sense, the budgeted ARMM's target of 1,032 for new classrooms is appropriation becomes automatically available actually the highest in the country, at year's end to the agencies in question, which should elim- the program had reported no projects com- mate one major source of delay to the execu- pleted, nor any even ongoing.02 Admittedly tion of central finds intended for regional use. this is against a very low 22 percent comple- However, this raises its own questions about tion rate nationwide, but the national average the procedures that can be adopted to con- for the percentage of allocated new classrooms trol expenditures intended for ARMVI e at least begun was 81 percent. While it is likely continued absence of a DBM presence in the the rate of zero is a reporting failure, it goes region implies downloads will still be handled 102 CY 2012 Basic Education Facilities Fund, Project Status (December 31, 2012). 103 lack of progress on education facilities construction is an issue of such concern that the Regional Governor has tasked staff from his Office of Special Concerns to physically verity each individual project. 104 Statements of Allotments and Obligations (FY 2012). through the DepEd Central Office. 63 An additional point of interest are the high The survey was implemented across an initial levels of "continuing appropriations" found sample of 180 schools drawn randomly from across these important parts of the education the full list of ARNIM schools. There were 140 sector. These figures reflect unobligated CO or primary and elementary schools (out of total MOOE funds that can be carried over from 2173 in the 2011 school list) and 40 second- the previous year, to be spent in the second ary (of 298 total in the 2011 school list) in year. However, in a major change the 2014 the sample. After replacement of inaccessible national budget policy specifies that there will sample schools and some elimination due be only "one year validity of appropriations", to data irregularities, there were 163 schools meaning this process is disallowed. Should remaining in the final sample for analysis. The these levels of unused appropriations continue, survey collected information through ques- the loss of resources to those programs that are tionnaires administered to the school principal slow to execute would be substantial.105 or ranking teacher, the Barangay captain where the school is located, and direct observation of Frontline service delivery: the schools survey the teachers and students present in the class- The efficiency with which budgeted funds are room on the day of the survey. executed shapes the levels and timeliness of education inputs available, but fiscal data alone The survey used a surprise visit and direct cannot confirm if the staff and other inputs are observation approach that is standard in actually available to deliver education servic- provider absence surveys.106 While the teams esat the frontline. The ARIM government had authorization by the regional education has had specific concerns in the area of human authorities, the list of sample schools was kept resources and school conditions. To investigate confidential until survey teams were deployed these issues, the project implemented a school to their respective divisions. Only 21 schools and community-level survey with local part- reported being forewarned in any way of the ners the Coalition for Bangsamoro Civil So- visit, with only 5 of these aware of the timing. ciety (CBCS), between 4 and 8 March 2013. The trade-off is that the use of surprise visits, The survey gathered representative data on: while essential to collect absence data, may undermine the quality of other types of data - * Patterns of teacher and pupil absence; such as financial or administrative information * Perceptions of school and community - collected due to the lack of preparation time leaders on education challenges; for school officers. * Textbook delivery; * Community and Local Government A striking result from the survey was the high Unit inputs (in cash or in kind) to basic number of schools that were not open at all education. on the day that the survey teams visited, and 105 NBM 115 (December 28, 2012), p 1. 106 Rogers and Koziol (2011), p. 11. 