REALIGNING THE UNION BUDGET TO MYANMAR’S DEVELOPMENT PRIORITIES MYANMAR: PUBLIC EXPENDITURE REVIEW SEPTEMBER 2015 ii CURRENCY EQUIVALENTS (Exchange Rate Effective as of May 15, 2015) Currency Unit = Kyat (MMK) US$1.00 = MMK 1,082 FISCAL YEAR April 1 – March 31 Vice President: Axel von Trotsenburg (EAPVP) Country Director: Ulrich Zachau (EACTF) Practice Director: Satu Kähkönen (GMFDR) Practice Manager: Mathew Verghis (GMFDR) Task Team Leader: Douglas Addison (GMFDR) Khwima Nthara (GMFDR) iii CONTENTS ACKNOWLEDGEMENTS VII ACRONYMS AND ABBREVIATIONS VIII GLOSSARY OF KEY TERMS IX EXECUTIVE SUMMARY: X CHAPTER 1 INTRODUCTION 2 A. BACKGROUND AND OBJECTIVES 2 B. ECONOMIC CONTEXT 2 C. SCOPE OF PUBLIC SECTOR 3 D. SCOPE AND OUTLINE 5 E. APPROACH AND METHODOLOGY 6 CHAPTER 2 SUSTAINABILITY OF AGGREGATE FISCAL POLICY 9 A. BACKGROUND 9 B. AGGREGATE FISCAL DEVELOPMENTS 11 C. PUBLIC DEBT 12 D. MEDIUM-TERM FISCAL FRAMEWORK 15 E. TREATMENT OF UPSTREAM HYDROCARBON REVENUES 18 F. FISCAL RISK MANAGEMENT 19 G. CONCLUSIONS AND RECOMMENDATIONS 24 CHAPTER 3 REBALANCING THE COMPOSITION OF THE UNION BUDGET 26 A. BACKGROUND 26 B. REVENUE EFFORT AND COMPOSITION 26 C. REVENUE POLICIES AND INSTITUTIONS 29 D. REVENUE ADMINISTRATION 30 E. FUNCTIONAL COMPOSITION OF EXPENDITURE 31 F. ECONOMIC COMPOSITION OF EXPENDITURE 33 G. CONCLUSIONS AND RECOMMENDATIONS 38 CHAPTER 4 IMPROVING THE COVERAGE, QUALITY AND EQUITY OF EDUCATION 40 A. BACKGROUND 40 B. THE EDUCATION SECTOR 40 C. EDUCATION CHALLENGES 42 D. POLICY PRIORITIES IN EDUCATION 45 E. SECTOR FINANCING 47 F. MINISTRY OF EDUCATION SPENDING 48 G. PUBLIC FINANCIAL MANAGEMENT IN EDUCATION 53 H. CONCLUSIONS AND RECOMMENDATIONS 56 CHAPTER 5 FROM MORE TO BETTER GOVERNMENT SPENDING ON HEALTH 58 A. BACKGROUND 58 B. THE HEALTH SECTOR 58 C. HEALTH CHALLENGES 59 D. POLICY PRIORITIES IN HEALTH 63 E. HEALTH SECTOR FINANCING 64 F. MINISTRY OF HEALTH EXPENDITURES 67 G. EFFICIENCY OF GOVERNMENT SPENDING ON HEALTH 69 H. EQUITY IN GOVERNMENT SPENDING ON HEALTH 72 I. CONCLUSIONS AND RECOMMENDATIONS 75 CHAPTER 6 FISCAL FRAMEWORK FOR SUB-NATIONAL SERVICE DELIVERY 78 A. BACKGROUND 78 B. SUB-NATIONAL ADMINISTRATION IN MYANMAR 78 C. FISCAL DEVELOPMENTS 81 D. EXPENDITURE DECENTRALIZATION 82 E. FISCAL TRANSFERS 86 F. SUB-NATIONAL PLANNING AND PUBLIC FINANCE MANAGEMENT 89 G. REVENUE DECENTRALIZATION 91 H. ACCOUNTABILITY FOR RESULTS 95 I. CONCLUSIONS AND RECOMMENDATIONS 96 REFERENCES 98 ANNEX 1: DATA AND ACCOUNTING ISSUES 104 ANNEX 2: CROSS-COUNTRY DATA COMPARISONS 107 ANNEX 3: ROLES AND RESPONSIBILITIES OF VARIOUS ORDERS OF GOVERNMENT 110 ANNEX 4: BUDGET OUTCOMES OF THE STATES AND REGIONS 115 iv BOXES, FIGURES, AND TABLES Boxes Box 1.1: Data Quality 7 Box 3.1: Size of the Wage Bill – Lessons from ECA Countries 35 Box 4.1: 10 Points Education Policy 45 Box 4.2: Data Challenges 46 Box 4.3: Important Gaps to Address through Higher Government Spending on Education 49 Box 5.1: Gaps in the Coverage and Comprehensiveness of Budget Data on Health 63 Figures Figure 1.1: Myanmar Public Sector 4 Figure 2.1: Size of General Government and Growth, 2009-12 9 Figure 2.2: Government Expenditure, 2009-13 11 Figure 2.3: Government Revenue, 2009-13 11 Figure 2.4: Fiscal Balances 12 Figure 2.5: Overall Fiscal Balance 12 Figure 2.6: Public Debt in Myanmar (Percent of GDP) 13 Figure 2.7: Public Debt (Percent of GDP) 13 Figure 2.8: Deficit Financing 14 Figure 2.9: Deficit Financing in Peer Groups 14 Figure 2.10: European Gas Prices, US$/MMBTU 17 Figure 2.11: Planned Borrowing, by Ministry 19 Figure 2.12: Planned Borrowing, by SEE Sector 19 Figure 2.13: Debt Service Obligations 20 Figure 2.14: Debt Servicing Costs (Kyat Million) 20 Figure 2.15: Gas and Oil Prices, 2001-2013 21 Figure 2.16: Gas Prices and Revenues, 2001-2013 21 Figure 3.1: General Government Revenues, 2010/11 26 Figure 3.2: Tax Revenues and Per-capita GDP, 2010/11 26 Figure 3.3: Regional Revenue Effort (Percent of GDP) 27 Figure 3.4: Revenue Composition 27 Figure 3.5: Customs Duties Collection 28 Figure 3.6: Logistics Performance Index 28 Figure 3.7: Government Function 2009/2010 32 Figure 3.8: Government Function 2013/14 32 Figure 3.9: Social Protection Spending 32 Figure 3.10: Military Expenditures 32 Figure 3.11: Economic Classification, 2009/10 33 Figure 3.12: Economic Classification, 2013/14 33 Figure 3.13: Capital Expenditure 34 Figure 3.14: Employee Compensation 34 Figure 3.15: Wage Bill and Civil Service Employment 36 Figure 3.16: Ratio of Investment to Wage Bill 36 Figure 3.17: Goods and Services 36 Figure 3.18: Social Benefits 36 Figure 4.1: Educational Attainment by Age Cohorts 2009/10 42 Figure 4.2: Estimates of Net Primary Enrolment Rates 42 Figure 4.3: 11th Graders Who Passed School Leaving Test in 2010 43 Figure 4.4: Early Grade Reading Assessment, Yangon Region 2014 43 Figure 4.5: Composition of Financing for Education 48 Figure 4.6: Composition of Private Spending 48 Figure 4.7: Union Education Expenditures, 1972-2011 48 Figure 4.8: Government Spending by Country 48 Figure 4.9: Block Grants Allocated Per School, for Goods and Services 52 Figure 4.10: Budget Calendar for FY2012/13 Budget Preparation 54 v Figure 5.1: Per-capita GDP and HDI 60 Figure 5.2: Public Health Spending, 1975/76 - 11/12 60 Figure 5.3: Age-standardized Death Rates 2012 (Per 100,000 Population) 61 Figure 5.4: Out-of-Pocket Financing (Percent of Total Health Care Expenditures) 65 Figure 5.5: Health Financing Expenditure Shares 65 Figure 5.6: Health Financing (% of GDP) 66 Figure 5.7: Health Financing (% Total Financing) 66 Figure 5.8: MoH Expenditures 67 Figure 5.9: MoH Capital and Current Spending 67 Figure 5.10: MoH Spending on Health Resource Use 2009/10 – 2013/14 69 Figure 5.11: Ministry of Health Spending of Public Goods 70 Figure 5.12: Health Spending by Function, Total 71 Figure 5.13: Health Spending by Function, MoH Only 71 Figure 6.1: Overview of Government Hierarchy in Myanmar 79 Figure 6.2: State/Region Structure and Population Density 79 Figure 6.3: Spending and Revenue Decentralization 83 Figure 6.4: Decentralization and Transfers 83 Figure 6.5: Expenditure Shares, by Economic Categories, 2013/14 RE 84 Figure 6.6: Budget Execution Rates for Expenditures 84 Figure 6.7: Poverty and Expenditures Per-capita 85 Figure 6.13: Degree of Equity in Distribution of State/Region Receipts Per-capita 87 Figure 6.8: Sub-national Revenues (Percent of Total Public Sector Revenues) 92 Figure 6.9: Own-source Revenue, Composition 93 Figure 6.10: Own-source Revenue, States/Regions 93 Figure 6.11: SEE Non-tax Contributions (Share of State/Region Own-Revenue) 93 Figure 6.12: Poverty and Own Current Revenues Per-capita 94 Tables Table 2.1: Fiscal Accounts, as Shares of GDP, 2009/10 – 2013/14 10 Table 2.2: Budgetary Deviations, as Percent of GDP 12 Table 2.3: Alternative Fiscal Scenarios for the Union Government 16 Table 2.4: SEE Net Flows (Percent of GDP) 23 Table 3.1: SEE Contributions to Union Revenues and Grants 29 Table 3.2: Government Employment 2009 34 Table 4.1: Education Expenditure by Different Ministries 2013/14 41 Table 4.2: Education Outcomes in 2010, by States and Regions 44 Table 4.3: Education Outcomes, by Income Quintiles, Cross-Country 44 Table 4.4: Total Education Expenditures 2009/10- 2013/14 49 Table 4.5: Basic Education Expenditures, 2011/12 to 2013/14 50 Table 4.6: Monthly Teacher Salaries by Rank 50 Table 4.7: Estimates of the Cost of New Initiatives Introduced Since 2011/12 51 Table 5.1: Health Status Indicators ASEAN Countries 62 Table 5.2: Selected Indicators of Health Status and Access, 2009/10 62 Table 5.3: Financing Sources and Financing Agents, Percent of Total, 2011/12 64 Table 5.4: Ministry of Health Capital Spending on Construction 68 Table 5.5: Health Care Personnel 2009/10 - 2012/13 68 Table 5.6: Urban-Rural Public Spending on Health at Facility Level (MMK Millions) 73 Table 5.7: Illness and Health Care Seeking Behavior 2009/10 73 Table 5.8: Distribution of Health Expenditures, Public 2009/10 74 Table 5.9: Incidence and Intensity of Catastrophic Health Payments 2009/10 75 Table 6.1: Total Expenditures as Shares of Total Revenues and Transfers 81 Table 6.2: Budget Components as Shares of Total Revenues and Transfers (Percent) 82 vi ACKNOWLEDGEMENTS This public expenditure review would not have been possible without the extraordinary degree of cooperation from His Excellency Minister U Win Shein, Director-General (Budget) Dr. Maung Maung Win, and Deputy Director-General (Budget) Daw Nwe Nwe Win in the Ministry of Finance; Lieutenant Ko Ko in the Ministry of Home Affairs; Her Excellency Minister Dr. Khin San Yee and His Excellency Minister Dr. Mya Aye (Retired) in the Ministry of Education; and His Excellency Minister Dr. Than Aung, His Excellency Minister Dr. Pe Thet Khin (Resigned) in the Ministry of Health. Their help is deeply appreciated. They are not, however, responsible for any of the conclusions in this report. Any errors are the responsibility of the authors. Special thanks are extended to the staff of the Budget Bureau in the Ministry of Finance who helped establish a comprehensive database for this review. The report was prepared by a multi-donor team led by Khwima Nthara and Douglas Addison with financial assistance from the Australian Department for Foreign Affairs and Trade. Editorial assistance was provided by Eric Palladini. The work was conducted under the guidance of Ulrich Zachau, Satu Kähkönen, Mathew Verghis, Shabih Mohib, Luis Benveniste, Toomas Palu, and Rob Taliercio. The team is grateful for their ongoing support and guidance provided. Chapter 2 on the sustainability of aggregate fiscal policy was led by Khwima Nthara and Habib Rab. The team included Valentina Flamini (IMF), Sarah Somoudi (DfID), and Douglas Addison. Useful comments were received from May Thet Zin, Lars Sondergaard, Kenichiro Kashiwase (IMF), Kelly Eckhold (IMF), and Declan Magee (DfID). Chapter 3 on rebalancing the composition of the Union Budget was led by Douglas Addison. Valuable assistance in data preparation was received from the Budget Bureau in the Ministry of Finance. Useful comments were provided by May Thet Zin, Lars Sondergaard and Declan Magee (DfID). Chapter 4 on improving coverage, quality and equity of education was led by Lars Sondergaard. The team included Victor Levin, Serif Sayin, Emiko Naka, Dilaka Lathapipat, Ei Kalya Moore. Nilar Tun prepared additional material. The report draws on work carried out as part of World Bank project preparations for “Decentralized Funding to Schools Project” and a “Rapid Assessment of the Financing to Education.” The rapid assessment was carried out as part of phase 1 of the Comprehensive Education Sector Review process. This rapid assessment was prepared jointly by Yee Yee Cho (Ministry of Education), Lars Sondergaard, Tsuyoshi Fukao, Dilaka Lathapipat, Victor Levine, Ei Moore, Vachraras Pasuksuwan, Emiko Naka (DFAT), and James Stevens. Comments were received from: Pedro Cerdan-Infantes, Mar Mar Thwin, Nilar and Edwina Betts. Chapter 5 on more to better spending on health was led by Mukesh Chawla. The team included Sutayut Osornprasop and Rocio Schmunis. The team is grateful to Theepakorn Jithitikulchai for assistance with the analysis of survey data, and to Caryn Bredenkamp for advice on conclusions and recommendations. Useful comments were provided by Hnin Hnin Pyne, and Lars Sondergaard. The chapter 6 on intergovernmental relations was led by Robert Boothe. The team included Anwar Shah (consultant), Miki Matsuura and Mark Miller. This work takes a collaborative approach with government, and builds on a consultative workshop held in February 2014, in which union government officials, state and region finance ministers met with the World Bank team to discuss priority issues and options for reform. Useful comments were received from Jonas Frank and Asbjorn Wee Peer reviewers for the PER were Antonio Nucifora, Jose Calix, Toby Linden, Karima Saleh, Owen Smith, and William Dillinger. vii ACRONYMS AND ABBREVIATIONS ADB Asian Development Bank ASEAN Association of Southeast Asian Nations BAU Basic Accounting Units BE Budget Estimate CBM Central Bank of Myanmar CESR Comprehensive Education Sector Reform DBE Departments of Basic Education, Ministry of Education DEPT Department of Education Planning and Training, Ministry of Education DFAT Department of Foreign Affairs and Trade (Australia) DfID Department for International Development (United Kingdom) ECCD Early Childhood Care and Development ECD Early Childhood Development EGMA Early Grade Math Assessment EGRA Early Grade Reading Assessment EMIS Education Management Information System FESR Framework for Economic and Social Reform FY Fiscal Year GAD General Administration Department, Ministry of Home Affairs GDP Gross Domestic Product HDI Human Development Index HQ Headquarters IHLCS Integrated Household Living Conditions Survey INGO International Non-governmental Organization IRD Internal Revenue Department, Ministry of Finance JICA Japan International Cooperation Agency LECS Lao Expenditure and Consumption Survey LMIC Low and Middle-Income Countries MEB Myanmar Economic Bank MMK Myanmar Kyat MNPED Ministry of National Planning and Economic Development MOGE Myanma Oil and Gas Enterprise MoE Ministry of Education MoF Ministry of Finance MoH Ministry of Health MoHA Ministry of Home Affairs NHA National Health Accounts NGO Non-governmental Organization PA Provisional Actual (the preliminary figure for government expenditure) PER Public Expenditure Review PETS Public Expenditure Tracking Survey RE Revised Estimate (revised budget estimate) REO Regional Education Officer SAA Special Administrative Area SEE State Economic Enterprises SEO State Education Office TEO Township Education Offices TF Treasury Forms UNICEF The United Nations Children's Fund viii GLOSSARY OF KEY TERMS Authority Constitutional, legal or regulatory mandate to undertake some activities or functions. Autonomy Authority to act and take decisions independently of central control. Decentralize Process of transferring power, authority and responsibility for public functions from central to sub-national tiers of government. Deconcentrate Process of transferring functions and responsibilities to administrative units of central government which operate at the sub-national level, which retain upwards accountability to the center. Delegation Allocation of power by the center to other levels of government in a unitary state, in which the center retains authority to withdraw the delegated power or to direct its use. Devolve/Devolution Conferral of power to other levels of government in a manner that gives them substantial autonomy, without the complete surrender of formal control by the center. Dual subordination The use of dual evaluation mechanisms to reinforce accountability for service delivery performance –to a Union ministry and to the Executive Secretary of a State or Region. Equalization Broadly defined as reducing disparities – which may be vertical (between tiers of government), horizontal (across different jurisdictions within the same tier), or based on some other objective (for example, reducing disparities in service delivery opportunities or outcomes). Federalism The governmental or constitutional structure found in a federation. A federation is a political entity characterized by a union of sovereign states or regions which have agreed to operate under a central government. The component states or regions reserve certain powers that may not be exercised by the central government. Grant Transfer of resources from central to sub-national government without expectation of repayment. Hluttaw Myanmar term for Parliament, Union State and Regions. Local Typically defined as the second tier of sub-national government – districts, self- administered divisions and zones, municipalities, townships, et cetera. Own-source revenues Revenues for which the responsibility for setting policy, administering, or both, lie with sub- national government Public service Service provided by government to citizens, such as healthcare, education, waste management, infrastructure, et