Ass ssin B sic Educ tion S rvic D liv r in th Philippin s The Philippines Public Educ tion Expenditure Tr ckin nd Qu ntit tive Service Deliver Stud Ass ssin B sic Educ tion S rvic D liv r in th Philippin s Ass ssin B sic Educ tion S rvic D liv r in th Philippin s The Philippines Public Educ tion Expenditure Tr ckin nd Qu ntit tive Service Deliver Stud Report No: AUS6799 | June 2016 Standard Disclaimer: This volume is a product of the sta of the International Bank for Reconstruction and Development/ The World Bank. The ndings, interpretations, and conclusions expressed in this paper do not necessarily re ect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the O ce of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. T ble of Contents Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Basic Education Spending Trends and Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Policy Note 2: Developing a Pro cient and Motivated Teacher Workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Policy Note 3: Building Better Learning Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Policy Note 4: Assessing School-based Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Policy Note 5: Providing Schools with Enough Resources to Deliver Quality Education. . . . . . . . . . . . . . . . . . . . . . 87 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education . . . . . . . . . . . . . 103 Policy Note 7: Understanding the Drivers of Public School Performance and E ciency . . . . . . . . . . . . . . . . . . . . 115 Conclusions and Policy Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Annex 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Fi ures Overview Figure 1: Annual GDP per capita growth rate (1960-2000) and student learning outcomes across countries . . . . . . . . . xiii Figure 2: Government spending on basic education, 2003 to 2015 (in 2014 constant prices) . . . . . . . . . . . . . . . . . . . xiv Figure 3: National achievement test scores and school annual revenue per student, 2013/14 . . . . . . . . . . . . . . . . . . . xvi Figure 4: Primary/elementary teacher absenteeism rates in selected countries, various years . . . . . . . . . . . . . . . . . . .xvii Figure 5: Percentage of teacher posts allotted for the 2013 school year where newly hired teachers were in post by the nal quarter of the 2014 calendar year . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii Figure 6: Percentage of questions answered correctly by the median teacher (binary scoring method), 2014 . . . . . . . . xviii Figure 7: Appropriations, allotments, obligations, and utilization rates for HRTD funds (PHP billions in 2014 constant prices), 2005-2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Figure 8: Indicators of classroom quality from direct classroom observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi Figure 9: Completion and satisfaction rates for 2013 school infrastructure projects, 2014 . . . . . . . . . . . . . . . . . . . . . . xxi Figure 10: Annual school revenue by type and source, 2013/14 school year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii Figure 11: Percentage of parents of elementary and high school students who know about the SGC and the SIP, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii Figure 12: Percentage of school principals reporting an engagement with Check My School (CMS) in 2013 or 2014 . . . . xxiv Figure 13: Share of MOOE allocation downloaded and received by schools, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . .xxv Figure 14: Percentage of schools that received input from stakeholders about MOOE funds, 2014 SY. . . . . . . . . . . . . . xxvii Figure 15: Percentage of schools that report receiving cash or in-kind contributions from local governments and average levels of support, 2013-14. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxviii Figure 16: Percentage of schools that failed to meet DepEd service standards, 2014 . . . . . . . . . . . . . . . . . . . . . . . . .xxx ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES iii Table of Contents Introduction Figure 1: Breakdown of the Department of Education (DepEd) budget, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Figure 2: Flow of Public Funds to Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Basic Education Spending Trends and Outcomes Figure 1: Government Spending on Basic education, 2003 to 2015 (in 2014 constant prices) . . . . . . . . . . . . . . . . . . . 12 Figure 2: Per Capita Government Spending on Basic Education by Region and Province and Poverty Incidence, 2012 . . . 13 Figure 3: Public Spending on Education as a Share of GNP and Total Government Spending, Selected Countries, 2012 . . 14 Figure 4: Government Spending on Education as a Share of GNP, Selected Countries, 2012 . . . . . . . . . . . . . . . . . . . . 15 Figure 5: School Revenue Per Student by Type of Revenue, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Figure 6: Public Student-teacher and Student-classroom Ratios, 2005-2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Figure 7: Student-teacher and Student-classroom Ratios and Per Capita Basic Education Spending, 2012 . . . . . . . . . . . 17 Figure 8: Gross and Net Enrollment Rates and Completion rates, 2005-2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Figure 9: National Achievement Test Scores by Region (%), 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Figure 10: Budget Execution Rates for Basic Education Spending, 2006-2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Figure 1: Student-teacher ratios (STRs) and teacher numbers in basic education, 2006-2014 . . . . . . . . . . . . . . . . . . . 26 Figure 2: Percentage of schools by student-teacher ratio, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Figure 3: Percentage of high school principals reporting subject specialist teacher shortages and magnitude of need, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Figure 4: Primary/elementary teacher absenteeism rates in selected countries, various years . . . . . . . . . . . . . . . . . . . 29 Figure 5: Teacher absenteeism rates by school location, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Figure 6: Percentage of absent teachers by reason of absence, 2014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Figure 7: Proportionate increase in the stock of school teachers in 2013 and 2014 compared to 2012 school-level student-teacher ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Figure 8: Percentage of schools by student-teacher ratio, 2012 and 2014 school years . . . . . . . . . . . . . . . . . . . . . . . 33 Figure 9: Percentage of teacher posts allotted for the 2013 school year where newly hired teachers were in post by the nal quarter of the 2014 calendar year . . . . . . . . . . . . . . . . . . . . . 35 Policy Note 2: Developing a Pro cient and Motivated Teacher Workforce Figure 1: Percentage of questions answered correctly by the median teacher (binary scoring method), 2014 . . . . . . . . . 40 Figure 2: Percentage of Grade 10 science teachers by their responses to easy and hard test questions, 2014 . . . . . . . . . 42 Figure 3: Average self-assessment scores on national competency-based standards, 2014. . . . . . . . . . . . . . . . . . . . . 43 Figure 4: Percentage of teachers receiving in-service training and duration of all training received, 2013 . . . . . . . . . . . . 44 Figure 5: Percentage of Grade 10 high school teachers by the type of additional support that they most need to improve their classroom teaching, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Figure 6: Percentage of school sta completing professional development plans and the time since one was completed . . 46 Figure 7: Appropriations, allotments, obligations, and utilization rates for HRTD funds (PHP billions in 2014 constant prices), 2005-2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure 8: Amounts of HRTD funds received by DepEd regional o ces and transferred to DepEd division o ces, 2013 . . . 48 Figure 9: Percentage of division o ces receiving 2013 HRTD funds by the quarter during which it was received . . . . . . 49 Figure 10: Elementary school teacher salaries as a percentage of per capita GDP, 2010 . . . . . . . . . . . . . . . . . . . . . . . . 50 Figure 11: Percentage of teachers whose salary payments were late, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Policy Note 3: Building Better Learning Environments Figure 1: Availability of key facilities in elementary and high schools, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Figure 2: Intensity of use and quality of classrooms, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Figure 3: Indicators of classroom quality from direct classroom observation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Figure 4: Trends in government school infrastructure spending, 2005-2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Figure 5: Infrastructure needs and project allocations, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 iv WWW.WORLDBANK.ORG/PH Figure 6: School infrastructure fund allotments, obligations, and utilization rates, 2005-2014 . . . . . . . . . . . . . . . . . . . 63 Figure 7: Completion and satisfaction rates for 2013 school infrastructure projects, 2014 . . . . . . . . . . . . . . . . . . . . . . 63 Figure 8: Receipt of project list and rst sub-allotment at DPWH district engineering o ces . . . . . . . . . . . . . . . . . . . 64 Figure 9: Common problems faced by the DPWH in implementing school infrastructure projects, 2014 . . . . . . . . . . . . 65 Policy Note 4: Assessing School-based Management Figure 1: Percentage of schools by latest self-assessed SBM implementation level, 2014 . . . . . . . . . . . . . . . . . . . . . . 72 Figure 2: Percentage of schools reporting weaknesses in key SBM dimensions, 2014 . . . . . . . . . . . . . . . . . . . . . . . . 73 Figure 3: Annual school revenue by type and source, 2013/14 school year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Figure 4: Percentage of SGCs by type of membership, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Figure 5: Percentage of parents of elementary and high school students who know about the SGC and the SIP, 2014 . . . 76 Figure 6: Frequency of PTA meetings and percentage of parents reporting that they attended regularly in 2013/14 school year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Figure 7: Percentage of PTAs providing support by type, 2014. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Figure 8: Percentage of schools with a transparency board and types of information posted on the boards, 2014 . . . . . . 78 Figure 9: Percentage of school principals reporting any involvement with Check My School (CMS) in 2013 or 2014 . . . . . 79 Figure 10: Average number of visits to schools in the rst half of 2014/15 school year and issues covered . . . . . . . . . . . . 80 Policy Note 5: Providing Schools with Enough Resources to Deliver Quality Education Figure 1: Total and per student appropriations for school MOOE, 2005-2015 (in 2014 constant prices) . . . . . . . . . . . . . 88 Figure 2: Percentage of stakeholders who are aware of MOOE funding and its components . . . . . . . . . . . . . . . . . . . 89 Figure 3: Share of MOOE allocation downloaded and received by schools, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Figure 4: Reasons why DepEd division o ces retain MOOE funds (% of divisions), 2013 . . . . . . . . . . . . . . . . . . . . . . 91 Figure 5: Timing of rst MOOE advance and share of overall allocation received, 2013 . . . . . . . . . . . . . . . . . . . . . . . 92 Figure 6: Challenges experienced by schools in spending MOOE funds, 2013/14 school year . . . . . . . . . . . . . . . . . . . 93 Figure 7: Average hours spent on MOOE administration per week by schools’ distance to division o ce, 2014 school year . . 94 Figure 8: Elementary school per student funding by source, 2013 and 2013/14 school years . . . . . . . . . . . . . . . . . . . 95 Figure 9: Percentage of total MOOE expenses by item of expenditure, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Figure 10: Percentage of schools that had transparency board displaying MOOE information in 2014 . . . . . . . . . . . . . . 96 Figure 11: Percentage of schools that received input from stakeholders about MOOE funds, 2014 SY. . . . . . . . . . . . . . . 97 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Figure 1: Total and per student public basic education spending, 2003-2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Figure 2: Total and per student local government education spending by region, 2013 . . . . . . . . . . . . . . . . . . . . . 105 Figure 3: Percentage of schools that report receiving cash or in-kind contributions from local governments and average levels of support, 2013/14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Figure 4: Composition of school funding for all schools and only schools that receive some LGU support, 2013/14 . . . . 107 Figure 5: Composition of LGU cash and in-kind contributions to schools (PHP), 2013/14 . . . . . . . . . . . . . . . . . . . . . 108 Figure 6: Total and per student LGU spending on basic education by region and poverty incidence, 2012-13 . . . . . . . 109 Figure 7: Average per student LGU funding (PHP), 2013/14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Figure 8: Percentage of schools where LGUs reported providing some funding but schools reported receiving none . . 111 Policy Note 7: Understanding the Drivers of Public School Performance and E ciency Figure 1: A Framework for Understanding School Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Figure 2: Average student enrollment and location for schools ranked in the bottom and top 20 percent of NAT scores, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Figure 3: Average proportion of parents who are members of the PTA and percentage of schools that do not receive all of their MOOE funds from the DepEd division o ces . . . . . . . . . . . . . . . . . . . . . . . 121 Figure 4: National achievement test scores and school annual revenue per student, 2013/14 . . . . . . . . . . . . . . . . . . 122 Figure 5: Average estimated e ciency scores and standard deviations for school attendance and NAT results in elementary and high schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES v Table of Contents T bles Overview Table 1: Summary of Major Findings and Policy Suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxxv Introduction Table 1: Examples of Education Tracking and Service Delivery Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Table 2: Survey Questionnaires, Contents, and Key Respondents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Table 3: Planned and Final Sample Sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Table 1: Benchmarks and Actual Timing of Key Steps in the Teacher Hiring Process, 2013 . . . . . . . . . . . . . . . . . . . . . 34 Table 2: Strengthening Systems to Hire and Deploy Teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Policy Note 2: Developing a Pro cient and Motivated Teacher Workforce Table 1: Strengthening Teacher Support Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Policy Note 3: Building Better Learning Environments Table 1: Percentage of Schools that Failed to Meet DepEd Service Standards, 2014 . . . . . . . . . . . . . . . . . . . . . . . . . 59 Table 2: Characteristics of School Infrastructure Projects, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Table 3: Strengthening Government Systems for School Infrastructure Improvements . . . . . . . . . . . . . . . . . . . . . . 69 Policy Note 4: Assessing School-based Management Table 1: Strengthening School-based Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Policy Note 5: Providing Schools with Enough Resources to Deliver Quality Education Table 1: Components of the MOOE Allocation Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Table 2: Allowed and Prohibited Spending Categories for MOOE Funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Table 3: Strengthening Systems to Provide Operational Funding to Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Table 1: Improving Local Government Support for Basic Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Policy Note 7: Understanding the Drivers of Public School Performance and E ciency Table 1: School National Achievement Test Results and Student Attendance Rates, 2014 . . . . . . . . . . . . . . . . . . . . 116 Table 2: Characteristics Used to Measure the Key Aspects of the Performance Framework . . . . . . . . . . . . . . . . . . . 118 Table 3: School Performance and Leadership Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 Annex 1 Table 1: Details of the Field Visits Carried Out during the Development of the Survey Instruments . . . . . . . . . . . . . . 132 Table 2: Description of Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Table 3: Planned and Final Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Table 4: Outline of the PETS-QSDS Sampling Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Table 5: Schedule and Coverage of Field Visits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Table 6: Other Data Sources Used in the Policy Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 vi WWW.WORLDBANK.ORG/PH Boxes Overview Box 1: The Philippines Public Education Expenditure Tracking and Quantitative Service Delivery Study . . . . . . . . . . . xv Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Box 1: Procedures for Measuring Teacher Absenteeism through Observation . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Box 2: The Current Process for Allocating and Hiring Teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Policy Note 2: Developing a Pro cient and Motivated Teacher Workforce Box 1: PETS-QSDS Approach to Assessing Teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Box 2: Indonesia’s Integrated Framework for Teacher Accountability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Policy Note 3: Building Better Learning Environments Box 1: Responsibilities for Joint DPWH-DepEd Projects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Box 2: Mexico’s Better Schools Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Policy Note 4: Assessing School-based Management Box 1: Measuring the Implementation of School-based Management in the Philippines . . . . . . . . . . . . . . . . . . . 73 Box 2: An Evaluation of School-based Management Training Activities in Indonesia . . . . . . . . . . . . . . . . . . . . . . . 81 Box 3: An Evaluation of School-based Management Training Activities in Uganda . . . . . . . . . . . . . . . . . . . . . . . . 83 Policy Note 5: Providing Schools with Enough Resources to Deliver Quality Education Box 1: Approach to Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Box 2: The Pupil Premium in England. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Box 1: Legal Framework for the Role Played by LGUs and their Local School Boards in Basic Education . . . . . . . . . . 106 Box 2: The Seal of Good Local Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Annex 1 Box 1: Finalizing Datasets for Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES vii Acknowled ements The team is grateful to o cials and sta in the Department Lynnette Perez (former World Bank Senior Education of Education for their overall support of this study. Specialist) began discussions on this study with the Special thanks go to Secretary Br. Armin A. Luistro and Government of the Philippines and was an invaluable source Undersecretaries Reynaldo Laguda and Dina Ocampo for of help and guidance at all stages. We are very appreciative their advice and guidance. We are also grateful to Roger of the support provided by Robyn Biti, Lea Neri, and Tess Masapol, Anna Bautista, Zaida Briones, Kat Lejara, and Mary Felipe at the Department of Foreign A airs and Trade of the Jane Feliciano from Planning Services and to Selwyn Briones Government of Australia. and Gilbert Murong from Finance Services for their support The World Bank team was led by Samer Al-Samarrai (Senior and help in data collection and survey implementation. Economist, GED02) and consisted of Vandana Sipahimalani More generally we would like to acknowledge the welcome Rao (ST Consultant), Jason Alinsunurin (ST Consultant), given to our team in many regional and divisional o ces Himdat Bayusuf (Education Specialist, GED02), Corinne and schools across the Philippines. We are indebted to all Bernaldez (Team Assistant, EACPF), Vicky Catibog (ST DepEd sta for their patience in answering the many, many Consultant), Marie Jo Cortijo (ST Consultant), Jesus Galvan (ST questions that we had. Consultant), Judith Kom (ST Consultant), Alma Mariano (ST Our thanks also go to Undersecretary Laura Pascua and her Consultant), Unika Shrestha (ST Consultant), and Marianne team at the Department of Budget and Management for Joy Vital (ST Consultant). Juan Munoz provided assistance in developing the overall sampling methodology and their feedback throughout the study. computing the sampling weights used in the study. The The Public Education Expenditure Tracking and Quantitative report and policy notes were edited by Fiona Mackintosh Service Delivery Survey was implemented by Taylor, Nelson, and laid out by Bill Pragluski at Critical Stages. and Sofres (TNS). The team are grateful for the enormous We are grateful to the many DepEd sta and teachers who e ort of the team at TNS (J. Flores, C. de Jesus, Y. Tolentino, provided useful comments in the workshops that were M. Raquedan, C. Gemaol, G. Bautista, J. Adeja, A. Agregado, J. held to share the study’s preliminary ndings. Comments dela Cruz, J. Virginio, D.A. Diaz, R. Olivares, J. Amin, G. Orpilla, were also provided on various parts of the report by and S Kakade) in planning, implementing, and coding this Honesto Nuqui (University of the Philippines) and Rosario complex survey. Thanks also go to the army of enumerators Manasan (Philippines Institute of Development Studies). The who interviewed thousands of respondents across the peer reviewers for the report were Fadila Caillaud (Senior Philippines. Their dedication and commitment to conducting Economist, GED01), Deon Filmer (Lead Economist, DECHD) the survey was central to the depth and breadth of the and Ken Vine (SiMERR, University of New England, Adelaide). results reported. Finally, the team would like to acknowledge the unswerving Thanks also go to the team at the Research Center for support and guidance that we received from the late Teacher Quality (RCTQ) at the Philippine Normal University Undersecretary Francisco Varela. His insights and enthusiasm (Marilyn Balagtas, Jenny Jocson, John Pegg, and Ken Vine) for the study were invaluable. for developing the teacher assessments used as part of the study and for supporting the implementation of the survey. The report was conducted by the Department of Education Our collaboration was an important part of the study, and and the World Bank with the support of the Australian RCTQ’s guidance and assistance on the analysis of teacher Government through the Australia-World Bank Philippines quality was vital to many of the main ndings of the study. Development Trust Fund. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES ix List of Abbrevi tions ANSA A liated Network for Social Accountability MOOE Maintenance and Other Operating Expenses APIS Annual Poverty Indicators Survey NAT National Achievement Tests AusAID Australian Aid NCR National Capital Region ARMM Autonomous Region in Muslim Mindanao NCBTS National Competency-Based Teacher Standards BOS Bantuan Operasional Sekolah NEDA National Economic and Development Authority BEFF Basic Education Facilities Fund NGO Non-Governmental Organization BSP Better Schools Program NOSCA Notice of organization, sta ng and BLGF Bureau of Local Government Finance compensation action CO Central O ce OFSTED O ce for Standards in Education, Children’s Services and Skills CMS Check My School PTA Parent-Teacher Association COA Commission on Audit PCF Performance Challenge Fund CAR Cordillera Administrative Region PBB Performance-based Bonus DEA Data Envelopment Analysis PHP Philippine Pesos DBM Department of Budget and Management PNU Philippines Normal University DBMR Department of Budget and Management Regional O ce PFSED Physical Facilities and Schools Engineering Division DepEd Department of Education PMT Proxy Means Testing DPWH Department of Public Works and Highways PETS-QSDS Public Education Expenditure Tracking and Quantitative Service Delivery Study DSWD Department of Social Welfare and Development PSDS Public Schools District Supervisor DILG Department of the Interior and Local Government PPP Public-private Partnership DO Division O ce RO Regional O ce DPFC Division Physical Facilities Coordinator RQA Registry of Quali ed Applicants EBEIS Enhanced Basic Education Information System RCTQ Research Center for Teacher Quality FC Field Coordinator SBP School Building Program FTL Field Team Leader SGC School Governing Council FY Financial Year SIP School Improvement Plan GAA General Appropriations Act SY School Year GF General Fund SBM School-based Management GPTA General Parent-Teacher Association SDS Schools District Superintendent GASTPE Government Assistance for Students and Teachers in Private Education SDE Schools Division Engineer GDP Gross Domestic Product SARO Special Allotment Release Order GNP Gross National Product SPED Special Education HUCs Highly Urbanized Cities SEF Special Education Fund HRIS Human Resource Information System SAOB Statement of Allotments, Obligations and Balances HRTD Human Resource Training and Development STR Student-teacher Ratio IU Implementing Unit Sub-ARO Sub-Allotment Release Order IPPD Individual Plan for Professional Development TPMS Teacher Professional Management System INSET In-Service Training TSNA Teachers’ Strengths and Needs Assessment KVP Kindergarten Volunteer Program TALIS Teaching and Learning International Survey LGC Local Government Code TIMSS Trends in International Mathematics and Science Study LGU Local Government Unit TNS Tylor, Nelson, and Sofres LSB Local School Board WATSAN Water and Sanitation ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xi Assessin B sic Educ tion Service Deliver in the Philippines Overview Introduction Building a strong and inclusive basic education system has E orts to address persistent and high levels of poverty and been a central strategy for improving economic growth in income inequality can also be enhanced by investing in many East Asian countries. Research from around the world good quality basic education. Di erences in educational has shown that increasing access to good quality basic attainment among socioeconomic groups are a key education raises productivity and can improve long-term driver of income inequality. The private and social returns per capita economic growth rates (Figure 1). Moreover, fast to education are high in the Philippines and, as a recent growing countries in East Asia such as China, Singapore, and World Bank study showed, can provide individuals with an South Korea have invested heavily in their education systems e ective route out of poverty. Making it possible for schools as part of their successful economic development strategies. to o er good quality education opportunities for all can be key to ensuring that the bene ts of future economic growth are shared more widely. Figure 1: Good Quality Education Can Improve Long- term Economic Prospects Annual GDP per capita growth rate (1960-2000) and student Over the last ve years, the Government of the Philippines learning outcomes across countries has embarked on an ambitious education sector reform program aimed at reducing poverty and increasing 7 Singapore national competitiveness. The 2013 Basic Education Act GDP per capita growth rate (%) 6 Taiwan was passed in response to growing concerns about how 5 Hong Kong the short length of basic education in the Philippines was Rep. of Korea Thailand constraining national competitiveness. The Act extended 4 China Indonesia Malaysia the basic education cycle from 10 to 13 years, an increase 3 which, in other countries, has been associated with large Japan economic bene ts. 2 1 Recent reforms have been backed up by substantial Philippines 0 amounts of new investment (Figure 2). Between 2010 and 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 2015, public spending on basic education increased by 60 Score on international assessment of learning outcomes percent in real terms, while per student funding levels also Source: Hanushek, E. and L. Woessmann (2016). “Knowledge Capital, Growth, increased considerably. and the East Asian Miracle. Hanushek, E. A. and L. Woessmann (2016).” Science 351(6271): 344-345. The reform program halted a long-term decline in public Note: The growth rate is conditional on the initial level of per capita GDP and basic education services. Massive school infrastructure initial years of school attainment. improvements and nationwide teacher recruitment have ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xiii Overview Figure 2: Public Education Spending Has Risen Rapidly in Recent Years Government spending on basic education, 2003 to 2015 (in 2014 constant prices) 400 18.6 20% 350 16.9 16.4 17.0 16.6 17.9 16.0 16.4 16.1 15.2 15.1 15.1 15.3 300 15% PHP billion (2014 prices) 250 200 10% 150 100 5% 2.4 2.1 1.9 1.9 2.1 2.0 2.2 2.2 2.3 2.1 2.3 2.2 2.6 50 0 0% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 National government National govt. as % of GDP (RHS) Local government National govt. as % of total govt. spend Sources: National government spending—DBM National Expenditure Program. Local government education spending—Bureau of Local Government Finance. GDP and GDP de ator—Philippines Statistics Authority. Notes: Basic education spending data are obligations except for 2015 where adjusted amounts are reported. Basic education refers to the DepEd budget and obligations for the Philippines Science High School. Local government spending gures are obligations but from 2014 only include spending from the Special Education Fund. The total government expenditure gures used to calculate the share of education spending exclude interest payments. improved school conditions. Key indicators of the adequacy percent of households. The latest completion rates also of the number of teachers (the student-teacher ratio) and show that only three-quarters of the children that start classrooms (the student-classroom ratio) in the school successfully complete elementary or high school. system have improved. For example, between 2010 and 2013, student-teacher ratios in public high schools fell from 38:1 to Improvements in the overall quality of basic education 29:1, while student-classroom ratios fell from 64:1 to 47:1. also need to be made if the potential economic bene ts are to be realized. The Philippines last participated in Increases in the availability of key inputs have also an international learning assessment in 2003 when the signi cantly increased access to basic education. results showed that only one-third of its elementary Kindergarten enrollment almost doubled in absolute terms and secondary school students reached the lowest between 2008 and 2014, and the proportion of children of international benchmark in mathematics. More recent school age attending basic education also increased. Poorer results of national examinations suggest that there have families bene tted most: the share of school-aged children been only limited improvements since then. from the poorest 20 percent of households who were attending school increased from 86 percent to 94 percent Key to tackling these challenges will be further increases in between 2002 and 2013. education spending. Despite impressive recent increases, the Philippines still spends less on education than many However, signi cant challenges remain before the neighboring and middle-income countries, and recent government’s education goals can be achieved. Access analysis has con rmed the need for more spending to to high school continues to be relatively unequal. In 2013, meet national education norms and standards. only 53 percent of high school aged children in the poorest 20 percent of households were attending high school However, the impact of further increases in education compared to 81 percent of children in the wealthiest 20 spending will depend crucially on how e ectively public xiv WWW.WORLDBANK.ORG/PH Box 1: The Philippines Public Education Expenditure Tracking and Quantitative Service Delivery Study The aim of the Philippines Public Education Expenditure and Quantitative Service Delivery Study has been to answer four main questions on the use of the public education budget: 1. Resource ow, management, and control. What factors prevent resources from reaching their intended destination in a timely and transparent manner? 2. Existence, use, and nancing of inputs at the school level. Do schools have access to essential inputs and how e ective are the systems that govern their use? 3. Equity. How do the resources available to schools and the systems that manage these resources di er among regions and socioeconomic groups? 4. School performance and resources. How and why does the performance of schools di er and what drives those di erences? The study has tracked over 80 percent of the national government education budget (including teacher salaries and training, school maintenance and operating expenses, construction, and learning materials) as well as local government spending on basic education. In order to assess how funds ow and how they are used at the school level, the study team conducted a nationally representative survey of government institutions and public schools in the last quarter of 2014. The Autonomous Region in Muslim Mindanao was excluded from the study because government funds for this region are managed separately and ow to schools through a di erent mechanism. In addition, integrated schools (which o er both elementary and high school education) and schools that did not have nal grade elementary and high school students were excluded from the sample, primarily because the study aimed to measure outcomes at the end of elementary school and at the end of high school. The sample for the survey included all regional o ces of the Department of Education (DepEd) and the Department of Budget and Management (DBM), 51 division and 113 district o ces of DepEd, 54 district engineering o ces of the Department of Public Works and Highways, 74 provincial and city/municipality local governments, 249 public elementary schools, and 200 public high schools. At the school level, interviewers administered a questionnaire to each parent-teacher association, assessed the competencies of approximately 1,500 teachers, and interviewed 2,200 student households. The data collected were used to explore the systems that govern the use of public funds and to assess how the availability of resources di ered among schools. The study team combined information on the ow of funds to schools with information on school characteristics and quality to evaluate how nancing and governance a ected school performance. nancial and resource management systems allocate and Figure 3 shows that on the whole, higher levels of annual use these resources. Low budget execution rates driven by school funding were associated with better school delays in fund releases, di culties in managing funds and performance, but this relationship is not very strong. Some inadequate coordination have led to widespread concerns schools appeared to use their resources more e ectively about the e ectiveness of public education spending. These than others to improve student learning outcomes. In Figure issues and more generally the e ectiveness of government 3, each panel is divided into quadrants that show where systems to manage and govern the use of education schools fell in relation to the average achievement score resources are key drivers of di erences in the ability of and average annual school spending per student. Schools in schools to translate resources into improved education quadrant A were the most e cient as they spent less than outcomes. the average school every year but had better than average ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xv Overview Figure 3: There Are Large Differences in Performance Even Among Schools with Similar Levels of Funding National achievement test scores and school annual revenue per student, 2013/14 Elementary schools High schools 100% 100% A A 90% 90% 80% 80% Grade 10 NAT scores Grade 6 NAT scores 70% 70% 60% 60% 50% 50% 40% 40% B B 30% 30% 0 5,000 10,000 15,000 20,000 25,000 0 5,000 10,000 15,000 20,000 25,000 School revenue per student (PHP) School revenue per student (PHP) Source: PETS-QSDS national survey – school-level data. Note: The horizontal line in the gure represents the average school NAT score while the vertical line marks the average school revenue per student. outcomes. The schools in quadrant B were the least e cient, elementary and high schools. The Bank developed the with levels of spending that were higher than average but design of the study in consultation with DepEd, the DBM, with below average performance. Levels of e ciency varied the National Economic and Development Authority (NEDA), greatly between schools that had similar levels of either the A liated Network for Social Accountability (ANSA), and performance or spending. experts on the Philippines education system (Box 1). Pressure on public nancial management systems will only grow with the introduction of senior high school and the The Te cher Workforce need for 1.5 million additional school places. It is therefore Ensuring that schools have enough teachers to provide vital that systems used to allocate resources, build new education in classes that are small enough to foster a good infrastructure, and recruit new teachers work e ectively. learning environment is an important rst step for improving education outcomes. While the optimal size of classes and The Philippines Public Education Expenditure Tracking and the impact of reducing class sizes continue to be debated, it Quantitative Service Delivery Study (PETS-QSDS) aimed to is clear that very large class sizes are detrimental to learning. gather detailed evidence of the quality of basic education Moreover, evidence from recent impact evaluations in services and the systems used to allocate and manage developing countries shows that, on the whole, reducing resources, to provide a snapshot of the key education inputs class sizes in schools improves learning.1 at the school level, and to evaluate whether these resources are being distributed equitably across schools. The study Te cher Distribution nd emerged from a request from the Department of Education Presence in Schools (DepEd) and the Department of Budget and Management (DBM) asking the World Bank to assess whether the large In the Philippines, the Department of Education (DepEd) budgeted increases in public education spending were has class size standards for each grade in elementary and being used e ectively to improve conditions in public high schools. Even though good progress has been made xvi WWW.WORLDBANK.ORG/PH towards meeting these standards, the distribution of low in the Philippines (Figure 4), at around 7.6 percent among teachers among schools remains unequal. Using DepEd elementary school teachers and 5.9 percent among high norms for school-level student-teacher ratios (STRs), it can school teachers. Absenteeism was more of a problem in be seen that 29 percent of elementary schools and 37 highly urbanized cities where almost 1 in 10 teachers was percent of high schools have too few teachers, whereas absent in 2014, a rate some 54 percent higher than the many other schools have a surplus. For example, about national average. This was probably a result of the larger size 52 percent of all elementary schools have STRs that imply of schools in highly urbanized cities, transport and tra c that they have too many teachers. These lower STRs are problems, and greater demands on teachers’ time. typically found in remote areas where the catchment population is too small to meet STR guidelines. Indeed, a S stems to Hire, Deplo , higher proportion of schools in rural municipalities tend nd Tr nsfer Te chers to have STRs that are below existing guidelines than city School-level data from the PETS-QSDS survey show that, while schools. However, a signi cant number of schools in more new teachers were generally allocated to schools with greater densely populated urban areas also have low STRs. It is need, there is room for improving new teacher allocations. The possible that transferring teachers from these schools to recent recruitment drive has reduced the number of schools schools with teacher shortages would be a more e cient with a teacher shortage but has also increased the number option than hiring additional teachers. with teacher surpluses. Also, while DepEd division o ces are now lling vacancies more rapidly (an average of 5 months Knowing how many teachers are allocated to each school down from 18 months in 2007), about one-third of the new is important for planning, but whether they turn up to teacher positions for 2013-14 remained un lled at the end of teach every day on time is even more important for student 2014 (Figure 5). It is likely that delays by teachers in submitting outcomes. Using unannounced visits and conventional their documentation and delays by division o ces in verifying methods to measure teacher attendance, the PETS-QSDS this information partly explain these delays in getting new study found that teacher absenteeism rates were generally teachers into the classroom. Figure 4: Teacher Absenteeism Rates in the Philippines Are Low Primary/elementary teacher absenteeism rates in selected countries, various years 30% 25% 20% 15% 10% 5% 0% Uganda India Pakistan Zambia Bangladesh PNG Ecuador Peru Mongolia Indonesia Philippines Cambodia (2013) (2014) (2004) Sources: Philippines - PETS-QSDS national survey – school-level data. Data on Indonesia come from Analytical and Capacity Development Partnership (2014). “Teacher Absenteeism in Indonesia: Policy brief.” Jakarta. Data for all other countries come from Rogers, F. H. and E. Vegas. (2009). “No More Cutting Class? Reducing Teacher Absence and Providing Incentives for Performance.” Policy Research Working Paper. No. WPS 4847. World Bank, Washington D.C. Notes: The Philippines study measured teacher absenteeism for teachers who were expected to be present at the beginning of the school day, whereas other studies checked teacher absenteeism at di erent times of the school day. All absenteeism rates are for 2002-2004 unless otherwise stated. . ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xvii Overview Figure 5: A Significant Number of Newly Hired possible to do so, only one-third of the divisions reported Teachers Were Not in Post at the Beginning of the receiving the amount reported by the DBM regional o ce. School Year Percentage of teacher posts allotted for the 2013 school year Transferring teachers between schools can redress where newly hired teachers were in post by the nal quarter of imbalances and increase the e ciency of public spending, the 2014 calendar year but teacher transfers are rare. A nationally representative 80% sample of Grade 6 and 10 teachers showed that fewer than 2 percent of teachers were transferred in 2013, most of them probably for personal family reasons rather than to redress 60% imbalances in teacher distribution. 40% The Qu lit of Element r nd Hi h School Te chers 20% According to the results of the PETS-QSDS assessments of teachers, knowledge of subject matter among elementary and high school teachers is low in most subjects. With the 0% exception of English at the elementary level, the average Kindergarten Elementary High School elementary or high school teacher could answer fewer than Source: PETS-QSDS national survey – DepEd division o ce data. half of the questions on the subject content tests correctly (Figure 6). Since these tests are closely aligned with the curriculum, the results suggest that teachers face signi cant When the study tracked fund ows for newly hired teachers, challenges in teaching a considerable portion of the current it found data inconsistencies and poor record keeping. curriculum. Information on the budget allocation for newly hired teachers was not available from either the DepEd division However, teachers rated their own competencies in o ces or the DBM regional o ces. This meant that the study the curriculum domain highly, including subject-matter team was largely unable to track the ow of funds between knowledge. Simple correlation coe cients showed a weak these two sets of o ces. In the rare cases where it was relationship between teachers’ self-assessment ratings and Figure 6: Teachers’ Performance on Content Knowledge Assessments Was Poor Percentage of questions answered correctly by the median teacher (binary scoring method), 2014 Grade 6 elementary school teachers Grade 10 high school teachers 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% English Math Filipino Science English Math Filipino Science Source: PETS-QSDS teacher content knowledge assessments xviii WWW.WORLDBANK.ORG/PH their subject-matter test scores. These ndings suggest that development plan, which school principals then aggregate teacher self-assessments may not be a good basis on which and submit to division o ces to organize relevant training. to plan professional development. However, many teachers and schools, including a quarter of high school teachers, had never prepared a plan, and in In-service Te cher Tr inin many other cases the plans were out of date. Providing teachers with good quality professional Support from school principals for teachers’ professional development opportunities is an e ective way of increasing development also varied. Only two-thirds of Grade 10 their competencies and improving student learning teachers in high schools reported that their school principal outcomes, but the professional development opportunities had spent a full period in their classroom over the course currently o ered to teachers frequently fail to meet even of the whole 2013 school year. When principals did observe minimum levels of quality and fall short of what teachers classroom teaching, most gave written comments focused want and need. Most teachers in the Philippines received on the teaching method that was used. Only around a some professional development training – approximately quarter of comments concerned the teacher’s knowledge of ve days of in-service training for the average Grade 10 the subject content. teacher in 2013 – but this is less than in other countries. Moreover, approximately 40 percent of teachers interviewed DepEd has increased the budget allocation for human for the PETS-QSDS study said that they needed more and resource training and development (HRTD), but utilization of better quality in-service training to improve their classroom these funds is often low, amounting to only 57 percent of the teaching. budget in 2014 (Figure 7). Systems at the school level to support teachers and identify The PETS-QSDS survey also found that only a small share their professional development needs are not working well. of the available funds is transferred from DepEd central Teachers are expected to develop their own professional o ce to its regional and division o ces. Also, almost half of Figure 7: The Budget for In-service Training Has Increased Recently but Utilization Rates Are Frequently Low Appropriations, allotments, obligations, and utilization rates for HRTD funds (PHP billions in 2014 constant prices), 2005-2014 Appropriations Allotments and obligations 2.0 2.5 57% 2.0 1.5 43% 43% 72% 1.5 34% 81% 1.0 57% 40% 1.0 30% 0.5 0.5 92% 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Allotments Obligations Source: Appropriations - Department of Budget and Management. All other data - DepEd Statement of Appropriations and Obligations, various years. Note: Allotments/obligations for a given year include current, continuing, and extended allotments/obligations. The utilization rate is indicated by the percentage gures in the graph and show the proportion of allotments obligated in a given year. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xix Overview HRTD funds were allocated for training activities provided urban areas.6 For example, in 2014 the average high school at DepEd’s central o ce, with the rest split equally between in a highly urbanized city had approximately 1,700 students regional and division o ces. Given that division o ces have compared with about 1,000 and 570 students in city and the closest contact with schools and are most likely to know municipality schools respectively. Around 30 percent of their in-service training needs, it is surprising that a larger high schools in these highly urbanized cities have student- share of HRTD funds is not downloaded to divisions. instructional room ratios in excess of 55:1. Little information was available on how DepEd’s regional While most schools meet DepEd’s service standards, a large o ces used their HRTD funds. The regional o ces retained proportion of those that do not are in highly urbanized cities. approximately 42 percent of all of the HRTD funds that they While 94 percent of elementary schools and 83 percent of received from DepEd central o ce rather than downloading high schools have student-classroom ratios that fall within or them to division o ces. DepEd expects regional o ces to below the DepEd standard, some 24 percent of elementary use these funds to organize mass teacher training on, for and 30 percent of high schools in highly urbanized cities example, the introduction of the new K to 12 curriculum. have student-classroom ratios well above the maximum set DepEd’s guidelines governing the use of these funds require out in DepEd standards. Many also fail to meet sanitation regional o ces to keep detailed records on who has been facility standards, including the 70 percent of high schools in trained and on the kind of training provided. However, when highly urbanized cities with more than 50 female students the PETS-QSDS study team visited all regional o ces to sharing each toilet. collect this information, no such information was available. This lack of records greatly reduces the transparency of the To get a better picture of the real learning environment use of these funds. faced by students, around 7,000 classes and classrooms were observed as part of the PETS-QSDS survey. These classroom School Infr structure observations revealed that class sizes were smaller than o cial statistics suggest but that many classrooms were in a poor state of repair (Figure 8). Evidence from around the world has shown that improving school infrastructure leads to better learning outcomes.2 S stems to Improve School For example, a 2011 review of the economics literature Infr structure over the last 20 years showed that the availability of basic school infrastructure (such as classrooms, desks, and chairs) Spending on school infrastructure has risen rapidly in recent and facilities (such as electricity, libraries, and blackboards) years, and many new classrooms have been built which has is frequently associated with better student learning eased congestion, particularly in high schools. The allocation achievement.3 These ndings have been backed up by of government infrastructure resources tends to be broadly a systematic review of recent impact evaluations, which in line with need but could be improved further: while half showed that infrastructure investments have a positive of all elementary schools with overcrowded classrooms impact on school enrollment rates, attendance rates, and were selected by DepEd for an infrastructure improvement learning achievement.4 Also, research in the Philippines has project, so were 30 percent of schools with adequate shown that reducing the number of students per classroom classroom space. is associated with better student learning outcomes, particularly in rural schools.5 Utilization rates for school infrastructure funds have been low over the last 10 years. In the 2013 budget, Congress The Current St te of School appropriated approximately PHP 27 billion for school Infr structure infrastructure improvements, but only around 70 percent of this appropriation was released in 2013. This is a particular Schools in more densely populated areas tend to be larger concern given the infrastructure shortages outlined in and have more congested learning environments, and high the previous section and given the large increases for schools in particular tend to be much larger in cities and infrastructure investment planned for the coming years. xx WWW.WORLDBANK.ORG/PH Figure 8: A Significant Proportion of Classrooms Were Observed to Be of Poor Quality Indicators of classroom quality from direct classroom observation 30% High School Percentage of observed classrooms 25% Elementary School 20% 15% 10% 5% 0% Poor state Unclean Poor Poor sound Without Inadequate of repair ventilation insulation electricity seating Source: PETS-QSDS national survey – classroom observations at the school level. As many as one in ve school infrastructure improvement were not recorded in the DepEd or DPWH o ces. It is clear projects started in 2013 had not been completed by late that existing information and monitoring systems do not 2014 mainly due to insu cient funds, other demands on accurately record the use of infrastructure funds. contractors’ time, and design issues (Figure 9). The initial cause of delays in these projects was often a delay in the release of funds by DPWH central o ce to their district Figure 9: Schools Report Low Levels of Satisfaction engineering o ces but other di culties included in exibility with Government-funded Infrastructure Projects Completion and satisfaction rates for 2013 school infrastructure in the funding of school projects (62 percent of DPWH projects, 2014 district engineering o ces reported receiving insu cient funds to remove debris while 33 percent complained of 100% impractical costing). Over 60 percent of DPWH o ces reported that there was not enough space allocated for the 80% new infrastructure. There were large discrepancies between the accounts given 60% by implementing o ces and schools of these infrastructure projects. First, elementary schools reported that fewer 40% DepEd-funded projects had been implemented than was reported by DepEd and the DPWH. Second, only around half 20% of the projects that the DepEd and DPWH o ces claimed to have implemented could be veri ed using information 0% provided by the schools. Only a quarter of projects reported Completion rate Satisfaction rate by schools as DepEd- nanced projects could be matched DepEd Local government Own funds with information from the relevant DepEd divisional or Source: PETS-QSDS national survey – infrastructure project information at DPWH district o ce. It is di cult to draw rm conclusions the school level for all elementary and high school projects. about whether infrastructure funds are reaching their Note: Data relates to all school infrastructure projects in 2013. Completion intended destinations since not all projects recorded at rates are the proportion of 2013 projects that were completed by the last the district or division level were implemented in schools, quarter of 2014. DepEd projects include those managed by the DPWH as well as major school repair and water and sanitation projects managed by while schools also reported projects funded by DepEd that DepEd Division o ces. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxi Overview School-b sed Autonom M n ement The main weaknesses identi ed by school principals related to school autonomy. On the whole, more In recent times, many countries have begun to devolve elementary school principals than high school principals authority and resources to schools in an e ort to highlighted weaknesses in their implementation of SBM. improve teaching and learning conditions. School-based However, more high school principals mentioned their management (SBM) reforms of this kind are a strategy to inability to raise su cient resources as a major impediment improve education quality by bringing decision-making to putting SBM into practice. It is likely that this also limits closer to local communities and by strengthening the ability of schools to carry out the activities included in accountability between schools and the parents and their school improvement plans. children that they serve. Schools have discretion over only a very small amount Recent studies from many di erent countries have shown of their total revenues. The PETS-QSDS survey of school that school-based management can have a positive revenue and expenditure showed that the average annual impact in terms of increasing access to education and revenue of an elementary school was approximately improving learning outcomes.7 Reforms of this kind have PHP 3 million, and for high schools it was PHP 8.1 million also had a positive impact on student learning outcomes (Figure 10). Over 80 percent of school funding was devoted in the Philippines.8 However, the type of SBM reforms to teacher and sta salaries, mostly for centrally hired being implemented varies from country to country, the government teachers and sta . While school principals reforms can take time, and their success depends critically have some control over the day-to-day supervision of their on parental participation, political support, and e ective sta , the hiring and resources used for school sta ng fall implementation. outside the school’s control. Only around PHP 188,000 or 7 percent of total funds is available to each elementary In the early 2000s, the Government of the Philippines school to use at its own discretion. High schools have a introduced school-based management reforms that have little more exibility in the use of their funds but only had a positive impact on student learning outcomes. A over around PHP 987,000 or 12 percent of each school’s key aspect of the success of these reforms has been the total funds are discretionary. provision of ever-increasing levels of operational funding to the school level coupled with the devolution to P rticip tion schools of greater autonomy over the use of these funds. By 2014, schools were managing signi cant amounts of E ective school-based management requires parents and resources and using these funds to implement their own local community members to play a strong supporting school improvement plans. Also, in 2015, DepEd issued role in school decision-making and oversight. The School new guidelines drawing on past experience that aimed Governing Council (SGC) for each school is a forum for to strengthen school improvement planning and make it parents, students, teachers, and community stakeholders more transparent. to participate in making school improvement decisions. It takes an active role in developing the school improvement Most schools in the Philippines have not yet gone very plan (SIP) and, together with the school principal, is far in implementing school-based management. In 2009, responsible for endorsing it and sending it on to the DepEd introduced a tool for schools to assess their own schools division superintendent (SDS) for approval. It is also progress towards SBM. In 2014, the PETS-QSDS survey expected to participate in monitoring the implementation found that most elementary and high schools had put in of the SIP. The number and type of the SGC’s members and place only the lowest level of SBM. This means that they the frequency of its meetings are decided at a rst meeting had only a minimum number of arrangements in place for attended by parents and other stakeholders where the community participation and for taking action to improve constitution and by-laws of the SGC and the election learning outcomes. process for initial council members are agreed. xxii WWW.WORLDBANK.ORG/PH Figure 10: Discretionary Resources at the School Level Are Low Annual school revenue by type and source, 2013/14 school year Composition of all school funding (PHP 000s) Percentage of discretionary funds by source 10,000 100% 8,000 80% 6,000 60% 4,000 40% 2,000 20% 0 0% Elementary School High School Elementary School High School Salary Infrastructure DepEd Other cash School Other in-kind Discretionary/cash Source: PETS-QSDS national survey – school level. Note: The averages reported in the table are for all schools even if they received zero revenue under a particular category. Other cash revenue includes cash provided to schools by parent-teacher associations, local governments, and barangays and from private sector or NGO contributions. Nearly three-quarters of the household heads who were Figure 11: Few Parents Know about the SGC or Are interviewed were unaware that their child’s school had an Aware of the School Improvement Plan Percentage of parents of elementary and high school students SGC (Figure 11), and fewer than half of all parents were aware who know about the SGC and the SIP, 2014 that their school had an improvement plan. Most of the parents who knew about the SIP seem to have been invited 40% to participate directly in its preparation, but awareness beyond this group was rare. These ndings raise questions about how e ectively the SGCs are ful lling their role as forums for school and community collaboration in improving 20% learning outcomes. Account bilit nd Tr nsp renc 0% The e ective implementation of SBM depends on parents, Aware of School Aware that the school has Governing Council an improvement plan PTAs, SGCs, and other stakeholders having access to the information needed to hold schools to account and to Kindergarten Elementary High School provide them with feedback. Source: PETS-QSDS national survey – household level. Note: The bars relate to the level of education attended by the student from the interviewed household. Around 70 percent of elementary and high schools have some kind of notice board or transparency board where ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxiii Overview Figure 12: Social Accountability Initiatives like Check My School Do Not Work in Many Schools Percentage of school principals reporting an engagement with Check My School (CMS) in 2013 or 2014 20% 15% 10% 5% 0% Elementary Schools High Schools Aware of CMS Visited by CMS (2013 or 2014) CMS report Feedback through CMS Source: PETS-QSDS national survey – school level. information was posted, but the study team found that many Beyond school-level institutions, social accountability were not visible to the public – including the one-third in initiatives like Check My School (CMS) appear to have a elementary schools that were located inside the principal’s limited impact. CMS uses mobile and internet technology o ce or the sta room. Schools rarely display or publish up- to enable communities to monitor the quality of basic to-date information on how they spend their discretionary education services, involving local facilitators who mobilize resources. local communities to verify information on the status of their school. Any discrepancies between the o cial government The limited availability of information provided by schools data and the data collected by the CMS are then brought to helps to explain why parents appear to be unaware the attention of DepEd. Although the initiative began in 2011, of the nancial resources that schools receive to fund its coverage appears to be quite small. Only 15 percent of improvements. Only 40 percent of parents of elementary and elementary school principals and 20 percent of high school high school students knew that schools received operational principals were aware of CMS, and only a small proportion of funding from the national government, and fewer than half these schools had had any direct dealings with the initiative of those had received information on how the funds had (Figure 12). For example, fewer than 5 percent of high schools been allocated in the previous year. had ever provided feedback through CMS. There are few other ways by which parents and other stakeholders can access information on school performance. N tion l Government As part of the school improvement planning process, schools are supposed to prepare an annual school report card that Support for School includes information on key school performance indicators Oper tions such as National Achievement Test (NAT) results and student progression indicators. Yet only around half of the parents Providing schools with direct funding to meet their of elementary and high school students said that they had operational needs is the cornerstone of the government’s been given such a card or even any information on the card e orts to strengthen school-based management in the in the previous two school years. A similar proportion of Philippines. Over the past ve years, funds provided directly parents reported that they had not even received a report to schools to support their maintenance and other operating card on their own child’s progress. expenses (MOOE) have increased by 45 percent in real xxiv WWW.WORLDBANK.ORG/PH terms, demonstrating the government’s commitment schools and for 82 percent received by high schools. both to increasing operational funding and to providing In the case of over 10 percent of schools, their MOOE schools with the funding that they need to implement their allocations were the only source of operational funding own improvement plans. Evidence from the Philippines that they received. and elsewhere shows that increased school funding and e ective school-based management (SBM) can lead to more Tr ckin MOOE funds e cient use of resources and, ultimately, to better education Only 77 percent of the total allocations for elementary outcomes. and non-IU high schools ultimately reached those schools in 2013 (Figure 13). In other words, elementary Over recent years, school MOOE funding has increased and high schools received only PHP 448 of the PHP 581 signi cantly in real terms even though its share in the overall allocated for each student. Given that the guidelines basic education budget has remained stable. For example, stipulate that all school MOOE funds should be between 2011 and 2014, per student appropriations for downloaded to schools, this nding is a concern. It elementary school MOOE increased from PHP 317 (US$7) to implies that about PHP 1.8 billion out of the PHP 8 billion PHP 566 (US$13). Despite these increases, a detailed costing MOOE budget intended to be used by schools in 2013 study revealed that existing levels of school MOOE are were not downloaded. insu cient and would need to more than double to ensure that existing service standards were met. The retention of school MOOE by DepEd division o ces is the main reason for the di erences between Public schools rely heavily on the school operating funds the initial allocations and the amount of funds actually provided by the national government. The PETS-QSDS downloaded to schools. Although division o ces are survey collected detailed information on all types of school required to download funds to schools in full, the PETS- revenues. In 2013, MOOE allocations accounted for 68 QSDS tracking exercise indicated that over 60 percent percent of all discretionary funding received by elementary of division o ces held onto some MOOE funds in 2013 Figure 13: A Significant Portion of the MOOE Allocations Do Not Reach Schools Share of MOOE allocation downloaded and received by schools, 2013 Share of original allocation Proportion of schools downloaded and received, 2013 100% 100% 84% % of original allocation 77% 75% 50% 25% 0% Central MOOE MOOE downloaded MOOE received by Elementary Schools High Schools allocation to schools by schools from Less than full allocation Allocated amount division Division Download exceeds allocation Sources: Data on allocations from DepEd central o ce. All other data from PETS-QSDS national survey – division and school levels. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxv Overview The di culties that school principals experience in managing MOOE funds can result in signi cant delays in their submission of liquidation reports as well as some misreporting of data. Although schools are usually required to submit liquidation reports within 35 days of receiving an MOOE transfer, only about half of all elementary school advances and one-third of high school advances were submitted within this period. Some schools submitted reports as long as three months after receiving the initial advance. Schools serving poorer and more marginalized groups are provided with higher levels of MOOE funding for each student. This is not because the MOOE funding formula provides additional funding to schools serving poorer groups. Instead, this is partly a re ection of the fact that wealthier students are more likely to attend schools in urban areas where student-classroom and student-teacher ratios are higher, thus leading to lower per student MOOE funding. in order to procure items for schools, pay their utility However, in elementary schools, the pro-poor allocation bills, or fund other services for schools. This was the case virtually disappears in terms of the amount of MOOE funding even though DepEd explicitly prohibits division o ces that the schools actually spend. The result is that, in practice, from procuring items using school MOOE funds except students from di erent socioeconomic groups attend in extreme cases where there is a demonstrated need for schools that receive similar levels of per student MOOE the division to do the procurement. This practice reduces despite the large di erences in their original allocations. This the amount of discretion that schools have over the use is a missed opportunity to use school MOOE as a way to of their own MOOE funds and has the potential to limit narrow the very large di erences in school revenue among the e ectiveness of ongoing school-based management schools serving di erent socioeconomic groups. reforms. Oversi ht nd Monitorin School liquidation (utilization) reports reveal that some schools nd it di cult to use all of the MOOE funds that While there are regulations in place that aim to make use they receive. The main reason why so many schools nd it of MOOE funds more transparent, compliance with these di cult to spend their entire MOOE allocation is because regulations is low. The Anti-Red Tape Act (2007) mandates they have insu cient capacity and time to meet their public schools to maintain a transparency board or bulletin reporting and procurement requirements. Over one- board detailing how they are using their MOOE funds. third of elementary schools and a quarter of high schools However, as of 2014, fewer than two-thirds of schools had reported encountering such problems in 2013, and the a transparency board containing information on MOOE most common reason they cited was lack of capacity to expenditures. Moreover, the information posted on about meet reporting requirements. The management of MOOE one-third of transparency boards was more than three funds is particularly demanding of school principals. In months old. the 2014 school year, principals reported spending six to eight hours a week on administrating MOOE funds, Decisions on the use of MOOE funds are largely con ned including procuring goods and services and preparing to the school principal and other teachers. In 2014, over 80 liquidation reports. percent of elementary schools reported that teachers had xxvi WWW.WORLDBANK.ORG/PH been consulted about how to use the schools’ MOOE education. While spending on basic education by local funds (Figure 14), but PTAs had an input in fewer than 30 government units (LGUs) has uctuated considerably percent of schools. This is not surprising since DepEd’s over the last 10 years, in real terms, it appears to have current guidelines explicitly prohibit PTAs from “interfering been declining since 2007. Combined with ever- in schools’ administrative management.” increasing levels of national spending, the share of public basic education funding accounted for by local Only 13 percent of elementary schools receive input from government spending declined from 11 percent in 2006 school governing councils (SGCs) even though SGCs are to only 6 percent in 2013. responsible for developing and implementing the school improvement plan (SIP), which includes activities nanced There are also large disparities in basic education by MOOE funds. This number is even lower among high spending by local governments across di erent regions schools, with SGCs, parents, and local school boards in the Philippines. These disparities are largely driven participating in MOOE decisions in fewer than 10 percent by di erences in the revenues that local governments of schools. The study’s interviews with parents revealed a collect for the Special Education Fund (SEF) from a 1 similar picture. Only about one-third of parents reported percent surcharge on property taxes. The National participating in discussions about the school’s use of Capital Region (NCR), the region with the highest overall nancial resources in general and fewer than a quarter levels of property tax revenue, accounts for over 44 reported participating in decisions about how to use the percent of total local government education funding in school’s MOOE funds. the Philippines. This is a very large and signi cant source of additional funding for the school system, equivalent Loc l Government Support to approximately PHP 3,500 in additional funding for each basic education student in the region or about 28 for B sic Educ tion percent over and above per student funding from the national government. Disparities between other regions Local government funding represents a small and are not as stark but still exacerbate funding inequalities declining share of overall public funding for basic between schools in di erent locations. Figure 14: There Is Little Participation by Local Stakeholders in Decisions about How MOOE Funds Are Used Percentage of schools that received input from stakeholders about MOOE funds, 2014 SY Elementary schools Non-IU high schools 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Teachers PTA School Governing Council Parents Local School Board Teachers PTA School Governing Council Parents Local School Board Source: PETS-QSDS national survey – school level ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxvii Overview Looking beyond aggregate spending gures, it can be Data on levels of annual local government education seen that fewer than 50 percent of schools receive any kind spending was often unavailable to the survey team at the of direct funding from local governments (Figure 15). Even local level, and what little information they were able to nd when schools do receive LGU funding, it tends to make was inconsistent. This clearly limits the ability of stakeholders up a very small share of overall funding. For example, LGU to hold local governments accountable. The PETS-QSDS funding represents less than 4 percent of the total cash and team collected nancial data on the Special Education in-kind funding that schools receive, with the remainder Funds (SEFs) from the Bureau of Local Government Finance largely coming from the national government. (BLGF) at the national level and from LSB o ces in cities, municipalities, and provinces. The team then attempted A comparison between levels of LGU direct school to match these two sets of data, which revealed signi cant funding and total LGU education spending reported at inconsistencies between the nationally and locally reported the national level suggests that signi cant amounts are spending aggregates. The team found that the information not being spent on activities that directly bene t schools. reported at the national and local levels was consistent in the Rough estimates by the PETS-QSDS team using nance case of only 16 percent of LSBs, while, for 22 percent of LSBs, data carefully collected during school visits suggest that there were inconsistencies in the levels of spending reported LGU spending on schools amounts to approximately PHP at the national and local levels, and half of all LSBs did not 9 billion. This represents only 58 percent of the PHP 16 report any SEF expenditure at the local level. billion that LGUs reported spending on education in 2013. There was little information on how the LGU funds that were not spent at the school level were used. The study Equit team made considerable e orts to collect such information from the local governments directly and from the DepEd Ensuring access to good quality basic education is necessary supervisors who sit on local school boards. However, little if children are to be equipped with the foundational skills information was available and, even where it was, the that they need in order to play a full and productive role in data were patchy and did not correlate with the o cial their communities. Despite recent progress in narrowing aggregate spending gures reported by the LGUs. disparities between socioeconomic groups, gaps in access Figure 15: Fewer than Half of All Schools Receive LGU Support, and the Amount of Funding Provided is Small Percentage of schools that report receiving cash or in-kind contributions from local governments and average levels of support, 2013-14 Percentage of schools receiving LGU support Average level of LGU support per school (PHP 000s) 60% 2,000 1,500 40% 1,000 20% 500 0% 0 HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools Elementary Schools High Schools Elementary Schools High Schools Source: PETS-QSDS national survey – school level. Note: The average level of support is only for schools that received LGU contributions in 2012/14. xxviii WWW.WORLDBANK.ORG/PH remain large, particularly at the high school level. The PETS- These patterns in school performance are partly driven by QSDS study also found large di erences in the quality of di erences in the quality and availability of facilities in rural education provision across regions, between rural and urban and urban areas. Average school size varies enormously areas, and among schools serving di erent income groups. between urban and rural areas. In 2014, the average high school in a highly urbanized city had approximately 1,700 Re ion l Inequ lit students compared to about 1,000 students and 570 students in city and municipality schools respectively. Signi cant di erences in levels of education spending and However, the facilities in highly urbanized city schools are the quality of the learning environment exist across regions frequently insu cient to accommodate these high student and provinces. For example, the national government’s numbers. For example, the proportion of schools that fail to spending per school-aged child in 2012 varied from PHP meet DepEd standards for classrooms and toilets is much 4,500 in the National Capital Region (NCR) to just over PHP higher in highly urbanized cities than in other locations 7,600 in the Cordillera Administrative Region. Moreover, (Figure 16). regional disparities in national government spending have been growing since 2005. Local government spending on Tackling de ciencies in the quality of the learning basic education tends to reinforce national government environment also appears to be more di cult in urban spending patterns. For example, Region II had one of areas. Two-thirds of DPWH district engineering o ces the highest national and local government per student and all o ces in urban areas reported nding it di cult to allocations in 2012, while Region VII had one of the lowest implement school infrastructure projects in urban schools. national and local government allocations. Lack of space and impractical uniform pricing policies were key reasons underlying these di culties. These funding disparities result in considerable variations in the quality of school learning environments across regions. The quality of other important education inputs also tends to This study shows that there is a strong association between vary between rural and urban areas. Fewer quali ed teachers appear to be available in urban areas, particularly in highly levels of per student education spending and the quality of urbanized areas, than in more rural areas. For example, there the learning environment. In other words, regions with low is an average of 39 students for every teacher in elementary levels of per student spending also tend to have the poorest schools in highly urbanized cities compared to only 29:1 in learning environments. For example, low levels of national schools located in municipalities. Di erences in STRs in high and local government funding in Region VII have resulted schools are not as stark, but large urban schools have higher in very poor student-teacher and student-classroom ratios STRs than schools in rural areas. compared to national averages. Teacher absenteeism rates also tend to be higher in cities, Rur l Urb n Disp rities particularly in highly urbanized cities. In high schools located in HUCs, almost 1 in 10 teachers was found to be absent in Even though urban schools tend to serve wealthier 2014. This absenteeism rate was 53 percent higher than the populations, they tend to perform poorly compared to rural national average. However, while their absenteeism rates schools. National examination results tend to be lower for were high, high school teachers in HUCs and other cities had students of urban elementary and high schools than for higher scores on subject content assessments conducted as students of rural schools. For example, in 2014 the average part of the study than rural high school teachers. score in the Grade 6 national achievement test was 66 percent in highly urbanized cities compared to 75 percent in Socioeconomic Disp rities rural municipality schools. The PETS-QSDS study also found that urban schools, particularly high schools, tended to be The PETS-QSDS survey included a nationally representative less e cient than rural schools in translating their generally sample of public elementary and high school student larger revenues into better learning outcomes. households. Using the information collected from these ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxix Overview Figure 16: A Greater Proportion of Urban Schools Fail to Meet Infrastructure Standards than Rural Schools Percentage of schools that failed to meet DepEd service standards, 2014 80% 60% 40% 20% 0% Students per Female students Students per Female students instructional room per toilet instructional room per toilet Elementary Schools High Schools HUCs Other Cities Municipalities Source: PETS-QSDS national survey – school level. survey households, it was possible to rank them by their reforms have had only a limited impact in schools that serve estimated levels of per capita household consumption. This poorer communities. The ndings from the PETS-QSDS study ranking was used to explore the characteristics of schools suggest that this may be due at least in part to the fact that attended by di erent socioeconomic groups. schools in poorer areas have less funding to support key elements of SBM, including the implementation of school Schools serving poorer students tended to be more improvement plans. resource-constrained than wealthier schools. Schools receive discretionary funds from the national government, Di erences in the availability and competency of teachers and these tend to be allocated relatively equitably. Schools by socioeconomic group present a more mixed picture. also generate their own funds, but their ability to do so is On the whole, poorer children attend schools with lower dependent on the socioeconomic composition of their rates of teacher absenteeism than wealthier children. This student populations. In 2014, students in the poorest quintile ranking reveals that students in the bottom 20 percent of attended elementary schools that collected about half as households in terms of consumption tend to attend schools much in discretionary funding as the schools attended by with lower rates of teacher absenteeism than students from students in the wealthiest quintile. the wealthiest 20 percent of households. This re ects the Lower levels of funding are a key reason why schools greater proportion of wealthier students attending schools serving poorer communities tend to have poorer ratings in highly urbanized cities where rates of teacher absenteeism for the implementation of school-based management. The are higher. However, high schools serving better-o students study found that students from the poorest 20 percent tended to have better-performing high school teachers than of households were more likely to attend elementary and high schools attended by poorer students. At the elementary high schools that had the lowest level of self-assessed level, no clear pattern emerged, although poorer students SBM implementation than students from the wealthiest 20 tended to be taught by teachers who performed better than percent of households. Other research has shown that SBM average on the Filipino test. xxx WWW.WORLDBANK.ORG/PH The mixed picture of the quality of basic education services However, the study also highlights some limitations in available to di erent socioeconomic groups highlights the the existing systems that manage public basic education need to look at all dimensions of inequality to improve basic funds. Allocation systems have managed to target funds education services. For example, a greater proportion of poor reasonably well, but there are still many examples of needy children than wealthy children are located in rural areas where schools missing out on additional support while other, schools are less crowded and the learning environment is better-equipped schools continue to receive additional better. However, at least in high schools, the best teachers are support. Moreover, delays in identifying school needs and in working in wealthier urban areas, which puts poor students the budget allocation process have meant that a signi cant in rural areas at a disadvantage. Addressing these di erent amount of funding goes unspent or has to be spent quickly dimensions of inequality is a necessary rst step to ensuring or at times that disrupt school operations. Even when that students across the many islands of the Philippines all schools bene t from improvement projects or resources, receive an education of similar quality. the study has also shown that they are frequently unhappy about the quality of the resources supplied. Conclusions nd In the absence of e orts to address these limitations, further Polic Directions increases in funding are likely to have only limited impact on the challenges faced by the education sector. Moreover, the The study has shown that, despite a renewed focus on successful introduction of the senior high school program in basic education by the Philippines government, further 2016 will be jeopardized if systems managing the allocation increases in both capital and recurrent public spending and use of public funds are not made more e ective. are needed. The study has shown that there have been signi cant improvements in the learning environment that Through a series of policy notes, this report has identi ed Filipino children experience every day but more still needs some of the key challenges and potential policy responses to be done. Many schools, particularly in urban areas, have in key areas of resource management. Instead of repeating insu cient and poor quality facilities and a shortage of these recommendations in full, this section draws together teachers. Operational funding still falls short of the amounts some common policy directions that have emerged that schools need to pay bills, undertake basic repairs, and from the ndings of the study as a whole. Table 1 also provide the day-to-day materials their students need. And summarizes the major ndings and policy suggestions there is rarely anything left over to fund school-level initiatives from each policy note. to improve student learning achievement. The study has also highlighted the lack of good quality opportunities for Improve Alloc tion Mech nisms teachers to improve their skills. Addressing these challenges Throu h Better Pl nnin will require further investments in the education sector. The report makes clear that the systems currently used to It also appears that the bulk of this funding will need to be allocate resources to schools could be strengthened by provided by the national government. The study has shown improved planning. While these allocation systems generally that local government funding, except in the National Capital target schools with greater need, they could be improved Region, is very small and that funding from other sources, further. For example, school infrastructure improvement including parent-teacher associations, is also negligible. projects are planned on an annual basis and project lists are Detailed estimates suggest that over 90 percent of school frequently nalized late. This shortens the time available to funding comes from the national government. Unless there implement them and compromises the ability of DepEd and are signi cant policy changes in the future, this means DPWH o ces to monitor projects adequately. Developing that increases in spending will need to be found from a two- or three-year plan that would forecast need and list the national government budget, either through greater all school improvement projects in each DepEd division revenue-raising e orts and/or by increasing the share of would reduce the work involved in putting together the basic education in the overall government budget from its annual project list and accelerate the transmission of funds existing level of 15 percent. to the responsible implementing o ces. It would also ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxxi Overview give implementing o ces time to do some advance Over recent years, schools have been given ever increasing planning and thus avoid some of the di culties associated amounts of discretion over how to spend their funding. with inadequate funding and site availability that were This has primarily been done to support school-based commonly cited as reasons for implementation delays. A management reforms by providing schools with the funds similar approach could be adopted for teachers and the to implement their own improvement plans. Evidence planning of other important school inputs. from the Philippines and other countries supports the idea that giving schools more autonomy can improve More input from division o ces and schools in planning education outcomes. could also improve allocation decisions. Funding decisions are largely taken at the central level and frequently do However, the study found that the existing system for not take into account local conditions. For example, in managing school operational funding is too onerous on the PETS-QSDS study, a considerable number of school school principals and that there is a risk that the bene ts principals and school division superintendents (SDSs) from reforms of this kind will not materialize. Simplifying the felt that existing procedures for allocating new teachers existing requirements for the management of these funds were suboptimal, and some SDSs felt that they had not would reduce the signi cant burden that this currently puts been adequately consulted. Similarly, schools had only on schools. It would also give school principals more time to limited input into the planning and implementation of focus on providing academic leadership in their schools. One infrastructure projects. This contributes to an impression approach that has been adopted in other countries would that division o ces and schools have limited in uence on be to treat operational funds as a grant. For example, this education decision-making. Strengthening the processes is the case in Indonesia, where these funds have much less for consulting schools and division o ces about the onerous reporting requirements than required by the usual decisions that a ect them and making information publicly government budget and implementation systems. Schools available on the decisions that are made would improve use simpli ed reporting templates to report on their use planning. of the funds and submit these forms to district o ces (the equivalent of division o ces in the Philippines) every quarter. Give Schools Gre ter Authorit nd Developing a similar system adjusted to take account of the Simplif Reportin Requirements speci c context of the Philippines could reduce the burden that using these funds currently puts on schools. Increasing the role played by schools and local DepEd o ces in deciding on the use of education funding may However, if division o ces, schools, and even local also improve the quality of public education spending. The communities are to get more involved in decisions about study has highlighted the limited authority that schools how education resources are allocated and used, these have over the implementation of school infrastructure spending decisions need to be more transparent. Across projects. It has also shown the weak links between the all of the spending areas that the PETS-QSDS survey individual professional development needs of teachers covered, information on the transfer and use of funds and the types of in-service training o ered. Providing was frequently missing, inaccessible, or in formats that schools and division o ces with more authority during the were not easy to understand. Even where information implementation of both infrastructure projects and teacher was available, it frequently was not consistent between training could yield signi cant bene ts. For example, giving di erent administrative levels. For example, there was a school principals the authority alongside DepEd and DPWH discrepancy of about 8 percent between the funds that o cials for certifying that infrastructure projects have been division o ces reported having downloaded to schools for satisfactorily completed might improve their quality. School operational expenses and the amounts that the schools principals in the Philippines have taken on roles of this kind reported receiving. In most cases, it was not possible for the in the past and with appropriate training could become study team to judge accurately the extent to which these a major driving force in improving the quality of school discrepancies arose from misappropriation of funding and/or infrastructure projects. incomplete reporting. xxxii WWW.WORLDBANK.ORG/PH Improve Tr nsp renc of Fund Setting basic standards for basic education provision Alloc tion nd Resource Use and ensuring that information about these standards is widely disseminated would enable parents and other Local government funding was an area that was particularly stakeholders to judge the fairness of allocations and to concerning. The study estimated that only around 60 percent hold the government more accountable for the provision of local government funding reached schools directly. of key education inputs like classrooms and teachers. It However, records on which schools received funds and what would also enable schools and other stakeholders to seek the funds were provided for were very weak. For example, 24 redress for unfair allocations or poor project implementation. percent of elementary schools to which local governments Policymakers could explore the feasibility of introducing a claimed to have provided in-kind support for salaries denied formal grievance system that schools could use to express ever having received this support. In contrast, LGUs themselves dissatisfaction with decisions related to all school funding, reported providing support to only approximately 40 percent including school infrastructure projects, as this would help to of all of the elementary schools that reported having received make government education spending more e ective and support from their LGU. Information on the 40 percent of improve education services. funds that did not go directly to schools was even less reliable. Stren then the Role of School Developing simple reporting formats in which to record Governin Councils nd P rent Te cher the funds allocated to divisions and schools and how Associ tions these institutions use these funds would go a long way to increasing transparency. A recent initiative by the Parents and other community members can also play a Department of the Interior and Local Government (DILG) to role in improving school decision-making and overseeing encourage greater accountability and transparency is one how schools use public funds and resources. However, useful example of what could be done. The DILG awards the study found that, while SGCs had been established in the Seal of Good Local Governance to local governments nearly all schools, their role was fairly limited. Giving SGCs that perform well in several areas as well as providing them greater responsibility for monitoring how schools use public with additional performance-based funds. This scheme has funds could provide a useful complement to DepEd’s own given LGUs a strong incentive to increase transparency and oversight role. For example, SGCs could become jointly to improve their use of the Special Education Funds. While responsible with the DepEd division o ces for signing o on the awarding of the Seal is already conditional on whether schools’ fund utilization reports. However, if the SGCs are to the annual plan of local school boards is aligned with the take on an expanded role, then their members will need to plans of the schools themselves, it could also be used to receive more training and support from DepEd. Schools also encourage better reporting. For example, the category of need to encourage greater participation from parents and “good nancial housekeeping” could be expanded to include others in the SGC to strengthen their planning and oversight more detailed reporting, a requirement to publish details of activities as the PETS-QSDS study found that most parents of how SEF funds are used, and a requirement to set targets for students were unaware of the existence of SGCs or of their the proportion of funds that should reach schools. school’s improvement plan. Introducing and widely disseminating a set of standards for A campaign to increase the awareness of parents and elementary and high schools would also increase transparency other education stakeholders of the role of SGCs and and encourage greater accountability for the use of public school improvement plans is needed. Recent evaluations funds. DepEd has a number of di erent guidelines that in Indonesia and Pakistan have shown that it is possible to set standards for elementary and high schools, which the use old and new technologies to disseminate information. PETS-QSDS study used to assess the adequacy of existing In Indonesia, an impact evaluation study showed that resource allocations. However, these guidelines are scattered well-designed information campaigns using simple SMS across di erent DepEd orders and regulations, and many text messages or school meetings can signi cantly increase stakeholders are unaware of their existence. public knowledge about schools’ funding levels and ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxxiii Overview responsibilities. It also found that this increased knowledge towards equalizing school funding and providing schools in led to higher levels of parental participation in school poorer areas with the additional resources that they need to a airs. In Pakistan, a similar strategy was used in a successful support the learning needs of marginalized students. Many pilot program to disseminate information about school other countries, like the United Kingdom, have introduced councils. The pilot set up a call center and used inbound funding components of this kind. The allocations of MOOE and outbound calls, robot calls, and SMS text messages to funds could also be adjusted to take into account price transmit important information about the role played by di erences between regions so that schools operating in school councils and their members. An assessment of the high-cost areas such as remote locations where travel and pilot showed that school council members’ knowledge transportation costs are high are able to purchase similar about their roles and responsibilities had increased, while amounts of goods and services as those in lower-cost school principals reported the councils had increased their locations. participation in school a airs. Further Rese rch Address Fundin nd Qu lit Inequ lities The study has highlighted a number of areas where current and existing resources could be used more e ectively. The study has shown that there are substantial di erences in However, there are a number of important gaps that further the quality of education services across the Philippines. The research could clarify. In particular, the PETS-QSDS study was factors associated with the distribution of quality vary, and not able to look at individual student performance and some there is no clear and consistent pattern. However, schools of its important determinants. In particular, the study did in highly urbanized areas, particularly high schools, tend to not explore the quality of classroom instruction and other be very large and, despite getting higher levels of funding, tend to have more stressed learning environments and important factors associated with the teaching and learning poorer levels of performance. The forthcoming extension process. The quantitative approach used in the PETS-QSDS of high school by two extra years is only going to add to study is not best suited to exploring issues of this kind, but the pressure that these schools face. A concerted e ort by qualitative work would be an appropriate way to add to what DepEd to address the speci c issues of these schools by, for is known about the quality of education in the Philippines example, establishing a task force to develop a complete and to help DepEd to develop appropriate pre-service and program of support for these schools could be an important in-service training. step forward. The PETS-QSDS study collected very detailed information In some cases, the distribution of education quality reinforces on the status of education service delivery across the existing inequalities. For example, poorer students tended Philippines. The report analyzed a wide range of priority to go to high schools that had teachers with more limited issues, but the data collected in the study could provide the knowledge of their subject areas. They also tended to go to basis of future research into an even wider range of issues. schools with lower levels of discretionary funding and those For example, further analysis of the teacher competency that reported having implemented only a minimal amount of assessments would be useful to provide further insights into school-based management. the professional development needs of teachers. The data could also be used to categorize schools for further research. The Philippines has a well-established funding formula For example, schools that have better school-community for school operational funding that could be extended relationships could be identi ed in the PETS-QSDS data to include an equity component. This would be one way as a starting point for more detailed qualitative work to to reduce the large school-level funding inequalities understand why these schools have been successful in this highlighted in this report. In order to compensate those respect. Finally, the data represent an important baseline for schools that receive lower funding from non-DepEd sources tracking future progress in increasing the e ectiveness of such as local governments, an equity component could be public education spending and improving student learning introduced in the allocation formula. This could go some way outcomes in the Philippines. xxxiv WWW.WORLDBANK.ORG/PH Table 1: Summary of Major Findings and Policy Suggestions Findings Policy suggestions 1. Strengthening Systems to Hire and Deploy Teachers Further improvements in teacher • Further develop guidelines/norms for school-level teaching needs, particularly for distribution are needed subject-speci c teachers in high schools • Monitor the distribution of teachers more closely to ensure e ciency and develop and implement the Human Resource Information System (HRIS) • Revise teacher transfer policy to make it a more e ective tool to improve teacher distribution Hiring and deployment systems • Strengthen the accountability of regional and divisional o ces for carrying out need to be strengthened processes in a timely way • Improve monitoring to ensure hiring timelines are met and that teachers are in post at the beginning of the school year • Improve coordination between the DBM and DepEd regional o ces High teacher absenteeism in • Increase the authority of DepEd division o ces and schools to deal with absenteeism highly urbanized cities • Link teachers’ attendance to their career development and performance bonuses • Explore the potential for increasing community involvement in monitoring teacher attendance 2. Strengthening Teacher Support Systems Teacher subject knowledge is • Strengthen pre-service and in-service training opportunities for all teachers (see below) generally low Professional development • Increase funding and opportunities for e ective professional development for teachers opportunities are limited and • Provide a greater share of this funding to division o ces and schools to enable them to delivery mechanisms are weak organize professional development activities • Transfer HRTD funds in a more timely manner • Increase the transparency of HRTD funds through improved reporting by DepEd region and division o ces and increase central o ce monitoring of fund use Systems to identify teachers’ • Revaluate existing teacher strength and needs assessments and teacher professional professional development needs development planning are weak • Develop diagnostic teacher competency tests to identify teachers’ in-service training needs • Develop content for professional development activities that will address existing weaknesses • Evaluate and scale up the use of school learning action cells as the main venue for in- service training A greater need to align teacher • Support a stronger alignment between professional development planning, career accountability systems to development and performance incentives raise teacher motivation and competencies ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxxv Overview Table 1 continued: Summary of Major Findings and Policy Suggestions Findings Policy suggestions 3. Strengthening Government Systems for School Infrastructure Improvements Allocation mechanisms for school • Establish transparent criteria for prioritizing schools based on existing facility infrastructure systems need standards strengthening • Validate project priority list at the division level • Increase transparency and accountability by disseminating project selection criteria and by establishing feedback/grievance mechanisms Large increases in infrastructure • Expand and establish partnerships to build and improve infrastructure spending are required but • Introduce multi-year planning for school improvement projects implementation capacity also needs to be strengthened • Strengthen coordination between implementing institutions signi cantly • Strengthen the role of schools and school governing councils in implementation Weaknesses in the monitoring • Increase the capacity of DepEd and DPWH to monitor projects regularly and evaluation of infrastructure • Increase the transparency of school infrastructure improvement activities at the spending need to be addressed division and school levels • Empower schools to monitor projects more closely and provide e ective feedback channels for all stakeholders to use 4. Strengthening School-based Management Schools assess their level of SBM • Provide further training to school principals, school governing councils, and parent- implementation as low teacher associations on school-based management • Build the capacity of district and division supervisors to mentor schools in the implementation of school-based management Only a small share of school • Increase central government MOOE funds to a level that will enable schools to meet funding is discretionary existing education service standards • Increase the authority of schools over other funds (such as school construction funding) SBM implementation is less • Introduce an equity component into the MOOE funding formula advanced and fewer discretionary • Provide additional supervisory support on school-based management to schools funds are available in schools that serving disadvantaged groups serve poorer communities SGCs are not functioning as they • Clarify the roles and responsibilities of SGCs and PTAs were designed to do • Establish a role for the SGC in overseeing the use of school discretionary funds • Increase the knowledge of SGC members on their roles and responsibilities through school-level training initiatives • Conduct an information campaign to increase public awareness of the role of SGCs and PTAs, the SIP, and the use of and the reporting requirements associated with discretionary school funds (such as MOOE funds and canteen revenues) Transparency at the school level • Strengthen monitoring by DepEd district and division o ces of the production of needs to be increased key information (such as school report cards and student report cards) by schools • Enforce regulations on making information publicly available on school performance and use of funds xxxvi WWW.WORLDBANK.ORG/PH Table 1 continued: Summary of Major Findings and Policy Suggestions Findings Policy suggestions 5. Strengthening Systems to Provide Operational Funding to Schools Not all schools receive their full • Enforce existing regulations on the transfer of funds MOOE allocation and the transfers • Introduce transfer targets for DepEd division o ces tied to performance bonuses are often not made on a predictable schedule • Make MOOE allocations and formula components for each school publicly available The management of MOOE funds • Treat school MOOE funds as a grant to simplify reporting requirements within schools is di cult and time- • Reduce the frequency of downloading and increase the time allowed for schools to consuming produce liquidation reports • Provide additional support to schools on MOOE management from district o ces and provide school-level training • Allow schools to deposit MOOE funds in bank accounts School-level institutions like the • Review and strengthen the role of SGCs in the planning and monitoring of MOOE school governing council are funds relatively weak • Disseminate information on the roles and responsibilities of SGCs to parents and the local community • Ensure that MOOE and other school-level funds are incorporated into school improvement plans Funding inequalities at the school • Introduce an equity component into the MOOE allocation formula level are large 6. Improving Local Government Support for Basic Education A low share of local government • Raise the share of local government funding provided directly to schools funding reaches schools and most is • Introduce a formula-based funding model to allocate more cash directly to schools provided in-kind to support their annual improvement plans Local government funding is • Reduce inequalities between local school board areas by adjusting the national inequitable funding formula • Introduce a simple and transparent local government funding formula to strengthen the link between funding and school needs Schools are rarely involved in • Make better use of schools’ annual improvement and procurement plans in LSB funding decisions planning • Ensure that the link to school improvement planning in the Seal of Good Local Governance is properly measured and monitored Transparency and accountability for • Improve reporting formats for the use of SEF funds and ensure that the results are the use of funds is weak publicly disseminated • Build on the Seal of Good Local Governance and associated performance-based funding by, for example, requiring LSBs to report planned school level funding allocations ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES xxxvii Overview Endnotes 1 See, for example, McEwan, P. (2013). “Improving Learning in Primary Schools of Developing Countries: A Meta-Analysis of Randomized Experiments.” Wellesley College, Wellesley, MA. 2 Woolner, P. Hall, E. Higgins, S. McCaughey, C. and K. Wall (2007). “A Sound Foundation? What We Know about the Impact of Environments on Learning and the Implications for Building Schools for the Future.” Oxford Review of Education 33(1): 47-70. 3 Glewwe, P. Hanushek, E. Humpage, S. and R. Ravina. (2011). “School Resources and Educational Outcomes in Developing Countries: A Review of the Literature from 1990 to 2010,” National Bureau of Economic Research, Cambridge, MA. 4 Krishnaratne, S. White, H. and E. Carpenter. (2013). “Quality Education for All Children? What Works in Education in Developing Countries?” Working Paper 20. New Delhi: International Initiative for Impact Evaluation (3ie). 5 Yamauchi, F. and S. Parandekar (2014). “School Resource and Performance Inequality: Evidence from the Philippines.” World Bank Policy Research Working Paper No. 6748, World Bank, Washington D.C. 6 Highly urbanized cities are cities with populations of more than 200,000 and with average revenues of at least PHP 50 million in 1991 prices. Other cities are urban areas that do not meet these criteria. Municipalities are the administrative units for all other areas in the Philippines. 7 AusAID ERF (2011). “Current Issues in Education: School Grants and School-based Management.” Canberra; Bruns, B., D. Filmer, and H. Patrinos (2011) “Making Schools Work: New Evidence on Accountability Reforms” World Bank, Washington, D.C.; and Carr-Hill, R., C. Rolleston, T. Pherali, and R. Schendel (2014) “The E ects of School-based Decision-making on Educational Outcomes in Low- and Middle-income contexts: A Systematic Review” Institute of Education, London. 8 Khattri, N., C. Ling, and S. Jha (2010). “The E ects of School-based Management in the Philippines: An Initial Assessment Using Administrative Data” Policy Research Working Paper Series. No. 5248, World Bank, Washington, D.C. and Yamauchi, F. (2014) “An Alternative Estimate of School-based Management Impacts on Students’ Achievements: Evidence from the Philippines.” Journal of Development E ectiveness. 6, no. 2: 97-110. xxxviii WWW.WORLDBANK.ORG/PH Introduction Introduction nd Purpose Over the last decade, the Government of the Philippines has embarked on an ambitious education reform program to en- sure that all Filipinos have the opportunity to obtain the skills that they need to play a full and productive role in society. The government has backed up these reforms, particularly over the last ve years, with substantial increases in investment in the education sector. Between 2010 and 2015, spending on basic education increased by 60 percent in real terms, and per student funding levels have increased considerably. The reform program has halted a long-term decline in public basic education services. Most of the increases in education While there have been improvements in education sector spending have been used to address major shortfalls in outcomes, signi cant challenges remain if the government’s education services that had built up during the early 2000s. goals for the education sector are to be realized. Despite In particular, a massive program for improving school some recent progress, socioeconomic disparities in access infrastructure and a nationwide teacher recruitment drive to basic education opportunities persist. For example, in have signi cantly improved school conditions. Between 2010 2013, 82 percent of 12 to 15-year-old children from the and 2013, student-teacher ratios in public high schools fell wealthiest 20 percent of households attended high school from 38:1 to 29:1, and student-classroom ratios fell from 64:1 compared with only 53 percent of children from households to 47:1 over a similar period. in the poorest 20 percent. Moreover, elementary and high school completion rates are relatively low, with the latest Increases in the availability of key inputs have also signi - data suggesting that almost a quarter of all enrollees fail to cantly increased access to basic education. Kindergarten complete. These challenges are compounded by the low enrollment almost doubled in absolute terms between quality of the basic education system. Employers frequently 2008 and 2014, and the proportion of children of school age highlight the limitations of recent basic education graduates, attending basic education also increased. Poorer socioeco- and this points to the need to improve the quality of nomic groups have bene tted the most from this trend. For teaching and learning in elementary and high schools across example, the share of children of school age from the poor- the Philippines. est 20 percent of households who were attending school increased from 86 to 94 percent between 2002 and 2013. Key to tackling these challenges will be further increases in public education spending. While recent increases have been However, the reforms have so far led to only modest impressive, the Philippines still spends less on education improvements in learning outcomes. The Philippines last than many of its neighbors and many other middle-income participated in an international learning assessment in 2003. countries. In 2014, the Philippines devoted just under 3 The results showed that the quality of education in the percent of its national income to public basic education country was low, with only around one-third of elementary compared with an average of nearly 5 percent for lower- and secondary school students being able to reach the middle-income countries as a whole. Recent detailed studies lowest international benchmark in mathematics. Since 2003, of education spending in the Philippines have con rmed the the results from national examinations suggest that there need for more spending to enable existing schools to meet have been only limited improvements in levels of learning. national education norms and standards. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 1 Introduction additional teachers and the building of a similar number of new classrooms. It is therefore vital that the systems that allocate resources, build new infrastructure, and recruit new teachers work e ectively. Otherwise the success of the senior high school program and the whole K to 12 reform program will be jeopardized. The purpose of this study is to assess the quality of basic education services and the systems used to allocate and manage public education resources. While small-scale qualitative studies have been conducted to look at particular public funding ows, there has been no previous attempt to comprehensively assess the strengths and weaknesses of the systems that manage and govern the use of public Ensuring that public systems allocate and use resources education funding. This study aims to ll this gap and to e ectively will be a vital element in translating increases in provide detailed evidence on the extent to which these public sector investment into improved education outcomes. systems are e ective in handling key items of public A number of recent studies have highlighted weaknesses in spending. The study also aims to provide a snapshot of the these systems that may reduce the e ectiveness of recent availability and quality of key education inputs at the school and future increases in sector spending. In 2013, budget level and to evaluate whether these resources are distributed execution rates in the education sector were lower than equitably across schools. they were in 2009 and were very low for some key education inputs such as capital outlays, school maintenance, and operating expenses. Several factors have been identi ed M in Rese rch Questions as being behind these low execution rates including bottlenecks in the ow of funds, di culties at the school nd Ke Educ tion Inputs level in managing public funds, and inadequate coordination Tr cked between the relevant government agencies. The Philippines Public Education Expenditure and Ine cient and ine ective spending has compounded Quantitative Service Delivery Study (PETS-QSDS) emerged low execution rates to further weaken the impact of from a request to the World Bank from the Department increased education spending on outcomes. A recent of Education (DepEd) and the Department of Budget and public expenditure review highlighted some possible Management (DBM). DepEd and the DBM wanted the Bank areas of ine ciency in government education spending, to assess whether the large budgeted increases in public including the complexity of and lack of transparency in education spending were being used e ectively to improve fund allocations, rigid allocation norms, and poor school conditions in public elementary and high schools. The governance.1 It also concluded that more detailed research detailed design of the study was developed during a series was needed to estimate the magnitude of spending of consultations between the Bank’s study team and key ine ciencies and to identify where the main bottlenecks in representatives of DepEd, the DBM, the National Economic the government’s nancial managements systems lie. and Development Authority (NEDA), and the A liated Network for Social Accountability (ANSA) as well as with a The introduction of senior high school will only add to the range of academics working on the Philippines education existing pressures faced by public nancial management system. systems. In 2016, senior high school will be introduced, and approximately 1.5 million additional school places will need Through these consultations, it was agreed that the study to be provided for children entering Grade 11 for the rst would explore four broad issues associated with the delivery time. This will require the hiring of approximately 60,000 of public basic education services in the Philippines: 2 WWW.WORLDBANK.ORG/PH 1. Resource ow, management and control. What factors Figure 1: The Study Tracks a Large Proportion of the prevent or delay resources from reaching their intended National Education Budget destination in a timely and transparent manner? Breakdown of the Department of Education (DepEd) budget, 2015 2. Existence, use, and nancing of inputs at school. Do schools have access to essential inputs and how e ective are Basic education systems that govern their use? facilities 17% 3. Equity. How do the resources available to schools and the systems managing these resources di er among regions School MOOE 4% Teacher and socioeconomic groups? salaries 62% Teacher training 4. School performance and resources. How and why does resources 1% the performance of schools di er and what drives those di erences? Other DepEd spending 16% Both the consultations with key stakeholders and practical considerations resulted in an agreement to track four key areas of the central government budget as part of the study:2 Source: DepEd Statement of Appropriations, Obligations, and Disbursements, 2015 • New teachers and teacher salaries Note: Shares are based on 2015 appropriations. • Teacher training resources, speci cally those from the Human Resource Training and Development Fund (HRTD) A number of other resource ows were excluded from and In-Service Training (INSET) the study for practical reasons. DepEd has a number of special programs (such as multigrade programs and special • School maintenance and other operating expenses programs for science, arts, and sports) that were mentioned (MOOE) during consultations for the study as areas that might be • Infrastructure spending from the Basic Education Facilities tracked. However, they were excluded on the grounds that Fund and the School Building Program. they are small in terms of the overall allocation of resources and that their impact on access and quality across the entire The funds tracked by the study cover the bulk of national school age population was likely to be limited. Furthermore, government funding for basic education (Figure 1). However, including these programs and their limited number of the tracking of teacher remuneration di ered from more bene ciary schools would have added to the complexity traditional public expenditure tracking exercises. It focused and size of the school sample for the study. Other research on whether teachers in schools were paid on time and in projects such as the series of program reviews that DepEd is full rather than on identifying whether all teachers on the currently undertaking may be better suited to assessing the payroll were teaching in their assigned schools.3 Given the e ectiveness of these programs. importance of recruiting new teachers to address existing shortfalls and to sta the senior high school program, the study also focused directly on hiring and deployment systems. Stud Appro ch In addition to looking at central government resource A clear diagnosis of the systems governing resource ows, the study also tracked local government support to allocation decisions is a necessary rst step to increasing basic education. The Special Education Fund (SEF) is the the e ciency of public spending.4 This study combined source of the majority of local government spending on the tracking of public resources with an assessment of the basic education and is funded by a 1 percent surcharge on quantity and quality of education services to: property taxes. In 2013, local governments contributed about 6 percent of total public education funding in the Philippines. • Examine the e ciency of public education spending and the incentives associated with service delivery. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 3 Introduction • Identify the main weaknesses in education governance of Public Works and Highways (DPWH). In most cases, and develop strategies to increase the e ectiveness of questionnaires were prepared for the central, regional, and education spending. division/district o ces of these departments (Table 2). The • Support e orts to make the resource allocation system questionnaires were based on an analysis plan prepared more accountable and performance-oriented. by the study team through a series of eld visits to many areas in the Philippines and interviews with key respondents • Act as a baseline to assess the impact of future education to identify the information needed to answer the study’s reforms.5 research questions. The team pre-tested and piloted the Many countries have undertaken similar public education questionnaires as part of the nal preparations for the survey. expenditure and quantitative service delivery studies (PETS- While the responses to this set of questionnaires were the QSDS) as can be seen in Table 1. This study draws on the main data source for the study, the team also collected tools and experiences from those many earlier studies as information from the central o ces of key government well as from a series of notes produced by the World Bank as agencies (such as DepEd and the DPWH) as well as from a guidance for implementing studies of this kind.6 number of other sources to present a fuller picture of the ow of education funds.7 A key objective of this study was to track the ow of national and local government education funding all the way down The Research Center for Teacher Quality at the Philippines to schools (Figure 2). In order to do this, it was necessary Normal University prepared a set of teacher assessments for rst to identify appropriate questions and then to prepare Grade 6 and 10 teachers for the study. These assessments questionnaires to collect data on the release of funds from included subject-based content assessments to measure all involved agencies, including the various o ces of the teachers’ knowledge of K to 12 curriculum content in Department of Education (DepEd), the Department of English, Filipino, Mathematics, and Science. They also Budget and Management (DBM), and the Department included a short version of the Philippine Government’s Table 1: Examples of Education Tracking and Service Delivery Surveys Country Year Objectives Zambia 2002, 2006, To determine the availability of key education inputs at the school level and track school and 2014 grants from the central government to schools Uganda 1996 To assess why increased public expenditure has not led to improved outcomes Mongolia 2006 To assess equity and regional disparities in the quality of education Cambodia 2005 To assess the e ectiveness of the Priority Action Program in delivering resources to schools in a timely manner Papua New 2012 To assess whether recent policy reforms and increases in spending have improved education Guinea outcomes Honduras 2001 To explore incentives for front-line service delivery workers and issues of teacher deployment, transfer, and attendance Bangladesh 2004/5 To track public expenditure from the central government to schools and students (for example, through stipend programs) and to explore the relationship between school resources, leakage of funds, and school performance Sources: Financial Management Reform Program (2005) “Social Sector Performance Surveys: Secondary Education in Bangladesh: Assessing Service Delivery” OPM and FMRP, Dhaka; Financial Management Reform Program (2006) “Social Sector Performance Surveys: Primary Education in Bangladesh: Assessing Service Delivery” OPM and FMRP, Dhaka; World Bank (2005). “Public Expenditure Tracking Surveys (PETS) in Primary Education” Washington D.C.; Howes, S, A. Mako, A. Swan, G. Walton, T. Webster, and C. Wiltshire (2014) “A lost decade? Service delivery and reforms in Papua New Guinea 2002 –2012”, The National Research Institute and the Development Policy Centre, Canberra; World Bank (2009) “Implementing Public Expenditure Tracking Surveys for Results: Lessons from a Decade of Global Experience. Washington D.C; and World Bank (2016). “Education Sector Public Expenditure Tracking and Service Delivery Survey in Zambia” Washington, DC. 4 WWW.WORLDBANK.ORG/PH Figure 2: Flow of Public Funds to Schools LEVEL BUDGET NATIONAL Department of Budget & Management Department of Education internal revenue operations budget GASTPE operations budget in-kind transfers allotment (textbooks) REGIONAL DBM Regional O ces DepEd Regional O ces PROVINCIAL/ city/municipality Local government unit DepEd School Division O ces own source DIVISION revenue province city/municipality own source SCHOOLS Private schools Public elementary Public secondary revenue Note: GASTPE - Government Assistance for Students and Teachers in Private Education. Teacher Strengths and Needs Assessment (TSNA) based municipalities out of 1,339 were sampled.10 With the exception on the National Competency Based Teacher Standards of Davao City, which was selected with certainty, all primary (NCBTS). The TSNA assesses teachers’ perceptions of their sampling units were selected with probability proportional pedagogical competence in the seven NCBTS domains; to their population size. In each city or municipality, up to (i) social regard for learning; (ii) learning environment; (iii) ve elementary schools and ve high schools were selected diversity of learners; (iv) curriculum; (v) planning, assessing, randomly from a list of eligible public schools.11 This sampling and reporting; (vi) community links; and (vii) personal growth approach yielded a nationally representative (excluding and development.8 ARMM) sample of 250 public elementary and 250 public high schools. Once the sample of elementary and high schools A nationally representative sample of public elementary had been selected, the team identi ed the relevant local and schools and high schools provided the starting point for national government o ces that supported these schools in the study.9 Schools in the Autonomous Region in Muslim order to interview them for the survey. Mindanao (ARMM) were excluded from the study because the public nancial management system in this region di ers from A random sample of kindergarten, Grade 6, and Grade 10 the norm, and a public education expenditure review had teachers was also selected to be interviewed. These sampled recently been completed for the region. A strati ed, clustered teachers also completed a subject content test and a sample was designed in which the primary sampling units strengths and weaknesses assessment. A random sample of were cities and municipalities. In total, 10 highly urbanized student households, strati ed by a student’s absence on the cities out of 34, (three in the National Capital Region, Davao day of the school visit, was also selected during the school City, and six others), 10 other cities out of 142, and 30 visits to verify student numbers and to collect information ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 5 Introduction Table 2: Survey Questionnaires, Contents, and Key Respondents Questionnaire Respondents Description of Types of Information Collected DBM regional o ces Chief budget specialist Fund releases for DepEd divisions and schools DepEd regional Regional director, payroll o cer, and The receipt, transfer, monitoring, and management o ces accountant of funds. Local governments Local school board member and Cash and in-kind resources provided to schools and in (provincial and city/ budget o cer coordination with DepEd. municipality) DepEd divisions School division superintendent, Division o ce facilities and personnel, characteristics of education supervisors, accountant, education services and nancing, teacher training activities, auditor, and other key sta new teacher hiring, detailed information on funding (such as MOOE) to the sampled schools. DPWH district Engineer in charge of DepEd projects Management and coordination with DepEd. Receipt engineering o ces and accountant of school infrastructure funds and status of projects in sampled schools. DepEd districts Public schools district supervisor Coordination and e ectiveness of local school boards, supervision and management of the provision of textbooks and learning materials. Schools School principal, custodian, physical School revenue and expenditure from all sources, facilities o cer receipt and use of school MOOE funds, details of school infrastructure projects, school characteristics and organization, student attendance, classroom conditions. Parent-teacher President Activities of the PTA and support to sampled school, associations revenue and expenditure accounts. Teachers (DepEd and Kindergarten, Grade 6, and Education, experience, training history, and salary LGU hired) Grade 10 teachers information. Teacher assessments Grade 6 and 10 teachers Subject knowledge (in English, Filipino, Mathematics, and Science) and teacher strength and needs assessments. Student households Household head Socioeconomic background, student education background, and the direct and indirect costs of education. on the socioeconomic status of students attending the team leaders were responsible for ensuring the quality of sampled schools. After the eldwork had been completed, the interviews and of the information collected by their appropriate sample weights were calculated for each level of teams. A series of quality control checks was included in each the survey. questionnaire to help enumerators and eld team leaders to identify any inconsistencies in the data collected. One eld Fieldwork for the study was undertaken between September supervisor was responsible for the survey in each of the three and December 2014. Enumerators were trained by a team island groupings of the Philippines, and they also checked from the World Bank and the survey rm over a two-week information from their teams regularly during the eldwork. period, which included training sessions held in the relevant In addition, a team of nine “back checkers” also carried out government o ces and schools in a pilot division. Field checks on 30 percent of all of the questionnaires completed 6 WWW.WORLDBANK.ORG/PH in the survey. Members of the survey team also conducted a consulted with key government agencies throughout the number of monitoring visits to sampled municipalities and implementation of the survey to facilitate their visits and to cities during the course of the eldwork to check the quality verify the collected data. of eldwork and to assess progress. During the preparation for the study, the team agreed with The nal sample ended up being similar in size to the original DepEd and the DBM to present its survey results in a series plan except in the case of public high schools and local of short policy notes that would each highlight a particular governments (Table 3). A number of sampled municipalities priority area for the government. It was also agreed that did not have ve public high schools and, in those cases, all the team would present the initial ndings and policy high schools in the municipality were selected. A number of recommendations in each policy note to stakeholders in local governments were also not interviewed, primarily either workshops in order to get feedback before nalizing each because they refused to be interviewed or because the note. These workshops took place between September 2015 key personnel were not available to provide the necessary and March 2016. They were attended by DepEd sta from the information. central, regional, and division o ces as well as by teachers and school principals. The DBM and NEDA sta responsible The rm implementing the survey encoded the data for education in the central and regional o ces attended collected from the eldwork using pre-prepared data entry the presentations and provided feedback. Other members of scripts. The resulting datasets went through a rigorous data the wider education community, including academics in the checking and cleaning process conducted jointly by the Philippines and researchers from other countries working on study team and the implementing rm. The nal cleaned education in the Philippines, also provided feedback on the datasets were available in November 2015. The survey team ndings contained in the notes. Table 3: Planned and Final Sample Sizes Questionnaire Planned sample size Final sample size DBM Regional O ces 16 16 DepEd Regional O ces 16 16 DepEd Division O ces 50 51 DepEd District O ces ** 113 DPWH District O ces ** 54 Provincial Local Governments* 30 27 City/Municipal Local Governments 50 47 Public Elementary Schools 250 249 Public High Schools 250 200 Teachers (Kindergarten and Grade 6) Max (1,250) 608 Teachers (Grade 10) Max (1,500) 946 Parent-teacher associations 500 449 Student households 2,500 2,189 Notes: * Since HUCs and other cities do not have a provincial-level local government unit (LGU), only 30 provincial LGUs were sampled. ** The sampling of district-level o ces depended on which schools were selected. All DepEd and DPWH o ces associated with the sampled schools were identi ed and interviewed. The nal sample size indicates the number that can be used for analysis after accounting for adjustments and non-responses. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 7 Introduction Outline of the Report The next chapter provides the context of the study and Policy Note 5: Providing Schools with Adequate Resources documents recent trends in basic education spending. It also to Deliver Quality Education reports on the shows in detail how recent spending increases have a ected provision and use of funds to cover school education outcomes. maintenance and operating expenses (MOOE). In line with agreements between the study team and key stakeholders while the study was being designed, the Policy Note 6: Assessing the Role Played by Local Government ndings are presented as a series of seven policy notes: in Supporting Basic Education reports on the funding provided for basic education by Policy Note 1: Assessing Systems for Hiring and Deploying local governments. Teachers reports on new teacher allocations, hiring processes, and salaries. Policy Note 7: Understanding the Drivers of Public School Performance and E ciency examines Policy Note 2: Developing a Pro cient and Motivated Teacher school e ciency and explores associations Workforce reports on the funds available between e ciency, performance, and the for and the provision of professional explanatory factors highlighted in the other development opportunities for teachers. policy notes. Policy Note 3: Building Better Learning Environments reports Finally, a short conclusion draws together the key ndings on school infrastructure spending. and policy recommendations contained in the separate policy notes. Policy Note 4: Assessing School-based Management evaluates how well school-based management institutions and processes are working. 8 WWW.WORLDBANK.ORG/PH Endnotes 1 World Bank and AusAID (2010). “Philippines Basic Education Public 6 The PETS-QSDS data portal is available at: http://pets.prognoz. Expenditure Review.” World Bank and AusAID, Manila. com/prod/Home.aspx. 2 School-based management grants were also included in the 7 See Annex 1 for a full list of these additional sources. study but relatively few schools receive these grants and the data 8 Full details of the instruments, their preparation, and eldwork that were collected as part of the study were not representative organization is available in RCTQ and SiMERR (2015) “PETS-QSDS of the grants as a whole. Final Report” Research Center for Teacher Quality, Manila. 3 This was partly the result of the team’s discussions with the 9 The study team developed the sampling approach in Department of Education and the Department of Budget and consultation with an international expert in sampling with Management on priorities for the study. Elementary school considerable experience of designing appropriate samples for teachers are also not allocated to speci c schools so it is not the Living Standards Measurement Survey and other public possible to track teacher posts in the way that other studies of expenditure tracking and service delivery surveys. this kind have done. 10 Highly urbanized cities are cities with populations of more than 4 Dehn, J., R. Reinnikka, and J. Svensson. (2003). “Survey Tools for 200,000 and with average revenues of at least PHP 50 million in Assessing Performance in Service Delivery” in F. Bourguignon and 1991 prices. Other cities are de ned as cities that do not meet L. A. Pereira da Silva Evaluating the Poverty and Distributional the criteria to be classi ed as highly urbanized. Municipalities are Impact of Economic Policies. World Bank and Oxford University administrative units for all other areas in the Philippines. Press, Washington D.C.; Gauthier, B. (2006). “PETS-QSDS in Sub- Saharan Africa: A Stocktaking Study.” World Bank, Washington D.C.; 11 Integrated schools were not included in the sample frame – there and Savedo , W. (2008). “Public Expenditure Tracking Surveys: were 164 schools of this kind in 2013/14 classi ed as elementary Planning, Implementation, and Uses.” Social Insight, Maine. schools and 568 at the high school level. Elementary or high schools that did not o er all grades were also excluded as were 5 A more detailed description of the survey methodology and 350 special high schools that followed a di erent curriculum. approach appears in Annex 1. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 9 B sic Educ tion Spendin Trends nd Outcomes Introduction Over the last decade, the Government of the Philippines has embarked on an ambitious education reform program to ensure that all Filipinos have the opportunity to obtain the skills that they need to play a full and productive role in society. The government has backed up these reforms, particularly over the last ve years, with substantial increases in investment in the sector. As a result, access to basic education has increased, particularly for the poorest households, and the overall learning environment has improved. While these improvements have halted a long-term decline in education sector outcomes, signi cant challenges remain. In particular, the share of national income devoted to basic education has only recently returned to the levels of the early 2000s, and spending per student in the Philippines is still low compared to spending in other middle-income countries. These relatively low levels of spending per student are likely to have limited the scope of the improvements in education outcomes over How H s Public Educ tion the last decade. High school completion rates remain Expenditure Ch n ed low compared with other countries in the region, and enrollment gaps among di erent socioeconomic groups Over the L st 12 Ye rs? persist. Stakeholders frequently express concerns about the quality of the education provided and about the level In the rst half of the 2000s, government spending on and breadth of skills that children possess when they leave basic education fell in real terms. Between 2003 and 2005 school. basic education spending fell from PHP 166 billion to PHP 152 billion (in 2014 prices). The share of national income The purpose of this note is to investigate whether the devoted to education also fell from 2.4 percent to 1.9 recent increases in the nancing for basic education have percent over the same period. Moreover, the decline in improved education outcomes and to identify any factors education spending led to signi cant drops in the level that may be constraining the e ectiveness of public of per student funding. Between 2003 and 2005 average spending. It concludes that further increases in public annual spending per public elementary and high school spending will be needed if the government’s ambitious student fell from PHP 9,500 to PHP 8,700 in real terms.1 goals for the education sector for the next ve years are to be achieved. However, increased investment will not be This downward trend was halted in 2005, and since enough on its own. It will also be necessary to ensure that then government spending on basic education has these resources are used e ectively. increased considerably. Between 2005 and 2014, national ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 11 Basic Education Spending Trends and Outcomes Figure 1: Public Education Spending Has Risen Rapidly in Recent Years Government Spending on Basic education, 2003 to 2015 (in 2014 constant prices) 400 18.6 20% 350 16.9 16.4 17.0 16.6 17.9 16.0 16.4 16.1 15.2 15.1 15.1 15.3 300 15% PHP billion (2014 prices) 250 200 10% 150 100 5% 2.4 2.1 1.9 1.9 2.1 2.0 2.2 2.2 2.3 2.1 2.3 2.2 2.6 50 0 0% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 National government National govt. as % of GDP (RHS) Local government National govt. as % of total govt. spend Sources: National government spending – DBM National Expenditure Program. Local government education spending – Bureau of Local Government Finance. GDP and GDP de ator - Philippines Statistics Authority Notes: Basic education spending data are obligations except for 2015 where adjusted amounts are reported. Basic education refers to the DepEd budget and obligations for the Philippines Science High School. Local government spending gures are obligations but from 2014 only include spending from the Special Education Fund. The total government expenditure gures used to calculate the share of education spending exclude interest payments. government spending on basic education more than impressive, the share of national income devoted to basic doubled in real terms. The share of national income devoted education has only recently returned to the levels of the to education also increased and, by 2014, had reached 2.2 early 2000s. Budget forecasts predict that the share of percent of GDP. Spending per student followed a similar national income devoted to basic education in 2015 will have pattern; by 2013 average spending per public elementary exceeded 2.4 percent for the rst time since 2003. and high school student was PHP 12,800 in real terms, an increase of almost 50 percent from the level in 2005. Spending by local governments makes up only a small proportion of overall basic education spending in the The underlying factors driving the increases in basic Philippines and has not changed much in recent times education spending can be broken down into two distinct (Figure 1).2 Local government spending mostly comes periods. Between 2005 and 2009, education spending rose from the Special Education Fund (SEF), which is nanced even though the share of education in total government from a 1 percent surcharge on property taxes.3 Since 2003, spending declined from 17 percent to 15 percent (Figure 1). local governments have contributed an annual average of This suggests that the budget increases were driven by rapid PHP 16 billion in basic education funding, but since central increases in the overall government budget rather than any government spending has increased rapidly over the renewed commitment to education itself. After 2009, the same period, local government funding as a share of total drivers of increased spending are less clear though overall education spending has declined rapidly. Between 2003 and budget growth and increases in the share devoted to the 2014 the share of local government spending in total public education budget both played a role. However, the share of education spending fell from 9 percent to 5 percent. the total budget devoted to basic education in 2014 was still lower than the share that prevailed in the early 2000s. Geographical disparities in public basic education spending are quite large and are not strongly associated with levels While the turnaround in government spending has been of poverty.4 For example, spending per school-aged child 12 WWW.WORLDBANK.ORG/PH in the Philippines varies from around PHP 4,500 in the Estimates for 2006 based on available sources suggest that National Capital Region (NCR) to just over PHP 7,600 in the household spending might be as high as one-third of govern- Cordillera Administrative Region (Figure 2). Moreover, on ment education spending.7 Partnerships between schools and some disparity measures, these regional di erences appear private sector corporations or NGOs may also be a signi cant to have increased between 2005 and 2012.5 Regional and source of funding, but little consistent information is avail- provincial poverty levels can sometimes provide a useful able on these contributions. A recent survey of elementary proxy for the size of educational de cits with those regions and high schools found that community and other non- or provinces with higher levels of poverty being likely to government sources contributed approximately 13 percent of have weaker education outcomes and thus to require greater average school-level nances in 2010.8 education investments. However, in the Philippines, levels of government spending on basic education are only weakly Despite recent increases in public education spending, the correlated with regional and provincial levels of poverty. Government of the Philippines still devotes a smaller share For example, Region 12 is one of the poorest regions in the of its national income to basic education than other middle- Philippines but receives a budget for basic education that is income countries and its regional competitors (Figure 3).9 below the national average (Figure 2). In 2014, it devoted just under 3 percent of GNP to public education spending compared to a 2012 average for lower- The public basic education system is also funded by contribu- middle-income countries of nearly 5 percent. In East Asia, the tions from households and the private sector. Only limited Government of the Philippines devotes a similar share of its information is available on household contributions to the national income to education as Cambodia but signi cantly public education system, but the available evidence suggests less than Malaysia, Thailand, and Viet Nam. The low share of that these contributions may be considerable. One limitation national income devoted to education is partly the result is that household surveys cannot distinguish spending on of the lower priority given to education by policymakers in di erent education levels or on public versus private schools.6 the Philippines than in some other countries in the region Figure 2: There are Large Disparities in Education Spending across the Philippines Per Capita Government Spending on Basic Education by Region and Province and Poverty Incidence, 2012 Regional Provincial 40% 60% XII VIII IX X 50% CARAGA Poverty incidence, 2012 (%) Poverty incidence, 2012 (%) 30% V average XI VII 40% poverty IV-B incidence VI 20% CAR 30% II average poverty incidence I 20% 10% II IV-A 10% average per-capita average per-capita NCR spending spending 0% 0% 4,000 5,000 6,000 7,000 8,000 3,500 5,500 7,500 9,500 Per capita basic education appropriations, 2012 Per capita basic education appropriations, 2012 Sources: Poverty incidence and population data come from the Philippines Statistical Authority. Data on appropriations come from the General Appropriations Act. Note: Per capita basic education appropriations are nominal appropriations per child aged between 5 and 19. Poverty incidence and appropriations data are for 2012 and population data is for 2010. Autonomous Region in Muslim Mindanao is not included. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 13 Basic Education Spending Trends and Outcomes (Figure 3). However, it is also due to the lower share of total education quality. An earlier study by the World Bank and government spending in national income in the Philippines AusAID also showed that it would take more than 6 percent of than in other comparable countries. GDP to implement a broad package of quality improvements coupled with similar enrollment expansion targets as the Low overall public education spending has also translated Department of Education study.11 into lower levels of per student spending than in other developing economies. The Philippines spends less per The resources that actually reach schools also tend to be lower student as a share of per capita GNP than most middle- than national spending aggregates show. Detailed information income countries (Figure 4). Per capita di erences also mask collected for the PETS-QSDS study at the school level show signi cant di erences in absolute levels of spending. For that annual school revenue in 2014 was approximately PHP example, based on purchasing power parity conversions, 11,400 (US$262) for elementary schools and PHP 12,400 the Philippines spent only US$380 per elementary student (US$286) for high schools (Figure 5). The vast majority of compared to US$760 in Viet Nam and US$2,350 in Malaysia. school level revenue comes from the national government. In 2014, 92 percent of elementary school and 94 percent of high A recent study by the Department of Education (DepEd) school funding came from the national government budget. explored the question of the resources that will be needed Conversely, only around 3 percent of total school revenue to fund various di erent expansion and quality improvement comes from local school boards or local governments. scenarios associated with achieving government goals for Comparing per student revenue at the school level that is the sector.10 The study highlighted the need for a larger share provided by DepEd and the overall national education budget of GDP to be spent on basic education if it is to be made suggests that about 80 percent of the national education accessible to all and to ensure modest improvements in budget actually reaches the school level. Figure 3: The Philippines Spends Less on Education than Many Other Countries Public Spending on Education as a Share of GNP and Total Government Spending, Selected Countries, 2012 Viet Nam Thailand Malaysia Republic of Korea Thailand Viet Nam High income Malaysia Higher middle income Singapore Republic of Korea Indonesia Lower middle income Philippines Low income Lower middle income Singapore Low income Lao Higher middle income Indonesia Lao Cambodia Cambodia Philippines High income 0 2 4 6 8 0 10 20 30 40 Percentage of GNP Percentage of total government expenditure Basic Tertiary Sources: Figure 1 and UNESCO Global Monitoring Report 2015 Notes: Information for all countries is 2012 or latest year available after 2009. Philippines data are for 2014. Total government expenditure includes interest payments. 14 WWW.WORLDBANK.ORG/PH Figure 4: Spending Per Basic Education Student in the Philippines Is Also Low Government Spending on Education as a Share of GNP, Selected Countries, 2012 30 Primary Spending as % of GNP per capita 25 Secondary 20 15 10 5 0 Low income (1,220) Philippines (4,380) Singapore (60,110) Indonesia (4,730) Lower middle income (3,875) Malaysia (16,270) Upper middle income (10,870) Thailand (9,280) High income (31,065) Viet Nam (3,620) Republic of Korea (30,180) Sources: Authors’ calculations and UNESCO Global Monitoring Report (2010). EFA Global Monitoring Report. Education For All 2000-2015, Achievements and Challenges. Paris, UNESCO. Notes: Information for all countries is from 2012 or the latest year available after 2009. Philippines data are for 2013 and include all spending (recurrent and capital) whereas other countries only cover recurrent spending. Figures in parentheses are GNP per capita in 2012 based on purchasing power parity. The proportion of funding that schools have discretion teacher and other sta salaries, infrastructure projects and over tends to be relatively low. Elementary schools in the other material support schools receive. While information Philippines control about 7 percent of the resources that at the school level from other countries is relatively rare, they receive compared to 12 percent for high schools. The Indonesian primary schools have control over about 18 remaining revenues are in-kind support either in the form of percent of the total revenue.12 Figure 5: Overall Spending at the School Level Is Lower than Aggregate Figures Suggest School Revenue Per Student by Type of Revenue, 2014 Per school (PHP 000s) Per student 10,000 12,500 7,500 5,000 10,000 2,500 0 7,500 Elementary School High School Elementary School High School DepEd Barangay Local school boards Other Source: PETS-QSDS national survey – school level. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 15 Basic Education Spending Trends and Outcomes Wh t H s the Incre sed Preliminary data from DepEd suggests that between 2011 and 2014 the proportion of schools operating more than one Investment in B sic shift fell from 11 percent to 6.5 percent.14 Nevertheless, the average student-classroom ratios in both elementary and Educ tion Been Used For? high schools remain high. A large proportion of the increased spending on basic While the overall share of the education budget devoted to education has been used to increase infrastructure salaries has declined, spending on teachers has increased investment and provide schools with essential inputs to in real terms. In 2005, approximately 90 percent of all basic improve the quality of education. These increases re ect education spending was spent on salaries, but by 2013 the government’s commitment to address the deteriorating salary spending accounted for only 81 percent.15 Despite condition of basic school infrastructure, the lack of adequate this declining share, salary spending increased in real terms teachers, and the limited availability of good quality from PHP 135 billion to PHP 217 billion between 2005 and textbooks and instructional materials. 2013 (2014 prices). These increases were largely the result of schools hiring new teachers to ensure that they had Capital spending increased from PHP 4 billion to PHP 24 adequate teachers and to reduce class sizes. New teacher billion in real terms (2014 constant prices) between 2005 hiring has had the greatest impact at the high school level and 2013, and its share in total spending increased from 2 where it has resulted in a decline in the student-teacher ratio percent to 9 percent. In practice, this yielded a large increase from 40:1 in 2005 to 27:1 in 2014 (Figure 6).16 in the resources available for school construction and repair. Between 2005 and 2013 spending on school construction There have also been large increases in spending on increased by a factor of ve in real terms from PHP 2.8 billion maintenance and operating expenses in schools. Between to PHP 14.6 billion.13 This trend mostly became evident 2005 and 2013 spending on maintenance and operating after 2010 and has been re ected in the declining student- expenses increased from PHP 12 billion to PHP 28 billion in classroom ratios in both elementary and high schools across real terms. The share of this spending in the total budget the country (Figure 6). Increased capital investment also increased from around 8 percent in 2005 to 10 percent reduced the proportion of public high schools that needed in 2013. At the same time, the government transferred a to schedule multiple shifts to accommodate students. greater share of operating funds directly to schools to enable Figure 6: Increased Education Spending Has Improved the School Learning Environment Public Student-teacher and Student-classroom Ratios, 2005–2013 Student-teacher ratio Student-classroom ratio 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 2004 2006 2008 2010 2012 2014 2004 2006 2008 2010 2012 2014 Elementary School High School Elementary School High School Sources: DepEd Fact Sheet and Enhanced Basic Education Information System (EBEIS) Notes: Figures only include public enrollment and teachers although over this period local government and volunteer teachers also taught in public schools. For example, prior to 2012, all kindergarten teachers in public schools were classi ed as volunteer teachers and were not included in the reported data. Classroom ratios are based on the total number of rooms used for teaching rather than on the total number of rooms in the school. 16 WWW.WORLDBANK.ORG/PH Figure 7: Regional Disparities Exist in Levels of Financing and Key Education Inputs Student-teacher and Student-classroom Ratios and Per Capita Basic Education Spending, 2012 Student-teacher ratio Student-classroom ratio 40 80 IV-A NCR average per-capita VII 70 spending 35 NCR XI 60 XII II V 30 50 XI CARAGA XII IV-A V VI II 40 average student- VIII I VI I CAR 25 VII teacher ratio 30 II CAR average student- VIII II 20 20 classroom ratio CARAGA average per-capita 10 spending 15 0 4,000 5,000 6,000 7,000 8,000 4,000 5,000 6,000 7,000 8,000 Average basic education spending per capita Average basic education spending per capita (students aged 5–19), 2012 (students aged 5–19), 2012 Sources: DepEd Education Basic Education Information System (EBEIS) and DBM GAA Notes: Per capita basic education spending is total regional appropriations (nominal) per child aged between 5 and 19. Autonomous Region in Muslim Mindanao (ARMM) is not included. them to make their own spending decisions. Speci cally, of the poorest input ratios as well as one of the lowest per the government introduced the Boncodin formula in 2013 capita allocations for basic education.17 by which maintenance and operating expenses are now allocated to schools on the basis of student numbers and other school characteristics and provided to them directly. How H s the Incre sed As a result, the amount of maintenance and operating funds allocated directly to schools nearly trebled in real terms Investment Affected between 2005 and 2013 from PHP 4 billion to PHP 12 billion. Educ tion Outcomes? Over the same period, spending on textbooks and other instructional materials also increased, indicating the high This section explores how the increased availability of key priority that DepEd has given to critical infrastructure and inputs has a ected educational outcomes. It is important quality inputs for the sector over the last decade. to keep in mind that it may take time before recent improvements in school learning environments are re ected The nationwide improvements documented in this section in national education achievement averages. For example, conceal signi cant disparities across regions in key education inputs (Figure 7). Region IV-A and NCR tend to have high the latest available information on national examination student-teacher and student-classroom ratios, while those in results is for students who started school at a time when CAR tend to be relatively low. This partly re ects di erences the education system was only beginning to recover from in population density, with NCR and Region IV-A being a long-term decline in academic outcomes. Later cohorts mainly urban areas with very large schools compared to of children are likely to bene t to a greater extent from CAR, which is mainly rural with many small schools. However, improved funding, and it is reasonable to expect that their di erences in key inputs are also driven by di erences in per learning outcomes will be better than those achieved by capita budget funding. For example, Region IV-A has some students leaving elementary and high school today.18 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 17 Basic Education Spending Trends and Outcomes Access nd School Completion R tes 2000s, elementary net enrollment rates continued to fall as population increases outpaced enrollment growth. However, Enrollment in kindergarten has expanded rapidly since since 2008 elementary enrollment rates have been on an the mid-2000s, which has largely been the result of upward trend (Figure 8). Between 2009 and 2012, more than increased public sector provision. In many countries, half a million additional children enrolled in elementary early childhood education has yielded higher learning school, and the net enrollment rate increased from 90 achievement and improved life outcomes for children percent to 95 percent. when they enter the labor force. In the Philippines, kindergarten enrollment almost doubled between Over a similar period, high school enrollment also expanded, 2005 and 2013, and approximately two-thirds of all and by 2013 some 300,000 additional children were continuing ve year olds are now attending kindergarten.19 The their education beyond elementary school compared to 2005. largest increases in kindergarten enrollment have been This resulted in the net enrollment rate increasing from 60 concentrated among the poorest and most vulnerable percent to 65 percent between 2009 and 2013.20 The Philippines sections of society. In 2008, the gross enrollment rate in compares favorably with other countries in terms of enrollment kindergarten for the poorest 20 percent of the population in the rst four years of secondary schooling. However, most was 33 percent, but this had increased to 63 percent by other middle-income countries have already extended 2013. Levels of kindergarten enrollment in the Philippines secondary education level to encompass six years, which the now compare favorably with rates in other middle- Government of the Philippines plans to follow in 2016. income countries both within the region and globally. Although most recent enrollment gains have been On the whole, elementary and high school enrollment concentrated among the poorest children, socioeconomic rates have been on an upward trend since the end of disparities remain large at the high school level. World Bank the rst decade of the 2000s. At the beginning of the estimates using data from the Annual Poverty Indicators Figure 8: Access and Attainment Indicators Have Shown Modest Improvements Recently Gross and Net Enrollment Rates and Completion rates, 2005-2012 Enrollment rates Completion rates 120% 80% 100% 75% 80% 70% 60% 65% 40% 60% 20% 55% 0% 2005 2006 2007 2008 2009 2010 2011 2012 2005 2006 2007 2008 2009 2010 2011 2012 50% 2005 2006 2007 2008 2009 2010 2011 2012 Elementary School High School Net enrollment rate Gross enrollment rate Elementary School High School Source: DepEd (2015). DepEd Fact Sheet 2013. DepEd. Manila. Note: The completion rate is the percentage of rst year entrants who complete the full education cycle. 18 WWW.WORLDBANK.ORG/PH Survey (APIS) show that high school net enrollment rates for which makes these increases di cult to interpret. High those in the poorest quintile increased from 45 percent in school achievement results have remained relatively stable 2002 to 53 percent in 2013, while rates remained relatively over the last 10 years, although overall scores are only slightly stable for the wealthier groups in the population. Di erences above 50 percent. While issues of comparability make it in net enrollment rates across the income distribution have di cult to make strong statements about trends, the national narrowed considerably at the kindergarten level and have achievement scores do not suggest that there has been any been virtually eliminated at the elementary level. However, signi cant improvement in overall learning achievement in socioeconomic disparities in enrollment at the high school recent times. level remain high. In 2013, the high school net enrollment rate for children in the poorest 20 percent of households was Di erences among regions in elementary and high school only 53 percent compared to a rate of 81 percent for those in student achievement are relatively small except in a limited the wealthiest household quintile. number of regions (Figure 9). Students in Caraga tend to perform better than most students in the Philippines while After some initial improvement, school completion rates students in ARMM score relatively poorly. Di erences among have remained relatively stagnant since 2007 (Figure 8). regions tend to be similar across elementary and high It is important to note that completion rates re ect past schools with some exceptions. Most notably, elementary investments in education. The 2012 data shown in Figure 7 school students in NCR rank 16 out of the 17 regions, while relate to students who entered the system in 2006 before high school students rank seventh. There is little correlation many of sector improvements outlined in the previous between levels of student achievement and per capita section came to fruition. The cohort survival rate is a more spending on basic education. In some cases (for example, synthetic measure of school completion and re ects Caraga), basic education spending and learning achievement changes in the education system more quickly.21 However, is high and in other cases achievement is low despite above cohort survival rates have also remained stable since 2005, average basic education spending (for example, in Region I). with approximately three-quarters of all students starting elementary or high school being expected to complete the cycle. Recent studies have pointed to high repetition rates, How Does the Public particularly in Grade 1, as a major explanatory factor for why one in four children drop out of school before completing.22 Fin nci l M n ement S stem Affect the Link Educ tion Qu lit between Spendin nd The Philippines last participated in an international learning assessment in 2003.23 The results showed that only around Educ tion Outcomes? one-third of elementary and secondary school students Whether the public nancial and resource management were able to reach the lowest international benchmark in systems allocate resources e ectively is a vital element mathematics. Di erences among socioeconomic groups were also stark. In secondary school, only 23 percent of the in translating increases in public sector investment into poorest children achieved the lowest benchmark compared improved education outcomes. These systems transform to 60 percent of the wealthiest children. sector objectives and policies into budget allocations that are used to nance the inputs and programs necessary to deliver Since 2003, national examination scores have improved but good quality education services. The overall level of funding, it is unclear whether these improvements have changed the level of administration or agency using resources, and the overall picture on learning signi cantly. Average scores the management and governance arrangements for these on the national achievement test rose signi cantly between resources are key determinants of education outcomes. A 2006 and 2008, reaching a high of 69 percent, but have number of recent studies have pointed to weaknesses in these remained stagnant since.24 However, it is not clear how systems in the Philippines that may limit the overall impact of comparable these achievement tests are from year to year the recent increases in sector spending. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 19 Basic Education Spending Trends and Outcomes Figure 9: Differences in Learning Achievement across Regions Are Generally Small National Achievement Test Scores by Region (%), 2011 Grade 6 Year 4 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 ARMM NCR IV-A I CAR VII VI XI II IX X III XII IV-B VIII CARAGA ARMM I IV-A II XII XI IX X CAR NCR VI III IV-B VII VIII CARAGA V V Source: DepEd Not all of the increased government allocation to the than three-quarters of the funds allocated to this category education sector has been spent in recent years because of in 2013 were actually spent. Given that this budget category low budget execution rates. The overall execution rate for provides local education o ces and schools with funding for basic education has uctuated over the last 10 years, but in important activities such as teacher training and the purchase 2013 it was lower than in 2004 (Figure 10). Execution rates of school supplies, this represents a missed opportunity to vary among key education inputs. First, rates for maintenance raise education outcomes further. Second, approximately one- and operating expenses have been falling since 2009. Less fth of all capital outlay resources, including those for school Figure 10: Budget Execution Rates Are Low for Some Key Categories of Expenditure Budget Execution Rates for Basic Education Spending, 2006-2013 100% 90% 80% 70% 60% 50% 40% 2006 2007 2008 2009 2010 2011 2012 2013 Personnel Capital Outlays Maintenance and Operating Expenses Total Source: DepEd Statement of Allotments, Obligations and Balances (SAOBs) for each year. Note: Budget execution rates are de ned as obligations as a percentage of allotments. 20 WWW.WORLDBANK.ORG/PH construction and rehabilitation, go unspent annually. This is a increased education spending on outcomes. A recent real cause for concern given the high student-classroom ratios, public expenditure review by the World Bank and AusAID particularly in high schools, and the general poor state of also highlighted some potential areas of ine ciency in much of the existing infrastructure (see Figure 6).25 government education spending that included:29 Several factors are likely to contribute to these low execution • Complexity and lack of transparency in administrative rates. A recent review by the World Bank and AusAID of procedures for fund allocation. Fund ows in the education existing studies both in the Philippines and in other countries system are complex, and key sta are frequently unaware has identi ed a number of factors that may lead to low of the rules used to allocate resources. For example, budget execution rates: 26 school principals are typically unaware of the formula that determines how their maintenance and operating • Delays in fund releases and allocations. Delays in the expenses are allocated. This limits accountability and release of allocations for some budget items mean that prevents schools from being able to plan e ectively. local education o ces and schools often receive funds late or not at all. When allocations are received late, they • Rigid norms for the allocation, release, and use of resources. have insu cient time to follow procurement rules and School principals frequently complain about the properly account for their use of funds before the end of restrictions imposed on the use of their operating funds. the scal year.27 The prohibition on using these funds for laptops, LCD projectors, and other instructional equipment because • Incomplete transfer of funds between levels. In other cases, they are classi ed as capital goods unnecessarily restricts there are delays in the transferring of funds from regional and division o ces of DepEd to schools. For example, in school autonomy. Rules for authorizing new classroom 2011, a small-scale survey showed that approximately construction are in exible and frequently do not a quarter of schools did not receive any resources for adequately identify the needs of schools in remote and maintenance and other operating expenses (MOOE) mountainous regions. These rules may be a signi cant despite increases in overall school allocations.28 factor in the regional inequalities in student-classroom ratios documented in Section 3. • Di culties at the school level in reporting on self-managed funds. Schools frequently nd it di cult to account for how • A lack of vertical coordination between national and local they have used the monthly funds that they receive to government education spending. There is relatively little cover their maintenance and operating expenses. This is coordination between central and local government particularly the case for schools in remote areas and with no funding of basic education. While relatively little is known dedicated accounting sta . When they are unable to pro- about this, there is signi cant potential for duplication duce these reports, the DepEd division o ce must withhold and wastage of resources due to this lack of e ective any subsequent funds, which are often left unspent. coordination, which is likely to be another factor driving regional inequalities in spending and ultimately in • Inadequate coordination between implementing units. A signi cant part of the basic education budget for school education outcomes. construction and repair is implemented through the • School governance. How schools manage their nancial Department of Public Works and Highways (DPWH). This and human resources has been shown to be a key has signi cantly increased the workload of the agency determinant of the e ectiveness of public spending. and has emphasized the need for close collaboration Schools that are managed and governed well tend to with DepEd. In some cases, the absence of this necessary make better decisions and have better student outcomes. coordination has led to di erences in budgeted and actual Evidence suggests that, while some elements of e ective funding needs, which have in turn led to project delays school-based management are in place in schools in the and poor budget execution. Philippines (such as school improvement planning), others Ine cient and ine ective spending can compound are still lacking (such as community participation and low execution rates to further weaken the impact of autonomy).30 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 21 Basic Education Spending Trends and Outcomes Summ r The Government of the Philippines has made signi cant e orts to increase public investment in education to expand access and raise education quality. This note has shown that recent increases have led to modest improvements in key education inputs and outcomes. However, it has also found that additional investment will be needed to build on these recent gains, particularly in schools and locations serving poor and disadvantaged communities. Moreover, recent studies have raised concerns about the e ectiveness of existing systems to allocate and manage the increased resources. In particular, the World Bank’s recent public education expenditure review concluded that merely increasing allocations to the sector is unlikely to produce signi cant improvements in outcomes. These ndings suggest that tackling spending ine ciencies has the potential to raise education outcomes both by using existing resources better and by ensuring that further funding increases are used e ectively. While recent studies have highlighted many of the potential constraints to more e cient resource use, more detailed analytical work is needed to explore the main drivers of e ciency and inequality and to identify the changes that will be needed to increase the e ectiveness of public education investments. 22 WWW.WORLDBANK.ORG/PH Endnotes 16 While di erences between regions in student-teacher and student-classroom ratios are wide, they have largely narrowed over the last 10 years. In particular, inequalities in student-teacher 1 Per student spending includes kindergarten. ratios at the high school level and student-classroom ratios at the 2 Local governments support the delivery of basic education elementary level have declined relatively rapidly since 2006. services by establishing and funding local school boards. 17 Previous analysis conducted by the World Bank and AusAID in 3 National guidelines require that SEF funds be used for school 2010 showed that adequate funding of teachers and classrooms maintenance and operating expenditures, construction, and is a key determinant of achievement in the Philippines context. sports. Some local governments also use their general funds See World Bank and AusAID (2010), “Philippines Basic Education to provide additional nancial support to local school boards. Public Expenditure Review.” World Bank and AusAID, Manila. 4 Assistance from development partners is not generally 18 These are the author’s calculations using household data included in the spending data used to construct regional from the Annual Poverty Indicators Survey (APIS) for both the aggregates. “kindergarten” and “preparatory” categories. Calculations by the authors using DepEd kindergarten enrollment gures and 5 The standard deviation of per capita regional spending UNESCO’s population estimates show that the overall gross increased from 783 to 1,833 between 2005 and 2013. kindergarten enrollment rate increased from 52 percent in 2009 6 In 2012/2013, 91 percent of elementary school students to 99 percent in 2013. attended government primary schools and 80 percent 19 Net enrollment rates estimated by the authors using APIS attended government high schools. household data suggest a more moderate increase in high school 7 World Bank and AusAID (2010). “Philippines Basic Education net enrollment from 66 percent in 2008 to 68 percent in 2013. Public Expenditure Review.” World Bank and AusAID, Manila. 20 The cohort survival rate is the percentage of a cohort of students 8 World Bank and AusAID (2013). “School-based Management in the rst grade in a given school year that is expected to survive in the Philippines: An Empirical Investigation.” World Bank and to the last grade. AusAID, Manila. 21 World Bank and AusAID (2010). “Philippines Basic Education Public 9 GNP is signi cantly larger than GDP in the Philippines, Expenditure Review.” World Bank and AusAID, Manila. but country rankings remain unchanged if spending as a 22 See, for example, Mullis, I., Martin, M., Gonzalez, E. and S. proportion of GDP is used, with the exception of Cambodia, Chrostowski (2004). “TIMSS 2003 International Mathematics which then moves below the Philippines. It is preferable to Report.” International Study Center, Boston College, Lynch School use GNP because it allows the inclusion of the averages for of Education, Boston. country income groupings in Figure 3. 23 There have been some notable increases in achievement in some 10 DepEd (2013). “Medium-term Expenditure Framework for subjects. For example, average scores in grade 6 elementary Basic Education 2014-2020: Enrollment Projections and Cost science and English scores increased by 8 and 5 percentage Simulations under Alternative Scenarios.” Department of points respectively between 2007 and 2012. Education, Manila. 24 The execution rates in Figure 9 may be underestimated because 11 World Bank and AusAID (2010). “Philippines Basic Education capital allocations can be carried over from one year to the Public Expenditure Review.” World Bank and AusAID, Manila. following year. However, they provide a good measure of what 12 These are appropriations and exclude an additional PHP proportion of an appropriation tends to be used in a single year. 14 billion of school construction funds included in the 25 World Bank and AusAID (2010). “Philippines Basic Education Public Department of Public Works and Highways (DPWH) budget Expenditure Review.” World Bank and AusAID, Manila. for 2013. 26 This appears to be a particular issue with centralized lump sum 13 Approximately 3 percent of all public elementary schools releases such as the School-based Management Grants. organize their schedules around multiple shifts, and this has not changed much in recent years. 27 World Bank and AusAID (2013). “School-based Management in the Philippines: An Empirical Investigation.” World Bank and 14 According to UNESCO statistics, the proportion of the total AusAID, Manila. basic education budget in the Philippines devoted to salaries is similar to those in other neighboring countries. 28 World Bank and AusAID (2010). “Philippines Basic Education Public Expenditure Review.” World Bank and AusAID, Manila. 15 Student-teacher ratios at the elementary level remained stable because increases in the number of teachers were 29 World Bank and AusAID (2013). “School-based Management o set by increases in student numbers during the in the Philippines: An Empirical Investigation.” World Bank and same period. AusAID, Manila. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 23 Polic Note 1: Assessin S stems for Hirin nd Deplo in Te chers Introduction Over the last decade, research from many di erent countries has demonstrated the important role played by teachers in improving students’ learning and increasing their competencies.1 Studies from countries as di erent as the US and Indonesia have shown the enormous bene ts that follow from having adequate and e ective teachers working in a country’s schools. In Indonesia, a value-added analysis of student learning outcomes found that the more teachers know, the greater the improvements in the learning competencies of primary and junior secondary students.2 In the US, better teaching in elementary and secondary schools has been shown to increase students’ college participation rates, raise their subsequent earnings, and improve other long-term outcomes.3 Ensuring that schools have enough teachers to provide educa- tion in classes that are small enough to foster a good learning environment is an important rst step. While the optimal size of classes and the impact of reducing class sizes continue to be debated, it is clear that very large class sizes are detrimental to learning. Moreover, evidence from recent impact evalu- ations in developing countries shows that, on the whole, reducing class sizes in elementary schools improves learning.4 In the Philippines, the Department of Education (DepEd) has that already have a su cient number while others continue class size standards for each grade in elementary and high to experience shortages is ine cient and potentially wasteful. schools. Providing enough teachers to ful ll these standards in Conversely, distributing the teacher workforce e ciently and all schools can contribute to a good learning environment for equitably can both improve student learning and optimize the basic education students throughout the country. use of scarce public resources. The distribution of public school teachers is also a key determi- The purpose of this policy note is twofold. First, it aims nant of the e ciency of overall government spending. Public to assess the adequacy, e ciency, and distribution of spending on teacher salaries has increased dramatically in existing allocations of teacher positions across and within recent years. Between 2010 and 2014, teacher salary spend- schools. Second, it evaluates current systems for hiring and ing increased from PHP 103 billion to PHP 172 billion (at 2014 deploying teachers and identi es areas in which they can constant prices). By 2014, teacher salaries represented about be strengthened.6 The ndings of this note are based on a 9 percent of the national government budget of the Philip- comprehensive survey of the public education system in the pines.5 This puts into sharp focus how important an e cient Philippines that tracked public education expenditure and distribution of teachers can be. Allocating teachers to schools assessed the quality of education services. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 25 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers The note shows that substantial progress has been made in Level nd Distribution of Te chers the Philippines in terms of reducing teacher shortages and improving the systems used to hire and deploy teachers. At The Government of the Philippines has made enormous the school level, class sizes have been reduced signi cantly, strides in recent years in providing schools with su cient and teacher absenteeism rates are low compared with other teachers. A key indicator of the adequacy of the number of countries. Despite these successes, some ine ciencies in the teachers in the school system is the student-teacher ratio (STR). distribution of teachers and bottlenecks in the hiring process It measures the number of students per teacher and provides remain. In particular, the systems for hiring and deploying a very rough proxy of average class sizes. Over the last decade, STRs have been reduced considerably in basic education teachers need to be strengthened and need to be focused on (Figure 1). For example, between 2011 and 2014, the number of those schools and regions with continuing teacher shortages. students for every high school teacher fell from 37 to 27.7 This Also, there is a need to develop a system to govern the transfer and other decreases in the numbers of students per teacher of teachers between schools in order to reverse the growing were the result of an aggressive hiring policy in those three number of schools with surplus teachers. years, during which an additional 123,000 elementary and high school teachers were hired, the equivalent of an increase of 24 The Av il bilit of percent in the total stock of teachers. Te chers in Element r Despite achieving signi cant improvements in recent years, nd Hi h Schools the Philippines still has some of the highest student-teacher ratios in East Asia and the Paci c and among countries at a In order to operate e ectively, schools require an adequate similar stage of economic development. For example, in 2014 number of teachers that turn up regularly and on time to the high school STR in the Philippines was 27:1 compared to teach. Using the PETS-QSDS data, this section looks at trends a region-wide average of 16:1. Even among the world’s lower- in the overall level and distribution of teachers across basic middle-income countries, the Philippines does not compare education and patterns of teacher absenteeism across the well, with an average STR in elementary schools of 36:1 Philippines. compared with 31:1 for all lower-middle-income countries. Figure 1: Student-teacher Ratios in Basic Education Have Fallen Significantly in Recent Years Student-teacher ratios (STRs) and teacher numbers in basic education, 2006–2014 Elementary schools High schools 50 500 50 500 40 400 40 400 Teachers (000s) Teachers (000s) 30 300 30 300 STR STR 20 200 20 200 10 100 10 100 0 0 0 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2006 2007 2008 2009 2010 2011 2012 2013 2014 Total teachers (right axis) Total teachers (right axis) Student-teacher ratio (left axis) Student-teacher ratio (left axis) Source: BEIS data on enrolment and number of teachers from DepEd factsheet, various years. Note: Data from the PETS-QSDS survey found similar student-teacher ratios for 2014. 26 WWW.WORLDBANK.ORG/PH Figure 2: Student-teacher Ratios Reveal that Many Schools Have Teacher Shortages Percentage of schools by student-teacher ratio, 2014 Elementary schools High schools 60% 60% 40% 40% 20% 20% 0% 0% >40 36–40 30 to 35 <30 >35 28–35 20–27 <20 Source: PETS-QSDS national survey – school-level data. Looking beyond national averages in the PETS-QSDS found in schools located in remote areas where there are not survey data, it can be seen that STRs tend to be worse enough students in a school’s catchment area to meet the in urban schools. For example, there is an average of 39 STR guidelines. Indeed, a greater proportion of schools in students for every teacher in elementary schools in highly rural municipalities tend to have STRs that are below existing urbanized cities compared to only 29 in schools located in guidelines compared to city schools. However, a signi cant municipalities.8 Di erences in STRs in high schools are not as number of schools in more densely populated areas also stark, but large urban schools have higher STRs than schools have low STRs. It is possible that transferring teachers from in rural areas.9 these schools to schools with teacher shortages would be a more e cient option than hiring additional teachers. The distribution of teachers throughout the Philippines is uneven, and many schools do not have enough teachers To get a better picture of the real learning environment faced according to DepEd’s standards. While DepEd does not have by students in the Philippines, the study team observed xed norms for school-level STRs, it does have guidelines on around 7,000 classes and discovered that class sizes were class size and on the ideal number of teachers per class for much smaller than student-teacher ratios suggest. Only each grade. According to these guidelines, the average STR 9 percent of classes in elementary and high schools had in elementary schools should be 35:1 while in high school more students than the relevant DepEd guideline (40-45 the average STR should be 27:1.10 The distribution of STRs for elementary and 40 for high schools). The apparent in the PETS-QSDS survey data reveals that approximately contradiction between these ndings and the proportion 29 percent of elementary schools and 37 percent of high of schools experiencing teacher shortages is explained by schools do not have enough teachers (Figure 2). Perhaps shifting. Approximately 30 percent of all elementary and unsurprisingly, the proportion of schools in urban areas high schools in the study reported operating more than one with teacher shortages is much higher than schools in rural shift in at least one grade. However, the use of multiple shifts municipalities. raises concerns about the amount of learning time each student receives and for this reason DepEd is slowly phasing While a signi cant proportion of schools have teacher out shifting in all schools. shortages, a large number of other schools, particularly at the elementary level, seem to have a surplus of teachers. High school principals report signi cant teacher shortages About 52 percent of all elementary schools have STRs below in some subject areas even though DepEd currently has existing standards (Figure 2). Lower STRs are frequently no speci c norms.11 Approximately one-third of all high ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 27 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers school principals interviewed during the PETS-QSDS survey up to teach every day on time is more important for student reported having a shortage of subject specialist teachers outcomes. The PETS-QSDS survey carried out unannounced (Figure 3). In the absence of any DepEd norms, it is di cult visits to the sampled elementary and high schools to record to assess how large these shortages are. However, the school teacher attendance (Box 1). Survey enumerators observed principals who reported a shortage were asked to estimate teachers arriving at school and recorded the time when they their needs in each subject. Comparing these self-reported arrived. Later on during the school visit, they recorded data numbers with a school’s existing stock of teachers showed on all of the teachers teaching in the school and their shift that there were substantial shortages of Filipino teachers pattern in order to calculate absenteeism rates. (Figure 3). Schools also reported that their shortages of mathematics, science, and English teachers were equivalent This exercise showed that overall levels of teacher absence to about 30 percent of their existing workforce. in the Philippines are low. The percentage of teachers that were absent on the day of the direct observations by the Te cher Attend nce survey teams was 7.6 percent for elementary school teachers and 5.9 percent for high school teachers. These rates are low Knowing how many teachers are allocated to each school compared with other countries that have similar measures of is important for planning purposes, but whether they turn teacher absenteeism rates (Figure 4). Figure 3: Many High School Principals Report Shortages of Subject Specialist Teachers Percentage of high school principals reporting subject specialist teacher shortages and magnitude of need, 2014 Schools reporting shortages Relative size of teacher shortage 50% 100% Teacher need as % of stock Percentage of schools (%) 40% 80% 30% 60% 20% 40% 10% 20% 0% 0% HUCs Other Cities Municipalities National Filipino Makabayan Math Science English Source: PETS-QSDS national survey – school-level data. Notes: Makabayan covers social studies, home economics, music, art, health and values. 28 WWW.WORLDBANK.ORG/PH Box 1: Procedures for Measuring Teacher Absenteeism through Observation The approach to directly observing teacher attendance that was adopted in the Philippines PETS-QSDS survey followed procedures and protocols used in a number of other well-known surveys of provider absence in health and education. DepEd’s regional and division o ces were not informed in advance about which schools would be visited by the PETS- QSDS survey teams. Where it was not possible to arrive unannounced on the rst day of their school visit, survey teams would complete all other survey activities in the school and then come back to the school at a later date, usually after they had nished surveying the other sample schools in the city or municipality. The survey teams would arrive at the sampled schools at least one hour before the start of the school day. An enumerator was posted at each school entrance and at the o ce where teachers o cially recorded their arrival (on ngerprint readers or in manual records). As the teachers arrived, the team would record their full name and time of arrival. After the teacher observation, eld team supervisors transferred information from each enumerator’s observation schedule to the teacher roster for the school. As part of the wider school survey, the team collected information on all school teachers and identi ed those who were scheduled to be at the school at the time when the teacher observation exercise was undertaken. This yielded information on the number of teachers who were supposed to be present at the time of the observation to compare with the number of teachers who actually arrived. Information on the punctuality of each teacher was also recorded. The survey team did not monitor the activities of the teachers once they were in school. This made it impossible to discover how many teachers were actually present and teaching in the classrooms after they had arrived at the school as some other studies have done. Other studies have also checked for absenteeism at di erent times of the day to get a more thorough picture of teacher attendance throughout the day. However, given the size of many of the schools visited by the survey teams in the Philippines and the large amount of other school-level information that they collected, it was not feasible to do this as part of the PETS-QSDS study. Source: Rogers, H. and M. Koziol (2012). “Provider Absence Surveys in Education and Health: A Guidance Note.” World Bank, Washington D.C. Figure 4: Teacher Absenteeism Rates in the Philippines Are Low Primary/elementary teacher absenteeism rates in selected countries, various years 30% 25% 20% 15% 10% 5% 0% Uganda India Pakistan Zambia Bangladesh PNG Ecuador Peru Mongolia Indonesia Philippines Cambodia (2013) (2014) (2004) Sources: Philippines - PETS-QSDS national survey – school-level data. Data on Indonesia come from Analytical and Capacity Development Partnership (2014). “Teacher Absenteeism in Indonesia: Policy brief.” Jakarta. Data for all other countries come from Rogers, F. H. and E. Vegas. (2009). “No More Cutting Class? Reducing Teacher Absence and Providing Incentives for Performance.” Policy Research Working Paper. No. WPS 4847. World Bank, Washington D.C. Notes: The Philippines study measured teacher absenteeism for teachers who were expected to be present at the beginning of the school day, whereas other studies checked teacher absenteeism at di erent times of the school day. All absenteeism rates are for 2002–2004 unless otherwise stated. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 29 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers While overall rates are low, teacher absenteeism appears to cities where rates of teacher absenteeism are higher. be higher in highly urbanized cities (Figure 5). In high schools located in highly urbanized cities (HUCs), almost 1 in 10 Levels of absenteeism varied according to a number of teachers were found to be absent in 2014. This absenteeism teacher and school characteristics.14 Absenteeism rates rate is 53 percent higher than the national average.12 The tended to be higher for more experienced teachers, although factors that may contribute to these higher than average this was only statistically signi cant for elementary school rates may be the larger size of schools in these areas, greater teachers.15 Schools that appeared to be more supportive di culties in terms of tra c congestion in getting to city of teachers also appeared to have lower absenteeism. For schools, and greater demands on teachers’ time. example, elementary schools where principals routinely observed teachers in their classrooms tended to have lower On the whole, poorer children attend schools with lower rates rates of teacher absenteeism. Moreover, schools where of teacher absenteeism than wealthier children. The PETS- teachers lled in their professional development plans more QSDS survey included a nationally representative sample regularly tended to have better teacher attendance rates but of public elementary and high school student households. the e ects were only statistically signi cant for high schools. Using the information on consumption and asset ownership Regular visits from the DepEd division o ce also improves collected from these survey households, it was possible to teacher attendance especially in high schools. On the whole, rank student households by their estimated levels of per schools with better facilities (such as electricity) also had capita household consumption.13 This ranking reveals that better teacher attendance, but this e ect was only statistically students in the bottom 20 percent of households in terms signi cant for high schools. of consumption tend to attend schools with lower rates of teacher absenteeism than students from the wealthiest 20 The majority of teachers who were recorded as absent were percent of households. This re ects the greater proportion out of school for permissible reasons (Figure 6). Relatively of wealthier students attending schools in highly urbanized few elementary school teachers and only 16 percent of high Figure 5: Absenteeism Rates in Highly Urbanized Cities are Higher than the National Average Teacher absenteeism rates by school location, 2014 Elementary schools High schools 15% 15% 10% 10% 5% 5% 0% 0% HUCs Other Cities Municipalities National HUCs Other Cities Municipalities National Source: PETS-QSDS national survey – school-level data. 30 WWW.WORLDBANK.ORG/PH Figure 6: Most Teacher Absence was for Permitted Purposes Percentage of absent teachers by reason of absence, 2014 Elementary schools High schools 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Sick leave Other leave O cial duties Training Uno cial Transferred O cial duties Sick leave Uno cial Other leave Traiing Transferred Source: PETS-QSDS national survey – school-level data. school teachers were absent for uno cial reasons. More than one-third of elementary school teachers and a quarter Te cher Hirin nd of high school teachers who were absent on the day of the Deplo ment survey team’s surprise visit were reported to be on sick leave. Another quarter of elementary school teachers and about Based on data from the PETS-QSDS survey, this section 15 percent of high school teachers were away from school assesses the e ectiveness of government systems at because they were taking o cial leave (casual, maternity, allocating teacher resources to the neediest schools and or earned leave). Teachers were also frequently absent from their ability to hire and place new teachers e ectively. their schools while undertaking o cial duties elsewhere, including attending division-led training, managing school Alloc tion of New Te chers maintenance and other operating funding, and participating The allocation of new teachers in 2013 and 2014 was in meetings at the division o ce. largely in line with need. In 2013 and 2014, approximately School principals appear to monitor teacher attendance 90,000 new basic education teachers were hired through closely. The survey collected information from o cial teacher an allocation process managed by DepEd (Box 2). School- attendance records in schools to assess their accuracy. level data from the PETS-QSDS survey show that, on These o cial records were similar to the PETS-QSDS direct the whole, new teachers were allocated to schools with observation data, which suggests that schools were keeping high STRs (Figure 7).16 However, it is clear that there is still good records of teacher attendance. On the basis of these room for improvement in the targeting of new teacher records, it appears that school principals take regular action allocations. For example, a large number of elementary to address teacher absenteeism and tardiness. Approximately schools with STRs higher than DepEd’s guideline of 35 17 percent of all elementary and 49 percent of high school students per teacher were allocated no additional teachers principals had taken action against teachers who were in either 2013 or 2014. Moreover, a number of schools regularly absent or late. However, their disciplinary actions that already had relatively low STRs were allocated extra tended to be limited to verbal or written warnings. teachers. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 31 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Box 2: The Current Process for Allocating and Hiring Teachers The Department of Education’s Central O ce decides on the number of new teacher positions that will be allocated to each elementary school and high school. It then submits the list to the Department of Budget and Management (DBM) Central O ce, which in turn informs all of its regional o ces and provides them with copies of the NOSCA (notice of organization, sta ng, and compensation action) for their region. The NOSCA speci es the number of new positions for each teacher classi cation and provides the authorization for making salary payments. The DepEd Central O ce similarly informs all of its regional o ces, which then inform the DepEd division o ces who are responsible for hiring all elementary and high school teachers. Until 2014, the NOSCA for each division included new teacher positions for all elementary schools in the division. However, in 2015 this practice was changed. Now the DepEd Central O ce allocates speci c positions to each elementary school in an addendum to its main memo and this is also re ected in the NOSCA. The hiring process begins when the DepEd division o ces advertise the vacancies. Once applications are received, the division o ce checks that candidates are included on the Registry of Quali ed Applicants (RQA), a list of teachers who are quali ed to ll these positions. The division o ces then interview candidates and select individuals to ll the new positions. They then submit the names of the teachers whom they have selected to the regional o ces of the DBM. There are special procedures for the DBM to release funds to DepEd division o ces to pay newly hired teachers in their rst few weeks on the job. These procedures were initiated in 2012 to enable newly hired teachers to start being paid as soon as they were hired because there were often delays in establishing them on the o cial payroll. Figure 7: Allocations of New Teachers Have Favored Schools with High Student-teacher Ratios Proportionate increase in stock of school teachers in 2013 and 2014 compared to 2012 school-level student-teacher ratios Elementary schools High schools 70% 90% approximate STR guideline New teachers as % of all teachers approximate New teachers as % of all teachers 80% 60% STR guideline 70% 50% 60% 40% 50% 30% 40% 30% 20% 20% 10% 10% 0% 0% 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 100 Student–teacher ratio, 2012 Student–teacher ratio, 2012 Source: PETS-QSDS national survey – school-level data. 32 WWW.WORLDBANK.ORG/PH Figure 8: Hiring New Teachers Has Reduced Teacher Shortages but Has Also Increased the Number of Schools with Surplus Teachers Percentage of schools by student-teacher ratio, 2012 and 2014 school years Elementary schools High schools 60% 60% 40% 40% 20% 20% 0% 0% >40 36–40 30 to 35 <30 >35 28–35 20–27 <20 2012 2014 2012 2014 Source: PETS-QSDS national survey – school-level data. While the recent hiring and allocating of a large number guidelines on class sizes and teacher-class ratios to identify of new teachers has signi cantly reduced the number of each school’s need for teaching sta . However, there is no schools with a teacher shortage, it has also increased the o cial DepEd order or memo to guide division o ces in number of schools with teacher surpluses (Figure 8).17 The how to make new teacher hiring decisions. This makes it proportion of elementary schools with teacher shortages di cult to monitor DepEd’s allocation decisions. dropped from 46 percent to 29 percent between 2012 and 2014. The proportion of high schools with teacher shortages More than one-third of high school principals and almost also dropped, from 70 percent to 38 percent over a similar a quarter of elementary school principals interviewed for period. However, the percentage of schools with teacher the PETS-QSDS survey felt that DepEd’s current methods surpluses has also increased. At the elementary level, the of allocating new teachers was unfair and suboptimal. The proportion of schools with STRs below 30 increased from high school principals felt that the lack of a mechanism 41 percent to 52 percent between 2012 and 2014. On closer to account for geographical di erences and the need for inspection, it is likely that some of these schools may be subject-speci c teachers were signi cant weaknesses. unable to achieve the STRs outlined in the DepEd guidelines, Similar opinions were expressed at the division level by but in most cases, while the surplus teachers are easing the school division superintendents (SDSs), about half of whom burden in their particular schools, they are not the most reported that they had not received an adequate number of e cient use of public education spending. new teacher positions in the previous two years. Moreover, one- fth of SDSs said that they had not been consulted by Until recently, DepEd’s process for allocating teacher their regional o ce on their teacher needs for 2014. positions among schools was based on a color coding system according to which schools with high STRs were Te cher Hirin Process prioritized.18 However, a recent assessment of this system revealed that the allocations did not always follow the The PETS-QSDS survey tracked the timeliness of the teacher color coding system.19 The current system now uses DepEd hiring process in 2013 and 2014. There are four key steps in ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 33 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Table 1: Benchmarks and Actual Timing of Key Steps in the Teacher Hiring Process, 2013 Receipt of Posting of registry Submission Receipt of SARO NOSCA of quali ed of list of lled applicants positions (1) (2) (3) (4) Benchmark date for completing each step March 2013 April 2013 June 2013 July 2013 Date on which the key steps were completed in the sample DepEd divisions: ¾ Earliest date December 2012 January 2013 January 2013 January 2013 ¾ Average date January 2013 May 2013 May 2013 July 2013 ¾ Latest date May 2014 May 2014 August 2014 - Percentage of divisions that completed the 56% 61% 75% 67% process on or before benchmark date Source: PETS-QSDS national survey – DepEd division o ce data. Notes: Numbers in parenthesis refer to the steps outlined in the text. the teacher hiring process that determine whether teachers The PETS-QSDS survey collected data from the DepEd are hired according to the DepEd guidelines, are ready to division o ces on the timing of these critical steps in the start teaching at the beginning of the school year, and are hiring process during the 2013 school year.20 paid on time: A signi cant proportion of DepEd division o ces continue 1. The DBM provides the list of teacher positions to DepEd’s to experience delays in completing the key steps of the division o ces and/or DepEd’s regional o ces. The hiring process. Table 1 compares the benchmarks that the DBM regional o ces release to DepEd division o ces DepEd Central O ce issued for the completion of key steps the NOSCA (notice of organization, sta ng, and with the actual completion dates achieved by the 51 DepEd compensation action), which contains information on the division o ces sampled in the PETS-QSDS study. It shows number and type of all teacher posts in each division. that nearly half of all division o ces receive the noti cation of their allocation of new teacher posts late. The late receipt 2. The DepEd division o ces post a list of quali ed applicants. of the NOSCA leads to further delays in the hiring process. DepEd division o ces verify and post the Registry of Around 40 percent of DepEd division o ces are late posting Quali ed Applicants (RQA) in the division o ce. The list the registry of quali ed applicants but are able to catch up a consists of all individuals with the required quali cations little by hiring teachers more quickly (Table 1). Only around to be a public school teacher in the division. two-thirds of division o ces received their allotment to pay teachers on time, with most schools receiving it on or after 3. The DepEd division o ces interview and appoint new the beginning of the school year. teachers. Each division o ce holds interviews for all new positions within its jurisdiction and then submits a list of However, there is evidence that there have been some new hires to the DepEd and/or DBM regional o ce. improvements in the overall timeliness of the teacher hiring process in the DepEd division o ces. The PETS-QSDS analysis 4. The DBM releases the allotment to pay new teachers. Once of the 2013 hiring process shows that it took on average ve the DBM regional o ce receives the list of new hires months from the time when the division o ces receive the from each division, it releases a special payment in the NOSCA to their submission of the list of lled positions to form of a special allotment release order (SARO), which DepEd and the DBM. In 2007, a similar analysis found that this division o ces then use to pay their newly hired teachers process had taken as long as 18 months to complete.21 until they are put on the regular payroll, usually in the following nancial year. 34 WWW.WORLDBANK.ORG/PH Despite these improvements, divisions reported that about of high school principals agreed with this statement. The one-third of the new teacher positions for the 2013–14 school principals who disagreed raised a number of reasons, school year remained un lled even by the end of the 2014 including the overly selective nature of the recruitment calendar year (Figure 9). The large number of new teachers process, the lack of weight given to demonstration teaching, unable to report for work at their new schools was partly the and the lack of focus on subject specialization. result of delays in the hiring process outlined in Table 1. For example, 65 percent of the divisions where teachers were not Tr ckin of Funds for in post at the start of the school year reported receiving the Newl Hired Te chers NOSCA late, while 22 percent reported posting the RQA late, and 19 percent reported submitting the list of lled positions The study team’s tracking of fund ows for newly hired late. However, the majority of divisions where teachers were teachers revealed inconsistencies in the data and poor record not in post at the start of the school year had completed all keeping. Information on the budget allocation for newly processes on time. It is likely that delays in the submission hired teachers was not available from either the DepEd by teachers and veri cation by division o ces of teacher division o ces or the DBM’s regional o ces.22 This meant documentation partly explains these delays in the nal step that the team was largely unable to track the ow of funds of getting new teachers into the classroom. However, further between these two sets of o ces. In the rare cases where research is needed to fully understand why these delays it was possible to do so, only one-third of the divisions occur and to identify ways to address them. reported receiving the same amount as the amount reported by the DBM regional o ce. About half of the remaining It was not the aim of the PETS-QSDS study to evaluate the divisions reported receiving substantially more in SAROs than hiring process at the division level, but SDSs and school the DBM had reported releasing. This may be because the principals did feel that these processes were on the whole fair. SAROs received by these division o ces contained funds About 90 percent of SDSs agreed that the process for hiring for additional payments. Fewer than a quarter of divisions new teachers results in the best teachers being hired, while reported receiving substantially less than the amount 86 percent of elementary school principals and 80 percent reported by the DBM regional o ce. This lack of consistency between the DBM and DepEd division o ces is concerning and highlights the need to strengthen monitoring and Figure 9: A Significant Number of Newly Hired reporting mechanisms. Teachers Were Not in Post at the Beginning of the School Year Percentage of teacher posts allotted for the 2013 Te cher Tr nsfers school year where newly hired teachers were in Transferring teachers from one school to another is a useful post by the nal quarter of the 2014 calendar year way to redress imbalances in the distribution of teachers and increase the e ciency of public spending. Decisions 80% about teacher transfers within each division are made by the SDS, and a transfer can be requested by the school 60% principal, a teacher, or the SDS based on needs of the schools in the division. However, it is more complicated to transfer a teaching post across divisions or regions as this requires 40% the approval of the central o ces of both DepEd and the DBM. DepEd’s recent medium-term expenditure plan (2014 20% to 2020) estimated that, in 2012, there were about 23,000 excess elementary school teachers and about 5,000 excess high school teachers. Even if only 25 percent of the surplus 0% teachers (a conservative estimate) were transferred to schools Kindergarten Elementary High School with shortages, this would enable DepEd to save about PHP Source: PETS-QSDS national survey – DepEd division o ce data. 3.8 billion annually.23 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 35 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers However teacher transfers are rarely used to address the in some schools is emerging as a key issue. The note has uneven distribution of teachers and the growing proportion also shown that, despite recent improvements in the formal of schools with teacher surpluses. The PETS-QSDS interviews processes governing the hiring of teachers, there are still with a nationally representative sample of Grade 6 and 10 frequent and signi cant delays in getting teachers into the teachers showed that fewer than 2 percent of teachers were classroom. transferred in 2013. It is also likely that the majority of these Having a clear set of guidelines that outline the size, teacher transfers were made for personal family reasons rather allocation, and hiring of the required teaching force in the than to redress imbalances in the distribution of teachers. Philippines based on expected class sizes and curriculum Any e orts to use teacher transfers to redistribute the load would be a useful tool to support the more equitable allocation of teachers across the country are currently allocation of both new and existing teachers. The last DepEd constrained both by the 1968 Magna Carta for Public School memorandum on teacher allocation processes was issued in Teachers and by the localization law. The Magna Carta gives 2010, and its guidance is now outdated. Speci c guidelines teachers the right to refuse a transfer without being disciplined for specialist subject positions will also be necessary to or red. However, the latest DepEd order on teacher transfers ensure that schools have enough specialist teachers and can in 2013 indicates that teachers can be transferred without the plan e ectively. Ensuring that these guidelines are widely teacher’s consent in certain cases. These include situations disseminated throughout all levels of DepEd and to the wider where a teacher’s existing school has an STR lower than the public would enable better decision-making within DepEd DepEd guidelines of 35:1 for elementary schools and 27:1 and increase the transparency of all decisions made about for high schools. The new clause is a clear attempt to give teacher deployment. SDSs greater authority to use transfers to manage teacher DepEd Central O ce needs to monitor the hiring process distribution.24 Nevertheless, the Magna Carta states that more closely, ensuring that regional and divisional o ces teachers can contest transfers even in such cases by appealing are adhering to the hiring timetable and providing them to the Director of Public Schools. The localization law gives with additional support where necessary. However, further priority to appointing and assigning teachers to residents research is needed to understand why so many newly of the barangay, city, or municipality in which the school is hired teachers are unable to take up their posts before the located. While this has the potential to increase the diversity start of the school year, even when all of the administrative of the teaching force, support mother tongue policies, and formalities appear to have been completed. ensure that schools in remote areas have teachers in place, it does constrain teacher management. Making more information available on the existing teaching force would improve initial teacher allocations and facilitate Policies for Stren thenin future planning. It is currently di cult for DepEd division o ces to obtain anything more than basic information Te cher Hirin nd on teacher allocations in their division without having to Deplo ment go through large numbers of paper les. Given the large number of schools and teachers in most divisions, this is a barrier to the e cient planning of division-level teacher Over the last ve years, the strong focus on improving school workforces. Plans are underway within DepEd to develop learning environments has led to a substantial increase in the a Human Resource Information System (HRIS), which will availability of teachers in basic education schools across the provide planners at all levels of the department with the Philippines. Moreover, teacher absenteeism rates are low and necessary information and capacity to make well-informed compare favorably with other countries. Systems to allocate decisions about teacher allocations. and hire teachers have also improved, and the time taken to complete key steps in the process has been shortened If DepEd were to adopt a more exible and e ective considerably. However, this policy note has shown that deployment and transfer policy (taking into account many schools still do not have enough teachers or enough special circumstances such as remote schools), this would of the right mix of teachers, and the over-supply of teachers have the potential to greatly improve the distribution of 36 WWW.WORLDBANK.ORG/PH Table 2: Strengthening Systems to Hire and Deploy Teachers Findings Policy suggestions Further improvements in teacher • Further develop guidelines/norms for school-level teaching needs, distribution are needed particularly for subject-speci c teachers in high schools • Monitor the distribution of teachers more closely to ensure e ciency and develop and implement the Human Resource Information System (HRIS) • Revise teacher transfer policy to make it a more e ective tool to improve teacher distribution Hiring and deployment systems need to • Strengthen the accountability of regional and divisional o ces for be strengthened carrying out processes in a timely way • Improve monitoring to ensure hiring timelines are met and that teachers are in post at the beginning of the school year • Improve coordination between the DBM and DepEd regional o ces High teacher absenteeism in highly • Increase the authority of DepEd division o ces and schools to deal with urbanized cities absenteeism • Link teachers’ attendance to their career development and performance bonuses • Explore the potential for increasing community involvement in monitoring teacher attendance teachers and ultimately increase the learning opportunities of the study suggest that school principals and division available to all children. The 2013 DepEd order on teacher supervisors need to have a wider range of sanctions and transfers and its alignment with school sta ng guidelines incentives to address attendance issues. For example, was an important development, but further e orts will be attendance rates could be a key results indicator in teachers’ needed to reconcile this policy with existing regulations and performance plans and ultimately be linked to their agreements on teacher employment conditions if it is to be career development and to performance-based bonuses. e ective. Introducing additional incentives such as cost-of- Research from other countries has shown that school living adjustments within teacher salaries and reevaluating governing councils and parents can help to reduce teacher hardship allowances in supporting transfers of teachers to absenteeism. A systematic global review concluded that “a remote schools would help to make transfers more attractive. combination of close monitoring (community/parents level) However, incentives of this kind need to be introduced and attractive incentives (teacher level) showed the highest carefully and evaluated quickly to ensure that they are potential to reduce teacher absenteeism.”25 delivering the desired outcomes. Over the last ve years, the Philippines government has Systems at the school level for monitoring teacher greatly increased the size of the basic education teacher attendance appear to work well, but sanctions for workforce. Ensuring that the teacher workforce is serving absenteeism and lateness are weak. On the whole, teacher the neediest schools and redressing imbalances in the absenteeism rates are low in the Philippines, but in highly existing distribution of teachers will ensure that these urbanized cities as many as one in ten teachers were absent increases will translate into better student learning during the study team’s unannounced visits. The ndings outcomes. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 37 Policy Note 1: Assessing Systems for Hiring and Deploying Teachers Endnotes 1 See, for example, Bruns, B. and J. Luque (2014). “Great Teachers: 14 The full results of the regression analysis reported here is included How to Raise Student Learning in Latin America and the in additional annexes and tables accompanying the main PETS- Caribbean.” World Bank, Washington, D.C. QSDS report. 2 De Ree, J. (2016). “How Much Teachers Know and How Much It 15 Years of experience were also positively related to absenteeism, Matters in Class: Analyzing Three Rounds of Subject-speci c Test but the results were not statistically signi cant. Score Data of Indonesian Students and Teachers.” Policy Research 16 A similar analysis was also carried out at the division level, which Working Paper No. WPS 7556. World Bank, Washington, D.C. con rmed that there was a positive relationship between high 3 Chetty, R., J. Friedman, and J. Rocko . (2011). “The Long- STRs and new teacher allocations. term Impacts of Teachers: Teacher Value-added and Student 17 The gures for shortages and surplus teachers are based on the Outcomes in Adulthood.” National Bureau of Economic approximate student-teacher ratios estimated from the DepEd Research, Cambridge MA. guidelines. 4 See, for example, McEwan, P. (2013). “Improving Learning in 18 DepEd order number 77 (2010). “Guidelines on the Allocation/ Primary Schools of Developing Countries: A Meta-Analysis of Deployment of New Teaching, Teaching-related, and Non- Randomized Experiments.” Wellesley College, Wellesley, MA. teaching Positions for FY 2010.” Department of Education, Manila. 5 This includes salaries for technical positions. 19 Albert, J. R. (2012). “Improving Teacher Deployment Practices 6 A second note in this series focuses on teacher quality and in the Philippines.” Policy Note 2012-02, Philippines Institute of assesses the systems used to strengthen the competencies of Development Studies, Manila. teachers in the existing workforce. 20 Similar data were collected for 2014, but, since the survey was 7 In this section, years refer to school years. For example, 2014 refers conducted in the last quarter of 2014, this information does not to the 2014/15 school year. cover all steps. 8 Highly urbanized cities are cities with a population of more than 21 World Bank and AusAID (2010). “Philippines Basic Education Public 200,000 and with average revenues of at least PHP 50 million in Expenditure Review.” World Bank and AusAID, Manila. 1991 prices. Other cities are de ned as cities that do not meet 22 Unfortunately, it was not possible to track payments for newly the criteria to be classi ed as highly urbanized. Municipalities are hired teachers from the DepEd or DBM central o ces down to administrative units for all other areas in the Philippines. the regional o ces because these funds are comingled with 9 While STRs for each grade are not discussed in detail in this note, funds for all newly created positions, including non-teaching they show, on the whole, that STRs tend to be higher in the positions, at all levels. earlier grades. 23 DepEd (2013). “Medium-term Expenditure Framework for 10 These estimates were calculated by combining the ideal class size Basic Education 2014-2020: Enrollment Projections and Cost and the ideal number of teachers per class as recommended in Simulations under Alternative Scenarios.” Department of the DepEd service standards. Education, Manila. 11 Elementary school teachers are expected to be generalists. 24 In 2013, DepEd issued an order that granted transferring teachers a one-o redeployment allowance of PHP 18,000. However, this 12 Di erences between absenteeism rates in HUCs and other cities was discontinued after a year. and municipalities are not statistically di erent from the national average. 25 Guerrero. G., J. Leon, M. Zapata, and S. Cueto (2013). “Getting teachers back to the classroom. A systematic review on what 13 The household questionnaire included a short module on works to improve teacher attendance in developing countries.” consumption and a set of questions on assets that have been Journal of Development E ectiveness 5(4): 466-488. used by the Department of Social Welfare and Development (DSWD) to undertake a proxy means testing (PMT) approach to estimating household consumption per capita. The results reported here are based on information gathered using the PMT approach, and a full description is included in a separate note. 38 WWW.WORLDBANK.ORG/PH Polic Note 2: Developin Proficient nd Motiv ted Te cher Workforce Introduction Over the last decade, research from many di erent countries has demonstrated the important role played by teachers in increasing students’ learning and improving their academic performance.1 Studies from countries as di erent as the US and Indonesia have shown the enormous bene ts that follow from having adequate and e ective teachers working in a country’s schools. In Indonesia, a value-added analysis of student learning outcomes found that the more teachers know, the greater the improvements in the learning competencies of primary and junior secondary students.2 In the US, better teaching in elementary and secondary schools has been shown to increase students’ college participation rates, raise their subsequent earnings, and improve other long-term outcomes.3 Providing teachers with good quality professional development opportunities has been shown to be an e ective way of increasing their competencies and improving student learning outcomes in many di erent settings. A series of systematic reviews have been undertaken recently to assess the impact of di erent interventions on student learning outcomes in developing countries.4 One of the most consistent ndings from these reviews has been the positive and signi cant impact that interventions to strengthen teaching practice, introduce The purpose of this policy note is to provide an overall innovative instructional methods, and strengthen teachers’ picture of teacher competencies in the Philippines and to subject knowledge can have on student learning. However, assess whether the country’s professional development in many countries, such professional development systems deliver the training opportunities that teachers need opportunities frequently fail to meet even minimum levels of to be e ective.5 The ndings reported in the note are based quality and fall short of what teachers want and need. on a comprehensive survey of the public education system that tracked public education expenditures and assessed the These reviews also highlight the importance of teachers’ quality of education services. motivation in determining their e ectiveness and ultimately in improving the learning outcomes of their students. They The note shows that elementary and high school teachers’ found some evidence that introducing nancial incentives for subject knowledge is weak and a major constraint to teachers based on individual teacher or school performance improving student performance. While public spending on can increase the amount that students learn. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 39 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce professional development activities has been increasing, Assessment (TSNA). This self-assessment tool gives a picture these ndings highlight the need to signi cantly of a teacher’s pedagogical competence according to DepEd’s expand in-service training opportunities and tailor them own National Competency Based Teacher Standards (NCBTS). more closely to teachers’ needs. Moreover, systems for These assessments have been used by the PETS-QSDS study professional development, teacher performance monitoring to provide evidence on levels of teacher competency in and incentives need to be aligned to improve teacher public elementary and high schools. competencies and motivation. According to the results of the PETS-QSDS exercise outlined The Qu lit of Element r in Box 1, knowledge of subject matter among elementary and high school teachers is low in most subjects. With the nd Hi h School Te chers exception of English at the elementary school level, the average elementary or high school teacher could answer The competency of teachers both in terms of their fewer than half of the questions on the subject content tests subject matter knowledge and instructional methods has correctly (Figure 1). For example, the median mathematics been shown to be an important determinant of student teacher in high school was able to answer only 31 percent learning outcomes in the Philippines.6 However, getting an of the questions completely correctly. Since these tests are accurate measurement of the competencies of teachers is closely aligned with the curriculum, the results suggest challenging. The Department of Education (DepEd) along that teachers face signi cant challenges in teaching a with experts at the Philippines Normal University (PNU) considerable portion of the current K to 12 curriculum. In and their partner university in Australia have developed other words, the ndings point to the need for signi cant a set of teacher competency assessments for the new K improvements in teachers’ skills and subject knowledge if to 12 curriculum (Box 1). All the assessments have gone they are to have a full command of the curriculum and to through a rigorous design and validation process to ensure provide e ective instruction in the classroom. that they accurately measure a teacher’s knowledge of the subjects required to teach in elementary and high schools. To explore why so many teachers were unable to answer In addition to subject content tests, the PNU also designed the assessment questions correctly, the PETS-QSDS study a shorter version of DepEd’s Teacher Strengths and Needs used a partial credit model to look more closely into the Figure 1: Teachers’ Performance on Content Knowledge Assessments Was Poor Percentage of questions answered correctly by the median teacher (binary scoring method), 2014 Grade 6 elementary school teachers Grade 10 high school teachers 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% English Math Filipino Science English Math Filipino Science Source: PETS-QSDS teacher content knowledge assessments. 40 WWW.WORLDBANK.ORG/PH incorrect answers that teachers gave. For example, teachers The results of the partial credit scoring model showed that were given a choice of four possible answers to each many teachers had some understanding of the subject area multiple choice question, and the three incorrect choices but lacked the higher order problem-solving skills necessary re ected di erent levels of understanding on the part of the to teach the curriculum e ectively. The model showed that, respondents (Box 1). in most cases, teachers do have some understanding, skill, Box 1: PETS-QSDS Approach to Assessing Teachers As part of the PETS-QSDS study, a nationally representative sample of 377 Grade 6 (elementary) teachers and 946 Grade 10 (high school) teachers took two assessments: 1. A subject-based content assessment for measuring teachers’ knowledge of the content of the K to 12 curriculum in English, Filipino, Mathematics, and Science. The assessments consisted of multiple choice and open-ended questions. Grade 6 teachers were randomly assigned to complete one subject assessment, whereas Grade 10 teachers completed the test in the subject that they taught. 2. A short form of the Philippine government’s Teacher Strengths and Needs Assessment (TSNA) based on the National Competency Based Teacher Standards (NCBTS). The TSNA assesses a teachers’ own perceptions of their pedagogical competence in the seven NCBTS domains: (i) social regard for learning; (ii) learning environment; (iii) diversity of learners; (iv) curriculum; (v) planning, assessing, and reporting; (vi) community links; and (vii) personal growth and development. Teachers were asked to rate their own level on a four-point scale from low to high on a set of statements related to the seven domains. The content tests were drawn from larger assessments used by the Research Center for Teacher Quality (RCTQ) at the Philippines Normal University. The results of the PETS-QSDS tests are consistent with ndings from a larger regionally representative study conducted by the RCTQ. Two scoring rubrics were used for the content tests: 1. Binary credit - questions were marked either correct or incorrect. 2. Partial credit - each question was scored on a four-category scale, with 1 being “incorrect,” 2 being “displays some understanding, skill, and knowledge,” 3 being “displays a higher level of understanding, skill, and knowledge,” and 4 being “completely correct.” Responses coded 1, 2 or 3 on the partial credit scale were recorded as 0 on the binary scale and those coded as 4 on the partial credit scale were recoded to 1 on the binary credit scale. In the case of the multiple choice items, the choices provided represented di erent levels of the teacher’s understanding of the competency being tested. For open-ended questions, the graders marking the tests interpreted written responses according to the four-point scale. The binary credit scores provided a clear picture of whether teachers displayed a thorough understanding of the competencies included in the curriculum. The partial credit scale, on the other hand, provided a more detailed assessment of how far teachers were from a complete understanding of the underlying competencies and from having the necessary knowledge to teach e ectively. Given the assessment categories used by teachers to assess their own competencies, the TSNA was scored using a rubric similar to the partial credit scale. To account for di erences in the di culty of test items, a one-parameter Rasch model was used for both the subject content tests and the TSNA to transform the raw test scores into scores (logits) that ranked teachers according to their underlying ability. Full details of the instruments used and the methods for analyzing the data are available in RCTQ and SiMERR, (2015) “PETS-QSDS nal report”. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 41 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce and knowledge in all curriculum areas but not necessarily QSDS survey included a nationally representative sample enough to answer all of the question correctly. For example, of public elementary and high school student households. the scoring for a relatively di cult test question showed that, Using information on consumption and asset ownership that while only 20 percent of high school science teachers got was collected in the survey, it was possible to rank student the question completely correct, another 64 percent chose households by estimated levels of per capita household an answer that demonstrated some useful knowledge in the consumption.8 Using this indicator, it was possible to look curriculum area covered by the question (Figure 2). These at di erences in the performance of teachers who were more detailed results can inform e orts by DepEd to develop teaching di erent kinds of students. The key nding was that, professional development activities that are more closely in high school, poorer students tended to be taught by less aligned with the existing knowledge and abilities of the competent teachers except in English.9 At the elementary country’s teachers. level, no clear pattern emerged although poorer students tended to be taught by teachers that performed better on The study team performed a simple regression analysis to the Filipino test. explore associations between teachers’ scores on the subject knowledge assessment and their characteristics.7 Overall, Teachers themselves generally assess their levels of skills as they found that teachers’ performance was not associated satisfactory. A shorter version of the TSNA self-assessment with their levels of education or experience. For example, was given to the same teachers who completed the subject the scores of teachers who had obtained a postgraduate matter tests. These teachers were asked to rate their own quali cation did not di er from those of teachers with skill level in seven domains that included social regard for only a bachelors’ degree in a statistically signi cant way. learning (including punctuality and the use of information Although newly hired teachers tended to do less well than from a variety of sources for learning), the learning existing teachers in most of the subject-speci c tests, these environment (maintaining a safe and conducive learning di erences were not statistically signi cant either. environment and setting high expectations for learners), and the curriculum (demonstrating a mastery of the subject The better-performing high school teachers tended to and communicating learning goals).10 The teachers were teach in schools that serve better-o students. The PETS- given a set of statements related to each domain and were Figure 2: Despite their Overall Poor Performance, Teachers Have Some Relevant Skills and Knowledge Upon Which to Build Percentage of Grade 10 science teachers by their responses to easy and hard test questions, 2014 60% 50% 40% 30% 20% 10% 0% Completely Some Greater Completely Completely Some Greater Completely incorrect knowledge knowledge correct incorrect knowledge knowledge correct Relatively easy Relatively hard Source: Results of the PETS-QSDS Grade 10 teacher content knowledge assessment. Note: The easy question is in the lowest quartile of Rasch model scores while the hard question is in the highest quartile of di culty. 42 WWW.WORLDBANK.ORG/PH asked to rate themselves on a scale of 0 (low) to 3 (high). On Despite having had relatively low scores in the subject matter average, teachers rated themselves above 2 (satisfactory) on tests, teachers rated their competencies in the curriculum all domains. Moreover, there were no signi cant di erences domain, including subject-matter knowledge, very highly in the self-reported levels of strengths and weaknesses (Figure 3). Competencies in this area include demonstrated between teachers in elementary schools and those in high mastery of the subject area and the appropriateness of schools. the teaching methods used for di erent learning activities. Teachers rated their competency in this area higher than on The teachers generally felt that they were weakest in the any of the other competencies in the national standards. planning and the learning environment domains. The study These results are in stark contrast with the same teachers’ team developed a common scale that accounted for the results on the objective subject matter tests (Figure 1). di culties involved in obtaining the various competencies This mistaken perception of their own skills on the part of measured by the statements in the assessment. This made it teachers highlights the di culty they are likely to have in possible to compare the relative strengths and weaknesses assessing their students’ performance as well. of teachers (Figure 3) and revealed that both elementary and high school teachers tended to rate their competencies in The results of the PETS-QSDS assessments suggest that self- planning as the weakest. This domain includes competencies assessments of teachers’ strengths and weaknesses may not associated with instructional planning, the use of di erent be a good basis on which to plan professional development assessment strategies to evaluate students’ learning, and activities. Simple correlation coe cients among the providing feedback to learners. The teachers also felt that elementary school teachers who took these tests showed they were weak in the competencies associated with the a weak relationship between the teachers’ self-assessment learning environment such as communicating high learning ratings and the subject-matter test scores. In most cases, expectations and the ability to deal with students’ behavioral these correlations were not statistically signi cant. There was issues. more variation in the picture for high school teachers. The self-assessment ratings for Filipino and Mathematics teachers Figure 3: Teachers Assess Their Weakest Skills as Those Associated with the Learning Environment and Planning and Assessment Average self-assessment scores on national competency-based standards, 2014 4 3 2 1 0 Grade 6 teachers Grade 10 teachers Social regard for learning Learning environment Diversity of learners Curriculum Planning, assessing, and reporting Community links Source: Results of the PETS-QSDS Grade 6 and 10 teacher strengths and needs assessment. Note: The mean of the Rasch scores has been shifted by ve units to report positive averages to aid interpretation. Higher scores are related to higher self- assessed skill levels in the particular domain. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 43 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce were generally positively correlated with their subject Incidence nd Intensit of Tr inin matter test scores, and these associations were statistically signi cant. However, similar correlations for English and The percentage of teachers in the Philippines who receive Science teachers were generally not statistically signi cant. some annual in-service training is high even when compared In all cases, the strength of the association was relatively to levels in high-income developed economies. The PETS- weak, which casts doubt on the usefulness of existing self- QSDS survey collected detailed data on the in-service assessment tools to plan teachers’ in-service training. training received by sampled teachers in 2013 and 2014. The data revealed that more than three-quarters of all basic education teachers had received some in-service training Supportin Te chers in (Figure 4). More high school teachers had received in-service School throu h Profession l training than elementary or kindergarten teachers. The team compared these rates with the rates of professional Development development activities for teachers in 34 OECD countries collected in the Teaching and Learning International Survey High quality and regular professional development (TALIS).12 On average, 89 percent of public school lower opportunities are needed to address the weaknesses in the secondary teachers in these countries participated in competencies of the existing teacher workforce. Studies from professional development activities in 2013. Participation both developing and developed countries have shown that, rates in the Philippines were comparable although they were when well-designed, in-service teacher training can increase higher in Malaysia, an East Asian country with similar levels of teacher’s content knowledge, improve their methods student learning as the Philippines. of instruction, and ultimately improve student learning outcomes.11 This section looks at in-service teacher training While the majority of teachers in the Philippines received in the Philippines and assesses the systems that plan and some professional development training, it was less than nance these opportunities. in most other countries. In 2013, the average Grade 10 high school teacher received approximately ve days of in-service Figure 4: Most Teachers Receive Some In-service Training but Only for Short Periods Percentage of teachers receiving in-service training and duration of all training received, 2013 Participation, % teachers Duration, days 90% 10 8 6 80% 4 2 70% 0 Kindergarten Elementary: High School: Kindergarten Elementary: High School: Grade 6 Grade 10 Grade 6 Grade 10 Source: PETS-QSDS national survey – teacher-level data. 44 WWW.WORLDBANK.ORG/PH Figure 5: A Large Proportion of Teachers Felt They Needed More In-service Training Opportunities Percentage of Grade 10 high school teachers by the type of additional support that they most need to improve their classroom teaching, 2014 80% 60% 40% 20% 0% More and better teaching materials More and better in-service training More and better physical facilities More support from principal Smaller class sizes More support from division supervisors More support from district supervisor Change to class routine training. The PETS-QSDS survey was conducted at the end and 39 percent of sessions for high school teachers were of 2014, which was approximately two-thirds of the way conducted by DepEd division, region, or central-level sta . through the school year, and there were signs that the length School principals were also involved, conducting 32 percent of in-service training had increased over the 2013 level. of all training for elementary school teachers and 44 percent For example, Grade 10 high school teachers had already of training for high school teachers. Outside experts were received seven days of in-service training by the time of the used sparingly, accounting for less than 15 percent of training survey. However, even this level is relatively low. The average sessions. lower secondary teacher in the OECD TALIS study received approximately eight days of professional development a Teachers were generally positive about the training that year.13 Given the higher levels of teacher competency in they had received, but a signi cant proportion felt that they OECD countries, it might be expected that the duration of needed more. Elementary and high school teachers ranked in-service training in the Philippines should be longer. over 80 percent of the training that they received in 2013 as extremely useful.14 However, approximately 40 percent The most common kind of training received by elementary of teachers interviewed said that they needed more and and high school teachers in the Philippines was in subject better quality in-service training to improve their classroom content. Approximately, 40 percent of elementary and 30 teaching (Figure 5). percent of high school teachers reported attending some subject-based training during the 2013 school year. Training Systems at the school level to support teachers and identify in methods of instruction and teaching was also relatively their professional development needs do not seem to be common with around 12 percent of elementary and high working well. Each teacher is expected to complete an school teachers attending this kind of training in 2013. Individual Plan for Professional Development (IPPD) to outline their professional development needs. In preparing these The training usually took place in schools and was frequently plans, teachers are expected to use a toolkit prepared by conducted by school principals. About a half of all DepEd which utilizes a teacher’s own strengths and needs elementary teacher training and two-thirds of high school assessment using the TSNA. School principals are then teacher training conducted in 2013/14 took place in schools. expected to aggregate the needs of individual teachers DepEd division o ces were also commonly used to conduct into a School Plan for Professional Development (SPPD) and training, while DepEd district-level o ces were frequently to submit it to the DepEd division o ce, which prepares a used for training for elementary school teachers. Just over division-level master plan for professional development. half of all training sessions for elementary school teachers ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 45 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce In the PETS-QSDS team’s interviews with teachers, it became part of the PETS-QSDS survey. This concurs with other ndings clear that a signi cant proportion of teachers and schools that have shown that most teacher training in the Philippines had not developed professional development plans. For is based on a mass training model that provides all teachers example, a quarter of high school teachers had never with very similar training. Only 17 percent of school division prepared an individual professional development plan superintendents (SDSs) said that division o ces based their (Figure 6). Even the plans that had been prepared were over decisions about teacher training on the actual needs of a year old on average for both elementary and high school teachers within the division. The majority of SDSs said that teachers.15 School plans were even less common, with a most division-level teacher decisions about training were quarter of elementary school principals and nearly one-third based on the SDS’s own assessment of needs or on directives of all high school principals reporting that they had never from DepEd’s central o ce. prepared one. While the teachers and school principals that had completed plans agreed that their training was related Support from school principals for teachers’ professional to the goals set out in their plans, the PETS-QSDS training development also tended to vary considerably among data show that the type and duration of training was more or elementary and high schools. Only two-thirds of Grade 10 less the same for those teachers who had lled out an IPPD teachers in high schools reported that the school principal and those who had not. had spent a full period in their classroom over the course of the whole 2013 school year. (The equivalent gure for Grade Other mechanisms to link teacher training to need also did 6 elementary teachers was much higher at 82 percent.) not appear to di erentiate training between teachers with When principals did observe classroom teaching, most of di erent needs. The length and type of training that teachers them gave written comments to the teacher, mostly focused received did not vary signi cantly by their characteristics such on the teaching method used. Only around a quarter of as levels of experience or quali cations. Moreover, the number comments concerned the teacher’s knowledge of the of days of training received by teachers did not di er in line subject content even though the PETS-QSDS assessments with their performance in the assessments administered as indicated that teachers have major weaknesses in this area.16 Figure 6: The Implementation of Systems to Identify Professional Development Needs is Weak Percentage of school sta completing professional development plans and the time since one was completed Completing plan Average duration since last completed (months) 100% 20 75% 15 50% 10 25% 5 0% 0 Teacher plan School plan Teacher plan School plan Elementary High School Elementary High School Source: PETS-QSDS national survey – teacher-level and school/principal data. Notes: Information on teacher plans taken from interviews with Grade 6 (elementary) and Grade 10 (high school) teachers. 46 WWW.WORLDBANK.ORG/PH Figure 7: The Budget for In-service Training Has Increased Recently but Utilization Rates Are Frequently Low Appropriations, allotments, obligations, and utilization rates for HRTD funds (PHP billions in 2014 constant prices), 2005–2014 Appropriations Allotments and obligations 2.0 2.5 57% 2.0 1.5 43% 43% 72% 1.5 34% 81% 1.0 57% 40% 1.0 30% 0.5 0.5 92% 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Allotments Obligations Source: Appropriations - Department of Budget and Management. All other data - DepEd Statement of Appropriations and Obligations, various years. Note: Allotments/obligations for a given year include current, continuing, and extended allotments/obligations. The utilization rate is indicated by the percentage gures in the graph and show the proportion of allotments obligated in a given year. Fundin In-service Te cher Tr inin training activities.17 This process of downloading has often been delayed, which has resulted in relatively low utilization In an e ort to provide more and better in-service teacher rates. For example, between 2005 and 2014, the average training, DepEd has begun to increase the budget allocation utilization rate for HRTD funds was only 55 percent (Figure for human resource training and development (Figure 7). 7). Even in 2014 after the signi cant increase in the HRTD Human resource training and development (HRTD) funds appropriation, only 57 percent of the budget was used are the main professional development resources provided because of delays in downloading these funds. by DepEd. Most HRTD funds are spent on in-service training for teachers, but they are also used to provide training for The PETS-QSDS data revealed that DepEd’s allocations of non-teaching personnel. After having risen gradually since HRTD funds to each region are positively correlated with the 2005, the budget for all in-service training nearly doubled to number of public school teachers in each region. Regions PHP 1.9 billion in 2014 to allow for more in-service training with a larger number of teachers received a larger HRTD related to the new K to 12 curriculum. While all these funds allocation. However, it was not possible to assess whether are not solely for the use of in-service teacher training they HRTD funds were also allocated according to the di erent represent approximately PHP 3,000 (US$70) for every public professional development needs of teachers in each region. basic education teacher. The PETS-QSDS survey also tracked the ow of HRTD funds However, the utilization of these funds uctuates and found that a relatively small share is downloaded to considerably. HRTD funds are managed by DepED’s central DepEd division o ces. In 2014, about half of all HRTD funds o ce, and a portion are downloaded or transferred to were allocated for training activities provided at DepEd’s DepEd’s regional o ces and onward to division o ces for central o ce. The remainder of the HRTD funds was split ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 47 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce relatively equally between DepEd’s regional o ces and HRTD funding and do not download any funding for division division o ces. Given that division o ces and their sta o ces, while other regions download all of their HRTD funds have the most contact with schools and are thus most likely and the responsibility for their use to division o ces. to know their in-service training needs, it is surprising that a larger share of HRTD funds is not downloaded to divisions. Little information is available on how DepEd’s regional o ces used their retained HRTD funds. The regional o ces retained On the whole, the tracking exercise also found that all HRTD approximately 42 percent of all of the HRTD funds that they funds that DepEd central o ce downloaded to regional received from DepEd central o ce rather than downloading o ces were received in full. The speci c amounts released them to division o ces. DepEd expects regional o ces to by DepEd central o ce closely matched the amounts use these funds to organize mass teacher training on, for that the regional o ces reported receiving. However, two example, the introduction of the new K to 12 curriculum. regional o ces did not report having received any funds DepEd’s guidelines governing the use of these funds require even though records in the central o ce showed that funds regional o ces to keep detailed records on who has been had been transferred to them. This is more likely to be due to trained and the kind of training provided. However, when the poor record keeping in DepEd regional o ces than to be a PETS-QSDS study team visited all regional o ces to collect leakage of funds. this information, no such information was available. This lack of records greatly reduces the transparency of the use of The share of HRTD funds downloaded from DepEd’s regional these funds. o ces to division o ces varies enormously (Figure 8). In total, around 58 percent of the regional HRTD funds that are Signi cant delays in the allotment process account for the received by DepEd regional o ces are downloaded to DepEd relatively low utilization rate of HRTD funds. Fewer than 15 division o ces. However, the proportion of funds that regional percent of regional o ces had received their HRTD allotment o ces retain varies enormously. Some regions retain all of their from DepEd’s central o ce by the end of the rst quarter of Figure 8: The Share of Regional HRTD Funds Transferred to Division Offices Varies Enormously Amounts of HRTD funds received by DepEd regional o ces and transferred to DepEd division o ces, 2013 Received by DepEd regional o ces (PHP millions) Share downloaded to division o ces 80 100% 60 75% 40 50% 20 25% 0 0% Region 4 Region 6 Region 3 Region 7 Region 5 NCR Region 8 CARAGA CAR Region 1 Region 2 Region 4B Region 12 Region 11 Region 3 Region 8 Region 4B NCR Region 6 Region 5 Region 4 CARAGA CAR Region 7 Region 1 Region 2 Region 12 Region 11 Source: PETS-QSDS national survey – DepEd regional o ce level. Note: The information on the HRTD funds received by Region 7 from the central o ce is taken from the central o ce database because some information was missing from the regional questionnaire. Only regions reporting receipt of HRTD funds are included. 48 WWW.WORLDBANK.ORG/PH Figure 9: A Significant Number of Division Offices Receive HRTD Funds Very Late in the Financial Year Percentage of division o ces receiving 2013 HRTD funds by the quarter during which it was received 100% 80% 60% 40% % of regional o ces receiving allotment from central o ce % of regional o ces that released allotment to division o ces 20% % of division o ces that received allotment from regional o ce 0% 1st Quarter 2nd Quarter 3rd Quarter 4th Quarter 1st Quarter 2nd Quarter 2013 2014 Source: PETS-QSDS national survey – DepEd region and division o ce levels. Note: Only those sampled divisions that reported receiving HRTD funds are included. 2013 (Figure 9). However, by the end of the second quarter, all regional o ces had already received their allotments Te cher Remuner tion and most had downloaded funds to at least some of their The level and timeliness of salary payments can be an division o ces. However, 19 percent of the divisions that important factor in motivating teachers, which in turn can received HRTD funds received their allotment a year after the a ect levels of student learning. The PETS-QSDS study did regional o ce rst received its transfer of funds from DepEd’s not formally track teachers’ salaries but did ask a nationally central o ce. representative sample of teachers some questions about the adequacy of their salary payments and whether they These delays can happen because the regional o ces of the received their salaries on time and in full. Department of Budget Management (DBM) are required to carefully check the funds requested by DepEd’s central and Cross-country comparisons indicate that teachers in the regional o ces on the Sub-Allotment Release Orders (SARO). Philippines are relatively well paid but their salary scale is Further delays occur because of the late release of cash more compressed. Comparable information on other East allocations. One- fth of the DepEd division o ces that had Asian countries shows that the starting salary of teachers in received their HRTD allotments for 2013 at the time of the the Philippines is relatively high. For example, the average PETS-QSDS survey had not received the associated release of earnings of a newly hired elementary school teacher in funds (notice of cash allocations) needed to use or obligate the Philippines is equivalent to 150 percent of per capita the funds. The di culties in using the allotted funds are GDP compared with only around 50 percent in Indonesia, particularly concerning given the clear need for in-service Malaysia, and Thailand (Figure 10). However, earnings do not training that was highlighted by the PETS-QSDS teacher increase very quickly after teachers are hired. After 15 years, content knowledge assessments. salaries increase by only 15 percent in the Philippines, which ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 49 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce Figure 10: Teacher Salaries Tend to be Higher in the Philippines than in Other Countries in the Region Elementary school teacher salaries as a percentage of per capita GDP, 2010 300% 200% 100% 0% Indonesia Malaysia Thailand Sri Lanka Japan Republic of Korea Philippines Starting salary Salary after 15 years of teaching Salary at top of scale Source: UNESCO (2012). “Global Education Digest,” UNESCO Institute for Statistics, Montreal. is low compared to other countries. These relatively small government. However, the length of the delay was much salary increases over the course of a teacher’s career are longer for elementary school teachers than for high school unlikely to motivate them signi cantly. The PETS-QSDS study teachers. Elementary school teachers had to wait an average found that high school teachers with a post-graduate degree of six months for their salary payments compared with a felt that their remuneration was not adequate. Moreover, four-month wait for high school teachers. about one- fth of teachers reported that they had to have other jobs or sources of income to supplement their earnings from teaching. Figure 11: Some Teachers Experienced Delays in Receiving their Salaries and Others However, teachers also receive a bonus based on their Were Still Owed Back Payments school’s ranking on a set of performance-based indicators Percentage of teachers whose salary payments including how well the school’s students score on the were late, 2014 National Achievement Test. Teachers’ performance-based 60% bonuses in 2015 ranged from PHP 5,000 to PHP 35,000, which was equivalent to between 1 and 9 percent of a teacher’s average annual salary.18 40% Teachers’ motivation can also be a ected by the timeliness of their salary payments. The PETS-QSDS study found that, while the majority of teachers were paid on time, 40 to 50 percent of newly hired kindergarten and elementary school 20% teachers received their salaries late (Figure 11). Most teachers also reported that they had been paid 0% correctly and were not owed any salary payments. Only Kindergarten Grade 6 Grade 10 around 11 percent of elementary and high school teachers reported that they were owed salary payments from the All teachers Newly hired teachers Source: PETS-QSDS data from DepEd teacher questionnaires for sampled teachers, 2014. 50 WWW.WORLDBANK.ORG/PH Polic Directions for grade learning outcomes. These cells, or groups, of kindergarten to grade 3 teachers meet regularly and Stren thenin S stems provide opportunities for members to learn from each other and to collaborate on strategies to improve to Support Te cher teaching and learning. These school learning action cells Development have the potential to increase the quantity and improve the quality of in-service training opportunities for teachers. However, it is vital that these cells are monitored This note has shown that teacher competencies are and evaluated closely by DepEd to ensure that their weak and that systems to support teacher development potential is fully realized. are inadequate. While some of the gaps in teacher competencies are partly due to weaknesses in pre-service This note has also shown that the monitoring of the use of teacher training and induction, more e orts need to be HRTD funds has been weak and needs to improve if these made to increase the support available to teachers who resources are to be used more e ectively. Information are already teaching in schools. on who is being trained and the type of training being provided is often unavailable. This makes it impossible A rst step would be to raise levels of funding to increase to monitor how the funds are being used and whether both the duration and quality of in-service training for teachers’ professional development needs are being met. basic education teachers. As this note has shown, a Developing simple reporting formats for DepEd’s regional signi cant proportion of HRTD funds are used for training and division o ces to record how they use HRTD funds that is organized at the regional level and above. Previous would greatly facilitate the monitoring and evaluation of studies have found that the most successful professional in-service training provision by DepEd. development models are provided within schools or at the local level. They also provide opportunities The existing system for identifying and planning for teachers to collaborate and support each other in professional development activities needs to be implementing new knowledge and techniques at the strengthened. This note has shown that the TSNA may school level.19 not be adequate for identifying teachers’ professional Providing divisions and schools with a larger share of development needs. Developing a diagnostic teacher professional development funding would make training competency assessment to better identify teachers’ more relevant and thus result in more substantial in-service training needs could support improvements improvements in teachers’ competencies. DepEd could in this area. Moreover, neither teachers nor schools seem revise its guidelines on the use of HRTD funds to require to be routinely preparing their professional development DepEd’s regional o ces to download a larger proportion plans, and these plans do not seem closely related to the of these funds to divisions and schools. Currently the in-service training opportunities currently being provided. DepEd guidelines encourage the provision of mass Greater e orts need to be made to ensure that teachers training at the regional level, which makes the training and schools use these plans to help division o ces to plan too remote from the actual needs of teachers at the in-service training opportunities better. school level. If more funds are provided to DepEd’s division o ces and schools, it will also be vital to improve Stronger links between teacher performance appraisals, the timeliness of fund release to give the recipients professional development opportunities and career time to properly plan and implement their professional development for teachers may create stronger incentives development activities. for teachers, school principals and administrators to utilize these systems and strengthen professional development DepEd is currently establishing “school learning opportunities. For example, individual and school plans for action cells” in elementary schools to strengthen early professional development and their implementation could ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 51 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce be used more speci cally in decisions over promotion or in improvements have been made in terms of the number and determining performance bonuses (see Box 2). distribution of teachers, there is an urgent need to provide greater support to teachers to improve their competencies Increasing the number and e ectiveness of teachers has and e ectiveness in the classroom. Only then will e orts to been a central component of e orts in the Philippines provide adequate teachers to all schools translate into better to raise the quality of basic education. While signi cant quality and improved learning outcomes. Box 2: Indonesia’s Integrated Framework for Teacher Accountability The Ministry of Education in Indonesia is starting to implement its integrated Teacher Professional Management System (TPMS). The system consists of three main elements: • Competency Testing. Teachers are expected to undergo a competency assessment to identify their strengths and weaknesses. The competency test is focused on teachers’ subject knowledge, which has been shown to be strongly related to student learning outcomes. Performance on the competency test is expected to be used as one of the key criteria for teachers’ career advancement. • Performance Appraisal. A school-based scheme has been piloted that links the outcomes of teachers’ annual performance appraisal to increments in the salary scale in order to give them an incentive to improve their performance. • Continuous Professional Development. This component covers a number of di erent aspects of in-service teacher training including the induction of teachers into schools, mentoring, and the activities of local teacher working groups. The aim of the TPMS is to link these three components together and make teachers more accountable for both their work performance through the annual appraisal and for taking advantage of in-service training opportunities. By scoring well on their appraisals and by participating in training, teachers can earn credit points that are ultimately linked to salary increases and promotion. While the TPMS is not yet being fully implemented, it represents a major development in terms of the provision of incentives and is expected to increase the motivation of teachers to improve how they teach. Integrating continuous professional development into the TPMS framework has also provided the government with the information necessary to target teachers’ professional development activities better. It is expected that the TPMS will sustain the momentum towards improving the quality of teachers in Indonesia and will establish a quality assurance mechanism that will ensure higher education standards well into the future. Sources: Ragatz, A. (2015). “Teacher Quality and Management. Background Study for the Preparation of the Education Chapter of the National Development Plan, 2015–2019.” Jakarta, Ministry of Planning, Republic of Indonesia., and Chang, M.C., S. Shae er, S. Al-Samarrai, A. Ragatz, J. de Ree, and R. Stevenson (2013). “Teacher Reform in Indonesia: The Role of Politics and Evidence in Policy Making.” Directions in Development. World Bank, Washington, D.C. 52 WWW.WORLDBANK.ORG/PH Table 1: Strengthening Teacher Support Systems Findings Policy suggestions Teacher subject knowledge is • Strengthen pre-service and in-service training opportunities for all teachers (see below) generally low Professional development • Increase funding and opportunities for e ective professional development for teachers opportunities are limited and delivery mechanisms are • Provide a greater share of this funding to division o ces and schools to enable them weak to organize professional development activities • Transfer HRTD funds in a more timely manner • Increase the transparency of HRTD funds through improved reporting by DepEd region and division o ces and increase central o ce monitoring of fund use Systems to identify teachers’ • Revaluate existing teacher strength and needs assessments and teacher professional professional development development planning needs are weak • Develop diagnostic teacher competency tests to identify teachers’ in-service training needs • Develop content for professional development activities that will address existing weaknesses • Evaluate and scale up the use of school learning action cells as the main venue for in-service training A greater need to align • Support a stronger alignment between professional development planning, career teacher accountability development and performance incentives systems to raise teacher motivation and competencies ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 53 Policy Note 2: Developing a Proficient and Motivated Teacher Workforce Endnotes 1 See for example, Bruns, B. and J. Luque (2014). “Great Teachers: 11 See, for example, McEwan, P. (2013) “Improving Learning in How to Raise Student Learning in Latin America and the Primary Schools of Developing Countries: A Meta-Analysis of Caribbean.” World Bank, Washington D.C. Randomized Experiments” Wellesley College, Wellesley, MA; 2 De Ree, J. (2016). “How Much Teachers Know and How Much It Glewwe, P. W., E. A. Hanushek, S.D. Humpage, and R. Ravina (2011) Matters in Class: Analyzing Three Rounds of Subject-speci c Test “School Resources and Educational Outcomes in Developing Score Data of Indonesian Students and Teachers.” Policy Research Countries: A Review of the Literature from 1990 to 2010” National Working Paper No. WPS 7556. World Bank, Washington D.C. Bureau of Economic Research, Cambridge MA; and the Council of Chief State School O cers (2009) “A Meta-Analysis Study of 3 Chetty, R., J. Friedman, and J. Rocko . (2011). “The Long-Term the E ects of Teacher Professional Development with a Math or Impacts of teachers: Teacher Value-added, and Student Science Content Focus on Improving Teaching and Learning.” Outcomes in Adulthood,” National Bureau of Economic Washington D.C. Research, Cambridge MA. 12 OECD (2014). TALIS 2013 Results: An International Perspective on 4 See, for example, McEwan, P. (2013). “Improving Learning in Teaching and Learning, OECD Publishing, Paris. Primary Schools of Developing Countries: A Meta-Analysis of Randomized Experiments.” Wellesley College, Wellesley MA. 13 OECD (2014). “TALIS 2013 Results: An International Perspective on Glewwe, P. W., E. A. Hanushek, S.D. Humpage, and R. Ravina. Teaching and Learning.” OECD Publishing, Paris. (2011). “School Resources and Educational Outcomes in 14 The sampled teachers were asked to rank each training they Developing Countries: A Review of the Literature from 1990 to attended from 1 (least useful) to 7 (most useful). Grade 6 teachers 2010.” National Bureau of Economic Research, Cambridge MA. rated 88 percent of all training as a 6 or 7 with 84 percent of 5 Another note in this series assesses the systems used to hire and Grade 10 teachers doing the same. deploy teachers in the Philippines. 15 Just under 90 percent of school principals reported that teachers 6 See for example, Yamauchi, F. and Y. Liu (2013). “Impacts of were expected to complete an IPPD annually. an Early Stage Education Intervention on Students’ Learning 16 Similar results were found when school principals were asked the Achievement: Evidence from the Philippines.” Journal of same questions. Development Studies 49(2) and Tan, J. P., J. Lane, and P. Coustere. 17 Downloading refers to the issuance of a sub-allotment release (1997). “Putting Inputs to Work in Elementary Schools: What Can order from DepEd central o ce to DepEd region and division Be Done in the Philippines?” Economic Development and Cultural o ces. Sub-allotments are authorizations issued by the central Change 45(4): 857-879. o ce of DepEd transferring a portion of an available allotment 7 The full results of the analysis reported here is included in a set to DepEd region or division o ces. See DepEd order No. 25 of additional annexes and tables accompanying the main PETS- (2014) “Guidelines for the Utilization of HRTD Funds,” Department QSDS report. of Education, Manila and DepEd order No. 66 (2010) “Policies 8 The household questionnaire included a short module on and Guidelines on Planning and Administration/Management consumption and a set of questions on assets that have been of the Human Resource Training and Development Program,” used by the Department of Social Welfare and Development Department of Education, Manila. (DSWD) to undertake a proxy means testing (PMT) approach 18 Based on the average annual basic salary of a Teacher 1 position to estimating household consumption per capita. The results on Grade 18 Step 8. reported here are based on information gathered using the PMT 19 Blank, R. and N. Alas (2009). “E ects of Teacher Professional approach, and a full description is included in a separate note. Development on Gains in Student Achievement.” Report 9 These results are statistically signi cant for Filipino and prepared for the Council of Chief State School O cers (CCSSO), Mathematics but not for Science. Washington D.C. 10 Box 2 describes all seven domains contained in the teacher competency standards. All teachers rated their skills in the professional development and growth domain highly. These results are not discussed in this note. 54 WWW.WORLDBANK.ORG/PH Polic Note 3: Buildin Better Le rnin Environments Introduction Evidence from around the world has shown that improving school infrastructure leads to better learning outcomes.1 For example, a 2011 review of the economics literature over the last 20 years showed that the availability of basic school infrastructure (such as classrooms, desks, and chairs) and facilities (such as electricity, libraries, and blackboards) is frequently associated with better student learning achievement.2 These ndings have been backed up by a systematic review of recent impact evaluations, which showed that infrastructure investments have a positive impact on school enrollment rates, attendance rates, and learning achievement.3 Also, research in the Philippines has shown that reducing the number of students per classroom is associated with better student learning outcomes, particularly in rural schools.4 Ensuring that schools have adequate infrastructure of good quality is a central pillar of government e orts to improve education outcomes in the Philippines. Over the last ve years, the government has been conducting a renewed reform e ort to increase access to good quality basic education. It has supported these reform e orts with signi cant increases in public spending on education. In real terms, school infrastructure spending in 2015 was ve In assessing government systems, it primarily focused on times higher than in 2010 and has been used to reduce projects managed by the Department of Public Works and classroom congestion and the proportion of schools Highways (DPWH), which is responsible for the bulk of school operating multiple shifts. infrastructure projects. The purpose of this policy note is to provide a snapshot of The note nds that, while there is a continuing need to invest the state of existing school infrastructure in the Philippines in school infrastructure, any increases need to be combined and to assess the government systems that build and with e orts to improve allocation mechanisms and project repair school buildings and facilities. The ndings come implementation. The note also nds that existing monitoring, from a comprehensive survey of the public education coordination, and accountability mechanisms are weak. The system carried out for the Philippines Public Education note argues that increasing the involvement of schools in Expenditure Tracking and Quantitative Service Delivery the planning, implementation, and monitoring process is a Study (PETS-QSDS) that tracked public education promising route to ensuring that all students are provided expenditure and assessed the quality of education services. with good quality schools and learning environments. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 55 Policy Note 3: Building Better Learning Environments The Current St te of detrimental to attendance and learning, particularly when students and teachers have to leave the school premises to School Infr structure collect water. Schools in municipalities tend to have more limited facilities The bulk of public school infrastructure in the Philippines was built using funds from national and local governments, than those in cities. Municipalities include almost all rural with only a limited proportion built using funds from private areas in the Philippines, and rural schools tend to be less sector sources. For example, in 2014 approximately three- well equipped than their urban - and generally wealthier quarters of all instructional rooms in elementary and around - counterparts. For example, elementary schools in highly two-thirds of rooms in high schools were built by the central urbanized cities are almost twice as likely as schools in government. Local governments were responsible for municipalities to have a health clinic (Figure 1). building around 10 percent of instructional rooms across the whole country, although wealthier city governments have However, schools in more densely populated areas tend to played a much more signi cant role than their counterparts be larger and have more congested learning environments. elsewhere with approximately one-quarter of all instructional Schools, particularly high schools, tend to be much larger buildings being built with local government funds in highly in city and urban areas. For example, in 2014 the average urbanized cities.5 high school in a highly urbanized city had approximately 1,700 students compared to about 1,000 students and While high schools tend to be better equipped than 570 students in city and municipality schools respectively. elementary schools, around one-third have no safe source These large numbers translate into more students in each of drinking water (Figure 1). This is partly because more high classroom in highly urban areas (Figure 2). On average, schools than elementary schools are located in urban areas 51 high school students share each classroom in highly where they rely on piped water and do not have access to urbanized cities compared with only 39 students per alternative sources such as drinking wells. When schools classroom in municipal schools. Moreover, around 30 percent do not have access to safe drinking water, this can be of high schools in these highly urbanized cities have student/ Figure 1: High Schools Tend to Have Better Facilities than Elementary Schools Availability of key facilities in elementary and high schools, 2014 Facilities by school type Elementary school facilities by location 90% 90% 80% 80% % of elementary schools 70% 70% 60% 60% % of schools 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Library Working Drinking Health Library Working Health internet water* clinic internet clinic High School Elementary School HUC Other Cities Municipalities Source: PETS-QSDS national survey – school level data. Note: * Drinking water only includes piped, well, and natural sources. 56 WWW.WORLDBANK.ORG/PH Figure 2: School Facilities Tend to be Used by More Students in Urban Areas Intensity of use and quality of classrooms, 2014 Students per instructional room State of instructional rooms 20% Students per instructional room 60 16% % of instructional rooms 50 15% 14% 40 13% 13% 12% 30 10% 20 6% 5% 10 0 0% Elementary School High School HUCs Other Cities Municipalities HUCs Other Cities Municipalities HUC Other Cities Municipalities Elementary School High School Condemned Condemnable Source: PETS-QSDS national survey – school-level data. Note: DepEd’s de nitions of condemned and condemnable instructional rooms are used. instructional room ratios in excess of 55. While this is not Aggregate statistics on the availability and quality of school a measure of class size because double shifting is more infrastructure often do not adequately represent the true commonplace in cities, it does point to the greater intensity learning environment in schools. Most statistics use student- of use of school facilities in urban areas. This is likely to result classroom ratios as a proxy for the average number of in the need for larger upkeep and maintenance resources for students taught together. However, many schools operate schools in these areas. more than one shift during the school day, and student groupings and timetables often mean that class sizes are In 2014, the study found that approximately one in very di erent to the simple averages recorded in o cial seven of all elementary and high school instructional statistics at the school level. For example, approximately 23 rooms in the Philippines to be unsuitable for teaching percent of both elementary and high schools in the study and learning. On the whole, a larger share of elementary reported operating more than one shift in at least one grade. school infrastructure than high school infrastructure was To get a better picture of the real learning environment of poor quality (Figure 2). Given the much larger numbers faced by students, around 7,000 classes and classrooms of elementary schools in the Philippines, these ndings were observed as part of the Philippines Public Education suggest that a larger share of resources would be needed Expenditure Tracking and Quantitative Service Delivery Study at this level to address shortcomings in the existing stock of (PETS-QSDS). classrooms. Fewer classrooms in cities tend to be classi ed as condemned or condemnable than in municipalities despite These classroom observations revealed that class sizes were their greater intensity of use. This may re ect their greater much smaller than o cial statistics on student-classroom ability to maintain their facilities because city schools tend ratios suggest. The average class size observed by the study in to receive more resources from local governments and their 2014 was 34 in high schools and 27 in elementary schools. This communities that can be used for this purpose. is below DepEd’s guidelines on ideal class sizes, which range ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 57 Policy Note 3: Building Better Learning Environments Figure 3: A Significant Proportion of Classrooms Were Observed to Be of Poor Quality Indicators of classroom quality from direct classroom observation 30% High School 25% Elementary School % of observed classrooms 20% 15% 10% 5% 0% Poor state Unclean Poor Poor sound Without Inadequate of repair ventilation insulation electricity seating Source: PETS-QSDS national survey – classroom observations at the school level. from 40 to 55 students. According to the study, only While most schools meet DepEd’s service standards, 5 percent of classrooms in elementary schools and 12 a signi cant proportion do not, particularly in highly percent in high school had more than 45 students being urbanized cities. The government’s medium-term taught at the same time. expenditure framework highlights a number of key service standards that DepEd is seeking to achieve.6 In A signi cant number of classrooms used by schools for particular, DepEd aims to eliminate multiple shifts in teaching were judged by the enumerators of the PETS- all schools and has set a target for student-classroom QSDS survey to be in a relatively poor state of repair ratios of between 45 and 55 in most elementary and in 2014 (Figure 3). The survey enumerators found that high school grades.7 Over 94 percent of elementary over 20 percent of classrooms in both elementary and schools and 83 percent of high schools have student- high schools were in a poor state of repair. While the classroom ratios that fall within or below that range enumerators were not speci cally trained to assess the (Table 1). However, 24 percent of elementary and 30 quality of infrastructure, these ndings do raise concerns percent of high schools in highly urbanized cities have about the impact that this may have on student student-classroom ratios well above the maximum set learning. out in DepEd standards. The more intensive use of high school infrastructure also a ects the learning environment negatively. In most A large proportion of schools also fail to meet DepEd cases, high schools have poorer indicators of the overall standards regarding sanitation facilities (Table 1). In learning environment than smaller and less crowded particular, around one-third of high schools fail to meet elementary schools. For example, as of 2014, a greater the 50 students per toilet standard. The proportion of proportion of high school classrooms tended to su er schools failing to meet sanitation standards is much from poor ventilation and insulation and to be less higher in city schools than in municipal schools. clean than elementary classrooms (Figure 3). In contrast, For example, 70 percent of high schools in highly far fewer elementary classrooms than high school urbanized cities have ratios of female students to toilets classrooms appeared to have adequate seating for in excess of 50. all students. 58 WWW.WORLDBANK.ORG/PH Table 1: Percentage of Schools that Failed to Meet DepEd Service Standards, 2014 Elementary Schools High Schools Other All Other All HUCs Cities Municipalities schools HUCs Cities Municipalities schools Students per instructional room Fewer than 45 46 74 87 84 61 40 70 66 Between 45 and 55 30 19 9 10 9 16 19 17 More than 55 24 7 4 6 30 44 11 17 Students per toilet More than 50: Girls 41 22 21 22 70 56 32 38 More than 50: Boys 35 14 19 19 65 49 33 37 Students per seat More than 1 21 24 23 23 7 6 7 7 Source: PETS-QSDS national survey – classroom data at the school level. Notes: Assumes that a trough urinal can be used by two boys. Unisex toilets are counted in both boys’ and girls’ totals. All students, including kindergarten and special education students, are included in the calculations. Numbers of students per seat are taken from classroom observations and relate to the percentage of classrooms rather than schools. An Assessment of that infrastructure spending has increased nineteen-fold in real terms since 2005. Existin Government The increased priority a orded to school infrastructure in Efforts to Improve School the national budget has resulted in a large increase in the Infr structure number of classrooms built, which has eased congestion, particularly in high schools. National statistics show that between 2005 and 2013 an additional 122,000 elementary In 2013, approximately two-thirds of all infrastructure projects and high school classrooms were built, increasing the overall taking place in basic education schools were funded by the stock to 477,000. These increases have also reduced student- Department of Education. Local governments provided 13 classroom ratios particularly in high schools. Between 2005 percent of project funding, while the rest of the projects and 2013, the average number of students per classroom fell were largely nanced from congressional funds, by the from 70 to 47 in high schools. private sector, or by non-governmental organizations. While the Department of Education maintains overall control The government budget devoted to improving and expand- over the selection of projects, most projects have recently ing school infrastructure has risen rapidly in recent years been managed and implemented by the Department of (Figure 4). Recognizing the backlog in school infrastructure Public Works and Highways (Box 1). Since 2013 the Basic needs, the government has devoted an ever-increasing share Education Facilities Fund (BEFF) and the School Building of the budget to construction and rehabilitation. In 2005 only Program (SBP) have been the two main budget lines around 2 percent of the budget was used for infrastructure, for school infrastructure, with the BEFF accounting for but this had increased to 18 percent by 2015. Given the in- over 95 percent of the total funds. The DPWH manages creasing size of the overall education budget, this has meant approximately three-quarters of these funds, most of ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 59 Policy Note 3: Building Better Learning Environments Figure 4: Government Spending on School Infrastructure Has Been Rising Rapidly Trends in government school infrastructure spending, 2005-2015 60 20% 50 16% PHP billion (2014 prices) % of DepEd budget 40 12% 30 8% 20 4% 10 0 0% 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 DepEd Infrastructure Funds DepEd School Building Program % of total DepEd budget (right axis) DPWH local infrastructure support for schools and basic education facilities Source: Department of Budget and Management. Note: All gures are appropriations. DepEd school infrastructure funds include all DepEd infrastructure spending including funds under construction of elementary and high schools in areas experiencing acute classroom shortages prior to 2011 and the Basic Education Facilities Fund (BEFF) from 2011 onwards. In 2013, the Department of Public Works and Highways (DPWH) received an appropriation for the BEFF directly. The 2014 and 2015 gures also include separate DPWH programs for local infrastructure support to schools. Box 1: Responsibilities for Joint DPWH-DepEd Projects Each year, DepEd identi es the extent of school infrastructure needs using the Basic Education Information System (BEIS) and develops a list of school-level projects to be carried out by the DPWH using the infrastructure budget appropriations for the year. Once the DepEd central o ce has selected the projects, it informs the regional and divisional o ces who in turn notify the selected schools. The Schools Division Superintendent (SDS) and the Schools Division Engineer (SDE) within DepEd Division o ces are responsible for coordinating with their DPWH counterparts on the procurement and implementation of projects. The main stages where coordination occurs are: • The DPWH prepares the program of works, undertakes procurement, and implements the works in accordance with DepEd’s standard designs and speci cations. • The SDS approves the program of works prepared by the DPWH. • The sta of DepEd’s Physical Facilities and Schools Engineering Division (PFSED) and the Division Physical Facilities Coordinator (DPFC) supervise the DPWH’s procurement, monitor the implementation of the projects, and conduct joint inspections of the completed works. • The DPWH, DepEd’s Physical Facilities sta , and the school principals are expected to carry out a joint inspection of the completed works. • The DepEd SDS and the DPWH District Engineer both approve the Certi cate of Completion for each completed project. Sources: DepEd Memo No. 87 (2015) and DepEd/DPWH joint circular No. 2013-1 (2013). 60 WWW.WORLDBANK.ORG/PH which are allocated to new construction of classrooms In the second step through site visits, the o ce veri es the and water and sanitation facilities. The remaining funds are actual condition of facilities and the feasibility of the needed managed by DepEd and used for repair and rehabilitation, work and nally prepares a list of potential projects that additional water and sanitation projects, and the provision are aligned with the available budget. Once the national of school furniture. Funds managed by DepEd are also budget is approved by Congress, the Department of Budget devoted to infrastructure projects carried out through the and Management releases budget allocations to the DPWH Public Private Partnership scheme.8 In the 2013 budget, and DepEd central o ces, which in turn allocate these national government funding was allocated to support the funds to their local o ces in charge of the implementation construction of 16,000 classrooms and the repair of 8,000 of the various projects. Projects managed by the DPWH classrooms and a large number of toilets and washroom are governed by a set of guidelines that outlines the roles facilities (Table 2). and responsibilities of DepEd and the DPWH during their planning and implementation (see Box 1). The central o ce of DepEd identi es priorities and allocates funds for infrastructure projects using information collected Alloc tions annually in the Basic Education Information System (BEIS). The allocation of government infrastructure resources tends First, the o ce identi es schools with high student- to be broadly in line with need. The study has found that new classroom ratios and a lack of water and sanitation facilities. classroom construction as a proportion of the total existing Table 2: Characteristics of School Infrastructure Projects, 2013 Basic Education Facilities Fund School Building Water and Program Construction Sanitation Repair Projects Elementary Schools 3,597 8,794 1,656 505 High Schools 1,765 3,768 367 189 Classrooms built/repaired Elementary Schools 8,204 5,696 1,081 High Schools 6,524 1,749 496 Toilets (urinals and bowls) Elementary Schools 3,012 32,607 - - High Schools 4,795 18,355 - - Hand washroom counters Elementary Schools - 34,224 - - High Schools - 19,878 - - Allocation (PHP billions) Elementary Schools 7.1 3.5 0.9 0.6 High Schools 6.9 1.9 0.3 0.3 Other - 1.3 0.1 0.1 Allocation (%) Elementary Schools 51 52 69 60 High Schools 49 28 23 30 Source: DepEd central o ce construction project list, 2013. Note: Estimates drawn from school project lists for each funding source. Of the 1,081 elementary and 496 high school classrooms, 674 elementary and 325 high school classrooms were newly constructed while 325 and 171 respectively were repaired. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 61 Policy Note 3: Building Better Learning Environments stock of classrooms was generally higher in those regions with in excess of DepEd standards (in other words higher than more congested classrooms (Figure 5). For example, in the 55) were selected for a project, while 30 percent of schools 2013 budget, funds were allocated to build an additional 800 with student-classroom ratios well below 45 received school classrooms (equivalent to 5 percent of the current stock) in the improvement funding. A similar picture emerges regarding National Capital Region (NCR) where student-classroom ratios sanitation facilities, with 28 percent of elementary schools were far higher than any other region. Classroom construction with a female student to toilet ratio below 50 being selected in the Autonomous Region in Muslim Mindanao (ARMM) in for a project while only 55 percent of elementary schools 2013 was much higher than in other regions, which partly with ratios in excess of 60 received support. re ects the region’s need for additional school places to accommodate out-of-school children. Implement tion A closer look reveals that on the whole schools with Utilization rates for school infrastructure funds have been poorer quality facilities are more likely to be selected for relatively low over the last 10 years (Figure 6). Between 2005 a school improvement project (Figure 5). For example, a and 2014, only 54 percent of allotments were obligated in the higher proportion of the schools that were awarded an same year they were allocated. With the exception of 2013, infrastructure project in 2013 had no access to drinking water DepEd has experienced signi cant di culties in utilizing and had fewer classrooms and toilets relative to the size of the growing budget for school infrastructure development. their student populations compared to schools that did not For example, in 2014, the utilization rate was only around receive a project. two-thirds, suggesting that one-third of allocations were not obligated. However, there is still signi cant room to strengthen the link between allocation and need. Approximately half of Utilization rates mask large di erences in budgeted funds the elementary schools that had student-classroom ratios and the amounts actually released in any given year. In the Figure 5: New Classroom Construction is Broadly in Line with Need at the Regional Level Infrastructure needs and project allocations, 2013 New elementary school classroom construction under Di erences in facility quality between elementary schools BEFF and student-classroom ratio by region with infrastructure projects and those without, 2013 12% 51 School with project ARMM School without project New construction as % of stock 10% 40 8% 30 30 6% NCR 20% 4% CARAGA IX IVB XII 12% CAR VIII XI IVA 2% X 6% III 4% II VI V I VII 0% % of schools % age of Students per Girls per 20 30 40 50 60 70 without classrooms instructional toilet drinking condemned room Student classroom ratio 2013/14 water Source: Left panel = DepEd central o ce construction project list (2013) and DepEd EBEIS. Right panel = PETS-QSDS DPWH and DepEd records of school construction projects and school-level information on school facilities. Note: School Building Program construction is not included. 62 WWW.WORLDBANK.ORG/PH Figure 6: Utilization Rates for School Infrastructure Have Been Relatively Low School infrastructure fund allotments, obligations, and utilization rates, 2005-2014 35 Allotment Obligation 30 91% PHP billions (current) 25 58% 64% 20 15 63% 10 35% 35% 65% 48% 79% 5 0% 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Source: DepEd and DPWH Statement of Appropriations and Obligations, various years Note: Allotments/obligations for a given year include current, continuing, and extended allotments/obligations. The utilization rate is indicated by the percentage gures in the graph and show the proportion of allotments obligated in a given year. 2013 budget, Congress appropriated approximately PHP Figure 7: Schools Report Low Levels of Satisfaction 27 billion for school infrastructure improvements, but only with Government-funded Infrastructure around 70 percent of this appropriation was released in 2013. Projects Even taking into account the releases from this appropriation Completion and satisfaction rates for 2013 school infrastructure projects, 2014 in the following year (2014), only 74 percent of the 2013 appropriation was nally released. This is a particular concern 100% given the infrastructure shortages outlined in the previous section and the large increases planned for infrastructure 80% investment in the coming years.9 Even when national government projects have been 60% identi ed and funds have been released, a signi cant proportion of projects are not completed, which results in 40% low levels of satisfaction at the school level. School sta reported to enumerators of the PETS-QSDS survey that 20% one in ve DepEd projects that began in 2013 remained incomplete by the nal quarter of 2014 (Figure 7). They cited insu cient funds, other demands on contractors’ 0% Completion rate Satisfaction rate time, and design issues as the main reasons why projects were not completed. Moreover, completion rates were DepEd Local government Own funds much higher for local government and school-managed Source: PETS-QSDS national survey – infrastructure project information at projects than projects managed by DepEd and the DPWH. the school level The high completion rates of school-managed projects is Note: Data relates to all school infrastructure projects in 2013. Completion likely to re ect the greater control that schools have over rates are the proportion of 2013 projects that were completed by the last quarter of 2014. DepEd projects include those managed by the DPWH as contracts and contractors when they are managing their well as major school repair and water and sanitation projects managed by own resources. DepEd Division o ces. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 63 Policy Note 3: Building Better Learning Environments Figure 8: The Timeliness of the Downloading of Infrastructure Allotments from Central Offices Varies by Budget Source Receipt of project list and rst sub-allotment at DPWH district engineering o ces11 100% 80% % of DPWH o ces 60% 40% 20% 0% End of rst quarter End of second quarter End of third quarter After third quarter (March) (June) (September) BEFF construction SBP construction BEFF WATSAN Source: PETS-QSDS national survey – DPWH district engineering o ces. Approximately 40 percent of national and local government infrastructure spending, this is a positive nding. However, infrastructure projects undertaken in 2013 were rated as around 15 percent of engineering o ces did not receive unsatisfactory by schools (Figure 7). The schools reported these funds until the end of the second quarter of the year. being much more satis ed with the projects that they managed and nanced themselves. A major factor in this Also, the majority of DPWH o ces received other infra- dissatisfaction, particularly with projects nanced by the structure funds (such as the BEFF water and sanitation and national government, is the large number of projects that school building program funds) much later in the year. For are left incomplete by contractors. Moreover, approximately example, regarding funds from the School Building Pro- three-quarters of schools also mentioned the poor quality gram, only around three-quarters of the o ces had received of the repairs and new construction as a reason for their their allocations for BEFF water and sanitation projects and dissatisfaction. fewer than one-third of them had received their allocations for school construction by June. Receiving funds so late Delays in transferring project lists and nancial in the year limits the time available to implement projects allocations from DepEd to the local o ces responsible and means that they are still being implemented when the for implementing them are the reason for many project school year starts in June, thus inconveniencing school sta implementation delays. Figure 8 shows the proportion of and students.10 DPWH district engineering o ces that had received their infrastructure allocation by the end of each quarter in 2013. A majority of district engineering o ces reported that, It shows that over 80 percent of DPWH district o ces had even after they have received their resources, they have received their 2013 allocation to carry out work on the BEFF to overcome many di culties to implement their projects construction component by the end of the rst quarter. The successfully. Two-thirds of DPWH district engineering o ces prompt release of these funds gives district engineering and all o ces in urban areas reported nding it di cult to o ces time to plan out their infrastructure projects over the implement school infrastructure projects, particularly due course of the nancial year. Given that BEFF construction to a lack of exibility in the funding of school projects. For funding represented the bulk of national government example, 62 percent of o ces reported that they did not 64 WWW.WORLDBANK.ORG/PH receive su cient funds to cover hauling away debris and one-quarter of projects reported by schools as DepEd- 33 percent complained of the impractical costing of school nanced projects could not be matched with information projects (Figure 9). A further issue raised by over 60 percent from the relevant DepEd divisional or DPWH district o ce. of the o ces was a lack of space within the schools to locate Third, among those projects that appeared in both sets the new infrastructure. of data, about 40 percent had di erent descriptions. For example, schools reported the project as being a water and The PETS-QSDS survey asked sta of the DPWH and DepEd sanitation project while DepEd or the DPWH reported it divisional o ces about the infrastructure projects that they as a construction project. The project costs also frequently undertook within the surveyed schools. The study team then di ered. compared this information with similar information collected within the schools themselves to check whether the projects It is hard to draw rm conclusions about whether all were actually carried out and to assess the accuracy of infrastructure funds are reaching their intended destinations. systems used to monitor the use of infrastructure funds. On the one hand, a signi cant number of projects recorded at the district or division level were not implemented in There were large discrepancies between implementing schools. On the other hand, schools reported projects o ces and schools in their accounts of infrastructure funded by DepEd that were not recorded in the DepEd or projects. The analysis was carried out only on elementary DPWH o ces. However, it does seem clear that existing schools since the number of projects documented in high information and monitoring systems are not able to schools was not su cient for the team to be able to draw accurately record the use of infrastructure funds. any reliable conclusions. First, the elementary schools reported that fewer DepEd-funded projects had been Weaknesses in monitoring and feedback systems are also implemented than was reported by DepEd and the DPWH. highlighted by other ndings from the survey. Managers Second, only around one-half of the projects that the DepEd of projects in the DepEd and DPWH o ces reported that and DPWH o ces claimed to have implemented could be almost all infrastructure projects in 2013 had already been veri ed by information provided by the schools. However, completed even though the schools reported a completion Figure 9: Many DPWH Offices Reported Facing Difficulties in Implementing and Supervising School Infrastructure Projects Common problems faced by the DPWH in implementing school infrastructure projects, 2014 70% 60% % of DPWH o ces reporting 50% 40% 30% 20% 10% 0% Insu cient funds for hauling Impractical uniform pricing Attracting contract bidders Speci cations too rigid Political intervention Insu cient Late receipt of project list Site availability DPWH sta Source: PETS-QSDS national survey – DPWH district engineering o ces. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 65 Policy Note 3: Building Better Learning Environments rate of only 80 percent (see Figure 7). The widespread The ndings from the PETS-QSDS study also show dissatisfaction with projects at the school level also does not that government systems for delivering infrastructure appear to be re ected at higher levels of the administration. improvements need to be strengthened and made Nearly all respondents in DepEd and DPWH o ces reported more accountable if any increased funding is to be used that they were satis ed with the projects undertaken in the e ectively. First, the methods used to identify and target surveyed schools, whereas only 57 percent of school heads school improvement projects need to be strengthened to were satis ed. ensure that the neediest schools are prioritized. Second, implementation capacity needs to be increased to ensure Social accountability initiatives aimed at soliciting feedback that government systems are t for purpose and are able and providing oversight on the state of school infrastructure to absorb the funds needed to ll infrastructure gaps. also do not appear to be widespread. There are many civil Finally, monitoring and evaluation e orts need to be society initiatives in the Philippines that aim to ensure that stepped up and mechanisms to feed the ndings back to education funds reach their intended bene ciaries. For implementers are needed to improve the quality of repairs example, the Check My School initiative aims to promote and construction. citizen participation in the monitoring of public school performance.12 It mobilizes communities to validate school- Improvin the Alloc tion of Public related information from DepEd on many issues including Infr structure Funds school infrastructure. Any discrepancies with this information or issues with the relevant school infrastructure that citizens At the end of 2014, DepEd embarked on a school inventory raise are publicized and fed back to DepEd to be resolved. exercise, the results of which will help to improve the While such initiatives have the potential to improve DepEd’s allocation of infrastructure funds. The exercise provided projects and strengthen accountability for the delivery and division-level and school sta with guidance on how to quality of infrastructure improvements, they operate on a classify schools in terms of the state of their facilities with limited scale. Only around 20 percent of elementary and high the aim of improving the quality of data collected annually school principals had heard of the program and only around through the Enhanced Basic Education Information System 10 percent of schools had any interaction with the initiative (EBEIS). This information can then be used to identify schools in 2013 or 2014. that do not meet existing DepEd standards. Combined with a realistic assessment of the potential to improve or expand school facilities (for example, the amount of buildable Polic Directions for space), policymakers can use the information to develop and establish a clear and transparent set of criteria to prioritize Improvin School Le rnin infrastructure spending among schools whose facilities do Environments not meet existing standards. Allocation decisions could also be greatly improved by the Notwithstanding the signi cant investments made in school creation of a validation and nalization process at the division infrastructure over the last ve years, further investment level. The present centralized mechanism for allocating is needed. Funds are needed to address the remaining funding according to need requires validation at the local de ciencies in existing school facilities and to keep pace level to ensure that the central-level data re ect the real with the ever-growing student population. Between 2010 situation in the selected schools. This would make sure that and 2013, an additional 1 million children entered the basic the extensive knowledge that divisional sta have of schools education system, an average of around 350,000 students and communities is re ected in the nal allocation decisions annually, and the introduction of the senior high school and would ensure that infrastructure funds are put to the program in 2016 will put an additional burden on existing best use. high schools. 66 WWW.WORLDBANK.ORG/PH To ensure that allocation decisions are transparent, Coordination has improved between DepEd and the DPWH information on the criteria and data used to select schools for in recent years, but increasing the role played by schools infrastructure improvements could be widely disseminated. in project planning and implementation could also yield Feedback mechanisms could also be established to provide signi cant bene ts. In 2015, DepEd and the DPWH reissued a a channel for schools and their communities to voice any 2013 joint circular outlining their respective roles and respon- grievances that they may have about either the data used or sibilities in project management (see Box 1). However, the the selection process (see below). guidelines specify only a weak role to be played by schools and their principals in project management. For example, Stren thenin Implement tion C p cit the guidelines state that school principals should accompany DepEd and DPWH o cials on inspection visits but they are Building strong partnerships between government systems not empowered to approve the completed works. Giving and non-government partners would increase the capacity school principals authority alongside DepEd and DPWH for infrastructure improvement. While infrastructure budgets o cials for certifying that projects have been satisfactorily have been growing rapidly, the study’s ndings have raised completed might reduce the di erences in rates of reported concerns about whether government systems have enough completion between DPWH/DepEd managers and schools capacity to absorb these increases. There is a long tradition that were highlighted in the PETS-QSDS survey ndings. of NGO and private sector support for school infrastructure School principals in the Philippines have taken on roles of this in the Philippines. Many of these non-government initiatives kind in the past and with appropriate training could become have demonstrated that they can e ectively deliver a major driving force in improving the quality of school good quality classrooms at a comparable or lower cost infrastructure projects. Other countries in the region (such as than government systems and frequently in areas where Indonesia and Laos) and beyond (for example, Mexico) have government systems are particularly stretched (such as engaged local communities in project implementation with remote and hard-to-reach areas).13 Establishing or expanding signi cant success (see Box 2). partnerships between the government and these institutions is likely to reduce the stress on government systems and Developin Better Monitorin , speed up the process of improving and expanding school Ev lu tion, nd Feedb ck facilities. The ndings outlined in this policy note have shown that the Developing a medium-term planning framework for school monitoring and evaluation of school improvement projects infrastructure investment would have the potential to in the Philippines is weak. Users reported that the overall increase the timeliness and improve the quality of projects. quality of projects is low, and project managers do not The ndings of this study demonstrate that in many seem to be informed about the progress and completion of instances both project lists and funds are received well after projects on the ground. the start of the nancial year. This shortens the period for implementation and compromises the ability of DepEd and In order to address these weaknesses, it seems clear that the DPWH o ces to monitor projects adequately. Developing capacity of DepEd divisional and DPWH o ces to monitor a two- or three-year plan that would list all school projects e ectively requires strengthening. Trained engineers improvement projects in each DepEd division would reduce need to make frequent visits to project sites and make the work involved in putting together the annual project list detailed reports to DepEd and DPWH in order to address the and accelerate the transmission of funds to the responsible project quality issues highlighted in the PETS-QSDS survey. implementing o ces. It would also give implementing Allocating projects to speci c DepEd divisional or DPWH o ces time to do some advance planning and thus avoid sta and requiring them to visit these project sites on a some of the di culties associated with inadequate funding regular basis (for example, twice a month) has the potential and site availability so commonly cited as reasons for to improve the quality of projects for a relatively small implementation delays. investment. Given the signi cant increase in the number of ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 67 Policy Note 3: Building Better Learning Environments Box 2: Mexico’s Better Schools Program The Better Schools Program (BSP) in Mexico was a key component of a broader education reform program aimed at improving the quality of basic education. Between 2008 and 2012 the program carried out school infrastructure improvement projects in 19,400 of the neediest basic education schools in Mexico. The involvement and participation of each school community through its parent association was a major thrust of the BSP. A central government body was in charge of making payments to contractors and of overseeing the overall implementation of the program, but the schools’ parents associations acted as the legal client and was responsible for: • Commenting on, making decisions, and supervising the school project • Contributing to the transparency and presentation of nancial accounts • Verifying that the building materials and improvements were of good quality • Approving all stages of the planning and implementation of the project. After the Ministry of Education made an initial selection of schools in need of urgent repair, technical facilitators worked with each school to complete a technical assessment of its requirements that formed the basis of the project. Prospective contractors aiming to implement BSP projects were required to go through a rigorous pre-screening exercise, after which an approved list of contractors was drawn up and these companies were randomly selected to implement packages of projects. A recent assessment of the program noted: • The process of identifying priority schools and working with the schools’ parents associations to undertake the necessary work was e cient and exible enough to adapt to each school’s circumstances • The procedure of awarding contracts was relatively quick, and the payments process was transparent. It also ensured that contractors had incentives to deliver good quality projects • Overall, the project was cost-e ective in that it reached more schools than was initially intended and used resources e ectively and in line with the relevant regulations and guidelines. The necessary components for the e ective participation of local communities in school infrastructure improvement were summarized in the assessment report: “The BSP has demonstrated how partnerships between the community and government can result in substantial bene ts to the community and foster trust in the government’s capacity to deliver quality education services. A clear decision-making framework, clarity of roles and expectations, and well-de ned lines of responsibility have contributed to the successful engagement of parents and others in the BSP.” Source: Blyth, A., Almeida, R., Forrester, D., Gorey, A. and J. Zepeda (2012). “Upgrading School Buildings in Mexico with Social Participation: The Better Schools Program,” OECD, Paris. projects that have been undertaken in recent times, it is likely stronger role in project monitoring could improve the that current numbers of physical facilities sta in DepEd and quality of projects and also yield better information on the DPWH o ces will need to be reassessed to ensure that they e ectiveness of contractors. While school principals are are adequate. already nominally involved in inspection visits, their role could be expanded. For example, they could produce regular Monitoring and evaluation systems also need to make on-site reports about project progress to share with DepEd much better use of schools and local communities. and DPWH engineers as well as contractors. If they were also School principals, teachers, and local communities have a involved in approving inspection reports and completion strong interest in ensuring that projects are implemented certi cates, this would signal to contractors that the schools well. Giving school principals and governing councils a are the nal clients for their projects. 68 WWW.WORLDBANK.ORG/PH In order to increase transparency, information on ongoing take towards making government education spending more school projects could also be made more widely available. e ective and improving education services. Encouraging Allowing public access to divisional lists of school projects social accountability initiatives such as Check My School would enable schools and local communities to check could also help schools to use grievance systems as well selections and ensure that criteria have been applied as providing further external pressure on government to correctly for allocating projects. In a similar way to the school implement improvements. maintenance and operating funds provided by DepEd, school sta could also report on the project and its progress The government of the Philippines has made rapid on the school’s transparency board to inform parents and progress in improving school learning environments. It other stakeholders about progress and to give them greater has introduced a process to identify the infrastructure oversight over infrastructure improvements.14 needs of thousands of elementary and high schools and developed systems to fund and implement a huge school Increasing transparency is only one of the steps needed to building program every year. While these achievements increase accountability for project funds. Across the whole are impressive, more funding will be needed to meet the process for allocating and implementing infrastructure remaining infrastructure challenge. However, increased spending, mechanisms are needed to enable schools and funding will not be enough. A stronger focus on the other stakeholders to seek redress for unfair allocations governance and management of these resources will be or poor project implementation. Exploring the feasibility required if schools and students across the Philippines are of introducing a formal grievance system for schools to to get the good quality facilities that they need to improve utilize for all school funding, including school infrastructure the teaching and learning environment and raise learning projects, would be a useful step that policymakers could achievement levels. Table 3: Strengthening Government Systems for School Infrastructure Improvements Findings Policy suggestions Allocation mechanisms for • Establish transparent criteria for prioritizing schools based on existing facility standards school infrastructure systems • Validate project priority list at the division level need strengthening • Increase transparency and accountability by disseminating project selection criteria and by establishing feedback/grievance mechanisms Large increases in • Expand and establish partnerships to build and improve infrastructure infrastructure spending are • Introduce multi-year planning for school improvement projects required but implementation capacity also needs to be • Strengthen coordination between implementing institutions strengthened signi cantly • Strengthen the role of schools and school governing councils in implementation Weaknesses in the • Increase the capacity of DepEd and DPWH to monitor projects regularly monitoring and evaluation of • Increase the transparency of school infrastructure improvement activities at the infrastructure spending need division and school levels to be addressed • Empower schools to monitor projects more closely and provide e ective feedback channels for all stakeholders to use ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 69 Policy Note 3: Building Better Learning Environments Endnotes 1 Woolner, P. Hall, E. Higgins, S. McCaughey, C. and K. Wall (2007). 7 In kindergarten and Grades 1 and 2, DepEd aims for classes to “A Sound Foundation? What We Know about the Impact of consist of fewer than 45 students. Environments on Learning and the Implications for Building 8 In 2013, 62 percent (PHP 4 billion) of the BEFF releases managed Schools for the Future.” Oxford Review of Education 33(1): 47-70. by DepEd were allocated to the PPP program. 2 Glewwe, P. Hanushek, E. Humpage, S. and R. Ravina. (2011). 9 Appropriations for school infrastructure are set to grow from PHP “School Resources and Educational Outcomes in Developing 27 billion in 2013 to PHP 81 billion in 2016. Countries: A Review of the Literature from 1990 to 2010,” National Bureau of Economic Research, Cambridge, MA. 10 A similar pattern can be seen in the infrastructure funds that ow through DepEd divisional o ces. 3 Krishnaratne, S. White, H. and E. Carpenter. (2013). “Quality Education for All Children? What Works in Education in 11 Downloading refers to the issuance of a sub-allotment release Developing Countries?” Working Paper 20. New Delhi: order from DPWH central o ce to the DPWH district engineering International Initiative for Impact Evaluation (3ie). o ce. Sub-allotments are authorizations issued by the central o ce of DPWH or DepEd transferring a portion of an available 4 Yamauchi, F. and S. Parandekar (2014). “School Resource and allotment to the district o ce. Performance Inequality: Evidence from the Philippines.” World Bank Policy Research Working Paper No. 6748, World Bank, 12 World Bank (2012). “Check My School: A Case Study on Washington D.C. Citizens’ Monitoring of the Education Sector in the Philippines.” Washington D.C. 5 Highly urbanized cities are cities with populations of over 200,000 and with average revenues of at least PHP 50 million in 1991 13 Philippines-Australia Classroom Construction Initiative (2013). prices. Other cities are de ned as cities that do not meet the “Partnership Scoping Study - A Discussion Paper.” Manila. criteria to be classi ed as highly urbanized. Municipalities are 14 The transparency board is a publicly accessible notice board administrative units for all other areas in the Philippines. where schools post information on the receipt of government 6 DepEd (2013). “Medium-term Expenditure Framework for operating funds and school canteen funds. Basic Education 2014-2020: Enrollment Projections and Cost Simulations under Alternative Scenarios.” Department of Education, Manila. 70 WWW.WORLDBANK.ORG/PH Polic Note 4: Assessin School-b sed M n ement Introduction In recent times, many countries around the world have be- gun to devolve more authority and resources to schools in an e ort to improve teaching and learning conditions. School- based management (SBM) reforms of this kind are seen as a strategy to improve education quality by bringing decision- making closer to local communities and by strengthening accountability between schools and the parents and children that they serve. Recent studies from many di erent countries have shown that school-based management can have a positive impact in terms of increasing access to education and improving learning outcomes.1 However, these reviews have also shown that the type of SBM reforms being implemented varies enormously from country to country and that the reforms can take some time to yield results. Moreover, their success depends critically on parental participation, political support, and e ective implementation. In the early 2000s, the Government of the Philippines began to introduce school-based management reforms, which have had a positive impact on student learning outcomes.2 A key aspect of the success of these reforms has been the provision of ever-increasing levels of operational funding to the school level coupled with the devolution of greater school autonomy over the use of these funds.3 The note shows that the key elements of an e ective By 2014, schools were managing signi cant amounts of school-based management system are in place. However, resources and using these funds to implement their own schools report that they are not yet implementing many school improvement plans. Also, in 2015, the Department of of the key aspects of this system. Moreover, parents and Education (DepEd) issued new guidelines drawing on past local communities still play a very limited role in decision- experience that aimed to strengthen school improvement making and in holding schools accountable. Given the ever planning and make it more transparent. increasing amounts of resources that schools now control The purpose of this policy note is to assess the current and the need to give them more exibility over how to use status of school-based management in elementary and those funds, this note argues that the role of representative high schools in the Philippines. The ndings reported in the school governing councils could be expanded and e orts note are based on a comprehensive survey of the public made by DepEd to increase awareness among parents and education system that tracked public education expenditure education stakeholders of the useful role they could play in and assessed the quality of education services. supporting school-based management. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 71 Policy Note 4: Assessing School-based Management Implement tion of School- Figure 1: Most Schools Rated Themselves as Being at the Lowest Level of SBM Implementation b sed M n ement in the Percentage of schools by latest self-assessed SBM implementation level, 2014 Philippines 100% Level 1-developing Studies that have assessed the implementation of school- Level 2-maturing 80% Level 3-advanced based management have generally focused on three key dimensions: 60% ¾ Autonomy. The degree of autonomy that schools have to make their own decisions determines their ability to 40% a ect the educational outcomes of their students. ¾ Participation in decision-making. A central component 20% of SBM reforms has been to encourage greater participation by parents and other members of the community in supporting schools, usually by 0% Elementary School High School establishing or strengthening school governing bodies. ¾ Accountability and transparency. The extent to which Source: PETS-QSDS national survey – school-level data. a school is accountable to its local community is an important aspect of its management practices, the poorest 20 percent of households were more likely to including the transparency of the school decision- making process and the quality of information that the attend elementary and high schools that had the lowest school provides to its stakeholders. level of self-assessed SBM implementation than students from the wealthiest 20 percent of households.5 Other Most schools in the Philippines have as yet not gone very research has shown that SBM reforms have had only a limited far in implementing school-based management. In 2009, impact in schools that serve poorer communities.6 The DepEd introduced a tool for schools to carry out their own ndings from the PETS-QSDS study suggest that this may be assessment of their progress in implementing SBM (Box 1). due at least in part to the fact that schools in poorer areas are In 2014 the PETS-QSDS study collected information on the nding it di cult to put SBM into practice. results of these self-assessments and found that the majority of elementary and high schools reported having put in place The main weaknesses identi ed by school principals were only the lowest level of SBM (Figure 1). This means that they related to school autonomy. On the whole, more elementary had only a minimum number of arrangements in place for school principals than high school principals highlighted community participation and for taking action to improve weaknesses in their implementation of SBM. However, more learning outcomes. Fewer than 10 percent of schools high school principals mentioned their inability to raise reported being at the highest level of SBM implementation su cient resources as a major impediment to putting SBM and, thus, meeting all standards of community participation into practice (Figure 2). It is likely that this also limits the and school-based management. ability of schools to carry out the activities included in their school improvement plans. Approximately 20 percent of Poorer students were more likely than wealthier students elementary and high school principals also cited weaknesses to attend schools with lower self-assessed SBM ratings. in school improvement planning as an issue preventing the The PETS-QSDS survey included a nationally representative implementation of SBM. sample of public elementary and high school student households. Using information collected on consumption Most principals did not regard the participation of teachers and asset ownership in the survey, it was possible to rank and other internal stakeholders in decision-making at student households by estimated levels of per capita the school level as a weakness. However, 40 percent of household consumption.4 Using this indicator, it was possible elementary school principals felt that the engagement of to disaggregate the SBM ratings by schools serving di erent external stakeholders, including parents, local government, socioeconomic groups. The study found that students from and other community representatives, was relatively weak. 72 WWW.WORLDBANK.ORG/PH Box 1: Measuring the Implementation of School-based Management in the Philippines DepEd rst introduced a self-assessment tool in 2009 to help schools to improve their school-based management practices. In 2013, it introduced a new self-assessment tool that asked schools to rate the extent to which they had implemented SBM based on 22 indicators that could be converted into an aggregate three-point scale as follows: 1. Level 1 – developing: Schools have developed structures and mechanisms that have increased community participation and improved learning outcomes to an acceptable level. 2. Level 2 – maturing: Schools have introduced and sustained continuous improvements that have led to community participation and have signi cantly improved student learning outcomes. 3. Level 3 – advanced: Schools have fully implemented a school-based management system that fully involves the local community and is self-renewing and self-sustaining. In the PETS-QSDS study, schools were also asked to identify any weaknesses they might have in six aspects of school- based management de ned in the original 2009 assessment. These six aspects are related to the three dimensions of school-based management assessed in this policy note, which are: 1. Autonomy. School principals were asked about strengths and weaknesses in their school leadership, their ability to raise resources, and their school improvement planning process. 2. Participation. School principals were asked to identify strengths and weaknesses in their schools’ links with internal and external stakeholders. 3. Transparency and accountability. School principals were asked about strengths and weaknesses in the institutions that hold the school accountable for performance. Sources: Parandekar, S. (2014). “Benchmarking Public Policy: Methodological Insights from Measurement of School-based Management.” World Bank Policy Research Working Paper No. 6938, World Bank, Washington D.C.; DepEd Order No. 83 (2012) “Implementing Guidelines on the Revised SBM Framework, Assessment, and Tool,” Department of Education, Manila; and DepEd (2009). “A Manual on the Assessment of SBM Management Practices,” Department of Education, Manila. Figure 2: School Principals Cited Issues Related to School Autonomy as a Major Weakness in the Implementation of SBM Percentage of schools reporting weaknesses in key SBM dimensions, 2014 50% Elementary schools High schools 40% 30% 20% 10% 0% Leadership Improvement Resource raising Internal External Performance process stakeholders stakeholders accountability Autonomy Participation Accountability & transparency Source: PETS-QSDS national survey – school-level data. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 73 Policy Note 4: Assessing School-based Management Autonom approximately PHP 3 million and for high schools it was PHP 8.1 million (Figure 3). Over 80 percent of school funding While schools have only a limited amount of autonomy to is devoted to teacher and sta salaries, and most of these make decisions about their own a airs, most schools develop salary payments are for centrally hired government teachers their own improvement plans and implement them to and sta . While school principals have some control over the the extent that their authority and resources allow. In each day-to-day supervision of their sta , the hiring and resources school, the school improvement plan (SIP) is developed by used for school sta ng fall outside the school’s control. Only the principal with help and input from teachers and the local around PHP 188,000 or 7 percent of total funds are available community. It reviews the school’s performance (for example, to each elementary school to use at its own discretion. High in terms of student scores on the National Achievement Test) schools have a little more exibility in the use of their funds and identi es areas where improvements are needed. Each SIP but only around PHP 987,000 or 12 percent of each school’s covers a three-year period, and the school produces a more total funds are discretionary. detailed implementation plan every year. However, schools have very limited authority over such important areas as the The bulk of these discretionary funds are provided by DepEd hiring and deployment of teachers and school infrastructure and come with some conditions over their use. DepEd improvements.7 In 2014, almost all schools had current school provides 70 percent of discretionary funds for elementary improvement plans in place, and only 5 percent of elementary schools and 83 percent for high schools. The bulk of cash schools and 10 percent of high schools did not have an up-to- funds provided by DepEd is allocated on the basis of date annual implementation plan. student numbers and is provided to cover maintenance Schools have discretion over only a very small amount and other operating expenses (MOOE).9 Schools are of their total revenues. The PETS-QSDS study collected allowed to use these funds on many recurrent items (such detailed information on all sources of school revenue and as utility payments and small repairs) but are prohibited expenditure from a nationally representative sample of from spending them on other important items such as elementary and high schools.8 The ndings showed that learning materials and any capital assets (for example, IT the average annual revenue of an elementary school was equipment).10 Figure 3: Discretionary Resources at the School Level are Low Annual school revenue by type and source, 2013/14 school year Composition of all school funding (PHP 000s) Percentage of discretionary funds by source 10,000 100% 8,000 80% 6,000 60% 4,000 40% 2,000 20% 0 0% Elementary School High School Elementary School High School Salary Infrastructure DepEd Other cash School Other in-kind Discretionary/cash Source: PETS-QSDS national survey – school level. Note: The averages reported in the table are for all schools even if they received zero revenue under a particular category. Other cash revenue includes cash provided to schools by parent-teacher associations, local governments, and barangays and from private sector or NGO contributions. 74 WWW.WORLDBANK.ORG/PH Schools also generate their own funds, but their ability to do costing study found that existing levels of MOOE were so is dependent on the socioeconomic composition of their insu cient to provide education services up to the standard student populations. Schools collect student contributions of existing DepEd norms.14 The limited amount of funding to cover speci c activities (such as school projects, Red that schools have available to spend at their own discretion in Cross, uniforms, and school feeding programs) as well as e ect curtails the autonomy that they have to make their own running the school canteen and other income-generating decisions on how best to improve their operations. projects. About 27 percent of discretionary revenue comes from these sources in elementary schools compared with P rticip tion 16 percent in high schools.11 However, schools that serve a E ective school-based management requires parents and large proportion of children from poorer households are local community members to play a strong supporting restricted in terms of how much of this kind of revenue role in school decision-making and oversight. The School they can raise. For example, in 2014 students in the poorest Governing Council (SGC) for each school is a forum for quintile attended high schools that collected approximately parents, students, teachers, and community stakeholders to PHP 60 per student in canteen funds, while students in the participate in making school improvement decisions. It takes wealthiest quintile attended schools that collected PHP 116 an active role in developing the school improvement plan per student in canteen funds.12 While some of this di erence (SIP) and, together with the school principal, is responsible for may have been due to the quantity and quality of food endorsing it to the schools division superintendent (SDS) for provided, it may also have re ected the greater capacity of approval. It is also expected to participate in the monitoring wealthier schools to collect revenues to support their own of the implementation of the SIP.15 The number and type of school improvement planning. the SGC’s members and the frequency of its meetings are A lot of school discretionary resources are devoted to routine initially decided through an establishment meeting attended expenditures, which leaves only a limited amount available by parents and other stakeholders. In this meeting the to fund additional activities. The most detailed information constitution and by-laws of the SGC and the election process available on the use of school discretionary resources comes for selecting council members are agreed. from o cial school records on school-level spending on MOOE. In 2013, schools used approximately three-quarters Most schools have established school governing councils, and of their MOOE funding to pay for supplies and printing, to their composition tends to be representative of stakeholders undertake routine maintenance, and to pay their utility bills.13 within the school and in the local community (Figure 4). The While these are clearly important, it does not leave much room PETS-QSDS survey found that approximately 90 percent of for schools to invest in other activities that might support elementary schools and 80 percent of high schools had SGCs. better learning. For example, less than 5 percent of MOOE The PETS-QSDS survey was administered in the last quarter of funding was used to nance teacher training. A recent detailed 2014, and its results showed that, on average, SGCs had met Figure 4: School Governing Councils Are Representative of Most Key Stakeholders Percentage of SGCs by type of membership, 2014 100% 75% 50% 25% 0% Elementary Schools High Schools Teachers Barangay captain Parents/PTA representatives Student council members NGO members Alumni Local professionals Source: PETS-QSDS national survey – school level. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 75 Policy Note 4: Assessing School-based Management twice since the beginning of the 2014/15 school year. This school and community collaboration in improving learning implies that SGC meetings are usually held every quarter. The outcomes. survey found that SGCs generally include representatives of Given that parents tend not to know about the existence of most of the major stakeholders including parents, teachers, the SGC, it is perhaps not surprising that fewer than half of and students. In terms of overall numbers, parents and all parents interviewed were aware that their school had an students tend to be most heavily represented on elementary improvement plan. Most of the parents who knew about the school SGCs, whereas in high schools, teachers are more SIP seem to have been invited to participate directly in its heavily represented. preparation, but any awareness beyond this group of parents School principals reported that the most common types was rare. of support provided by SGCs is the time and labor that Parent-teacher associations (PTA) are another mechanism they devote to school activities. Two-thirds of elementary through which parents can support school improvement. All school principals and nearly half of all high school principals schools sampled as part of the PETS-QSDS study reported reported receiving this kind of support from SGCs. About that they had a functioning PTA. The role of PTAs in school half of all principals also cited nancial support for school decision-making and in the development of the school planning as another type of support received from the SGC. improvement plan in particular is less clearly articulated than When SGCs meet, the most common topics of discussion the role of the SGC. The DepEd guidelines governing PTAs are student discipline, school improvement planning, school focus almost exclusively on procedures for collecting funds nances, and student academic performance. and reporting on their use.16 However, PTAs are an important Very few parents of students were aware that their child’s way in which parents can raise any concerns about school school had an SGC. The PETS-QSDS study interviewed a issues and provide support for school operations. As with random sample of student households from the schools that SGCs, many PTAs include representatives of other local were sampled. Nearly three-quarters of the household heads stakeholders including barangay captains and other local that were interviewed were unaware that their child’s school government o cials. had an SGC (Figure 5). These ndings raise questions about Parents are moderately more aware of and active in PTAs how e ectively the SGCs are ful lling their role as forums for than in school governing councils. All schools and almost all parents reported that their child’s school had a PTA. Figure 5: Few Parents Know about the SGC or Are Interviews with the Presidents of the PTAs indicated that Aware of the School Improvement Plan most PTAs met on a monthly or quarterly basis (Figure 6). Percentage of parents of elementary and high school Only 10 percent of PTAs in elementary and high schools met students who know about the SGC and the SIP, 2014 less than twice a year. Almost a half of all parents participated 40% in PTA meetings and on average attended four meetings during the school year (Figure 6). On the whole, over 90 percent of parents were satis ed with the functioning of their local PTA. 20% Even though PTAs play only a limited formal role in school a airs, PTA Presidents reported that they had had some involvement in both the formulation and monitoring of 0% school plans. Approximately 85 percent of elementary Aware of School Aware that the school has and high school PTAs participated in the development of Governing Council an improvement plan the school improvement plan as well as in monitoring its Kindergarten Elementary High School implementation. The small proportion of PTAs that were not Source: PETS-QSDS national survey – household level. involved in the SIP process mentioned either that they were Note: The bars relate to the level of education attended by the student not asked to help or that the PTA did not have the time to from the interviewed household. devote to it. 76 WWW.WORLDBANK.ORG/PH Figure 6: PTAs Meet Regularly and Parents Play an Active Role Frequency of PTA meetings and percentage of parents reporting that they attended regularly in 2013/14 school year Frequency of PTA meetings Parents’ participation in PTA meetings 60% 5 100% 50% 4 75% 40% 3 30% 50% 2 20% 25% 10% 1 0% 0 0% Elementary schools High schools Kindergarten Elementary High school Monthly Quarterly Biannually Annually No. of meetings attended in last school year Percentage of parents participating (RHS) Source: PETS-QSDS national survey. Left hand panel from student household interviews and right hand panel from PTA interviews. The main kind of support that PTAs provide to schools is in to schools through their own PTA fees as well as organizing the form of additional nancial and human resources. PTA fund-raising activities for the school. Only 32 percent of Presidents mentioned that providing their labor and skills for elementary school PTAs and 41 percent of high school PTAs school projects was one of their most common activities in mentioned planning as one of their main areas of support. support of schools (Figure 7). Two of the most commonly cited kinds of support given to schools by their PTAs were physical These ndings suggest that PTAs are generally more active labor and participation in school maintenance week (Brigada in school a airs than SGCs and are the main mechanism Eskwela). PTAs were also active in providing additional funding through which parents and other stakeholders participate. Figure 7: PTAs Commonly Provide Schools with Labor, Finances, and Assistance during the Brigada Eskwela Percentage of PTAs providing support by type, 2014 80% 60% 40% 20% 0% Elementary Schools High Schools Physical Financial Brigada Eskwela Fund raising Planning Administrative Source: PETS-QSDS national survey – PTA interviews. Note: Brigada Eskwela, or school maintenance week, happens at the start of the school year and the local community provides labor and resources to repair and tidy up school facilities. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 77 Policy Note 4: Assessing School-based Management These ndings are con rmed by school principals who where information was posted (Figure 8). However, many of stated that one of the most common ways in which they these transparency boards were not publicly accessible. For receive feedback and complaints from parents and other example, around 40 percent of the transparency boards in stakeholders is through the PTA. Approximately 60 percent elementary schools were located inside the principal’s o ce of elementary and high school principals reported receiving or in the sta room. comments and complaints from the PTA, and, with the Most schools do not display or publish up-to-date exception of direct discussions with parents and students, information on how they spend their discretionary resources. this was their most common feedback mechanism. The proportion of schools, particularly high schools, that reported such information was low (Figure 8). The most Account bilit nd Tr nsp renc common information posted in schools was reports on how This note has already provided an assessment of the kinds they had spent their MOOE funds. However, in one-third of institutions that are available at the school level to hold of the cases where this information was included on the schools accountable for the resources they receive and transparency board, it was over three months old. Information ultimately for the education outcomes they deliver. This on how revenues from sources such as school canteen funds section focuses on the information that parents, PTAs, and were reported in slightly over half of all elementary schools SGCs can use to hold schools to account and ways in which but in fewer than one-third of high schools. parents and other stakeholders are able to provide feedback Information on school performance was also rarely provided to schools. on publicly accessible notice boards. Only 41 percent of While the PETS-QSDS enumerators were visiting schools, they elementary schools and 12 percent of high schools placed noted that approximately 70 percent of elementary and high information such as the results achieved by their students on schools had some kind of notice board or transparency board the National Achievement Test and school dropout rates on Figure 8: About a Half of Schools Have an Accessible Transparency Board but the Information Provided Is Frequently Limited Percentage of schools with a transparency board and types of information posted on the boards, 2014 Elementary Schools High Schools 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Has transparency board Board in public space MOOE liquidation Canteen funds PTA funds LGU funds School performance MOOE guidelines Has transparency board Board in public space MOOE liquidation Canteen funds PTA funds MOOE guidelines LGU funds School performance Board Information on board Board Information on board Source: PETS-QSDS national survey – school level. 78 WWW.WORLDBANK.ORG/PH transparency boards (Figure 8). While many schools do display in the previous two school years. A similar proportion of this information, they usually do so within the principal’s parents reported that they had not even received a report o ce, which is inaccessible to parents and other community card on their own child’s progress. stakeholders. Despite the limited provision of information by schools, The limited availability of information provided by schools is parents on the whole felt that schools were fairly responsive perhaps a key reason why parents appear to be unaware of and that there were a number of mechanisms in place for the kinds of nancial resources that schools receive to fund parents to provide feedback. Over 85 percent of parents improvements. Interviews with the parents of students in either agreed or strongly agreed that schools provided PETS-QSDS schools revealed that they had little knowledge opportunities for parents to le complaints on school-related of what resources the schools received. For example, only 40 issues. A similar proportion of parents of elementary and high percent of parents of elementary and high school students school students felt that their school worked well with other were aware that schools received MOOE funding from the stakeholders to respond to the needs of learners. national government. Even among the parents who were aware of this, fewer than half had received any information Beyond school-level institutions, social accountability on how the funds had been allocated in their school over initiatives like Check My School do not appear to have had a the previous year. School meetings and PTA meetings were signi cant impact. Check My School (CMS) is a project that the most common places where information on MOOE was uses mobile and internet technology to enable communities disseminated. Fewer than 5 percent of elementary and high to monitor the quality of basic education services. The school parents reported obtaining this information from the initiative involves local facilitators who mobilize communities transparency board at the school. to verify information on the status of their school. Any There are few other ways by which parents and other discrepancies between o cial government data and the stakeholders can access information on school performance. data collected by the CMS are then brought to the attention As part of the school improvement planning process, schools of DepEd. Although the initiative began in 2011, its coverage are supposed to prepare an annual school report card that appears to be quite small. Only 15 percent of elementary includes information on key school performance indicators school principals and 20 percent of high school principals such as the National Achievement Test (NAT) results and were aware of CMS, and only a small proportion of these student progression indicators. Yet only around a half of schools had had any direct dealings with the initiative (Figure parents of elementary and high school students said that 9). For example, fewer than 5 percent of high schools had ever they had been given such a card or even any information provided feedback through CMS.17 Figure 9: Social Accountability Initiatives like Check My School Do Not Work in Many Schools Percentage of school principals reporting any involvement with Check My School (CMS) in 2013 or 2014 20% 15% 10% 5% 0% Elementary Schools High Schools Aware of CMS Visited by CMS CMS report Feedback through CMS Source: PETS-QSDS national survey – school level. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 79 Policy Note 4: Assessing School-based Management Figure 10: Division and District Supervisors Carry Out School Monitoring Visits Regularly Average number of visits to schools in rst half of 2014/15 school year and issues covered Number of visits Issues discussed 3 80% 60% 2 % of visits 40% 1 20% 0 0% District supervisor Division supervisors Division supervisors Academic Planning and management Administrative Financial Academic Planning and management Administrative Financial Elementary School High Schools Elementary School High Schools Source: PETS-QSDS national survey – school level. Note: Only issues that were discussed with the public schools district supervisor are reported for elementary schools. Division and district supervisors from DepEd regularly visit schools to monitor their operations and to provide them Polic Directions to with administrative and academic support. Elementary Stren then School-b sed school principals reported that, since the start of the 2014/15 school year, they had received two visits on average from the M n ement public schools district supervisor (PSDS) and one visit from division o ce supervisors (Figure 10). Given that the PETS- The relatively low level of SBM implementation reported by QSDS survey was administered in the last quarter of 2014, this schools in their own self-assessments highlights the need means that the PSDS was visiting an average of once every for DepEd to provide more support to schools. Providing two months and the division supervisors were visiting every school-level sta and SGC and PTA members with training quarter. Similarly, high schools were being visited by division on their roles and responsibilities has the potential to enable supervisors every two months.18 These visits usually included them to support schools in implementing all of the aspects academic supervision of classroom teaching and discussions of SBM. However, it is also likely that schools will need with school principals on planning, management, and support from district and division-level supervisors to help administration. them to develop and implement better school improvement plans and to increase the e ectiveness of PTAs and SGCs. 80 WWW.WORLDBANK.ORG/PH This support could be provided during the existing visits that actions aimed at improving the learning outcomes of their district and division supervisors make routinely to schools. students. DepEd has already embarked on a strategy to A recent evaluation of training in Indonesia highlighted the increase the level of school MOOE funds signi cantly, and importance of supervisors playing this kind of mentoring role this will go some way towards providing schools with the in supporting school improvements (see Box 2). resources that they need. Providing schools with more discretionary funding could The government should consider increasing the authority support the implementation of SBM and increase the that schools have over the use of other school-level funding. e ectiveness of public spending. The small share of For example, DepEd is currently engaged in a massive school discretionary funding that schools currently receive limits infrastructure improvement program, but analysis of the their autonomy and ability to implement their school PETS-QSDS survey has shown that the quality of classrooms improvement plans. Less than 15 percent of school funding and water and sanitation facilities built or rehabilitated is controlled directly by schools, and much of this funding under this program has generally been poor. Giving is spent on utility bills and routine supplies instead of on schools and the SGC a greater role during the planning and Box 2: An Evaluation of School-based Management Training Activities in Indonesia In 2011, Indonesia embarked on a massive program to train school principals and school committee members on the core elements of school-based management, including planning, budgeting, and nancial management. Approximately 650,000 people attended a three-day training event held in di erent locations across Indonesia. An evaluation of the program came to a number of important conclusions: • A follow-up survey of participants found that the materials covered under the training program were relevant to the needs of schools in terms of school-based management. • The training program contributed to a number of changes at the school level particularly in terms of school governance. These positive ndings were con rmed by interviews with parents and community members, which revealed that schools were using discretionary funds in a transparent and accountable way. • Schools reported that they were better able to follow the central government’s guidelines for reporting on their use of discretionary funds, which increased transparency and accountability. • The follow-up survey highlighted overall improvements in the management of school operational funding, particularly by school committee members. • The quality of training varied greatly across regions. There was a lack of supervision of the training events at the district level, which resulted in di erences in the length of the training, class sizes, the capacity of trainers, and the quality of training facilities. The majority of trainees interviewed in the follow-up survey felt that more guidance and assistance was needed to help them to implement the training elements in their own schools. Source: Shae er, S. (2013). “BOS Training: Its Implementation, Impact, and Implications for the Development of Indonesia’s Education System. An Independent Review” Report prepared for AusAID Indonesia, Jakarta. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 81 Policy Note 4: Assessing School-based Management implementation of these improvement projects and giving Pakistan have shown that it is possible to use old and new schools the authority to certify project completion could technologies to increase awareness. In Indonesia, an impact improve the quality of these facilities in the future.19 evaluation study showed that well-designed information campaigns using simple SMS text messages or school This note has also highlighted the need for DepEd to provide meetings can signi cantly increase public knowledge greater support to schools that serve poorer communities. about schools’ funding levels and responsibilities.20 It also It has shown that poorer children generally attend schools found that this increased knowledge led to higher levels where SBM implementation is less advanced and where of parental participation in school a airs. In Pakistan, a fewer discretionary funds are available. The introduction of similar strategy was used in a successful pilot program to an equity component in the school MOOE funding formula disseminate information about school councils.21 The pilot would be one way to address these funding inequalities. set up a call center and used inbound and outbound calls, For example, schools in remote areas and/or that serve robot calls, and SMS text messages to transmit important disadvantaged groups could be given additional funding information on the role played by school councils and their to account not only for di erences in their own revenue- members. An assessment of the pilot showed that the raising abilities but also for the higher costs they incur to knowledge that school council members had of their roles purchase school equipment and supplies and to attend and responsibilities had increased, while school principals training sessions. While more training and support for SBM reported an increase in the participation of the councils in implementation is needed throughout the Philippines, it school a airs. is likely that additional help will be needed by the most disadvantaged schools. This note has found that PTAs seemed to function well, with high participation rates by stakeholders and good relations Greater e orts are required to revitalize the role played with school principals. These ndings suggest that parents by school governing councils in the school improvement and other stakeholders could participate to a greater extent planning process in schools across the Philippines. As DepEd in school improvement planning and in the management considers devolving more of the public education budget to and oversight of school funds. However, it is also vital that schools, it is vital that accountability structures and resource the roles and responsibilities of SGCs and PTAs should be planning processes are strengthened. In particular, SGCs clari ed by DepEd to ensure that there are no duplications in could be given greater responsibility for monitoring how functions and that each institution ful ls its particular role. schools use their MOOE funds as a supplement to DepEd’s oversight mechanisms. For example, SGCs could become Schools also need to make more information available to jointly responsible with the DepEd divisions o ce for signing their stakeholders on how they use their funds and on the o on schools’ MOOE liquidation reports. If the SGCs are to school’s overall performance. This note has shown that many take on an expanded role, then more training and support schools do not routinely share this kind of information with will need to be provided by DepEd to their members. parents even though there are regulations obligating them to do so. Keeping parents informed can encourage them This note has shown that most parents of students to expand their role in supporting schools and in holding are unaware of the existence of SGCs or their school’s schools to account. Schools in the Philippines are required to improvement plan. It seems clear that a campaign is urgently produce report cards on the performance of the school and needed to increase the awareness of parents and other of individual students, but this note has shown that this often education stakeholders of the role of SGCs and school does not happen. There is evidence from other countries improvement plans. Recent evaluations in Indonesia and that these report cards can have signi cant bene ts in terms 82 WWW.WORLDBANK.ORG/PH of, for example, higher student test scores and reduced enormous bene ts in terms of better education outcomes. absenteeism by teachers (Box 3). Therefore, DepEd needs Schools around the country have already embarked on many to ensure that its regulations requiring schools to produce of the reforms needed to realize these bene ts. However, report cards are enforced and to make parents aware that there is a need to revitalize school-level institutions and to they are entitled to expect to be given more information on increase the amount of information that schools provide their children’s schools. to parents and other stakeholders on the services that the schools are providing. Only then will the full potential of The evidence from the Philippines clearly shows that the school-based management to improve education outcomes introduction of school-based management can have be realized. Box 3: An Evaluation of School-based Management Training Activities in Uganda A randomized controlled trial in Uganda explored the impact of di erent interventions that used school report cards as a way for school managing committees to monitor performance. The two interventions that were tested were as follows: • A standard scorecard. School committee members were trained on how to use a school-level report card developed by education o cials and NGOs. • A participatory scorecard. School committee members were trained and helped to produce a school-level report card that they themselves had developed and that included indicators of school progress agreed upon with the school. Interventions like these can be expected to improve education outcomes by providing local communities with the information needed to hold their schools accountable. They can also encourage schools to improve their performance either through social pressure or through a closer collaboration between the school and community. The experiment found that the participatory scorecard approach had a statistically signi cant impact on education outcomes. In terms of student learning, schools where the participatory scorecard was introduced had a statistically signi cant advantage in primary school test scores of approximately 0.2 standard deviations over control schools. Teacher attendance also improved, with teachers working in participatory scorecard schools being 13 percentage points more likely to be present than teachers in control schools. The results for the standard scorecard approach were less promising. The experiment found no statistically signi cant e ects on student learning although it did appear to have a positive e ect on the attendance of some teachers. The authors argue that the better outcomes for the participatory scorecard were primarily the result of increased cooperation between the school and the local community rather than because of di erences in the information contained in the di erent scorecards. These ndings suggest that using participatory methods to develop school report cards may improve education outcomes and strengthen the supporting role played by school committees. Source: Barr, A., F. Mugisha, P. Serneels and A. Zeitlin. (2012). “Information and Collective Action in Community-based Monitoring of Schools: Field and Lab Experimental Evidence from Uganda.” Unpublished paper. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 83 Policy Note 4: Assessing School-based Management Table 1: Strengthening School-based Management Findings Policy suggestions Schools assess their level of SBM • Provide further training to school principals, school governing councils, and implementation as low parent-teacher associations on school-based management • Build the capacity of district and division supervisors to mentor schools in the implementation of school-based management Only a small share of school funding is • Increase central government MOOE funds to a level that will enable schools discretionary to meet existing education service standards • Increase the authority of schools over other funds (such as school construction funding) SBM implementation is less advanced • Introduce an equity component into the MOOE funding formula and fewer discretionary funds are • Provide additional supervisory support on school-based management to available in schools that serve poorer schools serving disadvantaged groups communities SGCs are not functioning as they were • Clarify the roles and responsibilities of SGCs and PTAs designed to do • Establish a role for the SGC in overseeing the use of school discretionary funds • Increase the knowledge of SGC members on their roles and responsibilities through school-level training initiatives • Conduct an information campaign to increase public awareness of the role of SGCs and PTAs, the SIP, and the use of and the reporting requirements associated with discretionary school funds (such as MOOE funds and canteen revenues) Transparency at the school level needs • Strengthen monitoring by DepEd district and division o ces of the to be increased production of key information (such as school report cards and student report cards) by schools • Enforce regulations on making information publicly available on school performance and use of funds 84 WWW.WORLDBANK.ORG/PH Endnotes 1 AusAID ERF (2011). “Current Issues in Education: School Grants 10 A more detailed assessment of the systems governing the use and School-based Management.” Canberra; Bruns, B., D. Filmer, of MOOE funding for schools is provided in a separate note- and H. Patrinos (2011) “Making Schools Work: New Evidence on Providing Schools with Adequate Operating Expenses to Deliver Accountability Reforms” World Bank, Washington, D.C.; and Carr- Quality Education in the Philippines. Hill, R., C. Rolleston, T. Pherali, and R. Schendel (2014). “The E ects 11 Despite these sources representing a lower proportion of total of School-based Decision-making on Educational Outcomes in discretionary funds in high schools than in elementary schools, Low- and Middle-income contexts: A Systematic Review” Institute high schools collected higher absolute amounts from these of Education, London. sources in 2014. 2 Khattri, N., C. Ling, and S. Jha (2010). “The E ects of School-based 12 The pattern is similar in elementary schools and for other Management in the Philippines: An Initial Assessment Using school-level revenues (for example, charges for school utilities). Administrative Data” Policy Research Working Paper Series. No. Full details are available in a set of additional annexes and tables 5248, World Bank, Washington, D.C. and Yamauchi, F. (2014) “An accompanying the main PETS-QSDS report. Alternative Estimate of School-based Management Impacts on Students’ Achievements: Evidence from the Philippines.” Journal of 13 A more detailed assessment of school MOOE funding is provided Development E ectiveness. 6, no. 2: 97-110. in a separate note- Providing Schools with Adequate Operating Expenses to Deliver Quality Education in the Philippines. 3 World Bank and AusAID (2013). “School-based Management in the Philippines: An Empirical Investigation.” World Bank and 14 World Bank (2016). “The Development and Implementation AusAID, Manila. of a Normative School MOOE Formula in the Department of Education in the Philippines.” Washington D.C. 4 The household questionnaire included a short module on consumption and a set of questions on assets that have been 15 SGCs also play a role in making policies regarding student welfare used by the Department of Social Welfare and Development and in managing council resources. See DepEd (2009). “A Manual (DSWD) to undertake a proxy means testing (PMT) approach on School Governing Councils.” Department of Education, Manila. to estimating household consumption per capita. The results 16 DepEd Order No. 54 (2009). “Revised Guidelines Governing reported here are based on information gathered using the PMT Parents-Teachers Associations at the School Level.” Department of approach, and a full description is included in a separate note. Education, Manila. 5 The full results are provided in a set of additional annexes and 17 Similar results were also found for Government Watch initiatives tables accompanying the main PETS-QSDS report. in the education sector like textbook count and textbook walk. 6 Carr-Hill, R., C. Rolleston, T. Pherali, and R. Schendel (2014). 18 High schools are not monitored by public schools district “The E ects of School-based Decision-making on Educational supervisors. Outcomes in Low- and Middle-income Contexts: A Systematic 19 A fuller discussion of the strengths and weaknesses of existing Review” Institute of Education, London. school infrastructure improvement systems is provided 7 Systems governing the use of school infrastructure improvement in a separate note in this series - Building Better Learning funds is discussed in detail in a separate note - Building Better Environments in the Philippines. Learning Environments in the Philippines. 20 Cerdan-Infantes, P. and D. Filmer (2015). “Information, Knowledge, 8 A full description of the information collected in the PETS-QSDS and Behavior: Evaluating Alternative Methods of Delivering study and the approach used to calculate total school revenue School Information to Parents.” World Bank Policy Research and expenditure is included in a set of additional annexes and Working Paper No.7233, World Bank, Washington D.C. tables accompanying the main PETS-QSDS report. 21 Cambridge Education (2014). “Review of Implementation of 9 In 2013/14, DepEd also provided school-based management School Council Policy 2013.” Cambridge, UK. grants of between PHP 100,000 - 200,000 to 7 percent of elementary and 17 percent of high schools in low income class municipalities. The revenue from these funds is included under revenue from DepEd in Figure 3. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 85 Polic Note 5: Providin Schools with Enou h Resources to Deliver Qu lit Educ tion Introduction Providing schools with direct funding to meet their operational needs is the cornerstone of the government’s e orts to strengthen school-based management in the Philippines. Over the past ve years, funds that the government has provided directly to schools to support their maintenance and other operating expenses (MOOE) have increased by 45 percent in real terms, demonstrating the government’s commitment both to increasing operational funding and to providing schools with the funding that they need to implement their own improvement plans.1 Evidence from the Philippines and elsewhere shows that increased school funding and e ective school-based management (SBM) can lead to more e cient use of resources and, ultimately, to better education outcomes. Evaluations of a school-based management program in these and similar ndings demonstrate the need to improve the Philippines have shown that this program coupled the systems that manage MOOE funds in division o ces and with the provision of school grants led to signi cant in schools. The ndings also highlight the need to strengthen improvements in school performance.2 Over a three- the accountability mechanisms associated with the use year period, the introduction of SBM and the provision of of MOOE funds, particularly at the school and community grants improved the scores of Philippines students on the levels. Addressing these weaknesses is likely to signi cantly National Achievement Tests (NAT) by 4 to 5 percentage improve education outcomes and to support recent school- points (approximately 0.25 standard deviations).3 These based management reforms in the Philippines. ndings echo similar results from other countries that, on the whole, show that school-based management reform, if implemented e ectively, can improve education outcomes Levels of nd Trends in over time.4 School Oper tion l Fundin The aim of this policy note is to assess the current systems Over recent years, school MOOE funding has increased that govern the allocation, transfer, and use of MOOE funds signi cantly in real terms even though its share in the overall for schools. It uses data collected as part of the PETS-QSDS basic education budget has remained relatively stable. study from a nationally representative sample of public The total school MOOE budget increased in real terms by elementary and high schools.5 It shows that a signi cant over 50 percent between 2011 and 2015 (Figure 1).6 The proportion of MOOE funds do not reach schools and that biggest increase during this period was associated with the schools serving better-o students tend to receive a larger introduction in 2013 of a new formula for allocating school share of their intended allocation than other schools. While MOOE funding. These large overall increases in the school funding for school MOOE has and will continue to increase, MOOE budget have translated into even larger increases ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 87 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education in per-student funding levels. For example, between 2011 and 2014, per student appropriations for elementary school Assessin the S stems th t MOOE increased from PHP 317 (US$7) to PHP 566 (US$13). Govern the Alloc tion nd Public schools rely heavily on the school operating funds Use of School MOOE Funds provided by the national government. The PETS-QSDS Alloc tion study collected detailed information on all types of school revenues. In 2013, MOOE allocations accounted for 68 As in other countries, MOOE allocations in the Philippines percent of all discretionary funding received by elementary are based on school and student characteristics. DepEd introduced a funding formula for school MOOE allocations schools and 82 percent received by high schools. In the case for the rst time in 2013 (Table 1). According to the new of over 10 percent of schools, their MOOE allocations were formula, in 2013 the average elementary school was the only source of operational funding that they received.7 allocated approximately PHP 170,297 (US$3,720) and the average high school was allocated PHP 744,358 (US$16,180).9 Despite the importance of MOOE allocations to schools and despite recent increases, a detailed costing study However, the amount of MOOE funds that is actually allocated has revealed that existing levels of school MOOE do not to each school is frequently di erent to the amount prescribed cover the full operating costs of public elementary and by the formula. For example, in 2013 the allocations of MOOE high schools.8 The study reviewed existing DepEd service funds to elementary schools di ered from the expected standards and norms for elementary and high schools and amount based on the formula by PHP 7,000 or approximately collected data and information on the funds needed to meet 4 percent of the average allocation for each school. In the each of these standards. After verifying its ndings with case of a small number of schools, the allocation di ered DepEd sta and school principals throughout the Philippines, from the formula by as much as PHP 50,000.10 Possible reasons the study concluded that overall levels of elementary and for these discrepancies include confusion over which types high school MOOE funding will need to more than double if of classrooms are included in the formula and DepEd’s “no existing service standards are to be met. diminution policy,” which prevents a school’s MOOE funds in Figure 1: National MOOE Funding for Schools Has Increased Significantly in Recent Years Total and per student appropriations for school MOOE, 2005-2015 (in 2014 constant prices) Total appropriations Appropriations per-student 16 10% 1,000 948 High School 876 Elementary School 825 792 % of total DepEd 8% 731 PHP billion (2014 prices) 12 budget (right axis) 750 % of DepEd budget PHP (2014 prices) 6% 566 482 8 500 442 397 4% 333 317 4 250 2% High School 107 Elementary School 0 0% 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Sources: National government appropriations from General Appropriations Act, various years. Enrollment data come from DepEd Fact Sheets, various years. 88 WWW.WORLDBANK.ORG/PH Table 1: Components of the MOOE Allocation Formula 2013 to 2015 2016 Elementary High Schools Elementary High Schools Allocation component (PHP) (PHP) (PHP) (PHP) 1. Fixed 40,000 80,000 50,000 96,000 2. Per Student 200 400 250 480 3. Per DepEd Teacher 4,000 8,000 5,000 9,600 4. Per classroom 3,000 6,000 3,750 7,200 5. Per graduating student 250 250 313 300 Notes: Prior to 2016, the xed component for technical and vocational high schools was PHP 160,000. DepEd follows a no diminution policy that, in any given year, prevents a school’s MOOE from being less than its level in the previous year. any given year from being less than the previous year’s level. The limited knowledge that education stakeholders have However, it is di cult to identify the main causes for each of the funding formula reduces accountability for MOOE school because school-level data on the formula components allocations. While over half of all school principals claim to (for example, on the number of students in each school) is not be aware of the MOOE formula, fewer than one in ten know published on DepEd’s website alongside the data on school- the formula’s components and hardly any know the funding level MOOE allocations. This prevents school o cials from associated with each component (Figure 2). Knowledge being able to understand any discrepancies in their MOOE of the allocation formula is also limited among DepEd’s allocations and reduces the bene ts, in terms of increased division o cials. The study team’s interviews with student transparency, that a funding formula can provide. households also revealed a lack of awareness among parents. Figure 2: Knowledge of the MOOE Formula among Education Stakeholders is Limited Percentage of stakeholders who are aware of MOOE funding and its components DepEd sta Parents of students 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% Elementary school High school Division O cials Parents of elementary principals principals and high school students Reportedly knows MOOE formula Know schools receive MOOE Actually knows components of formula Received information on MOOE allocation Actually know the correct formula (out of those who know) Source: PETS-QSDS national survey - division, school, and household levels. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 89 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Box 1: Approach to Tracking In order to track the ow and use of MOOE funds, the study team obtained from o cial records maintained by Division o ces the amount they received for school MOOE, and on the amount they reported downloading to each PETS-QSDS sample school and for each MOOE advance. The records also included details on the check number, the date of check, the date on which the check was collected by schools, and the date on which the corresponding liquidation report was submitted. The same set of information was collected from records maintained by schools. In addition, the team collected information on the types of expenditures for which the schools used their MOOE funds from school liquidation reports corresponding to each MOOE advance downloaded to the school from the division o ce. Only 40 percent of parents were aware that schools receive reporting on funds. The survey used o cial records (such MOOE funds, and the majority did not know how MOOE as notices of cash advancements, checks, and liquidation funds were allocated. With no knowledge of the formula, reports) to record the receipt and downloading of funds schools, division o ces, and parents cannot con rm whether from each administrative level and to schools (Box 1). the schools are receiving the correct allocation and are unable to seek redress in cases where their funding falls The main nding of this tracking exercise is that only 77 below expectations. percent of the total allocations for elementary and non-IU high schools ultimately reached those schools in 2013 (Figure 3).11 In other words, elementary and high schools received Tr nsfer of MOOE Funds to Schools only PHP 448 of the PHP 581 allocated for each student. The PETS-QSDS study tracked MOOE funds from their initial Given that the guidelines stipulate that all school MOOE allocation in the DepEd central o ce all the way down to funds should be downloaded to schools, this nding is a their receipt by a nationally representative sample of schools. concern. It implies that about PHP 1.8 billion out of the PHP 8 Implementing unit (IU) high schools were excluded from the billion MOOE budget intended to be used by schools in 2013 analysis since they follow a di erent system for receiving and were not downloaded.12 Figure 3: A Significant Portion of the MOOE Allocations Do Not Reach Schools Share of MOOE allocation downloaded and received by schools, 2013 Share of original allocation Proportion of schools downloaded and received, 2013 100% 100% 84% % of original allocation 77% 75% 50% 25% 0% Central MOOE MOOE downloaded MOOE received by Elementary Schools High Schools allocation to schools by schools from Less than full allocation Allocated amount division division Download exceeds allocation Sources: Data on allocations from DepEd central o ce. All other data from PETS-QSDS national survey – division and school levels. 90 WWW.WORLDBANK.ORG/PH These aggregate ndings can be explored further to identify Figure 4: Divisions Retain MOOE Funds to Spend on the key stages where school MOOE funds are diverted. The Behalf of Schools rst stage relates to di erences between DepEd central o ce Reasons why DepEd division o ces retain MOOE funds (% of divisions), 2013 allocations and how much DepEd division o ces actually report downloading to schools. The second stage relates 60% to di erences in the amount that DepEd division o ces report downloading to schools and the amount that schools, through their records (for example, liquidation reports), report receiving. 40% Di erences between DepEd central o ce allocations and the amount of MOOE funding that the division o ces 20% downloaded to schools account for the largest share of funds that fail to reach schools. These di erences account for 16 percent of the 23 percent of school MOOE funds that do not reach schools. The records show that DepEd division o ces 0% Procurement Payment of Payment for failed to download the full allocation to over 80 percent on behalf of utilities for division-organized schools some/all schools in-service of elementary and non-IU high schools in 2013 (Figure 3). teacher training In many schools, these di erences were large; in around Source: PETS-QSDS national survey – division and school levels. 20 percent of elementary schools and 15 percent of high Note: Only divisions that reported retaining funds are included. schools, the amount downloaded by division o ces was less than half of the schools’ original allocation. These under- allocations were partly o set by over-allocations to some other schools; approximately 13 percent of schools were by divisions also makes it di cult to monitor whether given amounts that exceeded their allocation.13 school MOOE funds are being used for their intended purpose as schools have no way of verifying how the The retention of school MOOE by Division o ces is the main divisions used the retained funds. factor behind di erences between initial allocations and the funds downloaded to schools. While a 5 percent deduction Smaller discrepancies are also evident in the second by the Department of Budget and Management (DBM) for stage of the ow of MOOE funds to schools. Of the 23 tax purposes explains part of the gap between the MOOE percent of school MOOE funds that do not reach schools, funds allocated and received, the bulk of the gap at this level 7 percent can be accounted for at this stage (Figure 3).15 is explained by division o ces withholding funds. However, the poor quality of records at the school level makes it di cult to make a full accounting of the extent Although division o ces are required to download funds of the funds diverted at this stage. While it is possible to to schools in full, the tracking exercise indicates that over 60 compare the aggregate annual amount of MOOE funds percent of divisions held onto some MOOE funds in 2013.14 downloaded by divisions to schools with the amounts They did so to procure items for schools, to pay their utility that the schools received, gaps in the records kept at the bills, or to fund other services for schools (Figure 4). This is the school level stymie e orts to identify the sources of the case even though DepEd explicitly prohibits divisions from discrepancies. For example, at the school level, one-third procuring items using school MOOE funds except in extreme of the entries recording the schools’ receipt of the MOOE cases where there is a demonstrated need for the division transfers do not include a check number, which makes to do the procurement. This practice reduces the amount of it impossible to compare them with each individual discretion that schools have over the use of their own MOOE check sent by the division. Such gaps in documentation funds and has the potential to limit the e ectiveness of emphasize the need to strengthen MOOE reporting ongoing school-based management reforms. Procurement processes at every level. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 91 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Figure 5: Schools Often Do Not Receive their Full Allocation of MOOE Funds Because of Delayed and Unpredictable Transfers by Division Offices Timing of first MOOE advance and share of overall allocation received, 2013 Cumulative share of schools by receipt of rst Share of total MOOE allocation received by month MOOE advance, 2013 of rst advance, elementary schools, 2013 100% 100% 95% 90% 75% 85% 80% 50% 75% 70% 25% Elementary Schools 65% High Schools (non IU) 60% 0% January February March April May June July August September October November December January February March April May June July August September October December Month of rst advance Source: PETS-QSDS national survey – school level. Any delays by DepEd division o ces in downloading MOOE School M n ement of MOOE Funds funds to schools not only prevent schools from receiving those funds in a timely manner but also reduces the total The late downloading of MOOE funds by divisions is amounts that they receive. For example, in 2013, while most compounded by delays on the part of schools in collecting division o ces downloaded the rst advances to schools their MOOE checks. A comparison of the issuing and at the beginning of the calendar year, around 10 percent of collection dates of MOOE checks shows that only 40 percent schools did not receive their rst advance of MOOE funds of elementary school checks and 20 percent of high school until the start of the school year in June (see Figure 5). Schools checks were collected from the division o ces on the day on that receive their rst advance late also tend to receive a which they are released. These collection delays can be quite smaller share of their total allocation. For example, elementary long. Over one-third of MOOE checks were collected more schools that received their rst advance in February 2013 than two weeks after they were issued. In some cases, when received approximately 80 percent of their full allocation transfers are delayed or unpredictable, school principals use whereas schools that received their rst advance after August their personal funds or raise money from teachers to cover received less than a quarter of their full allocation.16 the school’s operating expenses. School liquidation reports show that in 2013 over 60 percent of elementary school and Any irregularity in the downloading of MOOE funds makes 45 percent of high school principals paid for approximately school planning more di cult and potentially less e cient. 5 percent of their operating expenses from their own funds In 2013, 20 percent of elementary schools and 25 percent while they waited for their MOOE funds. of high schools received MOOE funds from division o ces every month, but 34 percent of elementary schools and 30 School liquidation reports reveal that some schools nd percent of high schools received only ve advances in the it di cult to use all of the MOOE funds that they receive. whole nancial year. In the remaining schools, downloading Although schools are required by law to spend all of the did not follow any regular pattern.17 MOOE funds that they receive, around 70 percent of schools 92 WWW.WORLDBANK.ORG/PH were unable to do so in 2013. On average, a quarter of MOOE Smaller elementary schools and those in rural areas are more funds received by these schools went unspent. The under- likely to struggle to spend their MOOE resources. For example, spending of available MOOE funds is especially concerning the elementary schools that reported facing such issues had because other studies have revealed the inadequacy of an average allocation of PHP 132,500 each compared with the current levels of MOOE to meet schools’ needs.18 However, average of PHP 191,600 for elementary schools that reported any e orts to persuade the government to raise the MOOE experiencing no di culties in spending their allocations. The budget will meet with only limited success unless these lower allocation re ects the smaller number of students in schools facing liquidation di culties. Small high schools were school liquidation di culties are overcome. also more likely to nd it di cult to spend their MOOE funds, The main reason why so many schools nd it di cult to but the di erences between small and larger high schools were not statistically signi cant.20 spend their entire MOOE allocation is because they have insu cient capacity and time to meet their reporting and The management of MOOE funds is particularly demanding procurement requirements. Over a third of elementary on school principals. In the 2014 school year, principals schools and a quarter of high schools reported encountering reported spending six to eight hours a week on administrating such problems in 2013, and the lack of capacity to meet MOOE funds, including procuring goods and services and reporting requirements was one of the most commonly preparing liquidation reports (Figure 7). It is possible that a cited challenges (Figure 6). Other salient problems included key factor that determines the time needed to process MOOE di culties locating suppliers and canvassing.19 School funds is related to the location of the school. For example, for principals also reported having insu cient time to spend schools far away from division o ces and markets, it takes MOOE funds in general. Although fewer schools reported more time to collect their MOOE checks and to nd suppliers. facing such issues in 2014, their primary challenges remained However, the study team’s analysis shows that this is not the the same – a lack of time and insu cient capacity. case nor is it related to whether schools are located in rural or Figure 6: Schools Do Not Have Enough Time or Capacity to Spend All of Their MOOE Funds Challenges experienced by schools in spending MOOE funds, 2013/14 school year 50% Elementary Schools 40% Non-IU High Schools 30% 20% 10% 0% Di culties with Insu cient time to Di cult to identify Canvassing di cult and Required items not meeting reporting liquidate funds suppliers/located time-consuming available from the DBM requirements too far away procurement service Source: PETS-QSDS national survey – school level. Notes: Only schools that reported experiencing di culties in spending their MOOE allocations are included. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 93 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Figure 7: Managing MOOE Funds at the School Level is Time-consuming Average hours spent on MOOE administration per week by schools’ distance to division o ce, 2014 school year Elementary Schools Non-IU high schools 12 12 10 10 9 hrs 9 hrs 8 hrs 8 hrs 8 8 6 hrs 6 hrs 6 hrs 6 6 4 hrs 4 4 2 2 0 0 Less than 30–90 More than Average Less than 30–90 More than Average 30 minutes minutes 1.5 hours 30 minutes minutes 1.5 hours Time to travel to Division o ce from school Time to travel to Division o ce from school Source: PETS-QSDS national survey – school level. Note: The label above the bars shows the average number of hours taken to administer MOOE funds by schools located at di erent distances from their division o ce. urban areas (Figure 7). The team’s ndings suggest that the bulk of the time that school principals spend on managing Equit of School MOOE funds goes on activities that are common to all schools such as reporting and canvassing. MOOE Fundin Schools serving poorer and more marginalized groups The di culties that school principals experience in managing are provided with higher levels of MOOE funding for each MOOE funds can result in signi cant delays in their submission student. The PETS-QSDS survey included a nationally of liquidation reports as well as some misreporting of data. representative sample of the households of public Although schools are usually required to submit liquidation elementary and high school students. Using information reports within 35 days of receiving an MOOE advance only collected on consumption and asset ownership collected about half of all elementary school advances and one-third in the survey, it is possible to rank student households by of high school advances were submitted within this period.21 Some schools submit reports as long as three months after the estimated levels of per capita household consumption.22 initial advance was made. This indicator of household welfare reveals that elementary students from poorer households are more Moreover, division o cials report that these delays are a likely than students from wealthier households to attend leading cause of their own delays in downloading MOOE schools that have higher per capita allocations of MOOE funds to schools. This is the case because schools are not funds (Figure 8). This is not because the MOOE funding allowed to receive their next MOOE transfer unless they formula provides additional funding to schools serving submit a complete and correct liquidation report on how poorer groups. Instead, this is partly a re ection of the fact they spent the previous transfer. Of the 37 percent of division that wealthier students are more likely to attend schools o cials who reported delays in downloading MOOE funds to in urban areas where student-classroom and student- elementary schools in 2014 and the 31 percent who reported teacher ratios are higher, thus leading to lower per-student delays in transferring MOOE funds to non-IU high schools, MOOE funding (Table 1). High schools serving the poorest over half of them cited late liquidation reports as the reason, students also have higher per-student allocations than while more than a third (for elementary schools) and a quarter (for high schools) mentioned problems with previous those serving the wealthiest students, but the di erence is liquidation reports. not statistically signi cant. 94 WWW.WORLDBANK.ORG/PH Figure 8: The MOOE Funds that Schools Receive Do Not Reduce Inequalities in Overall School Funding Elementary school per-student funding by source, 2013 and 2013/14 school year Per-student MOOE amount in elementary Per-student non-MOOE funds in elementary schools, 2013 schools, 2013–14 700 400 Allocated PHP 600 556 Received 300 500 PHP 433 200 400 PHP 344 PHP 321 100 300 200 0 Poorest 20 percent Wealthiest 20 Poorest 20 percent Wealthiest 20 of households percent of households of households percent of households Source: PETS-QSDS national survey - school and student household levels. Notes: Non-MOOE sources are all other cash contributions received by schools directly (e.g. local government, Parent Teacher Associations, fees) and school based management grants. A full description of the funding sources included and the approach to estimating overall school revenue is contained in additional tables to the full PETS-QSDS report. However, in elementary schools, the pro-poor allocation possible to establish whether poorer schools receive their virtually disappears in terms of the amount of MOOE funding full allocation after any procurement by the division o ces that the schools actually spend. The result is that, in practice, is factored in. This signi cantly reduces the transparency students from di erent socioeconomic groups attend of and accountability for MOOE funds and may be fueling schools that receive similar levels of per-student MOOE unintended inequalities in levels of school funding. despite the large di erences in their original allocations. This is a missed opportunity to use school MOOE as a way to narrow the very large di erences in school revenue among Oversi ht nd Monitorin schools serving di erent socioeconomic groups (Figure 8). of the Use of MOOE Funds The PETS-QSDS ndings also suggest that, among Schools’ liquidation reports show that they spend only elementary schools, a larger share of the MOOE allocations a small proportion of their MOOE funds on prohibited for schools in poorer areas is not reaching the schools. It items. Over 80 percent of school MOOE expenditures are might be expected that division o ces are managing a spent on allowable items (Table 2 and Figure 9). However, larger share of the resources of poorer schools given that in 2013, elementary schools used 5 percent of their MOOE they are likely to lack the capacity to manage the funds funding to purchase equipment and instructional materials themselves (Figure 4). Since there is little information on prohibited under current guidelines, while high schools how the division o ces spend school MOOE funds, it is not used 2 percent of their funding on prohibited items. Table 2: Allowed and Prohibited Spending Categories for MOOE Funds Allowable expenses Ineligible expenses Supplies for teaching, utility payments, minor repairs, travel, Textbooks and instructional materials (such as publications, wages for janitorial services, teacher training, graduation periodicals, or review materials), furniture and equipment, ceremonies, and other activities identi ed in the School and teacher salaries. Improvement Plan (SIP). ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 95 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Figure 9: Most MOOE Funds Are Spent in Accordance with Existing Guidelines Percentage of total MOOE expenses by item of expenditure, 2013 Elementary Schools Non-IU high Schools 35% 30% 30% 25% 25% 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% Supplies Repairs Utilities Printing/ copying Other Travel Instructional materials Teacher training Janitors/security wages Equipment Supplies Travel Repairs Utilities Other Printing/ copying Teacher training Janitors/security wages Equipment Instructional materials Source: PETS-QSDS national survey – school level. Approximately 10 percent of total school MOOE expenditures While there are regulations in place that aim to strengthen were not reported on liquidation reports so it was not the transparency over MOOE use, compliance with these possible to assess whether those expenses were eligible. regulations is low. The Anti-Red Tape Act (2007) mandates public schools to maintain a transparency board or bulletin The schools’ records also reveal that the proportion of board detailing how they are using their MOOE funds. funds that they used to purchase items directly related to However, as of 2014, fewer than two-thirds of schools had learning was relatively low. For example, elementary schools a transparency board containing information on MOOE spent only 3 to 4 percent of their total MOOE expenditures expenditures (see Figure 10). Moreover, the information on teacher training and only 4 percent on purchasing posted on about a third of transparency boards was more instructional materials such as periodicals or review materials. than three months old. Figure 10: Schools’ Compliance with the Transparency Board Requirement is Low Percentage of schools that had transparency board displaying MOOE information in 2014 Elementary Schools Non-IU high Schools 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Has transparency Transparency board Has transparency Transparency board board with includes DepEd orders board with includes DepEd orders information on on appropriate MOOE information on on appropriate MOOE MOOE liquidation expenditures MOOE liquidation expenditures Source: PETS-QSDS national survey – school level. 96 WWW.WORLDBANK.ORG/PH Figure 11: There Is Little Participation by Local Stakeholders in Decisions about How MOOE Funds Are Used Percentage of schools that received input from stakeholders about MOOE funds, 2014 SY Elementary Schools Non-IU High Schools 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Teachers PTA School Governing Council Parents Local School Board Teachers PTA School Governing Council Parents Local School Board Source: PETS-QSDS national survey – school level. Another factor that contributes to weak monitoring is participating in MOOE decisions in fewer than 10 percent the lack of knowledge among stakeholders of the rules of schools. The study’s interviews with parents revealed a governing MOOE funds. Although schools are legally similar picture. Only about one-third of parents reported required to publicly disclose how they use their funds, there participating in discussions about the school’s use of nancial is no speci c law requiring them to post information on resources in general and fewer than a quarter reported their expenditures on allowable and prohibited items. Not participating in decisions about how to use the school’s surprisingly, only about half of the elementary schools and MOOE funds. a quarter of the non-IU high schools that have transparency boards post DepEd’s orders on the appropriate use of MOOE funds (Figure 10). This lack of information makes it di cult for Policies to Stren then the parents and other local stakeholders to judge whether the funds are being used appropriately. Alloc tion, Implement tion, nd Account bilit of Decisions on the use of MOOE funds are largely con ned to the school principal and other teachers. In 2014, over 80 MOOE Funds percent of schools reported that teachers had been consulted about how to use the schools’ MOOE funds (Figure 11). PTAs This note has shown that the bulk of MOOE funds are have an input into how MOOE funds are used in fewer than transferred to schools and that schools use these funds 30 percent of schools. This is not surprising since DepEd’s for the purposes for which they were intended. However, current guidelines explicitly prohibit PTAs from “interfering in in light of the government’s plans to signi cantly increase schools’ administrative management.”23 school operational funding, it will be necessary to tackle the remaining weaknesses to ensure that these funds are used Only 13 percent of elementary schools receive input from in the most e ective way to improve education outcomes. school governing councils (SGCs) even though SGCs are It is crucial that the system for allocating and managing responsible for developing and implementing the school MOOE funds is strengthened to ensure that schools receive improvement plan (SIP), which includes activities nanced their allocations in full and in a timely manner. The current by MOOE funds.24 This number is even lower among non-IU problems that some schools face in spending these high schools, with SGCs, parents, and local school boards funds fully and in an e ective way also need to be solved. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 97 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Increasing the transparency and oversight of fund use by division o ces to inform schools about their allocation at the involving school governing councils and parents could be beginning of the scal year. Moreover, posting the criteria an e ective way of improving school management of MOOE used to determine the allocated amount for each school funds. The note also found signi cant funding inequalities (such as the number of teachers and classrooms) online between schools serving disadvantaged groups and those as well as on the division o ce transparency board at the serving children from better-o households. Adjusting the beginning of the year would strengthen accountability. MOOE allocation formula to include an equity component would narrow these funding di erences and thus improve Simplif in the Liquid tion Process education outcomes. nd Supportin the M n ement of MOOE Funds Improvin the Flow of MOOE Funds to Schools Simplifying the existing requirements for schools to prepare liquidation reports on how they used their MOOE funds Ensuring that division o ces download all allocated MOOE would reduce the signi cant burden that this currently puts funds to schools would be an important rst step in on schools. It would also give school principals more time to improving the use of MOOE funds. Tightening and enforcing focus on providing academic leadership in their schools. One the rules governing division o ces procuring material on approach that has been adopted in other countries would behalf of schools would reduce the extent to which division be to treat MOOE funds as grants. For example, this is the o ces retain MOOE funds rather than transfer them directly case in Indonesia, where these funds now have much less to the schools. Even in cases where centralized procurement onerous reporting requirements than required by the usual may be more e cient, the division o ces could still government budget and implementation systems.25 Schools download the funds to schools, who could then pay the use simpli ed reporting templates to report on their use division o ces for any goods and services procured on their of the funds and submit these reports to district (division) behalf. This would also ensure that these funds are properly o ces every quarter. The Indonesian system also encourages accounted for in school liquidation reports and make it easier much more involvement by school committees (school for all MOOE funds to be monitored. Setting division o ce governing councils) in overseeing the use of school MOOE targets for the full and timely downloading of school MOOE funds and tying these to performance-based bonuses (PBB) funds, a role that used to be played by the district o ce, would also facilitate the ow of funds. and this has strengthened monitoring. Developing a similar system, adjusted for the speci c context of the Philippines, Stepping up monitoring e orts will also be essential to could reduce the burden on schools and at the same time improve the ow of funds to schools. The current reporting strengthen oversight of the use of funds. requirements hinder the e ective monitoring of MOOE downloading to schools. While division o ces are required Reducing the frequency of transfers from once a month to submit monthly reports on the status of downloading and to once a quarter would make procurement processes utilization to the O ce of the Secretary, there is no mechanism more e cient and would reduce the time that school to verify whether each school in the division has received their principals would need to spend preparing liquidation full allocated amount. Establishing a system to provide this reports. Releasing larger sums on a less frequent basis information as well as a feedback mechanism for schools to would make it easier for schools to procure items in bulk for inform DedEd about the fund transfers that they have received both immediate and future activities instead of having to would go a long way towards strengthening monitoring. procure small quantities frequently. It would also reduce the time needed to undertake canvassing and the time spent In order to strengthen accountability, it is vital that schools travelling to make purchases. Less frequent transfers would and other stakeholders know the details of the funding also reduce the number of liquidation reports that schools formula and the correct MOOE allocation for each school. A would need to submit, which again would reduce the rst step would be to enforce the existing requirement for burden on schools. 98 WWW.WORLDBANK.ORG/PH Extending the current deadline for schools to submit them to sign o on liquidation reports and to ensure that liquidation reports on their MOOE expenditures would the reports are posted on school transparency boards. It will further ease their management burden. Giving schools more be vital to disseminate information about the current and than the current 35 day limit would give them more time future responsibilities of the SGCs in order to reinvigorate this to spend MOOE funds and to do so e ciently. For example, important school-level institution and to hold schools more they would have more time to identify the best available accountable for their use of MOOE and other funds. suppliers in terms of both cost and quality. Ensuring that MOOE and other funds that schools receive Schools will need more support and training from the DepEd from both DepEd and other sources are added together and district level o ces to build their capacity to manage MOOE treated as a single school budget in the SIP can also increase funds. It is not nancially feasible to provide a dedicated oversight over use of MOOE funds. Having a single budget in bookkeeper to each school to manage the administrative each school can help stakeholders such as teachers and SGCs requirements associated with MOOE funding. Instead, the to identify any potential duplication of expenses and enable role played by district o ces in assisting and supporting e ective monitoring of MOOE and other school funds. schools needs to be expanded. In particular, district o ces could encourage schools to spend their MOOE funds in Reduce Fundin Inequ lities b accordance with the priorities in their school improvement Includin n Equit Component plans and could help them with the administration of MOOE in the MOOE Fundin Formul funds. Providing stakeholders such as teachers, parents, and members of SGCs and PTAs with training on how to spend Introducing an equity component into the MOOE funding MOOE funds e ectively could also help schools to manage formula would be one way to reduce the large school-level their funds better. funding inequalities highlighted in this note. In order to compensate those schools that receive lower funding from Allowing schools, at least those with bigger allocations, to non-DepEd sources such as local governments, an equity deposit their MOOE funds in bank accounts would also help component could be introduced in the allocation formula. them to manage their funds safely. Many schools have to This could go some way to equalizing school funding and maintain large sums of MOOE funds in cash, which is clearly provide schools in poorer areas with the additional resources a security risk. For example, rough calculations suggest that that they need to support the learning needs of marginalized about 22 percent of non-IU high schools with a thousand students. Many other countries, like the United Kingdom, or more students would have to hold onto more than PHP have introduced funding components of this kind (Box 100,000 in cash at any given time if the division o ces were 2). MOOE allocations should also be adjusted to take into to transfer allocations to schools on a quarterly rather than account price di erences between regions so that schools a monthly basis. Allowing schools to deposit these funds in operating in high-cost areas such as remote locations where bank accounts would overcome this security issue. travel and transportation costs are high are able to purchase similar amounts of goods and services as those in lower-cost Stren then Oversi ht over the locations. Use of MOOE Funds b Schools Over the last ve years, the Philippines government has Strengthening the role played by school governing councils provided schools with ever greater amounts of operational (SGCs) in overseeing MOOE funds has the potential to lead funding. While this trend is set to continue, it is imperative to more e ective use of MOOE funds. While SGCs are already that the systems used to allocate and manage MOOE funds expected to play a role in school improvement planning, are strengthened. Only then will the full potential of funds of their role could be expanded, for example, by requiring this kind to improve education outcomes be realized. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 99 Policy Note 5: Providing Schools with Adequate Resources to Deliver Quality Education Box 2: The Pupil Premium in England In 2011, the United Kingdom government introduced a “pupil premium.” This initiative provides government-funded schools within England with additional per-student funding to raise the attainment of disadvantaged pupils and to narrow the gap between them and other students. In 2014/15, each school received an additional £1,300 (US$2,031) for primary-aged students and £935 (US$1,461) for secondary-aged students. Rough calculations suggest that an average sized secondary school would receive approximately £200,000 (US$312,500) in additional funding through the pupil premium, which is the equivalent of ve full-time teachers. The main criteria used to calculate schools’ eligibility for this extra funding is the number of students in the school that have received free school meals over the last six years. Head teachers and school governing bodies are accountable for the use of these funds in two ways. First, performance tables that outline the performance of disadvantaged students compared to their peers are made available to the public. Second, schools are required to publish details online each year of how they have used the premium and what impact it has had. Schools typically use the additional resources to hire more teachers and teaching assistants in order to introduce special programs for disadvantaged students. In addition, the resources are frequently used to allow eligible students to participate fully in after-school activities. A recent study of the implementation of the pupil premium found that: • Schools are using the funding e ectively to improve learning outcomes of disadvantaged students and narrowing learning disparities. • The best schools combine targeted interventions with robust tracking to evaluate their e ectiveness. • School governing bodies take strategic responsibility for student targeting and hold school leaders accountable. • Challenges remain in schools with weak leadership, including weak governing bodies. Sources: OFSTED (2014) and www.gov.uk/pupil-premium-information-for-schools-and-alternative-provision-settings. Table 3: Strengthening Systems to Provide Operational Funding to Schools Findings Policy suggestions Not all schools receive their • Enforce existing regulations on the transfer of funds full MOOE allocation and the • Introduce transfer targets for DepEd division o ces tied to performance bonuses transfers are often not made on a predictable schedule • Make MOOE allocations and formula components for each school publicly available The management of MOOE • Treat school MOOE funds as a grant to simplify reporting requirements funds within schools is • Reduce the frequency of downloading and increase the time allowed for schools to di cult and time-consuming produce liquidation reports • Provide additional support to schools on MOOE management from district o ces and provide school-level training • Allow schools to deposit MOOE funds in bank accounts School-level institutions like • Review and strengthen the role of SGCs in the planning and monitoring of MOOE funds the school governing council • Disseminate information on the roles and responsibilities of SGCs to parents and the are relatively weak local community • Ensure that MOOE and other school-level funds are incorporated into school improvement plans Funding inequalities at the • Introduce an equity component into the MOOE allocation formula. school level are large 100 WWW.WORLDBANK.ORG/PH Endnotes 1 MOOE is di erent to school grant funding because schools are 13 Occasional congressional insertions (additions to the originally required to use and account for the use of these funds in the approved budget) or additional MOOE funding for Special same way as any other recipient of budget funds would be. Education programs may partly explain why some schools 2 Khattri, N., C. Ling, and S. Jha (2010). “The E ects of School-based received more than their allocated amount. Management in the Philippines: An Initial Assessment Using 14 Interestingly, only a quarter of the divisions admitted to retaining Administrative Data.” Policy Research Working Paper Series. No. funds when asked. 