64 which therefore could not be included in the official teaching period and 6 schools reported statistical sample. Altogether 29 out of 163 that they had already completed the year. This schools visited (18 percent) were closed on raises questions around the degree of rigor in the day of the survey visit. A portion of this observing the annual education calendar. The school closure is related to the proximity of the reasons given by schools for these closures are survey to the end of the school year: the survey as follows (Table 10): took place three weeks prior to the end of the Table 10: Reason Frequency % Reasons for Closed for year 6 21 school closure during survey Staff meeting 6 21 Division meeting 3 10 Unknown 3 10 Classes finished for day 2 7 Classrooms in use for NAT 2 7 Local conflict 2 7 Biometric update 1 3 Exam preparation 1 3 LET exams 1 3 Postponed for feeding program 1 3 Students attending another school 1 3 Source: 2013 CBCS-Australian Aid-WB Total 29 100 Total 29 _ 100 ARMM School Survey While the number of closed schools is too expenditure management systems described small to permit statistical analysis of reasons in this report may create excess demands on by province or division, it is striking that 19 teachers and increased absences or school of the 29 closures were in Lanao del Sur. As closure, further damaging education quality. will be seen below, this province also regis- tered very high absence rates for teachers and Most dramatically, one school could not be pupils in some of its divisions. When gradua- located at all, and two more were closed due tion and year-end activities (including "closed to local violence. Even if some of the impact for year") are excluded, 23 schools were shut of the survey timing is removed, the sample for other reasons. Over half of these remain- implies that on average it is likely that 15 ing schools - 13 in fact - were reportedly percent or more of ARNIM schools may be closed for some form of education adminis- closed on a given day. trative function such as teachers taking LET exams or staff meetings. Such a high propor- Teacher absence andghost teachers tion of closure for administrative reasons Teachers' absence is a serious issue in the suggests the very inefficiencies in the public ARMM education system.10 Among 1,577 in- dividual teachers captured in the survey, only 69 average absence rate of 19 percent.109 As 65 percent were present on the day of the survey, shown below, there are considerable variations giving an average absence rate of 31 percent.10s in absence rates across provinces/divisions This absence rate may consist of a combination within ARIM. Teachers'absence is high- of working teachers who are absent on the day, est, nearing half of the teachers, in particular and some "ghost teachers" who are reported as parts of Lanao del Sur (IA and JIB), and staff by schools in their reporting. Sulu (11). This is reflected in the province- level variations. In general,absence is lower in These rates compare unfavorably with inter- city divisions. Further analysis and future sur- national comparators - for example, a review vey rounds may be able to provide additional of teacher absence in six countries for the insight into the reasons for this clustering of 2004 World Development Report found an absence problems. Province Absence Rate (re) Division Absence Rate v Te Basilan 11 Basilan 15 Provincial! division teacher Lamitan City 7 absence rates Lanao del Sur 35 Lanao del Sur IA 25 Lanao del Sur IB 25 Lana o del Sur (IA 49 Lanao del Surl111 48 Marawi City 25 Maguindanao 25 Maguinsana s F 26 ey Maguindanao b l 25 Sulu 38 SuWu D 37 Source: Suu al 46 2013 CiC-Austra- lian Aid-WB ARMM Tawi-Tawi 37 Tawi-Tawi na School Survey The ability to identify ghost teachers was collected prior to the survey. This data is at unfortunately limited in this representative school level, but is likely to reflect mainly survey. In the case of elementary schools the existing teachers, though some ghost workers payroll identifies teachers not by post, but by may be included. The ability to separate out school district, making it impossible to use the ghost absences from regular teacher absence list of teachers being paid as a reference for a was also constrained by the use of a single sample-based survey - a non-reporting teacher survey round in the year, which removed the listed in the payroll may be assigned to another possibility to discern persistent from occasional post in the district. Instead, absence rates were absence without a follow-up. A second survey calculated based on monthly administrative is planned for the 2013-14 school year. reporting data on staff from school divisions "'h teachers captured in the survey comprise teachers listed in the DepEd ARMM's administrative records as assigned to a given school, if that school was open when visited. u Rogers and Koziol (2011), p7 See also World Bank (2004): the countries are India, Bangladesh, Uganda, Indonesia, Ecuador and Peru. 66 Male teachers were more likely to be absent (39 Pupilabsence percent) than females (30 percent).1o Senior- Student absence is also highly prevalent in ity is weakly correlated with higher rates of ARMM. We calculated attendance rates at absence, and Principals have the highest absence the class level (each section of each grade) rate.'Ihe statistical significance of the difference based on the reported enrolment of the is low