cetera. Regional Typically defined as the first tier of sub-national government – provinces, states or regions. Sub-national Comprising all tiers of government below the central level. Subsidiarity Principle under which those responsibilities that can be undertaken efficiently by the lowest level of government should be kept at that level – with those activities which cannot be performed at any lower level being moved up. Tax base The good, asset, income stream or investment against which a tax is levied – for example, income, land, or other property. Tax rate The ratio or rate at which a certain base is taxed. Transfer Term encompassing resources which are shifted from the central government to sub- national governments, including grants and contributions. Excludes loans. Unitary A system of governance under which sub-national tiers of government are subsidiary to central government under a single integrated system, typically where local tiers play a much stronger role in public service delivery relative to regional tiers. ix EXECUTIVE SUMMARY: Realigning the Union Budget to Myanmar’s Development Priorities Introduction Fundamental shifts Union Budget policies have undergone fundamental shifts since 2011 to in Union Budget accelerate delivery of essential public services. This first ever Public policies since 2011. Expenditure Review (PER) for Myanmar tries to better understand these shifts and recommend ways to further align budget policies to development priorities. Decades of government underspending on the back of a very low revenue base with no tax culture have contributed to poor economic and social outcomes. A new government in 2010 sought to redress this by implementing reforms towards a service oriented approach. The PER aims to The PER is a first step in deepening policy dialogue on Union Budget issues. It deepen dialogue on focuses on general government, and touches on State Economic Enterprises Union Budget (SEEs) to the extent that they impact general government finances and fiscal policies. policy. The Myanmar PER 2015 is divided into five parts: (i) sustainability of aggregate fiscal policy; (ii) rebalancing the composition of the Union Budget; (iii) improving coverage, quality and equity of education services; (iv) going from more to better government spending on health; and (v) a sound fiscal framework for sub-national service delivery. PER findings and Findings and recommendations from each part will be discussed in greater recommendations detail with government counterparts to help develop a concrete plan of action. could form the basis The PER is expected to be the start of future systematic analysis of Union of a follow up action Budget policies. plan. Sustainability of aggregate fiscal policy Renewed fiscal Myanmar has maintained a relatively prudent fiscal stance, albeit aided by one- pressures as revenue off policies that generated revenue windfalls. As these begin to recede, there windfalls recede. will be renewed pressures on fiscal balances. The PER aims to identify short to medium-term options for strengthening macro-fiscal institutions to monitor and respond to the challenge of sustainably ramping up spending on the back of limited fiscal space. x The size of general government in Myanmar is small compared to other countries. This reflects its historically narrow revenue base (6.3 percent of GDP in 2009) and a limited role in public service delivery (general government spending 7.2 percent of GDP in 2009). Between 2009 and 2014, general government revenue went from 6 to 11 percent of GDP thanks to one-off measures (exchange rate devaluation, telecom license sales) and a widened tax base. General government expenditure nearly doubled to 13 percent of GDP. Public debt is within The overall level of public debt in Myanmar has declined from 77 percent of sustainable GDP in 2007/08 to 47 percent in 2013/14 thanks to arrears clearance. Public thresholds but there sector debt is within sustainability thresholds, but is vulnerable to lower real is a need to look GDP growth and fiscal slippages. more closely at the debt portfolio. Part of this vulnerability stems from the composition of Myanmar’s public debt portfolio, which includes high shares of short-term domestic financing and non- concessional external loans. This calls for greater attention to the composition of deficit financing. Myanmar’s medium-term fiscal outlook is subject to important risks, including falling gas prices. Spending growth in the baseline scenario remains relatively strong as falling gas receipts may be offset by other tax and non-tax receipts. A bigger shock would require significant expenditure adjustments. Closer attention to The PER highlights three areas of fiscal risks that warrant closer attention and fiscal risks is could be managed more actively: the decentralized nature of debt warranted: debt, management; the impact of commodity price volatility on government commodity prices, revenue; and the capacity to ensure fiscal oversight of the SEEs. and SEEs. The PER recommends to: (i) adopt and publish a borrowing plan for the Union Budget, moving eventually to a Medium-Term Debt Strategy; (ii) centralize authority in MoF for negotiation and approval of new loans; (iii) channel all resource rents from Myanmar Oil and Gas Enterprise to the Union Budget, consider fiscal rules and benchmarks, and integrate gas revenue scenarios in the MTFF; and (v) establish a comprehensive framework for fiscal oversight of SEEs. Rebalancing the composition of the Union Budget There was The government’s transition to a service oriented approach has come with significant major rebalancing along and across the Union Budget since 2010/11. Further rebalancing across spending growth needs to be balanced against financing and implementation and along the Union capacity constraints. Budget. xi The PER aims to identify short to medium-term options for strengthening revenue policies, institutions and administration; and to explore the scope for further rebalancing expenditure across the budget to create fiscal space for Myanmar’s development priorities. The revenue effort Overall revenue effort has historically been low. Economic recovery, one-off has been gradually measures, and reforms to expand the revenue base have improved general improving. government revenue from 6 to 11 percent of GDP between 2009 and 2014. Income tax, tax on goods and services and non-tax revenue each make up roughly just under a third each of total revenue. Performance across all types of revenue remains below international comparators, including for customs duties (0.3 percent of GDP). Tax exemptions Tax exemptions for investment promotion are widely applied through a range warrant review and of different legal instruments. There is however no assessment of tax consolidation. expenditures or impact on incremental investment. Two areas to focus on in tax administration are arrears management, and transparency and consolidation of receipts in extra-budgetary “other accounts.” Spending has been There were big changes in the composition of Union Budget spending between rebalanced towards 2009 and 2015. These include the shares of: general public services falling from social sectors. 45 percent to 11 percent; social services increasing from 10 to 33 percent; defense increasing from 21 to 32 percent; and capital falling from 63 to 38 percent. There is a need to Until recently, the authorities in Myanmar repressed the wage bill with limits monitor wage bill on the size of the civil service, salaries, and benefits. During the period under developments. review, the share of the general government wage bill in total recurrent spending was lower than the Lower-Middle and Middle-Income-Countries. The situation is changing. The 2015/16 budget included a near doubling of the wage bill. This may not pose an immediate concern in the medium-term because there may be scope for a reduction in the high share of capital spending. Even so, the authorities should link any further wage increase to a fiscal sustainability analysis and broader public administration reform. The PER recommends to: (i) review and rationalize tax exemptions through a tax expenditure analysis, and consolidate the granting of any new exemptions; (ii) prioritize management of revenue arrears and consider gradual elimination of extra-budgetary ‘other accounts’; (iii) initiate an expenditure review of military spending to identify potential efficiency gains; (iv) conduct a review of public investment management and efficiency, including scope for savings through reduced spending on commercial activities in economic services; (v) conduct a pay and grading exercise to help inform any future changes to the wage bill. xii Improving coverage, quality and equity of education Decades of Many decades of government under-spending on education and prohibitive government household financing requirements have left major gaps in education outcomes. underspending have The government has initiated several reforms since 2012/13 to improve the led to low education situation. outcomes. The PER aims to analyze government expenditures and policies for basic education between 2009/10 and 2013/14, and to propose options for policy reforms and spending to improve coverage, quality and equity of education. Myanmar has suffered from low school enrolment and completion rates (one third of 1.2 million students enrolled in grade 1 made it to grade 11, and only one third of those passed the school leaving exam); poor learning outcomes (9 percent of a third grade class in Yangon cannot read a single word); and inequalities in access and quality (net primary enrolment as low as 69 percent in poorer areas compared to 85 percent average nationally). Reforms are starting Reforms initiated since 2012/13 include elimination of primary and secondary to show results and school fees, introduction of compulsory primary education, hiring 79,000 more education spending teachers, expansion of the stipend program for poor students, and delegation has quadrupled. of some spending authority to district and township education offices and to school head masters through block grants. To implement these policies, the government quadrupled the budget on education between 2011/12 and 2013/14, albeit from a low base. The MoE is now grappling with pressures to deliver more despite: (i) limited administrative, budget and performance data as a basis to prioritize spending; (ii) fledgling capacity to analyze policy and budget linkages; and (iii) fiduciary risks arising out of large cash payments and weak procurement rules. New challenges are In addition, pressures are emerging for other important needs including emerging that need teacher training; adding more years of schooling; ensuring wider coverage to proper assessment include marginalized groups; and Early Childhood Development. Another and costing. emerging issue within education is how much to decentralize and how soon. To assist in policy making and monitoring, the PER recommends: (i) use standardized assessments of student learning outcomes to monitor the quality of the education system. As a starting point, MoE could use the Early Grade Reading Assessments piloted in 2014 in the Yangon region. To ensure that rapidly rising budgets are aligned with implementation capacity, the PER recommends: (ii) MoE and MoF to carry out thorough costing of new initiatives (e.g. more school years, more teacher training, modernized curriculum, coverage to disadvantaged groups, early childhood education) and propose resourcing needs with implementation plans; and (iii) strengthen xiii implementation through more training in procurement and contract management. To gradually delegate more responsibility for day-to-day management of schools, the PER recommends: (iv) to empower district officers or school principals to take decisions within a framework of accountability for results to be developed under the leadership of MoE; and (v) carry out a Public Expenditure Tracking Survey to ensure schools are receiving allocations in accordance with the Ministry’s guidelines (i.e. in accordance with the formula and the time table specified in those guidelines). From more to better government spending on health Historically very low Health services in Myanmar have been inadequate due to historically low levels government of government spending, and ensuing weaknesses in health systems and spending on health regulatory capacity. The government in 2012/13 initiated reforms meant to and prohibitive out- improve the situation. of-pocket payments. The PER aims to analyze government expenditures and policies for healthcare between 2009/10 and 2013/14, and to propose options for further reforms in policies and spending, in particular through more efficient and more equitable government spending. In 2012, government spending on public health was a mere $1.6 per-capita per year, with out-of-pocket payments by households accounting for almost 80 percent of total health spending. This has created serious obstacles to healthcare access, with rural and hard to reach areas most severely affected, and some of the lowest health outcomes among ASEAN countries. Health spending The Union government introduced policies to improve access to and increased nine fold affordability of healthcare services. More medical personnel have been hired. in four years. Beginning in 2012/13, essential drugs and selected health care services are provided free of charge to children, pregnant mothers and for patients needing emergency surgery under certain circumstances in some facilities. There are early signs that these are beginning to help reduce prohibitive out- of-pocket payments by households. They have, however, caused a dramatic increase in expenditure on drugs and medical consumables. Further costs have been incurred due to wage increases and heavy investments in facilities and equipment. These policies have been financed through a nine-fold increase in Ministry of Health spending between 2009/10 and 2013/14. Ongoing challenges include constraints on policy making and monitoring due to unreliable and incomplete data, and high costs due to expenditure inefficiencies. In the medium-term, the government may also want to consider xiv options to increase equity public service provision including more rural primary care and better financial protection policies. Budget adjustments can impact positively To assist in policy making and monitoring, the PER recommends to: (i) improve on efficiency and the coverage and comprehensiveness of data for public and external financing equity of service of the health sector and conduct household budget surveys for private delivery. financing. To reduce expenditure inefficiencies, the PER recommends to: (ii) increase the share of MoH budget resources to preventive care and health interventions with strong public goods characteristics and (iii) formulate a policy for the strategic procurement of pharmaceuticals. To make public services more equitable, the PER recommends to: (iv) formulate a policy on capital spending that better balances rural construction against acquisition of expensive tertiary care equipment; and (v) formulate financial protection policies to help achieve the goal of universal access to health services. Fiscal framework for sub-national service delivery The 2008 Myanmar has embarked on a transition from a centralized to a decentralized Constitution system of government based on its 2008 Constitution. New institutions of sub- provides the basis national governance including State/Region level Parliaments and line for decentralization, departments are slowly being established but, as yet, with limited authority which is moving at a because they coexist with centralized administration and fiscal management. gradual pace. The PER aims to provide an assessment of intergovernmental fiscal relations over the period 2009/10 through 2014/15 and suggests options for prioritizing administrative and fiscal reforms. These are provided within an overall framework