5248, World Bank, Washington, D.C. 15 Schools reported receiving only 77 percent of the funds that the 3 World Bank and AusAID (2013). “School-based Management division o ces reported downloading in the rst three-quarters in the Philippines: An Empirical Investigation.” World Bank and of 2014. AusAID, Manila. 16 The pattern is similar for high schools. 4 AusAID ERF (2011). “Current Issues in Education: School Grants 17 The available data for 2014 show a similar pattern. and School-based Management.” Canberra. 18 World Bank (2016). “The Development and Implementation 5 The ndings reported in this note are discussed in more detail in of a Normative School MOOE Formula in the Department of additional annexes and tables accompanying the main PETS- Education in the Philippines.” Washington D.C. QSDS report. 19 Canvassing relates to the requirement that schools obtain at least 6 Years refer to nancial years (January to December) unless three quotes for purchases that they make with MOOE funds. otherwise stated. 20 There were no di erences in the proportion of rural and urban 7 Discretionary funding refers to the funds that schools receive in high schools that reported nding it di cult to spend their cash, over which they have discretion to use as they please. MOOE allocations. 8 World Bank (2016). “The Development and Implementation 21 More speci cally, schools are required to submit a completed of a Normative School MOOE Formula in the Department of liquidation report to the division o ces by the fth day of the Education in the Philippines.” World Bank, Washington D.C. month after they have received an MOOE advance in order to 9 These averages are based on statistics on the 2013 MOOE receive their next MOOE advance. allocations from DepEd for the elementary schools in the PETS- 22 The household questionnaire included a short module on QSDS sample and from the DBM for high schools. consumption and a set of questions on assets that have been 10 Author’s calculations using eBEIS data on enrollment and on the used by the Department of Social Welfare and Development numbers of teachers, classrooms, and graduating pupils for all (DSWD) to undertake a proxy means testing (PMT) approach public elementary schools. to estimating household consumption per capita. The results 11 The word “allocation” here refers to the cash allocation released by reported here are based on information gathered using the PMT the DBM to DepEd divisions based on authorized allotments and approach, and a full description is included in a separate note. released to cover spending obligations. 23 DepEd Order 54, s. (2009). “Revised Guidelines Governing Parents- 12 The study team determined this amount by scaling up the sum Teachers Associations (PTAs) at the School Level.” Department of of the amounts allocated for public elementary and non-IU high Education, Manila. schools in the PETS-QSDS survey to represent nationwide MOOE 24 DepEd (2009). “A Manual on School Governing Councils.” funding. In addition, they estimated national-level funding for Department of Education, Manila. non-IU high schools as about one-quarter of the total funding 25 Al-Samarrai, Samer; Fasih, Tazeen; Hasan, Amer; Syukriyah, for high schools (PHP 5.06 billion) as suggested by the data from Daim. 2014. “Assessing the Role of the School Operational Grant the DBM on per-school funding for the schools in the PETS-QSDS Program (BOS) in Improving Education Outcomes in Indonesia.” sample. World Bank, Jakarta, Indonesia. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 101 Polic Note 6: Assessin the Role Pl ed b Loc l Government in Supportin B sic Educ tion Introduction Many developing countries have devolved the responsibility for education services to local governments in an e ort to improve educational quality and make public spending more e cient. Advocates of decentralization have argued that bringing decision-making closer to schools makes public policy more responsive to local needs, strengthens accountability, and fosters innovation. In some countries, decentralization has gone further in that schools have been given responsibility for developing their own improvement plans and a degree of autonomy over the use of their resources. While the Philippines decentralized some public services to local governments in the early 1990s, basic education remains largely under the control of the national government. The 1991 local government code devolved responsibility for many basic services to local government units (provinces, cities, municipalities, and barangays1) including primary health care, hospital care, social welfare services, and a range of environmental and agricultural services. Local governments’ responsibilities in the eld of education were largely limited to maintaining basic education infrastructure and funding sports activities. The funding that each local government uses to ful ll its basic education responsibilities comes primarily from a surcharge Philippines. Using data that the PETS-QSDS team has on local property taxes that goes into a Special Education carefully collected from a nationally representative Fund (SEF) managed by a local school board (LSB). Over time, sample of elementary and high schools, it explores the LSBs have used SEF funding for a growing range of purposes magnitude of the funding that local governments are including, for example, employing additional school-level giving to schools and assesses the systems that govern sta and paying school utility bills. However, responsibility the use of local government funds in the education for education policy, standards, curriculum, and the hiring of sector. The policy note shows that local governments’ teachers and the bulk of education nancing remains with contribution to overall public education funding is small the national government. and highly inequitable. It also shows that the systems used to allocate and manage these funds are weak and The purpose of this policy note is to assess the nancing that greater transparency and accountability over these of basic education services by local governments in the resources is urgently needed. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 103 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Trends in Loc l Government Local government trends in education spending have translated into declining levels of funding for each basic Fin ncin of B sic education student (Figure 1). Between 2003 and 2013, LGU spending per student declined from PHP 951 to PHP 744, a Educ tion fall of over 20 percent in real terms. This is in stark contrast to national government funding, which rose by 35 percent from Education makes up only a small share of local government PHP 9,500 to PHP 12,800 over the same period. spending and has not changed much over the last decade (Figure 1). The composition of local government spending Local government support for education has declined is aligned with their devolved responsibilities, with the even though local revenue for education has increased bulk of funds being allocated to the provision of general considerably. Each province, city, and municipality in the public and economic services as well as health, nutrition, Philippines has a Special Education Fund (SEF), which is the and population services. In 2014, only PHP 13.3 billion or source of the majority of its spending on basic education. 5 percent of local government spending was devoted to These SEFs are funded by a 1 percent surcharge on property education. taxes raised by the local government.2 Between 2007 and 2014 and partly as a result of increasing property prices, Local government funding also represents a very small and SEF income has risen in real terms by approximately 40 declining share of overall public funding for basic education. percent from PHP 17 billion to PHP 24 billion (in constant While spending on basic education by local government 2014 prices). However, over the same period, SEF spending units (LGUs) has uctuated considerably over the last 10 years, actually declined, from PHP 15 billion to PHP 12 billion. in real terms, it appears to have been on a declining trend The accumulated surplus of unspent SEF funds since since 2007 (Figure 1). Combined with ever-increasing levels 2007 amounts to over PHP 63 billion. While the reasons of national spending, the share of public basic education for this under-spending are likely to di er between local funding accounted for by local government spending governments, earlier studies have pointed to problems with declined from 11 percent in 2006 to only 6 percent in 2013. cash management and poor budgeting.3 Figure 1: Spending on Basic Education by Local Governments is Low and Has Been Declining Total and per-student public basic education spending, 2003–2013 Total public spending Public spending per student 18 20% 14,000 PHP billion (2014 prices) 17 15% PHP (2014 prices) 10,000 16 10% 11 11 6,000 15 10 10 10 9 8 7 7 5% 14 6 6 2,000 13 0% 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 LGU spending on basic education National government spending per student LGU as % of total public spending (RHS) LGU spending per student Sources: National government spending – DBM National Expenditure Program. Local government education spending – Bureau of Local Government Finance. Enrollment – DepEd FACT sheet. GDP and GDP de ator - Philippines Statistics Authority. Notes: Student numbers include kindergarten. 104 WWW.WORLDBANK.ORG/PH Figure 2: There are Large Differences in the Amounts that Local Governments Spend on Education Total and per student local government education spending by region, 2013 Total LGU spending LGU spending per student 8 4,000 6 3,000 PHP billions PHP 4 2,000 2 1,000 0 0 NCR Region IV-A Region III Region VI Region I Region VII Region X Region XI Region XII Region II Region IV-B CAR Region VIII CARAGA Region IX NCR Region IV-A Region I Region III CAR Region X Region VI Region XI Region VII Region XII Region II Region IV-B CARAGA Region IX Region VIII Region V Region V Sources: Bureau of Local Government Finance. Enrollment rates from DepEd FACT sheet. Notes: Student numbers include kindergarten. The aggregate patterns of local government education funding from the SEF.4 The LSB is supposed to meet on a spending mask large disparities among regions, which monthly basis and to make decisions on the basis of majority are administrative units that do not have their own voting, including the approval of the budget. The national elected governments but encompass provinces, cities, government sets broad priorities for how local government and municipalities (Figure 2). The National Capital Region funding should be used, largely related to supporting public (NCR), the region with the highest overall levels of property schools falling under the jurisdiction of the LSBs. tax revenue, accounts for over 44 percent of total local government education funding in the Philippines. This is a Interviews with members of local school boards conducted very large and signi cant source of additional funding for as part of the PETS-QSDS study revealed that they meet the school system, equivalent to approximately PHP 3,500 less frequently than outlined in the Local Government in additional funding for each basic education student in Code. The study found that boards meet, on average, every the region or about 28 percent over and above per student quarter rather than every month as the code suggests. funding from the national government. Disparities between These meetings often focus on the overall nances of the other regions are not as stark but still exacerbate funding board and the needs of schools, but the outcomes of these inequalities between schools in di erent locations. meetings are rarely communicated to schools and their principals. For example, it became clear from the study team’s interviews with elementary and high school principals Loc l Government Support that more than one-third of them were not even aware for Public Schools when or how often the school board meets. Moreover, it did not appear that school principals and other stakeholders are given an opportunity to provide feedback to their boards. Within each province, city, and municipality, a local school board (LSB) is responsible for the allocation and use of local Obtaining reliable and accurate information on the government education funds (see Box 1). Its main function is amount of nancial support that schools receive from local to develop an overall plan and budget for local government governments is di cult. While local governments report funding of basic education in their locality, including information on aggregate education spending in their ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 105 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Box 1: Legal Framework for the Role Played by LGUs and their Local School Boards in Basic Education The Local Government Code (LGC) issued in 1991 (Republic Act 7160) set out the legal framework for the role of LGUs in basic education. It initiated the creation of provincial, city, and municipal local school boards (LSBs) as the main bodies for making decisions over the use of local government education funds. Each LSB is co-chaired by the head of the local government and either the DepEd’s School Division Superintendent (for provincial and city LSBs) or the Public School District Supervisor (for municipal LSBs). Members of the parent-teacher associations and representatives of teachers and non-teaching sta are also represented on the LSB. The main functions of the LSBs are: (i) to develop the annual school board budget in accordance with the amount of revenue in the SEF and with their determination of the supplementary maintenance and operating needs of public schools covered by the LSB; (ii) to authorize SEF disbursements; (iii) to act as an advisory committee for the local council; and (iv) to recommend school name changes. The LGC and a series of joint circulars from DepEd, the Department of Management and Budget, and the Department of the Interior and Local Government have set out the main priorities for the Special Education Funds, including the provision of additional teachers, the construction and repair of school buildings, other capital investments for schools, educational research, the purchase of books and instructional materials, equipment purchases, sports, and other extra- curricular activities. Sources: Government of the Philippines Republic Act 7160, (1991). “An Act Providing for a Local Government Code.” Manila; Manasan, R. G., A.B. Celestino, and J.S. Cuenca (2011). “Mobilizing LGU Support for Basic Education: Focus on the Special Education Fund,” Philippine Institute for Development Studies, Manila. annual accounts, these reports do not contain any detailed from local governments (Figure 3). On the whole, local information on the type and level of support given to governments tend to provide direct nancial support to individual schools. This makes it impossible to assess how a slightly greater proportion of elementary schools than funds are being used and how much local government high schools. These di erences were large in urban areas education funding directly bene ts public schools. where 53 percent of all elementary schools receive support compared to only 28 percent of high schools. Although To overcome this constraint, the PETS-QSDS study collected a greater proportion of elementary schools receive LGU detailed information on all sources of school revenue and support, a much greater share of total LGU funding goes expenditure from a nationally representative sample of to high schools. Direct funding of high schools by LGUs is elementary and high schools. Based on the World Bank’s equivalent to PHP 331,000 for every high school compared to prior experience of collecting this information in the only PHP 206,000 for each elementary school. Philippines and other countries, the study team developed a questionnaire that would capture all cash and in-kind Levels of direct support to education from local contributions received by schools from all national and governments tend to be higher in highly urbanized cities local governments (provincial, city, and municipal LSBs (HUCs) than in other cities and municipalities, particularly and barangays), members of Congress, parent-teacher in the case of high schools (Figure 3). LGUs in HUCs provide associations (PTAs), NGOs and any other sources. The study approximately PHP 1.6 million for each high school while also collected information on contributions from parents as LGUs in municipalities spend only PHP 256,000 per high well as funds received from income-generating projects like school. This is not surprising given that LGUs in highly the school canteen.5 urbanized cities include the National Capital Region, which has the largest SEF revenues (see Figure 2). Incidence nd Over ll Level of Loc l Government Support On average, LGU contributions make up a very small share to B sic Educ tion of overall public school funding across the Philippines. A comparison of school revenue from all sources shows that The ndings of this detailed exercise revealed that fewer public schools receive the majority of their funding from the than 50 percent of schools receive any kind of direct funding national government through the Department of Education 106 WWW.WORLDBANK.ORG/PH Figure 3: Fewer than Half of All Schools Receive LGU Support, and the Amount of Funding Provided is Small Percentage of schools that report receiving cash or in-kind contributions from local governments and average levels of support, 2013–14 Percentage of schools receiving LGU support Average level of LGU support per school (PHP 000s) 60% 2,000 1,500 40% 1,000 20% 500 0% 0 HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools HUC Other Cities Municipalities All Schools Elementary Schools High Schools Elementary Schools High Schools Source: PETS-QSDS national survey – school level. Note: The average level of support is only for schools that received LGU contributions in 2013–14. (Figure 4). Other sources such as contributions from private other hand, represent less than 4 percent of the cash and in- individuals, NGOs, and school-level income-generating kind funding that schools receive directly from all sources. projects (such as school canteens) are signi cant sources of funding for elementary schools, having made up around 5 A comparison between levels of LGU direct school funding percent of their total funding in 2013/14. LGU funding, on the and total LGU education spending reported at the national Figure 4: The Contribution of LGUs to Direct School Funding is Also Low Composition of school funding for all schools and only schools that receive some LGU support, 2013–14 Composition of funding only in schools Composition of all school funding (PHP 000s) reporting LGU support (PHP 000s) 10,000 10,000 7,500 7,500 5,000 5,000 2,500 2,500 0 0 Elementary School High School Elementary School High School DepEd Barangay Local school boards Other Source: PETS-QSDS national survey – school level. Note: The left hand panel reports the composition of school funding across all schools while the panel on the right reports the composition only for schools that reported receiving LGU funding. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 107 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education level suggests that signi cant amounts are not being spent of local governments had paid the utility bills of some of on activities that directly bene t schools. Rough estimates the schools in their jurisdiction. However, once again, the by the PETS-QSDS study team using carefully collected magnitude of this support often went unrecorded, which nance data from school visits suggest that LGU spending makes it di cult to discover exactly where the majority of on schools amounts to approximately PHP 9 billion. This the reported LGU spending went. represents only 58 percent of the PHP 16 billion that LGUs reported spending on education in 2013 (see Figure 1). Use of LGU Support b Schools There is little information on how the LGU funds that are The bulk of local government support for schools is not spent at the school level are being used. The study provided in the form of goods and services rather than cash team made considerable e orts to collect such information (Figure 5). In-kind contributions account for 90 percent of from the local governments directly and from the DepEd total LGU support for elementary schools and 98 percent supervisors who sit on local school boards. However, little of LGU support for high schools. In providing speci c information was available and, even where it was, the data in-kind support, local governments must have detailed were patchy and did not correlate with the o cial aggregate knowledge of what schools need and of what support they spending gures reported by the LGUs. The study found that are already receiving from other sources. For example, in one destination for this indirect spending was additional the area of construction and rehabilitation activities, local allowances or bonuses provided by the LGUs to DepEd governments need to ensure that their e orts supplement teachers; one-quarter of provincial LGUs and 15 percent of rather than duplicate DepEd’s plans for school infrastructure city and municipal LGUs reported making these payments. development. When there is an overlap, this constitutes These payments are unlikely to have been recorded in the ine cient use of education funding. The alternative to local o cial school nancial records. While this may account for governments providing schools with in-kind contributions the di erences between actual and reported LGU spending, is to give them cash that could be factored into their annual it is not possible to estimate the magnitude of this kind of improvement and procurement plans. However, less than 10 expenditure. The study also found that around one-quarter percent of LGU contributions are provided in this form.6 Figure 5: Most LGU Funding is Provided In Kind Composition of LGU cash and in-kind contributions to schools (PHP), 2013–14 Elementary schools High schools 100,000 100,000 Revenue per school Revenue per school 80,000 80,000 60,000 60,000 40,000 40,000 20,000 20,000 0 0 In-kind In-kind In-kind Cash In-kind In-kind In-kind Cash construction other salaries construction salaries other Source: PETS-QSDS national survey – school level. Note: In-kind contributions refer to goods or services given to schools directly rather than as cash grants for the school to purchase the goods and services for themselves. Other in-kind contributions include school supplies, books, and instructional materials. 108 WWW.WORLDBANK.ORG/PH Few of these local government in-kind contributions relate these projects largely because of delays in implementation to the provision of teachers and other sta . In 2014, the and the poor quality of the completed projects.7 government decided to incorporate all local government funded teachers into the national teacher workforce funded Equit nd Loc l Government Support directly by the national government. When the study team elded their survey at the end of 2014, a signi cant number Local government education support is unevenly distributed of local government teachers had already been absorbed across regions and is highly inequitable. Comparing regional into the national network, and fewer than 2 percent of poverty rates and local government spending shows that, teachers in elementary and high schools were funded by on the whole, poorer regions spend less on supporting local governments. However, local governments continue basic education (Figure 6). For example, less than 10 percent to provide some support to schools to fund the provision of of the population is poor in Region IV-A in Luzon, and it ancillary sta such as janitors and security guards. spends an average of PHP 950 per basic education student. In contrast, over 35 percent of the population is poor in The construction and rehabilitation of school infrastructure Region XII in Visayas, and it spends only around PHP 370 per absorbs a much larger share of total LGU funding for high student. These large di erences are primarily the result of schools than of total funding for elementary schools. Around di erences in the amount of property tax revenue that local three-quarters of all LGU funding in high schools is devoted school boards are able to collect for their Special Education to construction and rehabilitation compared to only one-third Funds. These ndings are worrying since schools serving for elementary schools. Of all projects undertaken by local poorer children tend to need higher levels of funding in governments in 2013 and 2014, 68 percent involved the new order to compensate for the more limited support that construction of classrooms or water and sanitation facilities, disadvantaged children tend to receive outside the school. with the repair of existing infrastructure accounting for the Looking at the per-student distribution of local government remaining projects. Most of the projects were completed on funding within regions reveals a di erent picture. Schools time, but satisfaction rates were relatively low with 54 percent located in rural municipalities tend to receive signi cantly of school principals saying that they were dissatis ed with more per-student funding from LGUs than schools in urban Figure 6: The Distribution of Local Government Education Spending is Inequitable Total and per student LGU spending on basic education by region and poverty incidence, 2012–13 Total LGU spending (PHP billions) LGU spending per student (PHP) 8 4,000 LGU spending per student, 2013 3,500 Total LGU spending, 2013 6 3,000 2,500 4 2,000 1,500 2 1,000 500 0 0 0 10 20 30 40 0 10 20 30 40 Poverty incidence, 2012 Poverty incidence, 2012 Sources: Poverty incidence data from the Philippines Statistical Authority, LGU spending from the Bureau of Local Government Finance. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 109 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Figure 7: Patterns of Local Government Funding are Inequitable Average per-student LGU funding (PHP), 2013–14 Elementary schools High schools 400 400 300 300 200 200 100 100 0 0 Poorest 20% of students Poorest 20% of students Wealthiest 20% of students Wealthiest 20% of students Source: PETS-QSDS national survey – school and student household levels. areas. For example, the average LGU per-student contribution to high schools in highly urbanized areas is PHP 245 compared Tr nsp renc nd with PHP 369 to high schools in municipalities. The key driver Account bilit of Loc l of this is the smaller size of rural schools.8 School Bo rds LGU contributions vary according to the socioeconomic status of students in very di erent ways in elementary schools and Information on overall levels of annual local government in high schools (Figure 7). The PETS-QSDS survey included education spending was often unavailable to the survey a nationally representative sample of households of public team at the local level, and what little data they were able elementary and high school students. Using information to nd were inconsistent. This clearly limits the ability of collected by the survey on consumption and asset ownership, stakeholders to hold local governments accountable. it is possible to rank these student households by their The PETS-QSDS study collected nancial data on the estimated levels of household consumption per capita.9 This Special Education Funds (SEFs) from the Bureau of Local ranking shows that students from poorer households tend to Government Finance (BLGF) at the national level and from attend elementary schools that receive less LGU per-student LSB o ces in cities, municipalities, and provinces. The funding than is received by the elementary schools attended team then attempted to match these two sets of data, by students from wealthier households. In contrast, poor high which revealed signi cant inconsistencies between the school students tend to attend high schools that have higher nationally and locally reported spending aggregates. The levels of per-student funding than is received by the high team found that the information reported at the national schools attended by their wealthier counterparts.10 Taken as and local levels was consistent in the case of only 16 a whole, the ndings suggest that local government funding percent of LSBs, while, for 22 percent of LSBs, there were tends to widen funding inequalities in elementary schools inconsistencies in the levels of spending reported at the and to narrow them in high schools. However, given that LGUs national and local levels and one-half of all LSBs did not provide relatively low levels of education funding, their impact report any SEF expenditure at the local level to compare on overall funding remains small. with national data. 110 WWW.WORLDBANK.ORG/PH The information on those schools that received local government support and on the kind of support that they Polic Directions for received also revealed discrepancies. The PETS-QSDS study Improvin Loc l Government tracked the spending that local governments had reported providing to elementary and high schools within the schools Support for B sic Educ tion themselves. Depending on the type of funding in question, local governments reported providing funding to a large This note has shown that, in recent times, local governments number of schools under their jurisdiction. However, when have failed to spend the revenue that they have collected in the team checked this information at the school level, a their Special Education Funds. Other studies have shown that signi cant proportion of schools reported that they had not existing levels of public education spending are inadequate received these funds. For example, 24 percent of elementary to deliver the quality of services outlined in existing national schools to which local governments claimed to have norms and standards. While local government funding provided in-kind support for salaries denied ever having constitutes a small percentage of overall education spending received this support (Figure 8). in the Philippines, it is imperative that all of these funds should be used to support education improvements. The The team found additional inconsistencies in reporting national government needs to strengthen its monitoring of between schools and local governments. For example, of local government SEF surpluses and to consider putting a all the elementary schools that reported having received cap on the level of these surpluses to ensure that the vast support from their local government, the LGUs themselves majority of these funds is spent on education. only reported providing support to approximately 40 percent of them. While reporting de ciencies prevented the study team from making a completely accurate assessment, it is clear that a These ndings demonstrate that local government reporting relatively low share of local government education spending and accounting procedures for the use of education funds reaches schools. The funds that do reach schools are largely are weak. This hampers the ability of schools, parents, and provided in-kind, which limits the control and exibility that other stakeholders to assess the fairness and e ectiveness of schools have to use the support according to their own priori- spending priorities and resource allocations and to hold local ties. Providing a greater proportion of local government fund- governments accountable. ing to schools in the form of cash, as is the case with the nation- Figure 8: Comparing Expenditures between LGUs and Schools Revealed Many Discrepancies Percentage of schools where LGUs reported providing some funding but schools reported receiving none Elementary schools High schools 30% 30% 20% 20% 10% 10% 0% 0% In-kind salary Other in-kind Cash Construction In-kind salary Other in-kind Cash Construction Source: PETS-QSDS national survey – provincial, city and municipality local school boards. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 111 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education al government’s funding for school maintenance, operating, decisions. Unless LSBs take into account school planning and other expenses (MOOE), would increase the autonomy of processes, there is a signi cant risk that local government schools and provide them with greater exibility and predict- funding will not help schools to meet their most pressing ability, all of which would increase the positive impact of local needs. Increasing coordination between LSBs and schools government funding on educational outcomes. themselves would mitigate this risk and provide LSBs with more information on which to base their allocations of Both the revenue base and the allocations of local support to basic education. Moreover, it would allow for government education funding are inequitable. This greater coordination between local governments and policy note has shown that the revenues available to local other sources of school funding, which would prevent any governments vary widely and are negatively correlated with potential overlaps and enable more e cient and equitable poverty rates and other markers of disadvantage. Addressing use of all sources of funding. this issue will be di cult, but the national government The outcomes of the funding decisions of LSBs are should consider adjusting its own funding allocations not transparent, which makes it di cult for education to compensate for the inequalities arising from these stakeholders to hold local governments accountable. di erences at the local government level. For example, the There are no consistent reporting formats to document the national funding formula for school maintenance, operating, amount of funds that LSBs allocate to schools, the amount and other expenses (MOOE) could be adjusted to account for that the schools receive, and the amount that the LSBs di erences in SEF revenues between localities.11 allocate to other education-related activities. Developing a simple reporting format, to be issued as part of a revision of This note has shown that local government spending tends to SEF guidelines by the national government, and the passing favor better-o schools, at least in terms of support for facilities. of regulations requiring that these reports are published on Therefore, it is vital that the allocation of local government LSB bulletin boards and school notice boards would also funding is improved to ensure a closer t between funding increase transparency. levels and needs at the school level. Introducing a simple and transparent funding formula, similar to the national funding A recent initiative by the Department of the Interior and Local formula for MOOE, could go a long way towards making the Government (DILG) to encourage greater accountability and allocation of local government funding more equitable. The transparency has been an important rst step. LGUs that guidelines for LSBs on using local government funds also perform well in several areas receive the Seal of Good Local need to be consistent with national guidelines to ensure that Governance as well as additional performance-based funds, there is no duplication and that local government spending and this scheme has provided LGUs with a strong incentive complements national government funding. to increase transparency and to improve the use of SEF funds (Box 2). While the awarding of the Seal is already conditional on The ndings of the study also show that schools and school whether the annual plan of the LSB is aligned with the plans principals are rarely involved in local government funding of the schools themselves, it could also be used to address Box 2: The Seal of Good Local Governance Building on earlier attempts to promote greater local government transparency and accountability in the use of public funds, the national government introduced the Seal of Good Local Governance in 2014. In order to receive the Seal, local governments must demonstrate good performance in three core areas (good nancial housekeeping, disaster preparedness, and social protection) and in at least one essential area (business friendliness and competitiveness, peace and order, and environmental management). Support for basic education falls under social protection. In order for a local government to receive the seal, it must demon- strate that it has completed 70 percent of the local school board’s annual plan and that the plan is aligned with schools’ own improvement programs. The DepEd superintendent currently certi es the achievement of these two conditions. Recipients of the Seal are also eligible for the Performance Challenge Fund (PCF), which provides additional funding for local governments to pursue their own priorities. In 2015, 41 of the 80 provinces, 28 of the 143 cities, and 170 of the 1,491 municipalities were awarded the Seal. Source: Department of the Interior and Local Government (2014). “Seal of Good Local Governance.” Circular No. 2014-39. Manila. 112 WWW.WORLDBANK.ORG/PH some other issues highlighted in this note. For example, the of the existing joint circular on the SEFs, and a set of clear category of “good nancial housekeeping” could be expanded operational guidelines. These guidelines will need to clearly to include more detailed reporting and the publication of how articulate priorities for the use of SEF funds, to devise ways to SEF funds are used and to set targets for the proportion of record the outcomes of school-level improvement plans, and funds that should reach schools. to document clearly any elements of an LSB school funding formula to allocate resources. Moreover, the guidelines should Any changes in the way in which local governments use their stress the need for local governments to devote the majority SEFs will require agreement between DepEd, the Department of their funding to school-level activities and should include of Management and Budget (DBM), and the DILG, an updating indicators to measure this. Table 1: Improving Local Government Support for Basic Education Findings Policy suggestions A low share of local government • Raise the share of local government funding provided directly to schools funding reaches schools and • Introduce a formula-based funding model to allocate more cash directly to schools most is provided in-kind to support their annual improvement plans Local government funding is • Reduce inequalities between local school board areas by adjusting the national inequitable funding formula • Introduce a simple and transparent local government funding formula to strengthen the link between funding and school needs Schools are rarely involved in • Make better use of schools’ annual improvement and procurement plans in LSB funding decisions planning • Ensure that the link to school improvement planning in the Seal of Good Local Governance is properly measured and monitored Transparency and accountability • Improve reporting formats for the use of SEF funds and ensure that the results are for the use of funds is weak publicly disseminated • Build on the Seal of Good Local Governance and associated performance-based funding by, for example, requiring LSBs to report planned school level funding allocations. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 113 Policy Note 6: Assessing the Role Played by Local Government in Supporting Basic Education Endnotes 1 Barangays (or villages) are the smallest administrative unit in 8 Average school-level funding is indeed higher in urban areas. local government in the Philippines. Above them come the For example, the average high school in a highly urbanized municipalities and above municipalities come provinces/cities. city receives PHP 458,000 compared with the average of 2 In 2014, SEFs were the source of approximately 85 percent of approximately PHP 107,000 received by high schools located in municipality spending and 70 percent of province and city municipalities. spending on basic education. 9 The household questionnaire included a short module on 3 Manasan, R. G., A.B. Celestino, and J.S. Cuenca (2011). “Mobilizing consumption and a set of questions on assets that have been LGU Support for Basic Education: Focus on the Special Education used by the Department of Social Welfare and Development Fund,” Philippine Institute for Development Studies, Manila. (DSWD) to undertake a proxy means testing (PMT) approach to estimating household consumption per capita. The results 4 Within cities, the LSB allocates all of the funds within the SEF, reported here are based on information gathered using the PMT whereas municipal LSBs are required to give 50 percent of approach, and a full description is included in a separate note. their SEF revenue to their provincial LSB to provide resources for provincial spending. Some local governments also allocate 10 These ndings are supported by the study team’s analysis of money from their own General Fund (GF) and other sources to school facilities. Elementary schools receiving LGU support provide additional support to education. tend to have better school facilities (such as IT equipment, playgrounds, clinics, and libraries) than those that receive no 5 A full description of the information collected in the PETS-QSDS funding. The opposite is true for high schools. study and the approach used to calculate total school revenue and expenditure is included in a set of additional annexes and 11 A fuller discussion of national government funding of schools tables accompanying the main PETS-QSDS report. through the maintenance, operating, and other expenses (MOOE) budget is provided in a separate note- Providing Schools with 6 The e ectiveness of school-level planning and implementation Adequate Operating Expenses to Deliver Quality Education in the mechanisms is discussed in more detail in a separate note - Philippines. Assessing School-based Management in the Philippines. 7 The quality of school infrastructure and of national and local government projects is discussed in more detail in a separate note - Building Better Learning Environments in the Philippines. 114 WWW.WORLDBANK.ORG/PH Polic Note 7: Underst ndin the Drivers of Public School Perform nce nd Efficienc Introduction After a long period of stagnation, public spending on education in the Philippines has increased signi cantly. Between 2005 and 2015, education spending more than doubled in real terms, with much of it being devoted to increasing the number of teachers and improving school infrastructure. As a result, student-teacher and student- classroom ratios have improved signi cantly. For example, the average student-teacher ratio in high schools fell from 40:1 in 2005 to 27:1 in 2014. Despite increased spending and improvements in key input ratios, improvements in learning outcomes have been relatively modest so far. Studies have shown this to be the case in many other countries as well. Clearly spending is only one of a host of factors that can a ect the ability of schools to improve the academic outcomes of their students. Some of these factors fall within the control of the education system and schools such as the provision and quality of education inputs and the e ectiveness of school-based management. Others such as the characteristics of students and their households are beyond the control of schools, but all of these factors a ect the e ciency of resource use and education outcomes. school infrastructure. It also found that student and school This note attempts to identify the factors associated with characteristics, school-level accountability mechanisms, di erences in overall school performance and e ciency. and some teacher characteristics are associated with how It draws on data from the Philippines Public Education e ciently schools can convert their nancial, physical, and Expenditure Tracking and Quantitative Service Delivery Study human resources into improved education outcomes. (PETS-QSDS), which tracked public education spending and However, given the limitations of using cross-sectional data, assessed the quality of the systems governing the use of the ndings are not strong, and factors associated with good these funds. The note also draws on the ndings from other performance tend to vary a lot according to the level of policy notes in this series to identify the key factors that education and the performance measures that were being determine the links between public spending and education analyzed. While more rigorous research is needed, the policy outcomes. note provides tentative evidence that the factors often The note found that school performance is related to several associated with better performance are not functioning of these key factors including school governance and e ectively in the Philippines. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 115 Policy Note 7: Understanding the Drivers of Public School Performance and Efficiency Appro ch to Assessin as part of the PETS-QSDS survey directly from student atten- dance registers. The speci c measure used in this note was School Perform nce average attendance in all grades on four speci c dates in the last six months of 2014.3 The school averages showed that This policy note used the average National Achievement Test student attendance is generally high in the Philippines with (NAT) score for each school as a measure of its performance.1 relatively little variation between rural and urban schools Speci cally, it used a school aggregate of the di erent (Table 1). subject tests that individual students took in 2014 at the end of elementary school (Grade 6) and high school (Grade 10). In order to develop a framework for understanding the The NAT scores of municipal schools tended to be higher factors a ecting the performance of individual schools, the than those of schools in urban areas (Table 1). For example, study team consulted several international studies.4 These the average NAT score for Grade 6 students in municipalities studies broadly looked at three main categories of factors was 75 percent compared with 66 percent in highly associated with school e ectiveness and good education urbanized cities.2 outcomes (Figure 1). Student attendance was used as an additional measure of 1. Household and Student Characteristics. The early school performance. The data on attendance were collected childhood care received by children, their development Table 1: School National Achievement Test Results and Student Attendance Rates, 2014 Elementary Schools High Schools High High Urbanized Other All Urbanized Other All Cities Cities Municipalities schools Cities Cities Municipalities schools National achievement score (%) Average (mean) 66.2 70.6 75.4 74.3 54.5 57.7 57.6 57.3 Variation (standard 11.9 13.8 13.6 13.7 7.7 8.4 12.6 11.7 deviation) Minimum 38.8 36.7 32.1 32.1 41.5 41.7 34.8 34.8 Maximum 83.4 91.5 94.2 94.2 72.6 80.8 85.9 85.9 Student Attendance (%) Average (mean) 87.5 93.8 88.4 89.0 88.5 92.9 89.3 89.8 Variation (standard 14.1 4.4 12.7 12.2 9.4 3.3 10.3 9.6 deviation) Minimum 19.7 78.1 34.8 19.7 39.6 81.4 56.6 39.6 Maximum 99.0 100.0 99.6 100.0 99.6 99.4 99.6 99.6 Source: PETS-QSDS national survey – school level. NAT scores of PETS-QSDS sampled schools – National Education Testing and Research Center of DepEd. Notes: Attendance rates were calculated using the average of student attendance on four speci c days in the second half of 2014. NAT scores were calculated as the averages of all subjects for Grade 6 (elementary) and Grade 10 (high school) students in PETS-QSDS sampled schools who took the national examination in 2014. 116 WWW.WORLDBANK.ORG/PH Figure 1: A Framework for Understanding School Performance Supporting School School Inputs Factors Performance • Parent and • School • Average school community support characteristics results on the National • Education system • School climate Achievement Test (NAT) support • Average student • Enabling conditions • Material support • Teaching and attendance rates learning process Household and Student Characteristics history, and the investments made by their household in The PETS-QSDS study team compiled a large number of vari- their education in uence the performance of the schools ables from the PETS-QSDS survey data for each of the catego- that they attend. The level and type of nancial and ries in the framework. They then reduced this list to a smaller academic support that children receive outside of school group that they judged to be the best measures of each depends crucially on the socioeconomic status of their particular factor and that had been identi ed as important in parents. the other policy notes in the PETS-QSDS series (Table 2). 2. School Factors. How the school itself is organized to This note explores the association between these important provide learning opportunities for its students critically explanatory factors and school performance in a number of a ects their academic achievement. This includes the ways. First, the note compares di erences in these factors motivation, attitudes, and competency of teachers, between high-performing and low-performing schools. the quality of the teaching and learning process, the Schools with scores in the top 20 percent on the NAT were amount of learning time that the school provides, and classi ed as high performers whereas schools with scores the systems used for student assessment. The quality of in the bottom 20 percent were classi ed as low performers. school leadership is another important school-level factor A similar de nition was used to rank school performance in that a ects the school’s performance. terms of student attendance rates. Second, the note explores 3. Supporting Inputs. In order for schools to perform well, the extent to which the factors in Table 2 explain di erences they need the appropriate inputs and resources to in e ciency.5 Finally, it looks directly at the relationship operate e ectively. This includes su cient classrooms between the explanatory factors and school performance of good quality to teach children in groups that are not using a multivariate regression approach.6 too large and the availability of other education inputs such as textbooks and discretionary funds to support While this exercise highlighted some important associations, school improvement plans. They also need support the approach had a number of limitations. Given the cross- from parents, the community, and, in the case of the sectional nature of the data and the lack of any strategy for Philippines, local school boards (LSBs). Schools in the addressing selection bias, it was not possible to establish Philippines also depend on the DepEd division and causality between the explanatory factors and the selected district o ces for support on administration, nancing performance measures. School performance is also not and teaching. only determined by the current status of household ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 117 Policy Note 7: Understanding the Drivers of Public School Performance and Efficiency Table 2: Characteristics Used to Measure the Key Aspects of the Performance Framework Household Characteristics School Factors Supporting Inputs Average per capita Material support: Revenue per student School characteristics: the school’s location, household consumption by source, MOOE funds received per enrollment rates, school feeding program, student, textbooks per student, adequacy the principal’s years of experience, time spent of classrooms, days of teacher training, by the principal in the classroom observing infrastructure and condition of classrooms. classes, the school’s SBM self-assessment level, whether the principal develops a school Parent and community support: Frequency plan for professional development. of meetings of the school governing council and the PTA, parental participation in school Capable teaching force: teacher decision-making, transparency board competency test scores, teacher attendance, available in a public space. teacher quali cations. E ective support from education system: Teacher learning process: the percentage frequency of supervision visits from the of parents who receive a report card for their DepEd Division o ce. child’s performance, proportion of students that report doing homework. characteristics, school factors, and supporting inputs but also by how those factors have evolved over time. It is unlikely Ch r cteristics of that the current state of the education system is a good School Perform nce re ection of its past, given the large amounts of money that have recently been invested. The lack of any information Household F ctors on the status of key indicators in earlier years will a ect the Students who attend higher performing schools tended overall results. Also, the analysis focused on performance to be from slightly better-o families than those who at the school level, which did not take into account any attended low-performing schools, but these di erences possible variations between di erent classes and teachers were not large and were rarely statistically signi cant. within the schools. These issues were compounded by the For example, annual per capita household consumption small sample of teachers and households for each school was approximately PHP 16,300 for students in those that was used in the study, which may have introduced some elementary schools ranked in the top 20 percent for measurement error. In addition, while the PETS-QSDS survey student attendance compared with PHP 14,700 for collected information on many important areas related to students in schools ranked in the bottom 20 percent. school performance, there were some important gaps. For Schools that had better average NAT scores also tended to example, very little information was available on the amount include students who were slightly wealthier than average, of real learning time available in schools or on the quality of but again the di erences were relatively small and not classroom teaching. This may have led to biased estimates statistically signi cant. of the importance of the measured factors where they were correlated with omitted variables. It is important to keep these limitations in mind when interpreting the results presented in this note. 118 WWW.WORLDBANK.ORG/PH School F ctors associated with school principals and the processes for which they are responsible tended to be better for high-performing It is clear from looking at di erences between high- schools (Table 3). High-performing schools tended to have performing and low-performing schools that schools in principals who spent more time in classrooms and who urbanized areas tended to perform less well than those in were more likely to have developed teacher professional municipalities. For example, a higher proportion of poorly development plans. However, these di erences were not performing elementary and high schools were located in generally statistically signi cant. highly urbanized cities than in municipalities, whereas good performing schools tended to be located more frequently Simple comparisons of high-performing and low-performing in municipalities (Figure 2).8 Related to this nding, schools schools did not reveal any statistically signi cant di erences with higher enrollments also tended to have lower levels of in the capacity of the teaching force.10 The PETS-QSDS study student achievement than smaller schools.9 For example, the tested a sample of Grade 6 and Grade 10 teachers in all high schools ranked in the top 20 percent in terms of NAT sampled schools. However, there was no clear or consistent scores had an average of 446 students compared with an pattern in the relationship between the performance ranking average of 833 students in schools in the bottom 20 percent. of the sampled teachers in a school and the performance ranking of the school itself. More elementary schools with high student attendance rates had school feeding programs than schools with the Supportin Inputs lowest rates of student attendance. For example, 25 percent of elementary schools in the bottom quintile of student Simple comparisons between high-performing and low attendance had a school feeding program compared with 71 performing schools yielded some di erences in terms percent of schools in the top quintile of student attendance. of the quality of school infrastructure. Only 6 percent of This suggests that school feeding programs may provide elementary schools in the highest quintile in terms of NAT families with an incentive to send their children to school. scores operated multiple shifts compared to 9 percent of schools in the bottom quintile. However, a similar pattern The quality of school leadership also appears to be was not observed for high schools, which may re ect the associated with better school performance. Indicators di erences in shift patterns between elementary and high Figure 2: Lower Performing Schools Tend to be Bigger and Are More Likely to Be Located in Highly Urbanized Cities Average student enrollment and location for schools ranked in the bottom and top 20 percent of NAT scores, 2014 Percentage of schools in HUCs Total school enrollment 15% 1,000 750 10% 500 5% 250 0% 0 Elementary School High School Elementary School High School Low performance (NAT) High performance (NAT) Source: PETS-QSDS national survey – school-level data. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 119 Policy Note 7: Understanding the Drivers of Public School Performance and Efficiency Table 3: School Performance and Leadership Indicators Elementary Schools High Schools Attendance NAT scores Attendance NAT scores Low High Low High Low High Low High perf. perf. perf. perf. perf. perf. perf. perf. Number of hours principal 2.6 7.1 2.9 4.8 3.7 5.5 3.8 5.8 observes classes Principal’s number of years 2.6 3.3 3.7 3.0 2.7 3.6 2.9 2.7 of experience School plan for teacher 54.0 97.3 61.3 84.5 53.5 67.8 72.3 82.5 professional development exists SBM self-assessment level 1.4 1.6 1.5 1.4 1.5 1.8 1.6 1.7 Source: PETS-QSDS national survey – school level. Notes: Hours of observation are over a typical school week. School based management self-assessment is based on a three point scale (Level 1 = developing, Level 2 = maturing, and Level 3 = advanced). schools. For example, it is more common in elementary and urban areas, which also had large di erences in student- schools to operate independent shifts, which may reduce teacher ratios. School attendance tended to be better in the amount of instructional time received by each municipal schools, which also tended to have lower student student. Subject-based teaching is implemented in high numbers and lower student-teacher ratios. schools, so shifting is more likely to be associated with the organization of timetables and to have less impact on each There is some evidence that better performing schools had student’s learning time. greater support from the local community and from the education system generally. Participation rates in parent- The availability of key facilities such as electricity and teacher associations (PTAs) were much higher in better water supply and the quality of classrooms (their state performing schools (Figure 3). A similar pattern was also of repair) also tended to be positively related to school found between school performance and the frequency of performance. Thus, schools with better facilities tended its school governing council meetings. However, di erences to have better attendance and NAT scores, although this between high-performing and low-performing schools in was signi cant only for elementary schools and in the terms of SGC meetings were rarely statistically signi cant. case of attendance for high schools. Whether DepEd division o ces provide schools with their full allocations for maintenance and other operating expenses Student-teacher ratios also tended to be lower in schools (MOOE) is a good measure of the amount of funds over which that had better NAT scores although the di erences were schools have discretion to spend on their own improvement not very large. High schools that ranked in the top quintile plans and a good indicator of the support that division of NAT scores had an average student-teacher ratio of 24:1 o ces provide to schools. If division o ces do not ful ll their compared with an average of 26:1 in schools in the bottom obligation to download all of a school’s allocation of MOOE quintile of NAT scores. In contrast, schools with low student funds, then it may not be providing much other support. attendance tended to have lower student-teacher ratios than Simple comparisons did show that better performing schools schools with better student attendance. This may have been were more likely to be located in divisions that downloaded a function, in part, of di erences in attendance between rural their full MOOE allocations (Figure 3). 120 WWW.WORLDBANK.ORG/PH Figure 3: Schools with Higher NAT Scores Have More Parental Participation and Receive More of their Funding from DepEd Division Offices Average proportion of parents who are members of the PTA and percentage of schools that do not receive all of their MOOE funds from the DepEd division o ces Elementary schools High schools 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Percentage of Division retains a Percentage of Division retains a parents with PTA proportion of parents with PTA proportion of membership MOOE membership MOOE Low performance (NAT) High performance (NAT) Source: PETS-QSDS national survey – school-level data. There did not appear to be a consistent pattern between Some schools appeared to use their resources more schools’ levels of revenue and their performance. As part of e ectively than others to improve student learning the framework outlined in the previous section, the team outcomes. In Figure 4, each panel is divided into quadrants used a number of di erent measures of annual school that show where schools fell in relation to the average revenue to explore di erences in school performance. achievement score and average annual school spending They found no consistent association between revenue per student. Schools in quadrant A were the most e cient and performance. In some cases, schools with lower than as they spent less than the average school every year but average revenues per student had better than average had better than average outcomes. The schools in quadrant performance indicators. These results point to the weak B were the least e cient, with levels of spending that were relationship between spending and school performance higher than average but with below average performance. and the likelihood that other factors (such as how well funds Levels of e ciency varied greatly between schools that had are managed at the school level) in uence the e ciency of similar levels of either performance or spending. school spending. The team’s estimates of school e ciency suggest that Efficienc of School Spendin education outcomes could be signi cantly improved without the need for any additional funding. Using data envelopment analysis (DEA), the team estimated e ciency scores for The previous section showed that schools vary considerably school performance (as measured by attendance and NAT in their ability to translate their revenues into positive scores) to measure the relative e ectiveness of schools in education outcomes. The PETS-QSDS study collected transforming their annual per-student spending into better detailed information on all sources of school revenue from education outcomes.12 The results revealed considerable a nationally representative sample of elementary and high variation among schools in their ability to translate resources schools. On the whole, the higher the annual level of school into better education outcomes. The estimated e ciency funding, the better the school’s performance, but this scores can be interpreted as the percentage increase relationship was not very strong (Figure 4). in output that a school could achieve with its current ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 121 Policy Note 7: Understanding the Drivers of Public School Performance and Efficiency Figure 4: There Are Large Differences in Performance Even Among Schools with Similar Levels of Funding National achievement test scores and school annual revenue per student, 2013–14 Elementary schools High schools 100% 100% A A 90% 90% 80% 80% Grade 10 NAT scores Grade 6 NAT scores 70% 70% 60% 60% 50% 50% 40% 40% B B 30% 30% 0 5,000 10,000 15,000 20,000 25,000 0 5,000 10,000 15,000 20,000 25,000 School revenue per student (PHP) School revenue per student (PHP) Source: PETS-QSDS national survey – school-level data. Note: The horizontal line in the gure represents the average school NAT score while the vertical line marks the average school revenue per student. resources. The estimates revealed that school e ciency was 77 percent. This suggests that Grade 6 NAT scores could varied considerably and there is considerable potential for be improved by around 23 percent if all schools were able to improving education outcomes if schools were able to use use their resources as e ciently as the most e cient schools their resources more e ectively (Figure 5). For example, the in the system (Figure 5). While not too much should be read average e ciency score for elementary school NAT results into the exact magnitude of these estimated improvements, Figure 5: Increasing School Efficiency Can Improve Education Outcomes Average estimated e ciency scores and standard deviations for school attendance and NAT results in elementary and high schools Elementary schools High schools 100% 100% 75% 75% 50% 50% 25% 25% 0% 0% NAT e ciency Attendance NAT e ciency Attendance score e ciency score score e ciency score Mean Standard deviation Source: PETS-QSDS national survey – school-level data. Note: The e ciency score shows the average e ciency of schools relative to the most e cient schools in the system. E ciency scores were estimated using data envelopment analysis. School attendance and NAT scores were the output measures used, and the input measure was total annual school revenue per student. 122 WWW.WORLDBANK.ORG/PH this exercise does demonstrate the potential bene ts that e ciency in high schools as measured by attendance, might accrue from more e ective resource use. appeared to actually reduce school e ciency in some cases. Schools’ levels of e ciency are a ected by some factors that Di erences in the sources of school funding did not appear they can in uence and some that they cannot, such as the to be strongly associated with school e ciency. It might characteristics of the communities and households to which have been expected that schools with more discretionary their students belong. The team used regression analysis to funding would be more e cient because they were explore the associations between the estimated e ciency able to use these funds to more e ectively address their scores and the explanatory factors in the framework outlined needs. However, the results did not show this. Frequently, in Figure 1.13 The results of this second stage analysis the relationship between the amount of discretionary highlighted some of the same explanatory factors that were funds and the source of those funds was not statistically revealed in the comparison of high-performing and low- signi cant. And in some cases, the size of discretionary performing schools in the previous section. funds received by a school appeared to be associated with lower levels of school e ciency. Schools in urban areas tended to be less e cient than rural schools in using their resources to deliver better NAT scores. There is some evidence that the availability of key school For example, elementary schools located in cities were less facilities and the condition of the existing stock of e cient than schools in municipalities or in highly urbanized classrooms a ected e ciency. For example, better learning cities. Larger schools, and particularly high schools, also environments, measured by an index of classroom conditions tended to be less e cient even after controlling for school such as the state of repair, were positively associated location. This is likely to re ect the di culties involved in the with school e ciency. However, this association was only management and organization of larger schools. However, the statistically signi cant for elementary school NAT scores and di erences in e ciency between schools in di erent locations high school attendance rates. and of di erent sizes were not always statistically signi cant. Factors associated with greater parental and community While school leadership factors were associated with greater participation tended to be related to better e ciency, but e ciency, the results were not always statistically signi cant. the results were rarely statistically signi cant. More frequent Similar to the high and low performance comparisons, schools with more experienced principals who spent PTA and SGC meetings and greater parental participation more time observing teaching and who developed plans were associated with greater e ciency in terms of NAT for teacher professional development tended to be more scores. However, only the frequency of SGC meetings e cient. However, these factors were rarely statistically in elementary schools was statistically signi cant. The signi cant on their own. frequency of PTA meetings and the proportion of parents participating in these meetings were associated with lower Teacher characteristics also did not appear to be associated e ciency as measured by attendance. It is unclear what was with levels of school e ciency. A number of factors such as driving these results. quali cations and test results were included to understand whether teacher characteristics were associated with Schools located in divisions where the DepEd o ces e ciency.14 However, these factors did not appear to play provided schools with their full MOOE allocations tended much of a role in determining the e ciency of either to be more e cient. To the extent that this factor was a elementary or high schools in terms of either NAT scores proxy for more e ective and transparent division-level or attendance. Meanwhile, better teacher performance management, it highlights the important role played by on subject matter tests, while statistically signi cant for education system governance in increasing school e ciency. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 123 Policy Note 7: Understanding the Drivers of Public School Performance and Efficiency Determin nts of School schools and their local communities, particularly through the institutions (such as the SGCs) and processes (such as the Perform nce school improvement plans) related to the implementation of school-based management could be bene cial. The team tried taking a third approach to analyzing school The note has shown that school leadership may be a key performance by extending the simple bivariate approach explanatory factor for di erences in performance among reported in Section 2 to some simple multivariate regression schools. Schools where the principals observed classroom analysis. School performance in terms of attendance and NAT teaching regularly and developed professional development scores was regressed against the set of explanatory variables plans for their teaching force were often more e cient and outlined in Table 2. The results of this nal approach were had better performance indicators than schools where this very similar to the e ciency score results and highlighted did not happen. similar associations between explanatory factors, levels of school attendance, and national examination scores.15 There is also some tentative evidence to suggest that schools within the jurisdiction of the better managed DepEd Conclusion divisions tended to perform better themselves. Division o ces are obligated by DepEd regulations to transfer all This policy note has shown that many schools in the MOOE funds to schools, and it can be assumed that those Philippines are not using the resources at their disposal in that do not do so are less well managed. The evidence the most e cient way to improve their performance. The reported in this note has shown that schools that fall analysis has found that many schools could improve their within the jurisdiction of those divisions also tended to be performance quite substantially by using their existing level ine cient and to perform poorly compared to schools in of resources more e ectively. better managed divisions. Understanding the factors that underlie existing levels of It is surprising that teacher characteristics were not e ciency using the information from the PETS-QSDS survey associated consistently with overall school performance. The was more challenging. The cross-sectoral nature of the PETS-QSDS study included a detailed assessment of teacher’s data and the relatively small sample sizes made it di cult subject knowledge, and the results were not associated to identify the main factors that could drive improvements strongly with school performance. It is possible that this was in e ciency. And while the study team made considerable because only a small proportion of each school’s teachers e orts to collect information on the broadest possible set took the test or it may be that levels of teacher competency of explanatory factors, there were still signi cant gaps. In are universally low in the Philippines and do not vary particular, the lack of any information on the teaching and signi cantly across schools. learning process within classrooms may have a ected the Finally, the note’s ndings suggest that larger schools in validity of the results. urban areas are ine cient and perform less well than smaller Notwithstanding these limitations, the note does provide schools in municipalities. Other policy notes based on the some tentative evidence that participation by parents PETS-QSDS data have highlighted the fact that large schools and community members in school a airs can in uence in urban areas are under signi cant amounts of stress. The the e ciency with which schools use their resources and results presented here tend to con rm that these schools are overall school performance. The frequency of SGC meetings also less e cient. Schools in highly urbanized cities tended had a positive and sometimes statistically signi cant to have higher than average levels of funding but lower than association with school e ciency and performance in average levels of performance. Given that the socioeconomic the multivariate analysis. However, factors associated with status of children in these schools did not appear to be greater participation and support from PTAs tended to be signi cantly di erent from the status of those in other public negatively related to e ciency in school outcomes. Further schools, this suggests that the ine ciency stemmed from research is needed to understand the factors underlying the challenges associated with managing very large schools these results but strengthening the relationships between and the very intensive use of limited school infrastructure. 124 WWW.WORLDBANK.ORG/PH Endnotes 1 A detailed description of the methodology and the full 8 The proportion of highly urbanized schools in the poor results are available in World Bank (2016). “Understanding the performing quintile was much higher than the proportion of Drivers of School E ciency/Performance in the Philippines.” highly urbanized schools in the population as a whole. Washington, D.C. 9 Di erences were statistically signi cant at the 1 percent level. 2 Highly urbanized cities are cities with populations of more than 10 Di erences in teacher attendance were also explored but no 200,000 and with average revenues of at least PHP 50 million in statistically signi cant di erences were found between high and 1991 prices. Other cities are de ned as cities that do not meet low performing schools. the criteria to be classi ed as highly urbanized. Municipalities are administrative units for all other areas in the Philippines. 11 A full description of the information collected and the approach that the PETS-QSDS study took to calculating total school 3 Other performance indicators such as a school’s dropout and revenue and expenditure is included in additional annexes and completion rate were also analyzed, but these school-reported tables accompanying the main PETS-QSDS report. rates did not seem to be very accurate. 12 E ciency scores were estimated for each of the performance 4 See for example, Glewwe, P.W., E. A. Hanushek, S.D. Humpage, measures. An output-oriented approach that assumes variable and R. Ravina (2011). “School Resources and Educational returns to scale was used to estimate e ciency scores. Full details Outcomes in Developing Countries: A Review of the Literature of the approach and results are available in World Bank (2016) from 1990 to 2010” NBER Working Paper No. 17554, National “Understanding the Drivers of School E ciency/Performance in Bureau of Economic Research, Cambridge, MA and Heneveld, the Philippines.” Washington, D.C. W. and H. Craig (1996) “Schools Count: World Bank Project Designs and the Quality of African Primary Education.” 13 E ciency scores cannot exceed 100 percent so a Tobit model World Bank, Washington, D.C. was used to estimate associations between e ciency scores and explanatory factors. 5 Data envelopment analysis was used to explore school e ciency. See, for example, Coelli, T. J., Rao, D. S. P., O’Donnell, C. J., and 14 Teacher attendance did not have a signi cant impact on Battese, G. E. (2005) An Introduction to E ciency and Productivity e ciency scores. In the nal analysis teacher attendance was Analysis, Springer Science & Business Media, New York and dropped because it allowed more school observations to be Herrera, S. and G. Pang (2005) “E ciency of Public Spending in used. Developing Countries: An E ciency Frontier Approach” Policy 15 The full results are not presented here but are included in Research Working Paper. No. 3645, World Bank, Washington, D.C. additional annexes and tables accompanying the main 6 A hierarchical modelling (HLM) approach was used when PETS-QSDS report. appropriate to account for the grouping of schools in the PETS-QSDS study under DepEd division and regional o ces. 7 Average elementary (high) school attendance for the bottom 20 percent of schools was 71 (75) percent and 98 (98) percent for the top performing 20 percent of schools. Average elementary (high) school national achievement test score for the bottom 20 percent of schools was 52 (42) percent and 90 (75) percent for the top performing 20 percent of schools. The full results are available in a set of additional tables accompanying the main PETS-QSDS report. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 125 Conclusions nd Polic Directions The study has shown that, despite a renewed focus on schools bene t from improvement projects or resources, basic education by the Philippines government, further the study has also shown that they are frequently unhappy increases in both capital and recurrent public spending about the quality of the resources supplied. are needed. The study has shown that there have been signi cant improvements in the learning environment In the absence of e orts to address these limitations, further that Filipino children experience every day but more still increases in funding are likely to have only limited impact on needs to be done. Many schools, particularly in urban areas, the challenges faced by the education sector. Moreover, the have insu cient and poor quality facilities and a shortage successful introduction of the senior high school program in of teachers. Operational funding still falls short of the 2016 will be jeopardized if systems managing the allocation amounts that schools need to pay bills, undertake basic and use of public funds are not made more e ective. repairs, and provide the day-to-day materials their students need. And there is rarely anything left over to fund school- Through a series of policy notes, this report has identi ed level initiatives to improve student learning achievement. some of the key challenges and potential policy responses The study has also highlighted the lack of good quality in key areas of resource management. Instead of repeating opportunities for teachers to improve their skills. Addressing these recommendations in full, this section draws together these challenges will require further investments in the some common policy directions that have emerged from education sector. the ndings of the study as a whole. It also appears that the bulk of this funding will need to be Improve Alloc tion Mech nisms provided by the national government. The study has shown that local government funding, except in the National Capital Throu h Better Pl nnin Region, is very small and that funding from other sources, The report makes clear that the systems currently used to including parent-teacher associations, is also negligible. allocate resources to schools could be strengthened by Detailed estimates suggest that over 90 percent of school improved planning. While these allocation systems generally funding comes from the national government. Unless there target schools with greater need, they could be improved are signi cant policy changes in the future, this means further. For example, school infrastructure improvement that increases in spending will need to be found from projects are planned on an annual basis and project lists are the national government budget, either through greater frequently nalized late. This shortens the time available to revenue-raising e orts and/or by increasing the share of implement them and compromises the ability of DepEd and basic education in the overall government budget from its DPWH o ces to monitor projects adequately. Developing existing level of 15 percent. a two- or three-year plan that would forecast need and list However, the study also highlights some limitations in all school improvement projects in each DepEd division the existing systems that manage public basic education would reduce the work involved in putting together the funds. Allocation systems have managed to target funds annual project list and accelerate the transmission of funds reasonably well, but there are still many examples of needy to the responsible implementing o ces. It would also give schools missing out on additional support while other, implementing o ces time to do some advance planning better-equipped schools continue to receive additional and thus avoid some of the di culties associated with support. Moreover, delays in identifying school needs and in inadequate funding and site availability that were commonly the budget allocation process have meant that a signi cant cited as reasons for implementation delays. A similar amount of funding goes unspent or has to be spent quickly approach could be adopted for teachers and the planning of or at times that disrupt school operations. Even when other important school inputs. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 127 Conclusions and Policy Directions More input from division o ces and schools in planning However, the study found that the existing system for could also improve allocation decisions. Funding decisions managing school operational funding is too onerous on are largely taken at the central level and frequently do not school principals and that there is a risk that the bene ts take into account local conditions. For example, in the PETS- from reforms of this kind will not materialize. Simplifying the QSDS study, a considerable number of school principals existing requirements for the management of these funds and school division superintendents (SDSs) felt that existing would reduce the signi cant burden that this currently procedures for allocating new teachers was suboptimal, puts on schools. It would also give school principals more and some SDSs felt that they had not been adequately time to focus on providing academic leadership in their consulted. Similarly, schools had only limited input into the schools, which is their primary responsibility. One approach planning and implementation of infrastructure projects. that has been adopted in other countries would be to This contributes to an impression that division o ces and treat operational funds as a grant. For example, this is schools have limited in uence on education decision- the case in Indonesia, where these funds have much less making. Strengthening the processes for consulting schools onerous reporting requirements than required by the usual and division o ces about the decisions that a ect them and government budget and implementation systems. Schools making information publicly available on the decisions that use simpli ed reporting templates to report on their use are made would improve planning. of the funds and submit these forms to district o ces (the equivalent of division o ces in the Philippines) every quarter. Give Schools Gre ter Authorit nd Developing a similar system adjusted to take account of the Simplif Reportin Requirements speci c context of the Philippines could reduce the burden that using these funds currently puts on schools. Increasing the role played by schools and local DepEd o ces in deciding on the use of education funding may However, if division o ces, schools, and even local also improve the quality of public education spending. The communities are to get more involved in decisions about study has highlighted the limited authority that schools have how education resources are allocated and used, these over the implementation of school infrastructure projects. spending decisions need to be more transparent. Across all It has also shown the weak links between the individual of the spending areas that the PETS-QSDS survey covered, professional development needs of teachers and the types information on the transfer and use of funds was frequently of in-service training o ered. Providing schools and division missing, inaccessible, or in formats that were not easy o ces with more authority during the implementation of to understand. Even where information was available, it both infrastructure projects and teacher training could yield frequently was not consistent between di erent administrative signi cant bene ts. For example, giving school principals levels. For example, there was a discrepancy of about 8 the authority alongside DepEd and DPWH o cials for percent between the funds that division o ces reported certifying that infrastructure projects have been satisfactorily having downloaded to schools for operational expenses completed might improve their quality. School principals in and the amounts that the schools reported receiving. In the Philippines have taken on roles of this kind in the past and most cases, it was not possible for the study team to judge with appropriate training could become a major driving force accurately the extent to which these discrepancies arose from in improving the quality of school infrastructure projects. misappropriation of funding and/or incomplete reporting. Over recent years, schools have been given ever increasing Improve Tr nsp renc of Fund amounts of discretion over how to spend their funding. Alloc tion nd Resource Use This has primarily been done to support school-based management reforms by providing schools with the funds to Local government funding was an area that was particularly implement their own school improvement plans. Evidence concerning. The study estimated that only around 60 from the Philippines and other countries supports the idea percent of local government funding reached schools that giving schools more autonomy can improve education directly. However, records on which schools received funds outcomes. and what the funds were provided for were very weak. For 128 WWW.WORLDBANK.ORG/PH example, 24 percent of elementary schools to which local would also enable schools and other stakeholders to seek governments claimed to have provided in-kind support redress for unfair allocations or poor project implementation. for salaries denied ever having received this support. In Policymakers should explore the feasibility of introducing a contrast, LGUs themselves reported providing support formal grievance system that schools could use to express to only approximately 40 percent of all of the elementary dissatisfaction with decisions related to all school funding, schools that reported having received support from their including school infrastructure projects, as this would help to LGU. Information on the 40 percent of funds that did not go make government education spending more e ective and directly to schools was even less reliable. improve education services. Developing simple reporting formats in which to record Stren then the Role of School the funds allocated to divisions and schools and how Governin Councils nd P rent these institutions use these funds would go a long way to Te cher Associ tions increasing transparency. A recent initiative by the Department of the Interior and Local Government (DILG) to encourage Parents and other community members can also play a greater accountability and transparency is one useful example role in improving school decision-making and overseeing of what could be done. The DILG awards the Seal of Good how schools use public funds and resources. However, the Local Governance to local governments that perform well study found that, while SGCs had been established in nearly in several areas as well as providing them with additional all schools, their role was fairly limited. Giving SGCs greater performance-based funds. This scheme has given LGUs a responsibility for monitoring how schools use public funds strong incentive to increase transparency and to improve could usefully complement DepEd’s own oversight role. their use of the Special Education Funds. While the awarding For example, SGCs could become jointly responsible with of the Seal is already conditional on whether the annual the DepEd division o ces for signing o on schools’ fund plan of local school boards is aligned with the plans of the utilization reports. However, if the SGCs are to take on an schools themselves, it could also be used to encourage expanded role, then their members will need to receive better reporting. For example, the category of “good nancial more training and support from DepEd. Schools also need to housekeeping” could be expanded to include more detailed encourage greater participation from parents and others in reporting, a requirement to publish details of how SEF funds the SGC to strengthen their planning and oversight activities are used, and a requirement to set targets for the proportion as the PETS-QSDS study found that most parents of students of funds that should reach schools. were unaware of the existence of SGCs or of their school’s improvement plan. Introducing and widely disseminating a set of standards for elementary and high schools would also increase transparency A campaign to increase the awareness of parents and and encourage greater accountability for the use of public other education stakeholders of the role of SGCs and funds. DepEd has a number of di erent guidelines that school improvement plans is needed. Recent evaluations set standards for elementary and high schools, which the in Indonesia and Pakistan have shown that it is possible to PETS-QSDS study used to assess the adequacy of existing use old and new technologies to disseminate information. resource allocations. However, these guidelines are scattered In Indonesia, an impact evaluation study showed that across di erent DepEd orders and regulations, and many well-designed information campaigns using simple SMS stakeholders are unaware of their existence. text messages or school meetings can signi cantly increase public knowledge about schools’ funding levels and Setting basic standards for basic education provision responsibilities. It also found that this increased knowledge and ensuring that information about these standards is led to higher levels of parental participation in school widely disseminated would enable parents and other a airs. In Pakistan, a similar strategy was used in a successful stakeholders to judge the fairness of allocations and to pilot program to disseminate information about school hold the government more accountable for the provision councils. The pilot set up a call center and used inbound of key education inputs like classrooms and teachers. It and outbound calls, robot calls, and SMS text messages to ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 129 Conclusions and Policy Directions transmit important information about the role played by other countries, like the United Kingdom, have introduced school councils and their members. An assessment of the funding components of this kind. The allocations of MOOE pilot showed that school council members’ knowledge funds could also be adjusted to take into account price about their roles and responsibilities had increased, while di erences between regions so that schools operating in school principals reported the councils had increased their high-cost areas such as remote locations where travel and participation in school a airs. transportation costs are high are able to purchase similar amounts of goods and services as those in lower-cost Address Fundin nd Qu lit locations. Inequ lities Further Rese rch The study has shown that there are substantial di erences in the quality of education services across the Philippines. The The study has highlighted a number of areas where current factors associated with the distribution of quality vary, and and existing resources could be used more e ectively. there is no clear and consistent pattern. However, schools However, there are a number of important gaps that further in highly urbanized areas, particularly high schools, tend to research could clarify. In particular, the PETS-QSDS study was be very large and, despite getting higher levels of funding, not able to look at individual student performance and some tend to have more stressed learning environments and of its important determinants. In particular, the study did poorer levels of performance. The forthcoming extension not explore the quality of classroom instruction and other of high school by two extra years is only going to add to important factors associated with the teaching and learning the pressure that these schools face. A concerted e ort by process. The quantitative approach used in the PETS-QSDS DepEd to address the speci c issues of these schools by, for study is not best suited to exploring issues of this kind, but example, establishing a task force to develop a complete qualitative work would be an appropriate way to add to what program of support for these schools could be an important is known about the quality of education in the Philippines step forward. and to help DepEd to develop appropriate pre-service and in-service training. In some cases, the distribution of education quality reinforces existing inequalities. For example, poorer students tended The PETS-QSDS study collected very detailed information to go to high schools that had teachers with more limited on the status of education service delivery across the knowledge of their subject areas. They also tended to go to Philippines. The report analyzed a wide range of priority schools with lower levels of discretionary funding and those issues, but the data collected in the study could provide the that reported having implemented only a minimal amount of basis of future research into an even wider range of issues. school-based management. For example, further analysis of the teacher competency assessments would be useful to provide further insights into The Philippines has a well-established funding formula the professional development needs of teachers. The data for school operational funding that could be extended could also be used to categorize schools for further research. to include an equity component. This would be one way For example, schools that have better school-community to reduce the large school-level funding inequalities relationships could be identi ed in the PETS-QSDS data highlighted in this report. In order to compensate those as a starting point for more detailed qualitative work to schools that receive lower funding from non-DepEd sources understand why these schools have been successful in this such as local governments, an equity component could be respect. Finally, the data represent an important baseline for introduced in the allocation formula. This could go some way tracking future progress in increasing the e ectiveness of towards equalizing school funding and providing schools in public education spending and improving student learning poorer areas with the additional resources that they need to outcomes in the Philippines. support the learning needs of marginalized students. Many 130 WWW.WORLDBANK.ORG/PH Annex 1 This annex outlines the approach taken to designing the approach adopted in the design of this study is similar to PETS-QSDS study and describes the survey instruments and PETS and QSDS studies that have been undertaken in other sampling strategy used to collect nationally representative countries. data to answer the study’s main research questions. In addition, it documents the details of the eldwork and In order to answer the main research questions, the data processing phases of the study’s implementation. The team developed a detailed analysis plan with a list of key process of designing the study began in November of 2013 information and indicators to be collected. The study tracked and the survey was implemented from August to December the following items in the central government budget. of 2014. • New teachers and teacher salaries. • Teacher training resources, speci cally those from the Developin the Human Resource Training and Development Fund (HRTD) and In-Service Training (INSET). Stud Desi n • School maintenance and other operating expenses (MOOE). The PETS-QSDS study sought to answer four main research questions related to: (i) the ow, management, and control • Infrastructure spending from the Basic Education Facilities of resources; (ii) how resources were used at the school Fund and the School Building Program. level; (iii) how the availability and use of resources in uence In addition to looking at central government resource school performance; and (iv) whether the availability ows, the study also tracked the ow of local government and management of resources di er among regions and education funds to schools, including the Special Education socioeconomic groups. The study tracked over 80 percent of Fund (SEF).2 the national education budget. A set of funding maps was created that outlined the various The study was designed by the World Bank in response steps in the planning, budgeting, and implementation of to requests from the DBM and DepEd for a study to track each budget line to be tracked in the study to help the public expenditure in the basic education system. Prior team to understand fund ow mechanisms and design the to designing the study, the Bank held consultations with expenditure tracking elements of the study.3 The analysis key government agencies to identify the main issues to be plan also identi ed the di erent levels of government o ces covered as well as the most important budget items to be and the other respondents that would need to be surveyed tracked. Speci cally, an advisory group was set up consisting to gather such information as follows: of o cials from DepEd, the DBM, the National Economic and Development Authority (NEDA), and the A liated Network • Department of Budget and Management regional o ces for Social Accountability (ANSA) and of various academics (DBMR) working on the Philippines education system. This group • DepEd regional, division, and district o ces reviewed the overall approach of the study before seeking approval of the study’s concept note and methodology by • Local government units (LGU) at the province and city/ the government. In designing the study, the team and the municipality levels advisory group referred to guidelines posted on the World • Schools Bank’s PETS/QSDS web portal that were drawn from other • Teachers countries’ experiences in implementing similar studies. It also • General parent-teacher associations (GPTA) drew signi cantly from the design of a PETS and QSDS study undertaken in Bangladesh in 2005.1 Therefore, the broad • Student households ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 131 Annex 1 Surve Instruments Since an important aspect of the study was to explore how e ectively schools use their available resources to deliver good quality learning opportunities, the questionnaires Questionnaires were developed corresponding to each were designed to gather considerable information on the institution that covered all of the information needed characteristics of the sampled schools, such as their available to answer the main research questions. In order to track facilities and management systems. The study also collected the ow of funds from central government o ces to information on education inputs and outcomes such as schools, it was necessary to understand the procedures student attendance and dropout rates to supplement the and documentation used for receiving, allocating, and main indicators commonly used to measure school quality disbursing funds at each administrative level. Therefore, in the Philippines such as students’ scores on the National the team conducted exploratory eld visits in CAR region4 Achievement Test. Information was also collected on the in November, 2013. The purpose of these visits was to current management system used and monitoring and understand the type of records that existed on the relevant coordination activities at the division, district, and regional budget items and to identify key respondents within each institutions. o ce or school for di erent modules of the surveys. The eld visits covered all levels of the institutions where sta would The survey instrument for teachers included a teacher be interviewed for the study. assessment module that was designed by the study team in partnership with the Philippines National Research Center for After the questionnaires had been designed, the team Teacher Quality (RCTQ). Each questionnaire included a quality conducted several rounds of pre-testing in di erent regions control section to be lled in by the enumerator before and before nalizing them to ensure that they were able to track after the interview. The school and government-level surveys the timing and ow of these funds from source to nal were conducted using a combination of English and Filipino. destination. Pilot interviews of all levels of the questionnaires The household questionnaire was translated into Filipino. were conducted in Region III and were followed by further Table 2 presents details of the types of information collected checks for some questionnaires. Details of the sites visited for from respondents in each questionnaire. the exploratory visits, testing and pilots are shown in Table 1. Table 1: Details of the Field Visits Carried Out during the Development of the Survey Instruments Region O ce Interviewee Dates Exploratory eld visits Regional director, accountant, payroll CAR DepEd regional o ce services unit o cers School division superintendent (SDS), DepEd division o ce, Benguet accountant, supply/planning o cers DepEd division o ce, Baguio SDS, accountant, COA auditor November 17-21, CAR DPWH, Benguet District Engineer 2013 Benguet municipal LGU Mayor, budget o cer in charge of the SEF Longlong Elementary School in Benguet Principal Lucban Elementary School in Baguio Principal Benguet National High School Principal, accountant 132 WWW.WORLDBANK.ORG/PH Table 1 continued: Details of the Field Visits Carried Out during the Development of the Survey Instruments Region O ce Interviewee Dates Field Visits to Test Questionnaires DepEd division o ce, Taguig city SDS, accountant, planning o cer February 22-28, NCR Bagumbayan National High School Principal 2014 Sta. Teresa Elementary school Principal SDS, accountant, cashier, planning o cer, DepEd division o ce, Cainta Region IV-A human resources o cer, supplies o cer April 11-14, 2014 Sipsipin Elementary School Principal Pilot Interviews Chief budget management specialist in Region IV-A DBM regional o ce charge of DepEd Regional director, human resources Region III DepEd regional O ce manager, physical facilities o cer, accountant, payroll o cer SDS, accountant, cashier, human resources DepEd division o ce, Bulacan o cer, supplies o cer, physical facilities supervisor, planning o cer SDS, accountant, cashier, human resources DepEd division o ce, Angeles City o cer, supplies o cer, physical facilities June 23-27, 2014 supervisor, planning o cer Pampanga DPWH 2nd district engineering District engineer, accountant o ce Angeles City LGU City accountant Dulong Malabon Elementary School, Principal, selected teachers, selected Bulacan households, PTA President Principal, accountant, bookkeeper, selected Nepomuceno High School, Angeles City teachers, selected households, PTA President Further Checks of Implementing Unit High Schools, DPWH, and LGU questionnaires Valenzuela National High School Principal, accountant, property custodian Valenzuela DPWH district o ce District engineer, accountant NCR August 19, 2014 Councilor in charge of education, Valenzuela City LGU budget o cer ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 133 Annex 1 Table 2: Description of Questionnaires General Information Tracking Information Level Questionnaire Respondents Collected Collected • Amount and timing of allotment releases and cash allocations Chief Budget • Overall regional (MOOE, new teacher items, SBM Department Specialist(s) in charge allocations to DepEd grants, HRTD funds, and SBP and of Budget and of the Department of BEF funds) to DepEd regional Management Education o ces and sampled DepEd division o ces and implementing unit (IU) high schools. • Education management • Amount and timing of allotments Region received for HRTD, SBP, and BEF • Monitoring and funds from DepEd central o ce coordination activities and cash allocations from the Regional director, • Opinions on adequacy DBM regional o ce. Department of of budget and payroll o cer, and • Amount and timing of allotment Education e ciency of programs accountant releases (new teacher items, (such as SBP, BEF, and HRTD funds, and SBP and BEF new teacher items) funds) to DepEd division o ces • Financial reporting and implementing unit high procedures and schools. feedback • Interview with division superintendent on education School division management issues, • No tracking. superintendent coordination with local governments, monitoring, and budget management Division, Department Province, and of Education • Number of new teacher posts District division o ce • Division o ce facilities received from DepEd central and personnel o ce and DBM regional o ce Division supplies o cer, for the division and for the planning o cer, human • Size of school-level sampled elementary and non- resources o cer, and workforce and teacher implementing unit high schools. education program transfers supervisors • Information on enrollment • Teacher training and teacher numbers in activities sampled elementary and non- implementing unit high schools. 134 WWW.WORLDBANK.ORG/PH Table 2 continued: Description of Questionnaires General Information Tracking Information Level Questionnaire Respondents Collected Collected • Amount and timing of allotments releases, and cash allocations for MOOE, SBM, and new teacher items received from the DBM regional o ce. • Amount and timing of allotment releases and cash allocations for HRTD, SBP, and BEF funds Division accountant, • Management and received from DepEd central and bookkeeper, cashier, allocation of MOOE and regional o ces. physical facilities SBM funds coordinator, and • Amount and timing of resident Commission • Auditing and nancial downloading of MOOE and SBM On Audit (COA) auditor reporting funds to sampled elementary and non-implementing unit high schools. • Amount and description of construction, repair, and water and sanitation projects in sampled elementary and non- implementing high schools using Division, SBP, BEF, and SEF funds. Province, and District Board member in charge of education for • Support through SEF provincial local school and other local funds Provincial board and budget supporting education government o cer in charge of SEF (such as the General and other education Fund) • Cash and in-kind resources spending provided to sampled elementary • Uses of LGU funds for and non-implementing unit high Councilor for education education (for example, schools. for city/municipality for teachers and City/ sports) local school board municipality and budget o cer in • Coordination with government charge of SEF and other DepEd education spending • Amount and timing of allotment Department of releases and school project Public Works information on SBP and BEF funds Engineer in charge of • Management and and Highways from DPWH central o ce. DepEd projects and coordination with district accountant DepEd • Amount, description, and engineering o ce status of projects in all sampled elementary and high schools. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 135 Annex 1 Table 2 continued: Description of Questionnaires General Information Tracking Information Level Questionnaire Respondents Collected Collected Division, DepEd district Public schools district • Coordination and • Quantity and timing of teacher Province, and o ces supervisor e ectiveness of local guides and learner materials from District school boards DepEd central o ce. • Management and • Allocations and delivery oversight of textbook schedules of learner materials and learning materials and teacher guides to sampled elementary and high schools. • School characteristics, condition, performance, management, and gov- ernance (for example, the quality of SBM) • Principal career • Information on school information characteristics to compare with Public EBEIS information at DepEd • Teacher characteristics, elementary, division and central o ces. School principal attendance, and IU and non-IU management • Information on absenteeism and high schools • Student absenteeism learner materials for sampled students. • Selection of teachers from sample and selection of students from Kindergarten, and Grades 6 and 10 samples Schools • Amount and timing of allotment and their releases and cash allocations for communities MOOE, SBM, and new teacher items received from DepEd district o ce and DBM regional o ce (for IUs). • Amount and timing of in-kind transfers from local government • School revenue and units. Public School principal and, expenditure from PTA, • Training activities from DepEd elementary, for Section C custodian canteen, donations, district o ce and HRTD and SEF IU and non-IU and physical facilities barangay, private training funds. high schools o cer sector, student charges etc. • School SBP, BEF, and SEF construction project description and status to check with DPWH district o ce, DepEd regional o ce (for IUs) and division o ce records. • Receipt of learner materials and teacher guides from sampled DepEd district o ce. 136 WWW.WORLDBANK.ORG/PH Table 2 continued: Description of Questionnaires General Information Tracking Information Level Questionnaire Respondents Collected Collected • Education, experience, training Teachers • Career and training (DepEd and history LGU teachers teaching Teachers • Income • No tracking information kindergarten • Strengths and needs and Grades 6 assessment and and 10) subject-speci c assessment for Grade 6 and 10 teachers Schools • Activities and support and their to school communities General • Contributions • Cash and in-kind contributions to Parent-Teacher General PTA President and charges for school association membership • Revenue and expenditure accounts • Socioeconomic background • Verify school registers by Student • Student education identifying a random sample of Household head students households background • Education direct and • Receipt of learner materials indirect costs ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 137 Annex 1 S mplin Appro ch nd aimed to examine the extent to which local governments and DepEd administrative o ces in uence the performance Fin l S mple of schools. For example, do schools in DepEd divisions who receive their full MOOE allocations in timely manner perform better? Therefore, a further consideration was to include in The sampling approach for this study was based on the need the sample schools in the jurisdiction of a wide range of local to collect information on a nationally representative sample of governments and DepEd divisional o ces. public elementary and high schools. The team developed the sampling approach in consultation with an international ex- Based on these requirements, the study adopted a strati ed pert on sampling who had considerable experience of design- clustered sampling approach in which the primary sampling ing appropriate samples for the Living Standards Measurement units were municipalities and cities. A sample of schools surveys and other public expenditure tracking studies of this was then selected within each sampled municipality or city. kind. The Autonomous Region in Muslim Mindanao (ARMM) The team concluded that a sample size of 250 elementary was excluded from the study based on a review of recent schools and 250 high schools was nancially feasible and studies and after consultations with the advisory board. would provide the necessary precision for some of the key indicators to be studied. In addition, a random sample Since the main purpose of the study was to assess whether of teachers and student households within the sampled schools receive their budgeted resources and use them schools was selected to be interviewed by survey teams e ectively, ensuring that an adequate number of schools were with the relevant questionnaires. Table 3 presents the in the sample was crucial in order to yield accurate estimates planned and nal samples for each questionnaire. A detailed of key indicators at the school level (for example, the share of description of the sampling approach taken at various levels schools receiving their full MOOE allocation). The study also is given in Table 4. Table 3: Planned and Final Samples Planned Final Questionnaire Sample Sample DBM Regional 16 16 DepEd Regional 16 16 DepEd Divisions 50 51 DepEd District **  113 DPWH District  ** 54 Provincial LGU* 30  27 City/Municipal LGU 50  47 Elementary Schools 250 249 High Schools 250 200 Teachers (Kindergarten and grade 6) Max(1,250) 608 Teachers (grade 10) Max(1,500) 946  GPTA 500 449 Student households 2,500 2,189 Notes: * Since HUCs and other cities do not have a provincial-level LGU, 30 provincial LGUs were sampled. ** The sampling of district level o ces was driven by the schools selected. All DepEd and DPWH o ces associated with the sampled schools were identi ed and interviewed. The nal sample size indicates the number of observations that can be used for analysis after accounting for adjustments and non-responses. 138 WWW.WORLDBANK.ORG/PH S mplin of Cities nd Municip lities, The nal sample included 249 elementary and 200 high nd Associ ted A encies schools. One sampled elementary school could not be ac- cessed for an interview because the school was closed because Cities and municipalities in 16 regions were selected within of security concerns. In addition, an incorrect high school that strata of highly urbanized cities (HUCs), other cities, and was not in the original sample had to be dropped.9 municipalities.5 A total of 30 municipalities, 10 other cities, and 10 HUCs were sampled with probability proportional The samples of DepEd and DPWH district o ces were to population size.6 These numbers were determined by determined by the schools that had been selected. All the need to ensure an adequate sample of schools from DepEd and DPWH district o ces that were associated with di erent types of location. In addition, following the team’s the 250 elementary schools and 201 high schools in our consultations with DepEd, three HUCs from the National original sample were identi ed and visited for interviews. Capital Region (NCR) were sampled with certainty given The original and nal sample sizes for the survey of general the region’s large population and the high share of basic parent-teacher associations (GPTA) were also based on the education funds spent in the NCR. sample of schools. All division o ces and LGUs that had at least one sampled S mplin of Te chers nd elementary or high school under their jurisdiction were Student Households selected. There was some attrition due to key o cials refusing to be interviewed or being unavailable, and this The sampling of teachers and student households was done in resulted in nal samples sizes of 27 provincial LGUs and 47 the eld by survey teams. Within each elementary school, up city/municipality LGUs. to three DepEd-funded teachers and up to two LGU-funded teachers were sampled. Among the DepEd-funded teachers, While most cities and municipalities are under the up to two Grade 6 teachers and one kindergarten teacher jurisdiction of a single division, DepEd’s decision in 2014 were selected from each school. In order to select Grade 6 to split a particular division in two meant that a city in the teachers, two Grade 6 sections were rst randomly sampled study sample came under the jurisdiction of both divisions. from the pool of all Grade 6 sections within each elementary school, and from those sections, up to two DepEd-funded S mplin of Schools teachers were then selected. However, if there was only one DepEd-funded Grade 6 teacher in the sampled sections to Random samples of ve public elementary schools o ering choose from, then an additional teacher was selected from the full K to 6 grades and ve public high schools were among DepEd-funded teachers in the other non-sampled chosen with equal probability within each sampled city or Grade 6 sections. If there were no DepEd-funded Grade 6 municipality. However, some cities and municipalities had teachers in the other sections, then a DepEd-funded teacher fewer than ve high schools. In such cases, all of the high from another grade was sampled. If the selected sections had schools in the city or municipality were included in the LGU-funded or locally hired teachers, then one LGU-funded sample. Private schools were excluded from the sampling teacher was sampled from this pool. If there were no LGU- frame because, on the whole, the public funds tracked funded teachers in the selected Grade 6 sections, then an by the study do not ow to these types of schools.7 The LGU-funded teacher was chosen either from the pool of all funds tracked in the study represented over 80 percent LGU-funded Grade 6 teachers or from all LGU teachers in the of the total central government education budget in school depending on the availability of LGU-funded teachers 2014 and were mainly used to support public schools. in other Grade 6 sections. Integrated schools and incomplete schools (in other words, elementary schools not o ering all K to 6 grades) were also High school teachers were also sampled from a pool of excluded.8 In addition, 350 special high schools that follow teachers teaching in two randomly selected Grade 10 a di erent curriculum and 84 high schools that were new sections in each school, but they did not need to be LGU- and did not o er grade 10 classes were not included in the funded. Emphasis was placed on sampling at least one sample frame. teacher specializing in Math, Science, English, and Filipino. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 139 Annex 1 There were signi cantly fewer teachers in the nal sample Two of these students were selected from the pool of than had been estimated. A number of factors contributed students who attended school on the rst day of the survey to the gap between the planned and nal samples. Some while the other two students were chosen from among schools had only one section per grade or, in the case of the students who were not at school that day. A total of high schools, had one teacher teaching multiple subjects, two kindergarten students were also selected from each which reduced the actual size of the teacher sample. elementary school, one from among those attending school Some elementary schools did not have any LGU-funded on the rst day of the survey and the second from among teachers or DepEd-funded kindergarten teachers. The size those who were absent. Four Grade 10 students from each of the high school teacher sample was also lower than high school were also sampled using the same approach. planned as the original estimate was based on 250 high schools instead of the 201 that were actually sampled. In Attrition in the household sample was a result of the addition, only teachers in Grades 6 and 10 were used in smaller nal size of the school samples as well as of some the analyses. people refusing to be interviewed or the team’s inability to contact some households. Moreover, in some cases, For the household questionnaire, four Grade 6 students security concerns prevented eld teams from visiting the were selected from the sampled sections in each school. households.10 Table 4: Outline of the PETS-QSDS Sampling Strategy Sampling Number of Selection Units Stage Domain Strata Units Selected Method Total Systematic sampling with probability NCR Cities 3 3 proportional to population size Selected with Davao City 1 1 City and certainty 1 municipalities All other regions All other highly 6 Systematic sampling 6 excluding ARMM urbanized cities with probability Other cities 10 proportional to 10 population size Municipalities 30 30 Complete (all- grade) public Systematic All 50 cities and 5 per city/ elementary 2 Government equal-probability 250 municipalities municipality schools (including sampling kindergarten)               All 250 selected Elementary Simple random 3 complete All Grade 6 sections 2 ~500 Grade 6 sections sample elementary schools               140 WWW.WORLDBANK.ORG/PH Table 4 continued: Outline of the PETS-QSDS Sampling Strategy Sampling Number of Selection Units Stage Domain Strata Units Selected Method Total At least two DepEd Selected with Grade 6 teachers in 1-2 ~500 certainty selected sections 3a All 250 selected Elementary school Simple random complete teachers   sample (select elementary schools At least one LGU/ 1 Grade 6 teacher in 250 locally hired teacher selected section where possible) Grade 6 students in selected section Simple random attending school 2 500 sample Elementary All 250 selected on rst day of the 3b survey school student complete households   elementary schools Grade 6 students in selected section Simple random 2 500 absent on the rst sample day of the survey DepEd Kindergarten Simple random All 250 selected Volunteer Program 1 250 Kindergarten sample 3c complete (KVP) teacher teachers elementary schools LGU/locally hired Simple random 1 250 teachers sample   Kindergarten students attending Simple random 1 250 school on the rst sample Kindergarten 3d All 250 selected day of the survey school complete households   elementary schools Kindergarten students absent on Simple random 1 250 the rst day of the sample survey               Complete public All 50 cities and 5 per city/ Systematic equal- 4 Government ~250 high schools municipalities municipality probability sampling     High school All 250 selected All Grade 10 Simple random 5 2 ~500 section secondary schools sections sample               ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 141 Annex 1 Table 4 continued: Outline of the PETS-QSDS Sampling Strategy Sampling Number of Selection Units Stage Domain Strata Units Selected Method Total Grade 10 teachers in math, science, High school All 250 selected Grade 10 teachers in English and Selected with 5a ~1500 teachers secondary schools selected sections Filipino certainty teaching selected sections   Grade 10 students in selected section 2 from selected Simple random attending school on 500 section sample High school the rst day of the All 250 selected survey student 5b secondary schools households Grade 10 students in selected section 2 from selected Simple random 500 absent on the rst section sample day of the survey Surve Implement tion trained to conduct the interviews in a respectful manner and to mitigate any potential biases in the responses. After the survey instruments had been nalized, the study Two batches of 10-day training sessions were held in was implemented between August and December of 2014 September before the launch of the actual eldwork. The by a Philippines-based survey rm, Tylor, Nelson, and Sofres training sessions were split into ve days of classroom (TNS), which was also responsible for data encoding. Several training and ve days of training in the eld. The classroom steps were taken to ensure the quality of the interviews and training included an overview of the project and its design, the resulting outputs, including providing enumerators with a discussion of each question in the questionnaires, and comprehensive training and building in many quality control instruction on the proper way to use skips and to document checks. In addition, TNS secured authorization from the responses. Role playing and mock interviews and recording relevant school division superintendents before conducting of responses were also used to prepare enumerators for the school interviews. eldwork. In the training held in the eld, enumerators participated in test runs of interviews during which they Enumer tor Tr inin were accompanied by a eld supervisor. Finally, sessions were held to discuss and resolve any di culties that were The survey enumerators underwent extensive training encountered during test runs before the launch of the actual in how to administer each module of the questionnaires eldwork. and in how to record the information obtained during the interview from verbal responses, o cial records, and A comprehensive eldwork manual was prepared for personal observation. They were also given detailed the enumerators to use. The manual provided them with instructions on the sampled grade sections, teachers, and background information on the Philippines’ education student households before administering the teacher and system as well as an overview of how DepEd is organized. It household questionnaires. Enumerators were speci cally also included a description of each stage of the budgeting 142 WWW.WORLDBANK.ORG/PH process to give enumerators an overview of the ow of basic for carrying out sampling of grade sections, teachers, and education funds in the Philippines. It also contained detailed student households. instructions for completing each module of the PETS-QSDS questionnaires as well as decision trees to facilitate the In addition, each region was assigned an executive sampling of teachers and student households. enumerator. These executive enumerators have experience in conducting these types of interviews so were appointed In addition, the World Bank team prepared a manual on to interview senior o cials in the regional, province, and the public nancial management system in the Philippines, district o ces of the relevant government departments. which explained all phases of the government’s budget The executive enumerators also interviewed the school process. The manual included a detailed description of the division superintendents. These senior enumerators were ow of public funds tracked in the survey starting from also accompanied by an assistant enumerator on any visits appropriation and allotment to spending guidelines as well that entailed documenting a lot of information from o cial as brief descriptions of funds that schools often receive from records. Interviews at the school and household levels were other sources such as school canteen funds and other non- conducted by the eld survey teams. The survey teams also government sources such as Parents Teachers Association administered sections of the division questionnaires that funds. The manual also contained relevant government were not covered by the executive enumerator. orders and guidelines associated with public fund ows in the education sector. In case an interview was not possible at the time of the rst visit, the enumerators scheduled another visit to conduct the Fieldwork Or ni tion interview. The eld interviews began on September 24, 2014 and were Qu lit Control conducted in two phases. The schedule and coverage of the two phases are shown in Table 5. The utmost care was taken to ensure quality by carrying out checks at various points during the implementation of The eld team was led by a eld manager and three eld the study. Quality control checks were conducted by eld coordinators who each covered one of the three island coordinators (FC), eld team leaders (FTL), a team of back groups in the Philippines - Luzon, Visayas, and Mindanao. checkers, and by each enumerator. In order to facilitate The eld coordinators were in charge of monitoring the systematic checks in the eld, questions related to survey survey implementation and of creating eld reports on a quality were embedded in the questionnaires themselves. A weekly basis. Each region was assigned two to three eld core survey management team handled any quality control survey teams to conduct the interviews, with each team issues and reviewed the weekly progress reports submitted consisting of about three members. Each team comprised a from the eld. In addition, the World Bank team conducted eld team leader (FTL) who was responsible for conducting survey monitoring visits as did the RCTQ (for the teacher quality checks and ensuring that interviews were carried assessments). Furthermore, any completed questionnaires out and documented properly. FTLs were also responsible with a lot of missing data were identi ed during the data Table 5: Schedule and Coverage of Field Visits Field Visits Regions No. of Cities/Municipalities Schedule Luzon and NCR 19 September 24-November 13 1st Phase Visayas and Mindanao 9 September 30-November 13 Luzon 7 November 12-December 9 2nd Phase Visayas and Mindanao 5 November 17-December 9 ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 143 Annex 1 processing phase, and the school or agency in question was checkers also ensured that all levels of questionnaires were revisited to get complete responses to the extent possible.11 completed in each city or municipality after the completion of the eldwork. The quality checks by each type of eldwork member are as • Other Enumerators. Each enumerator carried out a quality follows: check on each interview that they conducted themselves by lling in the quality control section included in each • Field Coordinator (FC). A eld coordinator was assigned to questionnaire. Prior to the beginning of the interview, each of the three island groups. FCs were responsible for enumerators followed the instructions on choosing the supervising their eld teams and executive enumerators correct respondent and entered the respondent’s details to ensure that they complied with interview protocols, in the quality control section. After the interview was including proper data collection and encoding during completed, the enumerators were required to check and interviews and proper sampling of teachers and document whether each module in the questionnaire households in the eld. FCs also consolidated and was completed and whether any relevant information submitted weekly progress reports from the eld to TNS or identi ers (such as school identi cation codes) had central o ce. been transferred from one questionnaire to another (for • Field Team Leader (FTL). FTLs were generally responsible for example, transferring data on speci c teachers from the ensuring the quality of the interviews conducted by their teacher observation module in the school questionnaires team and the information that they collect. Immediately to the teacher and RCTQ assessment questionnaires). after their teams conducted the interviews, while still in the eld, the FTLs checked the completed questionnaires for correctness, completeness, and consistency and took Processin D t steps to rectify any errors. For example, they veri ed for An l sis information obtained from o cial records against the original sources (such as MOOE records) or ensured The data collected during the eldwork were encoded by that the skips in the questionnaires had been followed TNS using automated data entry scripts that had been pre- or checked whether information recorded multiple tested. The World Bank team gave regular feedback to the times in the survey is fairly consistent (such as whether data entry team regarding the need for any corrections of totals add up correctly). In order to ensure that these the data. TNS submitted a complete set of data to the World checks are carried out systematically, each questionnaire Bank in July of 2015. includes a quality control section for FTLs to ll in after the completion of the interview. FTLs also observed a number D t Cle nin of interviews conducted by other enumerators in the team and provided them with appropriate feedback. In addition, After the World Bank received the datasets from TNS, some FTLs were required to follow up on a random selection of further cleaning and veri cation was required. The major interviewed households to cross-check some of the basic changes and follow-ups for veri cations are outlined in Box 1. information with the respondents. Cre tion of S mple Wei hts • Back Checkers. A team of nine back checkers was appointed to conduct further cross-checks of the quality controls In order to strengthen the external validity of the ndings carried out by the FTLs on a sample of questionnaires. These of the policy notes, the analyses used sampling weights back checkers re-checked approximately 30 percent of based on the probability of being selected from the relevant questionnaires in TNS central o ce and during eld visits. populations and sub-populations. The World Bank team They were provided with a separate form to assist in the calculated the sampling weights using the nalized datasets checking of questionnaires and to identify inconsistencies. after the completion of adjustments described in Box 1. Similar to FTLs, back checkers also made follow-up visits to a Details of how the weights for each dataset were calculated random sample of households and observed some original are provided in a set of additional tables and annexes interviews, including some conducted by FTLs. The back accompanying the report. 144 WWW.WORLDBANK.ORG/PH Box 1: Finalizing Datasets for Analysis The World Bank study team made the following revisions to the datasets when conducting preliminary analyses using the data submitted by TNS in July 2015. All of the nal analyses for the policy notes were conducted using the revised datasets. • Six non-implementing high schools were accidentally administered the HSIU2 questionnaire meant for implementing (IU) high schools. For variables that overlap between the HS2 and HSIU2 questionnaires, the data from the HSIU2 questionnaire for these six schools were appended to the HS2 data. All information in the HSIU2 dataset for these six schools were then removed. • An IU high school that was not in the sample was accidentally interviewed in place of Baguio High School of Agriculture - Tambobong Annex (which was in the original sample). All of the variables corresponding to this school in the high school, teacher, GPTA, and household questionnaires were set to missing. • Non-zero values for MOOE amounts in the division-level data were recorded for ve IU high schools. Following phone calls to verify their IU status and whether they had received any MOOE funds from the division o ce or from the DBM, the study team set the variables for MOOE amounts in the division data for four of these schools to missing as they did not report receiving any MOOE from divisions. • Adjustments were made to the teacher and RCTQ assessment datasets. Teachers who do not teach kindergarten, Grade 6, or Grade 10 were dropped as the analyses on teachers were focused on teachers who teach these grades. In addition, some teachers who had not been included in the “teacher roster” in the school questionnaires but had been interviewed were also dropped. Use of Other D t Sources While most of the analysis was conducted using the information gathered in the study. Table 6 lists all of datasets from this PETS-QSDS study, the study team these other data sources that were used in all seven often had to use other data sources to supplement the policy notes. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 145 Annex 1 Table 6: Other Data Sources Used in the Policy Notes Data Source Year(s) Description/use DepEd’s Fact Sheet on Basic 2011, 2013, Enrollment gures for kindergarten, elementary and high school, and Education Statistics 2015 teacher ratios in all public schools 2011, 2012, Enrollment gures by grade (2003 to 2014), number of classrooms and DepEd eBEIS 2013 number of teachers for all public schools DBM General Appropriations Act 2013 MOOE allocations for each public high school in the Philippines Elementary school MOOE allocations 2013, parameters of MOOE funding DepEd (Correspondence) 2013–2016 formula 2013-2016. Standard regional costing estimates for classroom construction and water DepEd (Correspondence) and sanitation units DepEd Sta ng Summary 2012– 2016 DepEd obligations for teacher salaries Document DepEd Order 12, s.2014 2014 Implementing guidelines for school MOOE Guidelines on the Allocation/Deployment of New Teaching, Teaching- DepEd Order 77, s.2010 2010 related, and Non-teaching Positions for FY 2010 DepEd Order 102, s. 2011 2011 Benchmark dates on new teacher allocations and deployment DepEd Order 54, s. 2009 2009 Guidelines governing PTAs at the school level DepEd Statement of Allotments Allotments and obligations for Human Resources and Training 2005–2014 and Obligations Development funds, school MOOE funds. DepEd Basic Education Facilities School-level details of each classroom construction project and water/ 2013 Fund data sanitation unit construction project Total government spending, total education spending, DepEd’s spending National Expenditure Program obligations Bureau of Local Government 2009–2014 Data on Special Education Funds Finance Various Philippines Statistics Authority GDP de ator years ADB Teacher Absenteeism 2014 Teacher absenteeism rates in Indonesia Survey - Indonesia OECD Teaching and Learning 2013 Survey on the professional development of teachers in di erent countries International Survey See policy World Development Indicators   note 1 UNESCO Institution for Statistics 2012 Teacher salaries in various countries PETS - Cambodia 2004 Absenteeism rates for primary school teachers 146 WWW.WORLDBANK.ORG/PH Endnotes 1 FMRP (2006) “Social Sector Performance Surveys: Primary 7 The bulk of government funding to private schools is given Education in Bangladesh: Assessing Service Delivery” Oxford through the Government Assistance for Students and Teachers Policy Management and Financial Management Reform Program, in Private Education (GASTPE), which only represented about 3 Dhaka. percent of new appropriations in 2014. The study team decided 2 The SEF comes from a 1 percent property tax surcharge and is not to track these resources given their relatively small share of managed by local government units. the overall budget and the fact that other studies were planning to look at the GASTPE program in more detail. 3 The funding map along with the relevant regulations and guidelines were collated into a single report used to train survey 8 As of 2013 school year, there were 164 integrated elementary enumerators. schools and 568 integrated high schools in the Philippines. Thirteen percent of all elementary schools were incomplete. 4 The CAR region, the two division o ces visited, and other sites visited to pre-test the questionnaires were selected largely 9 Instead of Baguio High School of Agriculture-Tambobong Annex, because of convenience as determined by the sites’ proximity to another IU high school was interviewed by mistake. Manila. 10 Some areas where sampled households were located were 5 The strati cation followed the legal classi cation of cities and known to harbor insurgents. municipalities based on population size and annual income. 11 Revisits were completed for 35 out of the 58 institutions or 6 Population data used to select the sample was from the 2010 interviewees that had been identi ed as needing a revisit. In Census. some cases where only one school in a municipality or city was identi ed, no revisit was carried out. ASSESSING BASIC EDUCATION SERVICE DELIVERY IN THE PHILIPPINES 147 The World Bank 1818 H Street, N.W. Washington, D.C. 20433 USA www.worldbank.org/ph