but the finding is consistent with some school in monthly reporting collected before international studies, with the proposed expla- the survey. On the day of the surprise visit, nation that weaker accountability for more se- the average attendance rate was 71 percent.1 nior staff contributes to higher absence rates.1 Lanao del Sur has the highest absence rate for both male and female students, as with teachers' absence, though these figures are this time concentrated in the 1A and B divisions (Table 13). Table 12: Teaching Grade Absence Rate r(%) Presence bysurvey.rO t, Teachrgersednc rt ws 1pecet teachfionrbuenogihe benerae. LaagelSrhaahdhgetebecert teachers absneMaster Teachers 26 Source: 2013 1BS-Austra- Head Teachers 25 lian Aid-WB ARMM Principals 30 School Survey 110 There were 101 teachers for whom the surveyors did not clearly identify sex, 96 of whom were not present on the survey date, but for the purposes of the study sex has been imputed from given names. 111 For example, Kremer et al. (2005), found across several states in India that "[o]lder teachers, more educated teachers, and head teachers are all paid more but are also more frequently absent." 112 This rate is calculated from schools and classes that are in session on the day of the survey. 67 Table 13: Absence rates (%) of students in classes in session by province, division and gender Province Total Male Female Division Total Male Female Basilan 16 19 13 Basilan 16 19 12 Lamitan City 17 20 14 LanaodelSur 61 62 59 Lanao IA 61 58 63 Lanao IB 64 69 59 Lanao IIA 38 51 25 Lanao IlB 63 65 61 Marawi 63 69 57 Maguindanao 23 25 21 Maguindanao I 25 25 25 Maguindanao 24 24 24 II Sulu 26 28 24 Sulu I 21 23 19 Sulu II 44 44 44 Tawi-Tawi 28 28 28 Tawi-Tawi 28 28 28 Source: 2013 CBCS-Australian Aid-WB ARMM School Survey Figure 23: Prevalent pupil absence 68 Attendance rates do not vary greatly by the absence rates is that they are lowest in grades, but as a whole, absence is a greater the final year of either elementary or second- concern for secondary school than elemen- ary schooling, suggesting a "near graduation tary school (Table 14). In all grades, female incentive" for those who have not dropped students were more likely to be in classrooms out earlier. This data raises the importance of than males (74 percent for females versus 68 attention to early drop-outs in particular. percent for males). An interesting feature of Table 14: Grade Male Female Student absence Grade 1 39 33 (%) by grade and gender Grade 3 41 35 Grade 4 38 41 Grade 5 38 31 Grade 6 27 26 Elementary Average* 37 34 Year 1 37 34 Year2 50 43 Source: Year 3 35 29 2013 CBCS-Austra- Year 4 26 25 lian Aid-WB ARMM School Survey. Secondary Average* 37 33 There is a complex relationship between Teacher and student absence tend to vary municipality-level poverty incidence and together positively - that is high absence absence."' Attendance rates of both teachers rates among teachers corresponds to higher and students seem to show non-linear "U" type rates among pupils (Figure 24). It is evident relationships with poverty incidence. In other that students tend to attend their classes at words, attendance is higher when poverty in- a greater rate if teachers are likely to come, cidence is very low or very high, and absence is or equally, teachers tend to come to schools highest in the middle level of poverty. It is pos- where more students attend their classes.114 sible that more urbanized settings offer higher This finding suggests that efforts at increasing returns to schooling while remote areas with teacher presence in schools may contribute to high poverty incidence have lower opportunity a virtuous circle of attendance, and incentives costs for schooling (thus teaching), both of for attendance should be found for teachers as which strengthen the incentive to attend (and well as for students (as currently exist in the teach). However, the mechanisms behind the national conditional cash transfer program). relationship between locality and teacher and pupil absence require further study. hNSCB, Small Areas Poverty Estimates, 2010. 114 Though the causality is still unclear, the data suggests that teachers' presence rates hypothetically rise from 65 percent to 80 percent as students' attendance rate moves from 40 percent to 100 percent. This finding, as with all absence rates reported, includes only students in classes in session on the day of the visit. 69 Figure 24: Source: Teacher and student presence (%) by division 2013 CBCS-Australian Aid-WB ARMM School Survey 100 90 80 70 60 50 40 30 20 10 0 Tee ca tu 0 0 M Teacher Presence <0%) 9 Student Presence C%) 70 Map 3: Student and Teacher Absence Rates by School Division Student Absence Ru F E ALErSTUDENTS ABSeNCe Rate(%) 1 6%-22%. \ - -4.00% 57.00 - 6100%0 - 600% 0 12 2 % ¯ - 0 .00-- 26 35%2300 % ¯ * ~4.00' r