of which includes expenditure and revenue responsibilities and fiscal transfers. The 2008 Constitution provides the overarching framework for sub-national administration in Myanmar. Initial steps have been taken to establish legislative and executive bodies at State and Region level. There are functional tensions between newly established sub-national institutions and traditional arrangements of central control. This could lead to confusion over management of and accountability for public services. In addition, accountability for results in public service delivery will be made easier when outcome and output monitoring is systematic and publicly reported. States/regions have little impact on public sector borrowing because their expenditures are budgeted to match expected revenues and Union transfers. Most functional responsibilities in the 2008 Constitution are assigned to Union xv government, with smaller and more localized responsibilities assigned to States/Regions. The sub-national spending share in 2013/14 was therefore small (7 percent of public sector budget). A relatively large portion of State/Region spending is used for capital investment. Per-capita spending by States/Regions is weakly correlated with poverty incidence. Revenue Revenue assignments to States/Regions are very limited (6 percent of public assignments are still sector revenue). Own-source revenues financed 36 percent of total limited. There is expenditures in 2013/14, the remaining 64 percent was financed from Union scope to explore transfers. The distribution of own-source revenues is uneven: 68 percent of increased revenue own-source revenues in 2013/14 were collected in Yangon and Mandalay. The autonomy, linked to burden of State/Region revenue is moderately progressive when comparing spending revenue collection per-capita to poverty incidence. assignments. Intergovernmental fiscal transfers have grown rapidly in the past three years but the lack of a rules-based system for determining general-purpose grants has made them ad hoc and unpredictable on a year-to-year basis. The PER recommends to: (i) strengthen accountability of public administration to new institutions of sub-national governance by clarifying lines of authority of General Administration Department (GAD) officials to locally elected State/Region authorities; (ii) improve accountability for service delivery by introducing a systematic performance monitoring and evaluation framework to be designed and developed under the leadership of GAD in partnership with sector ministries; (iii) develop and adopt a policy on functional and expenditure assignments; (iv) develop and adopt a policy on State/Region revenues with options for potential sub-national revenue autonomy; and (v) adopt a simple formula-based system of intergovernmental fiscal transfers. xvi INTRODUCTION 1 Union Budget policies have undergone fundamental shifts since 2011 to accelerate delivery of essential public services. This first ever Public Expenditure Review for Myanmar tries to better understand these shifts and recommend ways to further align budget policies to development priorities. Decades of government underspending on the back of a very low revenue base with no tax culture have contributed to poor economic and social outcomes. A new government in 2010 sought to redress this by implementing reforms towards a service oriented approach. Efforts have been made to strengthen institutions of Public Finance Management. Economic controls have been gradually liberalized, steps to delegate fiscal responsibilities to sub-national authorities have been taken, and reforms to encourage private sector investment are under way. The PER is a first step in deepening policy dialogue on Union Budget issues. It focuses on general government, and touches on SEEs to the extent that the impact general government finances and fiscal policy. The Myanmar PER 2015 is divided into five parts: (i) sustainability of aggregate fiscal policy; (ii) rebalancing the composition of the Union Budget; (iii) improving coverage, quality and equity of education services; (iv) going from more to better government spending on health; and (v) a sound fiscal framework for sub-national service delivery. Findings and recommendations from each part will be discussed in greater detail with government counterparts to help develop a concrete plan of action. The PER is expected to be the start of future systematic analysis of Union Budget policies. 1 CHAPTER 1 INTRODUCTION A. Background and Objectives 1.1. The Myanmar Public Expenditure Review (2015) aims to take stock of the fundamental shifts in Union Budget policies since 2011, and recommend ways to further align those policies with the country’s enormous development needs. The Myanmar PER analyzes general government finances from 2009/10 to 2013/14, a period in which government spending almost doubled and was refocused on addressing major service delivery gaps. This marked the start of a long-term process to transform the Union Budget into an instrument for growth and poverty reduction. Initial efforts at increasing budget transparency have already fueled debate and demand for results in the executive, the legislature and the public at large. 1.2. The objectives of the Myanmar PER (2015) are to: (i) provide empirical analysis of, and establish a baseline on, the alignment of the Union Budget with the country’s development priorities; (ii) help prioritize reform of Union Budget policies and institutions to promote fiscal discipline, allocative efficiency, and value for money; (iii) offer a basis for future systematic analysis of Union Budget policies and thereby enhance transparency of and accountability for public service delivery. The findings and recommendations of this first PER should help to: further deepen policy dialogue between government and development partners; inform the provision of technical assistance to help implement recommendations; and enable stakeholders to monitor progress. B. Economic Context 1.3. Myanmar has been characterized by a centrally controlled government and economy. The country’s development has been stymied for decades by violent conflict. In an attempt to impose stability and promote development, the military took over power through a coup in 1962 and introduced the Burmese Way to Socialism, which became a blue print for Myanmar’s economic policy from 1962 to 1988. The framework was characterized by central planning and control of many aspects of the economy, an isolationist policy that sought to reduce foreign influence, and an increased role of the military in the economy including through ownership of companies. 1.4. These policies resulted in poor outcomes on macroeconomic stability and public service delivery. Infrastructure services were far below those in most poor and lower-middle-income countries. Health and education outcome indicators declined, as households were burdened with health care and education costs. Families were typically required to pay for the cost of medicines and materials needed for surgical procedures and had to pay for uniforms, writing supplies, and exercise books for their children’s schooling. Health facility and school usage was below potential because of these high “out-of- pocket” costs. Social protection programs for households falling into poverty were almost absent. 1.5. A new government elected in 2010 fundamentally reoriented policies towards public service delivery. The December 2012 Framework for Economic and Social Reform (FESR) set out a number of policy reforms that committed the government to achieving people-centered development, civic participation and human resource development, effective and transparent use of public financial resources, sustainable regional development, decentralization and greater autonomy for local government, and poverty reduction. 2 1.6. Economic controls were gradually liberalized with major impact on public finances. One of the most distorting elements of prior economic controls was the use of a highly overvalued official exchange rate, which created substantial resource misallocations in the economy. According to one estimate, the total efficiency loss caused by the multiple exchange rate regime was estimated at about 14–17 percent of GDP in 2006/07.1 The over-valued official exchange rate was unified under a managed float in 2012, thus effectively bringing the official rate to near parity with the market determined rate used by the rest of the economy. This had a major positive impact on government revenue thanks to revaluation of Kyat denominated receipts. 1.7. Policies on intergovernmental relations also began to evolve. Parliaments for States/Regions have been created and elections were held in 2010. According to the 2008 Constitution, States and Regions are assigned specific but limited responsibilities for public service delivery, particularly for the small and medium-scale infrastructure in their areas. The new government has made it clear that it intends to implement this change and has issued several edicts to that end. Even so, the Union continues to maintain a high degree of control over the fiscal policies of States/Regions. Moreover, the 2008 Constitution continues to assign to the Union responsibilities for providing most public services, including health and education, as well as large scale infrastructure. 1.8. To help encourage the private sector, the FESR commits the government to corporatize or privatize most SEEs. In 2009/10, the Union had ownership of 50 formally defined SEEs and the states and regions own still more. Many of these, in turn, operate multiple factories or other facilities in industries that do not warrant public sector involvement. SEEs also benefit from implicit and explicit government subsidies, which have restricted competition and private sector growth. C. Scope of Public Sector 1.9. A description of the scope of the public sector is important to establish the unit of analysis in this PER. Until 2011, the Union government controlled all aspects of the public sector, which included: central ministries and departments; sub-national authorities (States/Regions, cantons, development committees); and SEEs, most of which fall under the ownership of ministries and departments (Figure 1.1). The PER is focused primarily on the general government part of the public sector, which in a unitary state like Myanmar includes central and sub-national authorities that are engaged in non-market, non- production public service delivery. It excludes SEEs, which even though in Myanmar are part of general government ministries, are still engaged in either market or other production-based activities. 1.10. The scope of Union government coverage changed slightly during the period of preparation of this PER. With effect from November 2010, each State and Region has its own parliament that adopts its own annual budgets, subject to review by a Finance Committee led by one of the Vice Presidents, the Minister of Finance and selected ministers. Thus, in theory, the Union government is now equivalent to the central government. In practice, however, the concepts of central and general government in Myanmar were essentially equivalent for the period under review because of limits on any actual devolution of authority to States and Regions. For the purposes of this review, “Union” refers to general government, excluding SEEs, while the “public sector” refers to the broader definition of government including the SEEs. 1 Hori and Wong (2008), “Efficiency Costs of Myanmar’s Multiple Exchange Rate Regime,” IMF Working Paper, WP/08/199. 3 Figure 1.1: Myanmar Public Sector 1.11. The PER covers SEEs to the extent that they affect general government finances. In the case of Myanmar, during the period under review, the SEEs were funded through the Union Budget. All SEE transactions flow into and out of the same Union Fund Account (also referred to as the State Fund Account) used by the Union ministries as well as sub-national authorities. Analysis of macro-fiscal policies needs to take account of the links between SEEs and the general government (e.g. transfers to and from SEEs, contingent liabilities). This is covered in chapter 2, which looks at strengthening fiscal oversight of SEEs. But in fiscal accounting, SEEs that are market operators should be treated as entities with separate balance sheets.2 This is usually formalized through corporatization processes, which the government has already begun to explore for selected SEEs. 1.12. There are two other topics related to the comprehensiveness of public sector accounts and their impact on general government finances and macro-fiscal outcomes. These are: (i) extra-budgetary accounts; and (ii) the operations of two off-budget military holding companies. Extra-budgetary accounts, referred to as “other accounts,” are part of the consolidated Union Fund Account (together with Ministry and Development Accounts and SEE Accounts) but are not part of the Union Budget. There is little oversight or reporting on these accounts. They are touched on briefly in chapters 2 and 3, but are not analyzed in the PER. The two military holding companies (Union of Myanmar Economic Holdings Limited and Myanmar Economic Corporation) have wide-ranging economic interests. UMEHL for example has 38 wholly owned businesses and shares in 9 joint ventures with foreign companies, and employs about 14,000 workers. The conglomerate has a monopoly on the country's gems sector and also has a significant portfolio in various industries including banking, tourism, real estate, transportation, and metals. There is little reporting or publicly available information on these companies, and therefore they are not covered in this PER. 2 For more information, please refer to section 2.C of the IMF 2001 Government Financial Statistics Manual. 4 D. Scope and Outline 1.13. The Myanmar PER 2015 is divided into five chapters: (i) sustainability of aggregate fiscal policy; (ii) rebalancing the composition of the Union Budget; (iii) improving coverage, quality and equity of education services; (iv) going from more to better government spending on health; and (v) a sound fiscal framework for sub-national service delivery. 1.14. Chapter 2 on sustainability of aggregate fiscal policy aims to build on the government’s ongoing efforts to strengthen macro-fiscal institutions for monitoring and responding to the challenge of ramping up spending on the back of limited fiscal space. It starts by providing an overview of aggregate fiscal developments in Myanmar and public debt issues, including debt levels and sustainability, and deficit financing policies. It then looks at possible scenarios for Myanmar’s medium-term fiscal outlook, and the implications these may have for fiscal aggregates in the Union Budget. It ends with a discussion on options for managing fiscal risks that may arise from: the decentralized nature of debt management; the impact of commodity price volatility on government revenue; and the capacity to ensure fiscal oversight of State Economic Enterprises (SEEs). 1.15. Chapter 3 on rebalancing the composition of the Union Budget helps to take stock of major shifts in the Union Budget since 2009/10 and highlights possible measures that can further enhance the alignment of budget allocations with policy priorities. It reviews the current revenue effort and composition in Myanmar, and recommends short to medium-term options for strengthening revenue policies, institutions and administration so as to enhance collections in an efficient and equitable manner. It analyzes spending shifts across the Union Budget, as government tries to temper expenditure growth in line with financing and implementation capacity. The PER reviews the current functional and economic composition Union Budget expenditures to explore ways to create fiscal space for development priorities. 1.16. Chapter 4 on improving coverage, quality and equity of education services aims to assist in further prioritizing spending and policy reforms in the education sector. It focuses on basic education, which accounts for around two thirds of government spending on education as a whole. It provides an overview of how the sector is organized for the regulation and delivery of education services and highlights some of the big gaps in education outcomes. It looks at the government’s recent strategy to fill education gaps, and proposes measures that could further strengthen that strategy. It then analyzes how government spending has responded to the new strategy and suggests options for further prioritization. It ends by reviewing the current arrangements for budget preparation, implementation and reporting in the sector, with recommendations to manage the rapidly growing inflows of resources in the sector. 1.17. Chapter 5 on going from more to better government spending on health aims to assist in prioritizing spending and policy reforms to improve health outcomes, particularly through better targeted, more efficient and more equitable government spending. It starts with an overview of the health sector, which unlike education, has a strong private sector presence. It provides an overview of health challenges in Myanmar, how these have informed government priorities in recent years, and the implications these have had for health sector financing. It then looks more deeply at government spending on health and recommends possible changes to allocations that could improve the efficiency and equity of that spending. 1.18. Chapter 6 on a sound fiscal framework for sub-national service delivery looks at options for prioritizing administrative and fiscal reforms to inform ongoing discussions on the overall framework of 5 intergovernmental relations. It reviews the current arrangements for sub-national administration in Myanmar, including the structure of government and key features of public administration. It then analyzes the main pillars of fiscal decentralization, namely: what government functions and expenditures are assigned to whom; what local receipts help to fund those assignments; and on what basis are Union transfers helping to close fiscal gaps. In each area, the PER proposes options to strengthen current policies and institutions, and discusses their coordination and sequencing. It ends with a brief review of sub- national budget preparation and implementation. E. Approach and Methodology 1.19. The Myanmar PER is based on data covering the years 2009/10 through 2013/14 made available to the World Bank and its development partners in May and September 2013. Government expenditures are reported in a system based on paper ledgers that is difficult to analyze in detail on a timely basis. For purposes of this review, data in the ledgers was converted to an electronic database (BOOST).3 Aside from this, the review uses a mix of data sources. Chapter 2 utilizes aggregated data in various IMF reports published through 2014, complemented by World Bank staff estimates. The analysis in chapters 3-5 examines Union and SEE operations individually using the BOOST database. The BOOST database produces results highly similar to, but not the same as, those provided by the IMF. Importantly, for the years 2012/13 and 2013/14, the BOOST database numbers are revised estimates and budget estimates while the IMF data are preliminary outcomes and revised estimates respectively. Chapter 6 relies on state and region data provided by the Ministry of Finance. There is some uncertainty about the accuracy of other data, such as GDP, population, and poverty rates. Specific issues are summarized in Box 1.1 as well as Annex 1. 1.20. The PER was developed in very close collaboration and consultation with the Ministry of Finance, the Ministry of Education (MoE), the Ministry of Health (MoH), the Ministry of Home Affairs (MoHA) and other government agencies and departments. The findings and recommendations of the PER have been part of ongoing policy dialogue and will be discussed in greater detail to help develop a concrete plan of action. The PER is expected to be the start of future systematic analysis of Union Budget policies. 3 See Kheyfets, Mastruzzi, Merotto and Sondergaard (2011) for more details on BOOST. 6 Box 1.1: Data Quality Developing quality national data is one of the four priority areas of the government’s economic plan. Limited capacity and funding over the years eroded the country’s statistical database, while the use of some data became highly politicized. These constraints have caused many debates. Although an increasing amount of information is becoming available via government websites, a large portion is not published in a timely manner. Some of the key issues are summarized below. Poverty: Two Integrated Household Living Conditions Assessments (IHLCA) were conducted, one in 2005 and one in 2010. There are two weaknesses inherent in these surveys. First, given that the last population census was conducted in 1983, the sampling frame used for the survey may not be representative of the population. Second, due to conflict, the survey did not cover some of the border areas. These weaknesses have called into question the extent to which the survey data was representative of the country. Some of the results appear to be inconsistent with available data regarding public service delivery and infrastructure. It is hoped that the results of the 2014 population census will provide a more up to date sampling frame for the next full round of IHLCA, currently planned for 2015. National accounts: Historical data show high real growth rates that are inconsistent with economic fundamentals and available data from key sectors. Public sector activities were recorded at deeply over-valued exchange rates. Growth rates from recent years appear to be more realistic and consistent with other indicators but there remains considerable uncertainty about the actual level of GDP. In particular, if the historical growth rates were incorrect, then all subsequent estimates based on those growth rates will also be incorrect. The IMF (2014) Assessment of Data Adequacy for Surveillance observes several additional short-comings but also notes good progress in improving the quality of new estimates. Population: The last published census was conducted in 1983. Provisional results from the March 2014 census were made available in August 2014. Final results had not been published at the time of this public expenditure review. Population estimates range from 48 to 61 million, with an estimated 3 million people living abroad. Uncertainty about the true population size creates doubts about estimates of GDP per capita. Trade: There are substantial flows of illicit drugs, jade and timber from the border areas which is difficult to capture in officially produced export statistics. Merchandise imports are under-estimated: known omissions include military imports, imports linked to FDI under joint venture agreements, and some imports linked to exemptions from custom duties. Trade data are recorded at the time of entries by customs, causing serious volatility in values and incorrect time records. As in many countries, data recorded by the authorities and data recorded by trading partners are inconsistent. Recent data are, however, fairly well correlated over time. Balance of payments: According to the IMF Assessment of Data Adequacy for Surveillance, detailed data on services transactions and financial flows are generally not available and transactions that are not undertaken through the official banking system are usually not estimated. External debt: The evaluation of external obligations which are not nominated in U.S. dollars is conducted irregularly. Historical data are distorted by applying the exchange rate at the evaluation point. Consumer price statistics: Consumption basket weights represent urban households only. Some construction inputs are included; rentals of owner occupied housing are excluded; missing prices are not imputed; and the classification of items is outdated. Fiscal data: There is no comprehensive monthly or quarterly compilation of fiscal data. Annual comprehensive data are compiled with delays of up to 12 months after the end of the reference year. In addition, only consolidated data for state economic enterprises are available, and some transactions are recorded partly on an accrual basis and partly on a cash basis. It is difficult to know the true size of consolidated government because the full array of transactions between the Union government and its state economic enterprises are not well identified. Fiscal and monetary data are not consistent. The recording of government debt statistics is not comprehensive. Monetary and financial statistics. The monetary survey compiled by the CBM covers the central bank and all commercial banks (public and private). Reporting of monetary data in accord with IMF classification principles was established only by January 2012. Prior data may have problems associated with the use of the previously overvalued official exchange rate for valuing foreign currency-denominated balance sheet accounts. Sources: IMF (2014) Assessment of Data Adequacy for Surveillance and World Bank staff assessments. 7 SUSTAINABILITY OF 2 AGGREGATE FISCAL POLICY Background: Myanmar has maintained a relatively prudent fiscal stance, albeit aided by one-off policies that generated revenue windfalls. As these begin to recede, there are renewed fiscal pressures on fiscal balances. Objectives: To identify short to medium-term options for strengthening macro-fiscal institutions to monitor and respond to the challenge of sustainably ramping up spending on the back of limited fiscal space. Key findings: The size of general government in Myanmar is small compared to other countries. This reflects its historically narrow revenue base (6.3 percent of GDP in 2009) and limited role in public service delivery (general government spending 7.2 percent of GDP in 2009). Between 2009 and 2014, revenue went from 6 to 11 percent of GDP thanks to one-off measures (exchange rate devaluation, telecom license sales) and a widened tax base. Expenditure nearly doubled to 13 percent of GDP. The overall level of public debt in Myanmar has declined from 77 percent of GDP in 2007/08 to 47 percent in 2013/14 thanks to arrears clearance. Public sector debt is within sustainability thresholds, but is vulnerable to lower real GDP growth and fiscal slippages. Part of this vulnerability stems from the composition of Myanmar’s public debt portfolio: non-concessional external loans, and half of the portfolio in mostly domestic short-term financing instruments. This calls for greater attention to the composition of deficit financing. Myanmar’s medium-term fiscal outlook is subject to important risks, including falling gas prices. Spending growth in the baseline scenario remains relatively strong as falling gas receipts may be offset by other tax and non-tax receipts. But a bigger shock would require significant expenditure adjustments. The PER highlights three areas of fiscal risks that warrant closer attention and could be managed more actively: the decentralized nature of debt management, which leads to fragmented decisions on debt and increases the risk to the debt portfolio; the impact of commodity price volatility on government revenue; and the capacity to ensure fiscal oversight of State Economic Enterprises (SEEs). Recommended options: (i) adopt and publish a borrowing plan for the Union Budget, moving eventually to a Medium-Term Debt Strategy; (ii) centralize authority in MoF for negotiation and approval of new loans; (iii) channel all resource rents from Myanmar Oil and Gas Enterprise to the Union Budget, consider fiscal rules and benchmarks, and integrate gas revenue scenarios in the MTFF; and (v) establish a comprehensive framework for fiscal oversight of SEEs. 8 CHAPTER 2 SUSTAINABILITY OF AGGREGATE FISCAL POLICY A. Background 2.1. The Union government has maintained a moderately prudent fiscal stance between 2009/10 and 2013/14, with deficits below 5 percent of GDP, albeit aided by one-off policies on debt arrears clearance, exchange rate devaluation, and telecom and banking licensing. As these windfalls begin to recede, there are renewed pressures on fiscal balances. Greater access to external concessional finance and policies to develop domestic debt markets are gradually helping to reduce monetization of the fiscal deficit, which nonetheless remains high at 50 percent of net domestic financing. These forces will pose important challenges for the government’s emerging Medium-Term Fiscal Framework. 2.2. The PER aims to build on ongoing efforts to strengthen macro-fiscal institutions for monitoring and responding to these challenges. It provides an overview of aggregate fiscal developments in Myanmar and public debt issues, including debt levels and sustainability, and deficit financing policies. It then looks at possible scenarios for Myanmar’s medium-term fiscal outlook, and the implications these may have for fiscal aggregates in the Union Budget. It ends with a discussion on options for managing fiscal risks that may arise from: the decentralized nature of debt management; the impact of commodity price volatility on government revenue; and the capacity to ensure fiscal oversight of State Economic Enterprises (SEEs). Figure 2.1: Size of General Government and Growth, 2009-12 40 Government revenue (% of GDP, 09-12 avg) 35 Bubble size reflects 30 growth performance Vietnam (2009-2012 average) 25 20 Lao PDR Ghana 15 Averages 10 5 Myanmar Nigeria Government spending (% of GDP, 09-12 avg) 0 5 10 15 20 25 30 35 40 Sources: World Development Indicators and World Bank staff estimates. 9 Table 2.1: Fiscal Accounts, as Shares of GDP, 2009/10 – 2013/14 Prelim. Estimated 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 General (Union) Revenues and Grants 6.4 6.3 5.8 6.3 6.5 9.6 11.6 A. Revenues 6.4 6.3 5.8 6.3 6.5 9.5 11.4 1. Tax Receipts 3.3 3.1 3.1 3.3 3.9 7.1 7.2 Income Tax Receipts from SEEs 0.9 0.9 0.8 0.7 0.9 1.6 1.5 Commercial Tax Receipts from SEEs .. .. 0.6 0.4 0.4 1.2 1.1 Other Tax Receipts .. .. 1.6 2.1 2.6 4.3 4.5 2. Non-tax 3.1 3.2 2.7 3.0 2.6 2.5 4.2 Contribution Receipts from SEEs 2.5 2.5 2.3 2.0 2.3 1.6 1.5 Paid by Private Sector a/ 0.5 0.7 0.5 1.1 0.4 0.9 2.7 B. Grants, Recurrent and Capital 0.0 0.0 0.0 0.0 0.0 0.1 0.2 Expenditures 8.0 7.2 9.1 11.0 10.4 13.5 14.3 A. Recurrent 3.0 3.0 3.3 3.9 4.5 7.4 8.7 1. Interest Payments 0.5 0.6 0.8 0.9 1.1 1.4 1.5 2. Primary Recurrent 2.5 2.4 2.6 3.0 3.4 5.9 7.2 B. Capital 5.0 4.2 5.8 7.1 5.9 6.1 5.6 Overall Balance -1.6 -0.8 -3.3 -4.6 -3.9 -3.9 -2.7 plus SEE Balance b/ -1.6 -1.6 -1.7 -0.8 -0.7 2.2 1.1 plus Discrepancy .. .. .. .. .. -1.8 .. equals Consolidated Overall Balance -3.3 -2.4 -4.9 -5.4 -4.6 -3.4 -1.6 Primary Balance -1.1 -0.2 -2.5 -3.7 -2.8 -2.5 -1.2 Recurrent Balance 3.4 3.4 2.5 2.5 2.0 2.2 2.9 Net Financing 3.3 2.4 4.9 5.4 4.6 3.4 1.6 A. Foreign financing 0.0 0.0 0.0 0.0 0.0 1.2 1.0 B. Domestic financing 3.2 2.4 4.9 5.4 4.6 2.2 0.6 1. Central bank credit 2.4 2.0 3.4 3.5 1.8 1.3 0.8 2. Bank financing 0.3 1.1 1.3 1.1 1.5 0.9 -0.3 3. Other 0.6 -0.7 0.2 0.8 1.3 0.0 0.2 Total Public Debt Outstanding (% GDP) c/ 77 58 57 50 50 47 40 A. Domestic 16 17 20 21 23 23 21 B. External 61 41 37 29 27 25 19 o/w Arrears 41 28 26 20 19 9 0 Union Interest Paid as % Revenue and Grants 8.5 9.0 13.0 14.2 16.8 14.9 13.1 Prices: Market exchange rate (MMK/US$) d/ 1,110 992 1,004 861 824 864 959 Consumer Price Inflation Rate, Period Avg (%) 32.9 22.5 2.2 8.2 2.8 2.8 5.7 GDP Deflator Growth (%) 23.6 13.6 4.9 8.2 2.8 2.8 5.7 Interest Rate, Three Month Treasury Bills (%) 4.0 4.0 4.0 4.0 4.0 4.0 4.0 Inferred Average Nominal Interest Rate (%) e/ 4.6 4.6 5.4 5.4 5.6 5.2 6.2 a. Receipts in 2013/14 include one-off revenues from the auctioning of telecommunication licenses. b. Includes the Central Bank of Myanmar, the Myanmar Economic Bank, the Myanmar Foreign Trade Bank and three other banking institutions. c. Excludes any debt owed by the state economic enterprises. No SEE debt data were available at the time of the review. d. Period averages for fiscal years. e. Calculated from annual interest paid divided by the stock of domestic debt in each previous year. Source: IMF Country Reports No. 13/250, 14/91 and 14/307 and World Bank staff calculations. 10 B. Aggregate Fiscal Developments 2.3. The size of general government in Myanmar is relatively small compared to other countries (Figure 2.1). This reflects the government’s historically narrow revenue base on the one hand and limited role in public service delivery on the other. General government can play a pivotal role in Myanmar’s growth prospects by improving public services, infrastructure and the enabling environment for a healthy and productive population, and private investment. Significant spending needs have to be balanced with fiscal constraints. Strong real GDP growth, averaging around 7 percent between 2010/11 and 2013/14, was thanks to economic rebound and natural resource exports. This offers important scope to enhance fiscal space, which in turn can be used to eliminate poverty and promote shared prosperity. 2.4. The fiscal space needed for a near doubling in aggregate spending in Myanmar from 7.2 percent of GDP in 2008/09 to 13.5 percent in 2012/13 (Table 2.1 and Figure 2.2) was in part afforded by a jump in government revenue (Figure 2.3). While low relative to international standards, aggregate revenue went from 6.3 percent of GDP to 9.6 percent. This was largely thanks to the devaluation of the exchange rate, leading to a more accurate accounting of Kyat denominated receipts from gas exports. Receipts from the sale of telecom and banking sector licenses in 2014/15 contributed to a further 2 percent of GDP in receipts. Figure 2.2: Government Expenditure, 2009-13 Figure 2.3: Government Revenue, 2009-13 35 Percent of GDP (%) 30 Percent of GDP (%) 30 25 2009 2010 2011 25 2009 2010 2011 2012 2013 20 2012 2013 20 15 15 13.5 9.6 10 10 7.2 6.3 5 5 0 0 Georgia Vietnam Sri Lanka Pakistan Philippines Lao PDR Nigeria Myanmar Malaysia Thailand Vietnam Lao P.D.R. Philippines Indonesia Cambodia Myanmar Sources: World Development Indicators and World Bank staff estimates. 2.5. The general government deficit over this period expanded rapidly from 0.8 percent of GDP in 2008/09 to 4.6 percent in 2010/11 (Figure 2.4). This was driven by capital spending, partly destined for Nay Pyi Taw, but also due to rising interest payments, which went from 0.6 percent of GDP to 1.1 percent between 2008/09 and 2011/12. This was followed by a period of consolidation, as reflected by the growing gap between the overall and primary deficits (Figure 2.4). A notable achievement is that despite the rapid increase in spending, the government has maintained a recurrent surplus (i.e. recurrent revenue covering recurrent expenditure). Maintaining a positive current balance (i.e. no borrowing for recurrent spending) will be an important benchmark for fiscal sustainability going forward. 11 Figure 2.4: Fiscal Balances Figure 2.5: Overall Fiscal Balance 3 4 Union Government 2 SEEs 3.4 3.4 1 2 2.9 Consolidated Public Sector 2.5 2.5 Percent of GDP (%) Percent of GDP (%) 2 2.2 0 -1.1 -0.2 -2.5 -1.2 0 -1 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 -2 -2 -3 -1.6 -1.6 -0.8 -2.4 -3.7 -2.5 -2.8 -2.8 -4 -3.0 -3.3 -4 -2.7 -3.6 -3.4 -3.3 -4.0 -3.9 -3.9 -5 -4.6 -4.9 -5.4 -4.6 -4.6 -6 -6 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 Overall balance Primary balance Recurrent balance Sources: IMF and World Bank staff estimates. 2.6. Under-estimation of revenues has affected aggregate fiscal discipline. The ministries used larger than expected revenues to finance unplanned expenditures. Revenues were on average under-estimated by 1.3 percent of GDP between 2009/10 and 2010/11 (Table 2.2). Over-realized revenues financed extra, unbudgeted (mostly capital) spending. The data show, however, that the deviations in expenditures exceeded the revenue deviations so that the Union typically saw net borrowing exceed budgetary targets by no less than 0.9 percent of GDP in 2012/13 and as much as 2.2 percent of GDP in 2009/10. Table 2.2: Budgetary Deviations, as Percent of GDP a/ 2009/10 2010/11 2011/12 2012/13 Union -2.2 -1.8 -1.0 -0.9 Revenue 1.0 1.8 1.6 0.9 Expenditures 3.2 3.6 2.6 1.8 Recurrent Expenditures 0.5 0.2 0.3 0.7 Capital Investment 2.7 3.4 2.3 1.1 a. Excludes SEEs. Sources: MoF BOOST database and World Bank staff calculations. 2.7. The government has instituted a process of supplementary budgets to address revenue and policy uncertainty and is reviewing rules for changes to budget appropriations (i.e. allocations approved by parliament). More work is needed. The supplementary budget for example simply formalizes changes to budget appropriations that have already been made. The Union Budget Law has a clear appropriation structure. Though this could be complemented by further guidance on authority for approving changes to budget appropriations. Some countries allow reallocations from one category of spending to another up to a certain ceiling without having to seek Ministry of Finance or Parliament approval. However, it is advisable that any change to the overall level of government spending through for example over-realized revenue, be reviewed and approved by the legislature as this has direct implications for financing decisions. C. Public Debt 2.8. The overall level of public debt in Myanmar has declined rapidly from 77 percent of GDP in 2007/08 to an estimated 47 percent in 2013/14. By 2007/08, cumulative external borrowing by the 12 government was US$12.3 billion or approximately 61 percent of GDP (Figure 2.6). Two-thirds of that was attributable to debt service payments in arrears, most of which were owed to bilateral creditors. 2.9. External arrears were eliminated between 2012/13 and 2013/14 (Figure 2.6). In April 2012, Myanmar agreed with Japan, its largest creditor, on a debt restructuring plan to resolve its arrears through a bridge loan operation and cancelation of principal payments and overdue charges. Myanmar also resolved its arrears held by the World Bank and the Asian Development Bank in January 2013, with the help of a bridge financing operation by Japan. In late-January 2013, the Paris Club agreed to write off 50 percent of all arrears and reschedule the remaining arrears over 15 years with a seven-year grace period. The result is a sharp reduction in Myanmar’s external obligations, from 50 percent of GDP in 2011/12 to 40 percent of GDP by 2013/14. (Figure 2.7). Figure 2.6: Public Debt in Myanmar (Percent of GDP) Figure 2.7: Public Debt (Percent of GDP) 80 105 70 108 Domestic External External arrears 107 95 104 60 100 41 85 94 50 80 28 26 75 9 77 40 20 19 58 65 16 57 50 50 30 19 55 47 20 11 9 8 13 20 45 10 20 21 23 23 21 35 16 17 0 25 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2008 2009 2010 2011 2012 2013 Advanced Asia LAC MNA Myanmar Source: World Development Indicators. Sources: IMF and World Bank staff estimates. 2.10. Debt Sustainability Analysis by the IMF and the World Bank concludes that Myanmar is at low risk of debt distress following the above arrears clearance operations.4 All indicators of external debt solvency and liquidity remain below their indicative sustainability thresholds over the projection period, both under the baseline scenario and alternative scenarios and stress tests.5 Total public sector debt (i.e. external and domestic public debt) also remains within sustainability thresholds, but is vulnerable to lower real GDP growth and fiscal slippages. Part of this vulnerability stems from the composition of Myanmar’s public debt portfolio. Unlike other Low Income Countries, less than 20 percent of external debt is in the form of concessional loans from multilateral organizations. Over half of public debt is in the form of domestic, mostly short-term, financing instruments. With arrears clearance, Myanmar is regaining access to concessional resources, and should reduce its reliance on non-concessional financing. This, however, will require active debt management to ensure that short to medium-term deficit financing choices are consistent with longer-term objectives of ensuring debt sustainability. 4 IMF (2014), “Staff Report for the 2014 Article IV Consultation – Debt Sustainability Analysis.” 5 Real GDP growth assumed at 8.1 percent over the medium-term and 6.6 percent over the long-term; overall fiscal balance at -4.6 percent over the medium-term and -3.8 percent over the long-term; non-interest current account balance at -4.7 percent over the medium-term and -3.3 percent over the long-term. 13 2.11. Budget deficit financing has come from four main sources: (i) short-term Treasury Bill subscriptions by the Central Bank (main source of monetization; 3-month maturity, 4 percent interest); (ii) recently instituted Treasury Bill auctions (3-month maturity, market rate of 8 percent in recent auctions); (iii) medium-term treasury bond subscriptions by local banks (2, 3 and 5-year maturity); and (iv) foreign loans. The government has historically relied very heavily on monetization (Figure 2.8), due to lack of domestic debt markets and limited access to external loans. This fuelled double digit inflation for most years between 1988 and 2010 (peaking at 57 percent in 2002) and contributed to high debt servicing costs discussed earlier. Reducing the monetization of the budget deficit has therefore been a major priority of the government. 2.12. A number of measures are being taken to diversify financing sources. Firstly, the government has tried to rebalance the financing mix away from short-term Treasury Bill subscriptions by the Central Bank towards medium-term Treasury Bond sales to institutional investors. Secondly, starting January this year, the government conducted its first Treasury Bill auction, and is expected to conduct Treasury Bond auctions in January 2016. Although uptake of Treasury Bills has been lower than expected, this is a very important step in the long road ahead for developing domestic debt markets. Thirdly, the government is gradually accessing more external debt on concessional terms. This should help to rebalance the financing mix closer to the norm for Low-Income and Lower Middle-Income Countries (Figure 2.9). Financing from the Central Bank of Myanmar (CBM) is estimated to have declined from 3.5 percent of GDP in 2010/11 to 1.8 percent of GDP in 2011/12 because of some of these measures. Although this is expected to gradually reduce monetization, eliminating it completely is likely to take some time. Figure 2.8: Deficit Financing Figure 2.9: Deficit Financing in Peer Groups 3.5 (2007-14 average or latest available) 3.0 6 Foreign Loans Domestic Banks CBM Other 2.5 Percent of GDP (%) 2.0 3.3 4 1.0 1.5 1.2 4.5 3.5 1.4 4.6 4.4 1.8 2 3.6 3.4 2.4 3.4 1.0 2.0 0.5 1.2 1.3 0.8 1.0 0.7 0 0.3 0.0 Lower middle Middle income Upper middle Myanmar income income -2 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 Net domestic financing (% of GDP) Net external financing (% of GDP) Sources: IMF and World Bank staff estimates. Source: World Development Indicators. 2.13. Ensuring a sound financing mix for a balanced debt portfolio can be informed by an annual borrowing plan, first, and then gradually moving to a Medium-Term Debt Strategy. Myanmar currently has annual borrowing limits set out in the Union Budget Law. These limits, however, are set in absolute terms with no breakdown of the actual mix of debt instruments. For example the 2014/15 Budget Law set the gross borrowing limit at Kyat 3.6 trillion, going to Kyat 3.9 trillion in the 2015/16 Budget Law. There is currently no analysis of how the costs and risks of the government’s debt portfolio is affected by the proposed new borrowing. The Parliament approves external loans (usually associated with specific projects), which form part of this limit; domestic borrowing is the residual. Although there is a Treasury Bill auction calendar, given the limited scope of these auctions and uncertainty over uptake (including for 14 Treasury Bond subscriptions), Central Bank purchase of Treasury Bills make up for outstanding financing needs. 2.14. A borrowing plan or Debt Management Strategy would help implement the government’s objective of diversifying financing sources. The plan would help assess the overall cost of the debt portfolio, and potential risks (i.e. changes to costs) arising out of different factors including interest rate changes, exchange rate movements, refinancing needs, or other. In doing so, it can inform the government’s decisions on the most appropriate financing strategy given its borrowing constraints. Without systematically going through such an assessment and analysis, the government may end up with a costly debt portfolio that is highly vulnerable to economic shocks, as noted in the discussion on debt sustainability. During the period under review, for example, nearly half of the government’s debt portfolio is in short-term debt, which is subject to constant rollover and is not matched to gestation periods of public investments. 2.15. The government can prepare a borrowing plan and MTDS with joint Bank-Fund technical assistance. This assistance would look at: setting debt management objectives; analyzing the cost and risk of existing debt; identifying and analyzing potential funding sources; baseline projections including policy risks; reviewing long-term structural factors; assessing and ranking alternative borrowing strategies base on the basis of cost-risk trade-off; reviewing proposed strategies with fiscal and monetary policy authorities; and finally agreeing the borrowing plan.6 The Debt Management Office in the Ministry of Finance would lead this exercise, coordinating closely with the Central Bank and the Ministry of Planning in particular. D. Medium-Term Fiscal Framework 2.16. Myanmar’s medium term economic outlook through 2020/21 is strong, though with important downside risks. Growth is driven by investment and commodity exports, encouraged by trade liberalization, the suspension of sanctions and gas production. Expansion in the tourism and telecommunication services also holds up the outlook. Inflation is expected to trend down with the discontinuation of fiscal deficit monetization and a strengthened monetary policy framework. 2.17. The baseline medium term fiscal forecast is for real GDP growth to peak at 8.5 percent in 2015/16 following several years of accelerating growth (Table 2.3). The real growth rate is forecast to gradually taper down to a steady-state rate of 7.5 percent per annum by 2020/21.7 This is expected in part because of slower than expected progress in the large gas export projects with almost no volume growth after 2015/16. Growth is also expected to slow as initial productivity gains from the burst of reforms in 2012 through 2014 are gradually exhausted. Consumer price inflation is expected to accelerate to 8 percent in 2015/16, driven by supply side constraints, from 6.6 percent expected in 2014/15. Phased elimination of central bank financing should allow inflation to fall gradually to 5.6 percent by 2020/21. The exchange rate is assumed to move in accord with the differential between domestic and international inflation.8 6 IMF and World Bank (2009), Developing a Medium-Term Debt Management Strategy (MTDS) – Guidance Note for Country Authorities. 7 This follows the growth scenario found in IMF Country Report No. 14/307. 8 International inflation is forecast in the World Bank, April 2015, Commodity Markets Outlook. 15 Table 2.3: Alternative Fiscal Scenarios for the Union Government Prelim. Revised Forecast 2013/14 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2020/21 Baseline Scenario, as Percent of GDP General (Union) Government Revenues and Grants a/ 10.8 11.5 10.2 10.9 10.9 10.9 10.8 10.8 1. Tax 8.1 9.4 8.0 8.3 8.2 8.2 8.1 8.1 a. Gas (Income and Commercial) 1.2 1.2 1.3 1.2 1.1 1.1 1.0 0.9 a. Other 7.0 8.1 6.7 7.1 7.1 7.1 7.1 7.1 2. Transfers from SEEs 1.4 0.5 1.1 1.1 1.1 1.0 1.0 1.0 a. Gas 0.7 0.4 0.5 0.5 0.4 0.4 0.4 0.4 a. Other 0.7 0.1 0.6 0.6 0.6 0.6 0.6 0.6 3. Non-tax b/ 1.2 1.3 0.9 1.1 1.1 1.1 1.1 1.1 4. Grants 0.1 0.4 0.3 0.4 0.5 0.7 0.7 0.7 Expenditures c/ 14.2 16.7 15.9 16.5 16.4 16.3 16.2 16.1 Overall Balance -3.4 -5.2 -5.7 -5.6 -5.5 -5.4 -5.4 -5.3 State Economic Enterprises SEE receipts, net of transfers to Union 12.3 13.2 12.4 12.1 11.8 11.6 11.4 11.3 1. Gas 4.0 5.2 4.0 3.7 3.5 3.2 3.1 2.9 2. Other 8.3 8.0 8.4 8.4 8.4 8.4 8.4 8.4 Expenditures, net of transfers to Union 10.9 12.6 11.2 11.0 10.8 10.7 10.6 10.4 Overall Balance 1.4 0.6 1.2 1.1 1.0 0.9 0.9 0.8 Overall balance (Consolidated Public -2.0 -4.6 -4.5 -4.5 -4.5 -4.5 -4.5 -4.5 Sector) Net Financing 1.9 4.5 4.5 4.5 4.5 4.5 4.5 4.5 A. Foreign financing 0.6 0.8 0.5 1.6 2.0 1.9 2.0 1.8 B. Domestic financing 1.3 3.7 4.0 2.9 2.5 2.5 2.5 2.7 1. Central bank credit 0.8 1.5 1.8 0.4 0.1 0.0 0.0 0.0 2. Bank financing -0.1 2.2 2.2 2.5 2.4 2.5 2.5 2.7 3. Other 0.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Stock of Domestic Debt 20.6 21.5 22.5 22.4 22.0 21.8 21.6 21.8 Scenario Assumptions: Baseline Real GDP growth 8.3 8.5 8.5 8.2 8.1 7.7 7.6 7.5 of which: Gas exports 18.1 38.2 4.6 -0.5 0.0 0.7 1.5 1.7 Deflator Growth 5.7 6.5 7.6 6.6 6.3 6.1 5.8 5.6 CPI inflation (average) 5.7 6.6 8.0 6.6 6.3 6.1 5.8 5.6 Gas price (US$ per MMBTU) 9.9 9.9 8.4 8.5 8.7 8.8 9.0 9.1 Scenario Assumptions: Low GDP Growth Real GDP growth (%) 8.5 8.5 6.2 5.9 5.6 5.5 5.0 Scenario Assumptions: Low Tax Revenues Tax revenues (% of GDP) 9.4 8.0 6.8 6.6 6.3 6.1 6.0 Scenario Assumptions: Low Gas Prices Gas Price (USD per MMBTU) 9.9 8.4 7.7 7.2 6.8 6.5 6.2 Real expenditure reductions, Scenario Outcomes for Union Expenditures as percent of baseline values Low GDP growth 0 0 -1 -3 -5 -6 -7 Low tax revenues 0 0 -9 -10 -12 -12 -13 Low gas prices 0 0 -3 -4 -5 -6 -7 Current price spending reductions, as percent of baseline capital Low GDP growth 0 0 -4 -8 -14 -18 -22 Low tax revenues 0 0 -26 -29 -35 -37 -40 Low gas prices 0 0 -7 -12 -16 -18 -20 a. Excludes revenues collected by states and regions and thus is comparable to central government. b. FY2014/15 includes proceeds from sales of two telecom licenses awarded in June 2013. c. Excludes expenditures financed from revenues collected by states and regions. Sources: IMF and World Bank staff calculations. 16 2.18. Gas prices are expected to fall by 16 percent in 2015/16 and then slowly grow by 1-2 percent per year through 2020/219 (Figure 2.10). Historically, however, gas prices have been quite volatile. The long- run standard deviation in price growth rates between 1978 and 2014 is 18 percent. This creates upper and lower boundaries on the projected price (Figure 2.10). 2.19. The baseline scenario assumes that fiscal and monetary policies of 2015/16 are continued through 2020/21. Tax rates and SEE obligations remain unchanged. Union revenues as a share of GDP gradually fall, partly because gas export volumes do not grow as quickly as GDP and partly because gas prices are not expected to fully recover from the drop in 2015. It is also assumed that the government will maintain a consolidated public sector deficit target close to 4.5 percent of GDP with a Union deficit of close to 5.5 percent of GDP being offset by a net SEE surplus near 1 percent of GDP. The net external financing in the baseline scenario is conservative: could be as high as 2 percent of GDP in 2016/17 and 2017/18. Figure 2.10: European Gas Prices, US$/MMBTU 25 20 15 10 5 0 Optimistic (DEC growth + 1 SD 18%) Baseline Pessimistic (DEC growth - 1 SD 18%) Low Case a. Myanmar gas pipeline prices are positively correlated with European gas pipeline prices. Sources: World Bank 2015 Commodity Markets Outlook and World Bank staff projections. 2.20. The baseline scenario allows for healthy expenditure growth. Union revenues are expected to remain close to 11 percent of GDP. This implies that Union government expenditures will be kept at levels close to 16 percent of GDP, yielding an average 8 percent real increase in current spending per year. 2.21. Three alternative scenarios presented below include slower real growth, a less effective tax effort, and lower gas prices. Each alternative projection scenario includes the assumption that the government will maintain a consolidated public sector deficit target close to 5 percent of GDP. This implies that revenue reductions will need to be compensated by reducing expenditures below the baseline values.  A prolonged down-turn in gas prices is a risk to Myanmar’s growth prospects. If gas prices were to decrease in line with the low case depicted in Figure 2.10 above, then the real value of total Union expenditures in 2020/21 would need to be 7 percent lower than the baseline value for that year.10 9 This forecast is for European gas, mainly supplied by pipelines, from the April 2015 Commodity Markets Outlook. 10 This calculation assumes the gas share of GDP averages about 8 percent of GDP. 17 This would be equivalent to 20 percent of the baseline capital budget projected for 2020/21. The down-turn in 2015 may be offset to some extent by planned investments and increased production, but the government will need to monitor closely to ensure external and fiscal buffers for exchange rate stability and fiscal sustainability.  Another serious risk to the budget is low non-gas tax revenues. If the projected increase in the tax to GDP ratio to 8.1 percent of GDP by 2020/21 is not maintained and instead gradually falls back to 6 percent of GDP by 2020/21 (e.g. if more tax exemptions are granted), then the real value of total Union expenditures in 2020/21 would need to be 13 percent lower than the baseline value for that year. This would be equivalent to 40 percent of the baseline capital budget projected for 2020/21.  There are other fiscal risks as well. Real GDP growth could be slower than expected. If actual growth is two percentage points lower than forecast, then the real value of total Union expenditures in 2020/21 would need to be 7 percent lower than the baseline value. Intergovernmental transfers could continue to add to net spending if not off-set by slower growth in other expenditures. Off- budget external financing can complicate fiscal management and pose risks to debt control. Lastly, various SEE operations could generate large contingent liabilities for the government: this risk can be managed if their situation can be carefully monitored and ongoing reforms managed carefully. E. Treatment of Upstream Hydrocarbon Revenues 2.22. This public expenditure review proposes that all resource rents (royalties, corporate income taxes, profits) from the Myanmar Oil and Gas Enterprise (MOGE) be channeled directly and in their entirety to the Union Budget. Upstream hydrocarbon enterprises such as MOGE extract non-renewable natural resources that are also national assets. Therefore a proportionately larger share of their profits should accrue to the government for the broader benefit of the national population. 2.23. Different countries have adopted different models for managing rents from their state enterprises operating in the hydrocarbon sector.11 At one extreme (e.g. Timor-Leste, Ghana), all receipts are channeled to and spending is appropriated through the national budget. At the other extreme (e.g. Myanmar), all receipts are channeled through the enterprise, which then pays royalties and share of profits (tax and contributions) to the Union Budget. Even in more advanced cases (e.g. Norway), however, all resource rents are paid directly to the national budget. The difference for these more advanced cases is that the National Oil Company (e.g. Statoil in Norway) operates like any other commercial operator. The government allows the National Oil Company to do so because it is subject to corporate governance standards under a Companies Act consistent with international standards. This is not the case for SEEs in Myanmar. 2.24. There are several additional reasons for proposing that all resource rents from MOGE be directed to the Union Budget. First, they are macro-critical because of their size. Second, it enables the government to maintain oversight of major investment decisions by appropriating for these through the budget. Third, until MOGE and other upstream hydrocarbon SEEs are corporatized and are subject to standard international norms of corporate governance, their rents should be subject to budgetary discipline and oversight. A fourth critical benefit is increased transparency over natural resource rents. At the moment it is not possible to say from published documents how much MOGE and other hydrocarbon SEEs are retaining in their own accounts. 11 Norwegian Agency for Development Cooperation, (2015). 18 F. Fiscal Risk Management 2.25. The PER highlights below some areas of fiscal risk that may warrant closer attention going forward. Fiscal risks are factors (e.g. macroeconomic shocks, commodity price developments, extra- budgetary activities) that can cause major deviations to fiscal outcomes (e.g. government revenue, deficits, public debt) from what was planned in the budget, and therefore prevent achievement of service delivery objectives. The PER proposes possible institutional reforms to better manage fiscal risks in a few areas, namely: the decentralized nature of debt management; commodity price volatility; and the capacity for fiscal oversight of SEEs. 2.26. The decentralized nature of debt management refers to the lack of clarity in Myanmar’s legal framework on authority to approve borrowings and loan guarantees. According to the country’s 2008 Constitution, the Pyidaungsu Hluttaw (parliament) is empowered to make decisions on state borrowing of international, regional or bilateral agreements submitted by the President. However, the role of the Ministry of Finance is not specified and it is not clear if line ministries and SEEs have the right to undertake borrowing transactions directly and if so under what conditions and procedures. Lack of oversight and coordination in this area can lead to unsustainable levels of borrowing and derailment of fiscal objectives. 2.27. Line ministries are currently able to contract loans directly, with some Ministry of Finance oversight, which was instituted only recently. For example, in the 2014/15 Union Budget Law, the Ministry of Construction accounted for nearly three quarters of Kyat 697 billion in planned borrowing by Ministries and Departments, and nearly a third of Kyat 970 billion in the 2015/16 Union Budget Law (Figure 2.11). The Ministry of Finance accounted for less than 1 percent, as these are external, project-linked loans (though there is no detail in the Union Budget Laws on the source or terms and conditions of these loans). SEEs, which are part of line ministries, planned to borrow an additional Kyat 542 billion and Kyat 293 billion in 2014/15 and 2015/16 respectively (Figure 2.12). Half of this on average was borrowing by SEEs under the Ministry of Electric Power. Figure 2.11: Planned Borrowing, by Ministry Figure 2.12: Planned Borrowing, by SEE Sector Kyat billion Kyat billion NPED Finance 2015/16 BE 2014/15 BE 2015/16 BE 2014/15 BE Finance ICT Energy Transpotation Health Railways Education Home Affairs ICT Livestock and Fish Electric Industry Agriculture Co-operative Electric Power Construction 0 50 100 150 200 0 200 400 Source: Union Budget Laws 2014/15 and 2015/16. 2.28. Government policies also allowed line ministries to directly service loan obligations. Since many line ministries do not have their own-source of revenue (and even if they do, good practice requires these funds to be consolidated in the Treasury), the Union Budget Law appropriates funds to those ministries 19 for debt service payments. Based on data from the 2014/15 and 2015/16 Union Budget Laws, the Ministry of Finance was responsible for over 90 percent of debt service obligations. This is likely because: (i) these planned amounts include obligations on domestic debt, which the Ministry of Finance is responsible for; and (ii) external debt servicing obligations on line ministries’ project related loans may have grace periods or extended maturities and loan repayment schedules. Therefore debt service obligations of ministries and departments may grow over time. Between 2014/15 and 2015/16, debt service obligations of Ministries other than Finance went from Kyat 37 billion to Kyat 82 billion due to a jump in debt service obligations of the Ministry of Construction. Figure 2.13: Debt Service Obligations Figure 2.14: Debt Servicing Costs (Kyat Million) (Percent of total Ministry obligations) 45,000 40,000 Co-operative Transp. Agric. Other 1% 1% 1% Co- 35,000 2014/15 BE 2015/16 BE operative 30,000 4% 25,000 20,000 Transportation Electric 15,000 Agriculture Power 10,000 Finance and Revenue 5,000 93% - Co-operative Transpotation Agriculture Electric Construction Other Source: Union Budget Laws 2014/15 and 2015/16. 2.29. Efforts have been made since 2011 to ensure that borrowing by line ministries is adequately coordinated with the government’s overall objectives for debt management. There is for example the overall borrowing limit in the Union Budget Law – though as noted above, this measure is not very effective in the absence of any detail on the planned financing mix. The more important measure is the delegation of greater responsibility to the Ministry of Finance to oversee any new borrowing by line ministries and SEEs, even though the loan can actually still be signed by the line ministry. 2.30. The PER proposes to take this further by explicitly assigning the authority to approve borrowings and loan guarantees to the Ministry of Finance through a Public Debt Law, which was under consideration as this report was being written. It is critical to concentrate these functions in a single debt management entity that has the capacity and mandate to review borrowing plans and ensure that these are consistent with the government’s debt management objectives. The alternative of having fragmented decision-making over financing of the Union Budget deficit can impact severely on fiscal risks through lack of comprehensive information on borrowing across government, over-borrowing, heightened portfolio risks or other. The PER recommends that the Debt Management Office in the Ministry of Finance should assess and negotiate all loans on behalf of the government. In addition, to establish greater transparency over these borrowing decisions, particularly important now given concerns over rising debt levels, the Ministry of Finance should publish the key elements of its annual or medium-term borrowing plan. 2.31. Commodity price volatility is another source of fiscal risks. The price of Myanmar’s gas exports is indexed to the heavy fuel and a variety of production cost indicators. The price is adjusted every quarter by taking the last 12 months’ average of these variables. Available data shows that large price shocks in 20 2008 and 2009 (Figure 2.15) did impact negatively on gas revenues (Figure 2.16). For the period 2000/01 through 2012/13, the average elasticity of gas revenues measured in US dollars per million British thermal units (MMBTU) to gas prices in US dollars per MMBTU was 0.6. Thus, if world oil prices fluctuate by a maximum of 40 percent, up or down, net gas revenues per billion cubic feet could swing up or down by a maximum of 24 percent. This can significantly derail budget policy objectives, particularly given that over 18 percent of government revenue in Myanmar comes from gas. Figure 2.15: Gas and Oil Prices, 2001-2013 Figure 2.16: Gas Prices and Revenues, 2001-2013 14 120 14 12 100 12 10 10 US$ per MMBTU US$m/MMBTU 80 US$ per barrel 8 8 60 6 6 40 4 4 Gas price Gas Price (LHS) 2 20 Gas revenue Petroleum Price (RHS) 2 0 0 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Sources: Ministry of Energy, IMF and World Bank staff calculations. 2.32. One way to deal with the commodity price risk is to diversify further the tax base away from volatile oil and gas receipts (Chapter 3). Options under consideration by the government include excise tax reform, broadening the income tax base, a new value-added tax and a new property tax. Each of these should be less correlated with oil and gas receipts. In addition, the property tax, if enacted, would be uncorrelated with the income and value-added taxes, to the benefit of fiscal stability. 2.33. Establishing a sovereign wealth fund is another option for managing oil and gas revenue volatility. These funds can play several roles. They can accumulate savings for intergenerational equity, which may require rules for example on withdrawing only interest earnings so that the capital can grow (or be maintained) in real terms over the long-term. These funds could also help to smooth revenue flows into the budget, saving when revenues are high and allowing draw-downs during commodity price downturns. 2.34. The government, through the Ministry of Finance Fiscal Policy Department, could also consider fiscal rules or benchmarks to avoid pro-cyclical spending. These could be in the form of targets that aim to maintain fiscal discipline during commodity price booms and avert real exchange rate appreciation. For example, targeting: (i) the non-oil and gas fiscal balance (i.e. fiscal balance excluding revenue from gas) as a share of non-oil and gas GDP can help to ensure that the fiscal deficit excluding gas revenue does not become unsustainable as a result of spending during commodity price booms; (ii) growth in primary expenditure (i.e. recurrent expenditure excluding interest payments) can help to contain growth in government consumption and non-essential spending; or (iii) a positive recurrent balance (i.e. recurrent expenditure as a ratio of non-gas revenues) can help ensure that government consumption is funded entirely out of stable, non-gas revenues. 21 2.35. Finally, it is important to integrate gas revenue issues explicitly in the government’s emerging Medium-Term Fiscal Framework. The PER proposes that the Ministry of Finance Budget Department and Internal Revenue Department work with the Ministry of Energy to establish a protocol for data exchange on oil and gas production and price projections. The MTFF should look at various scenarios for medium- term gas revenues. These should be presented and discussed across relevant government departments during budget preparation to have an agreed forecast. It is generally recommended to base planned revenue in the budget on conservative price assumptions to avoid over-budgeted spending in case of a downturn. Throughout the fiscal year, it will be important to closely monitor developments on gas revenues, to ensure that timely adjustments can be made to government spending if and when needed. 2.36. It would be prudent to make SEEs subject to strong fiscal oversight as their operations make up over half of all public sector operations, and therefore constitute a potential source of fiscal risks. In Myanmar these risks are heightened by the fact that SEEs are controlled by the line ministries and their deficits are automatically financed directly from the Union Fund Account. This makes the general government explicitly liable for SEEs, unlike some other countries where state enterprises are semi- autonomous. In Myanmar, however, beginning in 1990/91, a Union Fund Account system was established in which the receipts from SEEs were integrated with line ministry receipts. All SEE operations were to be funded from the Union Fund Account, through the Union budget.12 2.37. The lack of autonomy -- combined with numerous implicit subsidies, a direct financial subsidy for SEE deficits, and an over-valued official exchange rate -- reduced the incentive for SEEs to adapt and grow under changing conditions. The over-valued exchange rate also distorted the valuation of SEE operations by reducing the Kyat value of their net foreign exchange positions relative to their net local currency positions. This meant, for example, that an exporting SEE might register a net loss because its domestic costs outweighed the value of its exports. In 2009/10, 31 out of 50 enterprises were found to be loss-making. 2.38. The exchange rate reforms in 2012/13 as a result had an overall positive impact on SEE finances. The number of enterprises that recorded negative overall balances fell from 23 out of 46 in 2011/12 to only 8 out of 44 in 2012/13.13 In 2013/14, the expected outcome was 14 out of 42 enterprises. Most of the gains were earned by MOGE, Myanma Petroleum Products Enterprise, and Myanma Petrochemical Enterprise.14 The overall balance of the SEE sector becomes negative when these enterprises are omitted (Table 2.4). This is largely due to a few large loss makers including Myanma Railways and Myanma Electric Power Enterprise. Myanma Railways was a loss-maker in all years under review.15 Myanma Electric Power suffered from exchange rate reform, having to purchase gas at a higher local currency price from MOGE. The authorities revised electricity tariffs upward in response, but not enough to cover the full costs of production. 2.39. Further reforms helped set a firmer budget constraint for SEEs. At the start of 2012/13, the Union government introduced a limit on the level of implicit subsidies by requiring SEEs to finance 78 percent of working capital requirements from their operating revenues. At the same time, SEEs were allowed to keep a larger share of their profits: 25 percent of net profits were to be paid as income tax and 20 percent 12 World Bank (1995), chapter 3. 13 The total number of enterprises changes over time due to consolidations and privatization. 14 The enterprises responsible for timber, gems and other mining output also recorded gains. 15 Revised MoF data from June 2014 indicate there were 12 loss-makers under 9 ministries. 22 of net profits were to be put into the State Fund as a contribution to general government. The remainder could be carried over in their “Own Fund Account” for their own use in the next fiscal year. In 2013/14, profitable SEEs were expected to self-finance 100 percent of their working capital from their “Own Fund Account” or borrow from state owned banks at 4 percent interest. The government will cover 20 percent of working capital expenses of loss-making SEEs. In 2014/15, four SEEs were moved off-budget.16 Table 2.4: SEE Net Flows (Percent of GDP) Preliminary Estimated 2009/10 2010/11 2011/12 2012/13 2013/14 Myanma Oil and Gas, Net Flow 0.0 0.0 0.0 1.8 1.4 Revenues 0.2 0.2 0.2 4.9 5.1 Expenditures 0.2 0.2 0.2 3.1 3.8 Myanma Petroleum Products, Net Flow 0.1 0.6 0.0 1.2 0.5 Revenues 1.6 1.7 1.5 1.6 1.4 Expenditures a/ 1.4 1.2 1.5 0.4 0.9 Myanma Petrochemical, Net Flow -0.2 -0.2 0.1 1.0 0.3 Revenues 0.7 0.5 0.7 1.6 1.4 Expenditures 1.0 0.7 0.5 0.6 1.1 All Other, Net Flow b/ -1.5 -1.2 -0.8 -1.8 -1.1 Revenues 4.7 4.6 5.4 7.2 6.8 Expenditures 6.2 5.8 6.3 9.0 7.8 Total, Net Flow -1.7 -0.8 -0.7 2.2 1.1 Revenues 7.2 7.0 7.8 15.3 14.7 Expenditures 8.8 7.8 8.5 13.1 13.6 Number of Enterprises c/ 50 49 46 44 42 In Overall Surplus 19 21 23 36 28 In Overall Deficit 31 28 23 8 14 a. Data for 2012/13 and 2013/14 indicate much reduced payments for taxes and contributions to the Union. b. From IMF total less MoF BOOST data. c. The total number of enterprises changes over time due to consolidations and privatization. Sources: MoF BOOST database and IMF staff estimates 2.40. To complement these efforts, the PER recommends that the SEE Department in the Ministry of Finance establish a comprehensive framework for fiscal transparency and oversight of SEEs. The objective would be to ensure that the government is exercising fiscal discipline in relation to SEEs, averting undue burden and risks. This could be done through an analysis of the financial status of SEEs and advice on the legal and regulatory framework for fiscal oversight of SEEs, taxation and dividend policies, and fiscal reporting. These should in turn help to better monitor fiscal risks from SEEs. 2.41. The analysis of the financial status of SEEs should help to develop a typology of SEEs for risk- based monitoring. This would look at characteristics of SEEs (in terms of sectors, size, line of business (commercial/non-commercial), turnover, employment, and ownership structure), and their financial links to government (preferential access to credit, government guarantees of SEE debt, preferential tax treatment, any SEE bond issuances, recapitalization of SEEs, post-privatization (or joint venture) obligations). The legal and regulatory review should aim to take stock of all SEE-related provisions in laws outside of the SEE Law (1989) (e.g. procurement, public debt, employment, public financial management, audit, competition, banking, bankruptcy). This can help establish existing arrangements for oversight and potential for reform including if necessary greater control over any new obligations that SEEs can enter into. This would also cover the existing framework for financial controls, and treatment of SEE profits including in the natural resources sector as discussed above. All these elements should be brought together with a review on further strengthening fiscal reporting standards on SEEs, and using this to manage fiscal risks. 16 Inland Water Transport, Myanma Airways, the Port Authority, and the Shipyards. 23 G. Conclusions and Recommendations 2.42. Myanmar faces the challenge of meeting large service delivery needs on the back of relatively limited fiscal space, as characterized by an already small general government, narrow revenue base and limited financing options. Revenue windfalls from one-off measures have enabled the government to rapidly increase spending, while maintaining fiscal deficits within 5 percent of GDP. Overall public debt levels are still within sustainability thresholds, though there is a need to focus greater attention on ensuring a sound financing mix for a balanced debt portfolio. Over the medium-term, fiscal balances are vulnerable to lower gas revenue (around 18 percent of total revenue) due to international commodity price developments. This may be offset by other tax and non-tax receipts thanks to strong projected growth, failing which there may be a need for significant expenditure adjustments. 2.43. The above points to the need for greater attention to monitoring and managing fiscal risks. The PER highlights three areas in this regard. The first is on consolidating debt management functions, in particular the authority to negotiate and approve new loans and guarantees, within the Ministry of Finance. This should help to avoid fragmented financing decisions and build-up of unsustainable debt. The second is on managing risks from commodity price volatility. This may require a review of how hydrocarbon SEE profits are managed and adoption of fiscal rules. The third is on strengthening the fiscal oversight of SEEs given their impact on general government operations and finances. Managing these risks should help ensure that budget policies are not derailed by unexpected shocks to fiscal aggregates. Issues Options Rebalancing the composition of Short-term: MoF Debt Management Office to develop and adopt an the public debt portfolio will annual borrowing plan to promote a sound financing mix for a reduce vulnerability to breaching balanced debt portfolio (ref ongoing ADB TA). debt sustainability thresholds. Medium-Term: Develop MTDS with joint IMF-World Bank technical assistance. A single authority to assess, Short-term: MoF to propose that it be explicitly assigned the authority negotiate and approve to approve borrowings and loan guarantees through a Public Debt borrowings and loan guarantees Law, which is currently under consideration. leads will reduce fiscal risks. Managing commodity price Short-term: channel all resource rents from upstream hydrocarbon volatility with promote more SEEs to the Union Budget, and integrate gas price developments in the stability in government MTFF. expenditure. Medium-term: MoF to consider fiscal rules and benchmarks linked to commodity revenues. Fiscal oversight of SEEs can Short-term: MoF to initiate TA on SEE classification, regulations, ensure early mitigation of fiscal revenue treatment, fiscal reporting and fiscal risk monitoring. risks linked to SEE operations. Medium-term: MoF to adopt a comprehensive framework for fiscal oversight of SEEs. 24 REBALANCING 3 THE COMPOSITION OF THE UNION BUDGET Background: The government’s transition to a service oriented approach has come with major rebalancing along and across the Union Budget since 2010/11. Further spending growth however needs to be balanced against financing and implementation capacity constraints. Objectives: To identify short to medium-term options for strengthening revenue policies, institutions and administration; and to explore the scope for further rebalancing expenditure across the budget to create fiscal space for Myanmar’s development priorities. Key findings: Overall revenue effort has historically been low. Economic recovery, one-off measures, and reforms to expand the revenue base have improved revenue from 6 to 11 percent of GDP between 2009 and 2014. Income tax, tax on goods and services and non-tax revenue each make up roughly just under a third each of total revenue. Performance across all types of revenue remains below international comparators, including for customs duties (0.3 percent of GDP). Tax exemptions for investment promotion are widely applied through a range of different legal instruments. There is however no assessment of tax expenditures or impact on incremental investment. Two areas to focus on in tax administration are arrears management, and transparency and consolidation of receipts in extra-budgetary “other accounts.” On Union Budget spending, big changes between 2009 and the 2015 Budget include the share of: general public services, falling from 45 percent to 11 percent; social services, increasing from 10 to 33 percent; defense increasing from 21 to 32 percent; and capital investment falling from 63 to 38 percent. The wage bill in Myanmar has historically been suppressed by limits on both the size of and pay in the public service. Recent increases in the wage bill, including near doubling in the 2015/16 Budget, should not pose major concern in the medium-term. Any further increase should be linked to broader public administration reform. Recommended options: (i) review and rationalize tax exemptions through a tax expenditure analysis, and consolidate the granting of any new exemptions; (ii) prioritize management of revenue arrears and consider gradual elimination of extra-budgetary ‘other accounts’; (iii) initiate an expenditure review of military spending to identify potential efficiency gains; (iv) conduct a review of public investment management and efficiency, including scope for savings through reduced spending on commercial activities in economic services; (v) conduct a pay and grading exercise to help inform any future changes to the wage bill. 25 CHAPTER 3 REBALANCING THE COMPOSITION OF THE UNION BUDGET A. Background17 3.1. The government’s transition to a service oriented approach has come with major rebalancing along and across the Union Budget since 2010/11. Along the budget, efforts are under way to improve revenue mobilization, which increased from 6 to 11 percent of GDP between 2011 and 2014. Across the budget, spending has shifted from general public services towards social sectors. A gradual decline in capital investments has provided fiscal space for salaries and wages of public servants, essential goods and services in the social sectors, and other public investments. 3.2. The PER takes stock of these shifts and highlights possible measures that can further enhance the alignment of the budget with policy priorities. After reviewing the current revenue effort and composition in Myanmar, the PER touches on options to strengthen revenue policies, institutions and administration. It then reviews the current functional and economic composition Union Budget expenditures to explore scope to further rebalance allocations and create fiscal space for Myanmar’s development priorities. Figure 3.1: General Government Revenues, 2010/11 Figure 3.2: Tax Revenues and Per-capita GDP, 2010/11 25 30 27.3 Vietnam 23.1 22.4 Tax revenue (% GDP) 25 20 21.3 Percent of GDP Thailand 20 18.3 17.0 17.0 16.8 Malaysia 15 Nepal Lao PDR 15 Philippines Indonesia 10 10 Bangladesh China 6.3 Cambodia 5 5 Myanmar 0 Vietnam Malaysia Indonesia China Lao P.D.R. Thailand Philippines Myanmar Cambodia 0 - 2,000 4,000 6,000 8,000 10,000 US$ GDP/capita Sources: World Development Indicators and IMF estimates. B. Revenue Effort and Composition 3.3. The overall revenue effort of the Union government has historically been below potential. General government receipts in 2010/11 were at 6.3 percent of GDP, compared to a regional average of around 21 percent (Figure 3.1). Even when controlling for per-capita GDP, Myanmar’s tax revenue remains 17 Much of the data in this chapter, especially for expenditures, come from the MoF BOOST database. They are similar to but not the same as the IMF estimates reported in Chapter 2 for all years except 2012/13 and 2013/14. The IMF data for 2012/13 and 2013/14 are staff estimates while the BOOST data portray the revised budget estimates for 2012/13 and the budget estimates for 2013/14. The BOOST data provide more detail on composition. 26 low compared to other countries at similar income levels (Figure 3.2). This was due to a combination of factors: a narrow tax base dominated by SEEs and the public sector; a closed economy meant little by way of trade taxes; and an overvalued exchange rate that suppressed government receipts from export earnings. 3.4. Economic recovery coupled with recent reforms have helped to improve revenue mobilization (Figure 3.3). The average ratio of the percentage change in general government revenue to the percentage change in GDP (buoyancy) for the years 2006/07 to 2011/12 was 1.0, indicating that, on average, revenues grew at the same pace as GDP.18 The average ratio for a sample of eight East Asian countries was 1.1 with a low of 1.0 and a high of 1.4.19 This is in part because most other countries in the region have a higher share of income taxation in their total revenues, which are typically more buoyant. Figure 3.3: Regional Revenue Effort (Percent of GDP) 30 25.6 25 25.6 24.1 25 22.9 22.1 21.6 20.8 20.2 20 18.6 17.9 17.1 17.5 17.1 16.5 15.8 15 11.6 10 5.8 5 0 Malaysia Vietnam Thailand China Lao P.D.R. Philippines Indonesia Cambodia Myanmar 2009 2010 2011 2012 2013 Sources: IMF Fiscal Monitor, Ministry of Finance BOOST database and World Bank staff estimates. 3.5. Income tax receipts in Myanmar in 2013/14 were close to 2.4 percent of GDP compared to 5.7 percent in a sample of nine low and middle-income East Asian countries (Figure 3.4). The income tax effort is of special concern not only for revenue generation but also for macroeconomic management. Progressive income taxes can act as stabilizers, automatically reducing burdens on business and families during recessions. Administrative and policy reforms are under way to enhance income tax collections. Figure 3.4: Revenue Composition 20 3.4 Non-Tax Percent of GDP (%) 15 1.0 Social Contrib. 0.6 Other Taxes 10 6.4 Goods and Services 2.2 1.3 Customs 5 0.3 2.1 Income Tax 5.7 2.4 0 Myanmar Asian Sample (Central Gov.) Sources: Ministry of Finance BOOST database and World Development Indicators. 18 Calculated from a regression of the natural log of revenue on the natural log of nominal GDP. 19 World Development Indicators: Cambodia, China, Indonesia, Korea, Lao PDR, Malaysia, Philippines and Thailand. 27 3.6. Taxes on goods and services and non-tax revenue each accounted for around a third of general government revenue. Taxes on goods and services, also referred to a commercial taxes, are indirect taxes. In 2013/14, they accounted for around 2.1 percent of GDP compared to 6.4 percent in a sample of nine East Asian countries mentioned above (Figure 3.4). Non-tax revenue, which includes SEE contributions and capital receipts such as sale of licenses to the banking and telecommunications sector accounted or 2.2 percent of GDP compared to an average of 3.4 percent of GDP in selected Asian countries. 3.7. Customs duties collected in Myanmar were around 0.3 percent of GDP in 2013/14 compared to 1.3 percent in the same sample of East Asian countries mentioned above (Figure 3.5). Customs duties often constitute a large share of government revenue at Myanmar’s stage of development. However, imports as a share of GDP in Myanmar (20.7 percent) is considerably lower than the regional average (50.4 percent of GDP). This will likely change with further integration. In the short run, however, there may be gains from customs administration reform. Myanmar scores poorly when it comes to the speed, simplicity and predictability of customs clearance in the World Bank Logistics Performance Index (Figure 3.6). Figure 3.5: Customs Duties Collection Figure 3.6: Logistics Performance Index 3.50 120 10 124 Philippines 3.00 9 100 2.99 2.50 Customs duties (% of imports) 8 2.73 80 7 Averages 2.00 Size of bubbles Score Rank 6 reflects Cambodia 2.07 60 customs revenue/GDP 1.50 5 Lao PDR Mongolia 40 4 1.00 44 Myanmar Korea 35 3 0.50 20 2 China Thailand 1 Indonesia 0.00 0 Malaysia China Indonesia Myanmar 0 18 38 58 78 LPI customs rank LPI customs score Imports (% of GDP) Sources: Ministry of Finance, World Development Sources: World Bank Logistics Performance Index (2014), Indicators, World Bank staff estimates. customs rank and score. 3.8. Social contributions do not make up a large share of revenues. These are revenues that finance social insurance schemes including retirement benefits. The amounts involved can be large, ranging from 1 percent of GDP for Thailand to 5 percent for mineral rich Mongolia. Although the amounts recorded by the authorities for this purpose are extremely small, with a GDP share close to zero, there are several social insurance schemes in operation. The Ministry of Labor operates a social insurance system through its Social Security Board established in 1954. A review of accounting practices (e.g. imputation from wages and salaries) could reveal the actual magnitude of resources collected for social insurance schemes. 3.9. SEEs have contributed importantly to Union revenues. Their combined payments to the Union increased from 3.7 percent of GDP in 2009/10 to 4.1 percent of GDP in 2013/14 (Table 3.1). Of this, payments by MOGE jumped sharply from an average of 0.1 percent of GDP through 2011/12 to 2.2 percent of GDP by 2013/14. MOGE payments thus accounted for 19 percent of total Union revenues in 2013/14. Payments by other SEEs fell from 3.7 percent of GDP in 2009/10 to 1.9 percent in 2013/14. The public sector had a net gain of 0.6 percentage point of GDP. 28 Table 3.1: SEE Contributions to Union Revenues and Grants Change Percent of GDP 2009/10 2010/11 2011/12 2012/13 2013/14 2011 to 2013 Revenue and Grants 5.8 6.3 6.5 9.6 11.6 5.1 A. State Economic Enterprise a/ 3.7 3.1 3.6 4.4 4.1 0.6 1. Myanmar Oil and Gas 0.1 0.1 0.1 1.9 2.2 2.1 2. Other SEE 3.7 3.0 3.5 2.5 1.9 -1.6 B. Other 2.1 3.2 3.0 5.2 7.4 4.5 SEE as Percent of Total 63.9 49.3 54.6 45.4 35.8 .. MOGE as Percent of Total 1.1 1.6 1.6 19.3 19.3 .. a. Includes commercial tax, income tax and transfers (contributions) to the Union. Sources: IMF and World Bank staff calculations. 3.10. The cumulative change from all other taxpayers was impressive, suggesting that the tax base has begun to diversify away from dependence on the SEEs. Non-SEE revenue increased from 2.1 percent of GDP in 2009/10 to 3.0 percent in 2011/12 and 7.4 percent of GDP in 2013/14. The cumulative change between 2011/12 and 2013/14 was 4.5 percentage points of GDP. Thus, the SEE share of total Union revenues and grants has fallen from 63.9 percent in 2009/10 to 35.8 percent in 2013/14. 3.11. Further diversification of the tax base would help the Union in two ways: (i) increase fiscal space; and (ii) help manage the risk created by the volatility in gas export receipts, as discussed in Chapter 2. Options under consideration by the government include excise tax reform, broadening the income tax base, a new value-added tax and a new property tax. C. Revenue Policies and Institutions 3.12. Two overarching and potentially competing objectives on revenue collections have guided Union Government efforts: first to increase the tax-to-GDP ratio while making tax administration more efficient and reducing the cost of tax payer compliance;20 and second to promote investment and create job opportunities by using tax incentives. The first objective targets an important cost of doing business in Myanmar, namely the burden of paying tax. There is however no clear evidence from other countries that tax incentives generate incremental investment that would eventually offset the short-term costs of incentives, particularly in countries like Myanmar that have more binding constraints to investment. 3.13. The PER therefore proposes to comprehensively review existing exemptions. In Cambodia, for example, such an analysis showed that the combined cost of exemptions in 2010 exceeded the value of FDI and was the equivalent to half of total investment. The first step in Myanmar should be to take stock of all exemptions, which are covered in various legislation: the Foreign Investment Law (2013), Myanmar Citizens Investment Laws (1994), Special Economic Zone Law (2011), the Dawei Special Economic Zone Law (2011), and sector laws such as the Myanmar Mines Law (1994). There may be discretionary channels to grant exemptions, such as for customs.21 The system for granting exemptions should be consolidated. 20 Strategic Reform Plan for the Internal Revenue Department, 2013 to 2018. 21 Tariff Law, section 7 (1992): “The Minister may, by notification exempt partially or wholly from levy of customs duties in respect of any of the following cases: (a) nature and type of goods exported from Myanmar or imported into Myanmar; (b) nature and type of goods exported from Myanmar or imported into Myanmar by any Government department or any organization.” 29 It can otherwise add significant burden on tax administration. The next step would to estimate the level of tax expenditures from current exemptions and recommend ways to rationalize these. 3.14. The legal framework for Myanmar’s tax system has been fundamentally revised and continues to evolve. The country’s Socialist period saw the promulgation of the Income Tax Law (1974), the Profit Tax Law (1976), and the Goods and Services Tax (1976).22 The Commercial Tax Law, enacted in 1990, replaced the goods and services tax and applied to all sectors in the economy. In April 2012, the Commercial Tax rate was harmonized at 5 percent for most goods and selected services. There are now plans to phase out the Commercial Tax and introduce a value added tax (VAT). Going forward, the government could consider adopting a unified tax code to ensure consistency in overall tax policy and avoid the problem of tax cascading. 3.15. Income tax legislation has also been reformed. Income tax includes corporate income tax, individual income tax, capital gains tax, and withholding tax. It applies to all types of taxpayers. In April, 2012, income tax bands were adjusted and extended to public officials. The latter were previously exempt, though the tax free income threshold has also increased from MMK 500,000 to MMK 2,000,000. In April, 2012, the corporate income tax rate was reduced from 30 percent to 25 percent for resident companies (one rate) while the corporate rate for non-resident companies remains at 35 percent. Before 2011, a separate profits tax, adopted since 1976, was imposed on certain types of income not covered under the income tax law. The profit tax was repealed in April 2012 as it was causing tax cascading. The current system is simpler, less distortionary and more progressive. D. Revenue Administration 3.16. The Internal Revenue Department (IRD) of the Ministry of Finance administers the largest share of domestic revenues. There are currently fifteen types of taxes and duties levied in Myanmar. The IRD collects personal and corporate income tax, commercial tax, state lottery tax, and stamp tax. The Customs Department handles customs duties while the Trade Directorate of the Ministry of Commerce takes care of fees for import licenses. The General Administration Department of the Ministry of Home Affairs is responsible for excise duty, land tax, water and embankment tax and minerals tax. The Road Transport Administration Department administers transportation taxes. The Forestry Department looks after a tax on rubber and on general forestry extraction. The Fisheries Department administers a tax on fisheries. 3.17. Several reforms are underway in tax administration within the Internal Revenue Department. These include the establishment of a large taxpayer office, the assignment of tax identification numbers, the introduction of tax-payer self-assessment, the forthcoming establishment of medium and small taxpayer offices, modernization of information and communications technology, organizational restructuring and a recruitment effort intended to increase staff size by 35 percent. 3.18. Management of arrears is one area that warrants closer attention. According to the Public Expenditure and Financial Accountability report of 2013, both the IRD and Customs Department have not to date introduced systematic arrears management systems. There is no analysis of tax arrears or arrears management plans based on systemic, well-defined classification of arrears by tax type, age, and subject taxpayers. The aggregate level of tax arrears have not been reduced over the past three years and remain stably high at about MMK100 billion (8 percent of total tax collections in 2010). The arrears collection of the two main taxes, income and commercial, shows the persistent low performance. The collection during 22 Source: http://www.myanmar.com/finance/dept_ird_06.html. 30 the year over the total tax arrears at the beginning of the year for income tax stands at just 62 in 2009/10 and 2010/11 and 52 percent. The collection of commercial tax arrears during the same two years is even lower: 24 and 26 percent. 3.19. Transparency of tax policy is gradually improving. In April 2012 the government published the revised personal income tax code and corporate tax code in both Myanmar and English on the IRD’s website. The IRD also published and distributed a simplified guide to taxpayers for calculating taxable income based on the revised tax code for all major tax instruments. This was followed by a countrywide taxpayer outreach program, including training to tax officials on application of the revised tax laws. 3.20. Transparency of revenue management could be enhanced through greater control over the use of off-budget “other accounts” held in the Myanmar Economic Bank (MEB) for ministry and SEE transactions. Other account receipts in 2011/12 were estimated at MMK 2.54 trillion (44 percent of total Union revenue), and expenditures at MMK 2.26 trillion (28 percent of Union expenditure). Although financial regulations only allowed the Ministries of Defense and Home Affairs to open such accounts, others have been allowed to do the same upon request. The Ministry of Defense held the largest number of other accounts, followed by the Ministry of Health. The Ministry of Health, for example, had 427 other accounts, which were used for receipts from community cost sharing, hospital equity funds, and interest on trust funds (each of the 330 townships has an other account). 3.21. It would be useful to MoF if these “other account” receipts could be accounted for in gross terms in the Union Budet in order to ensure a comprehensive picture of fiscal policy. Aside from criticality from a fiscal transparency perspective, this change could also enable greater budget flexibility by redirecting previously off-budget revenues to priority needs across the Union budget. and more efficient cash management. For example, liquidity constraints would not need to be alleviated through short-term borrowing. To achieve this, it would be effective if the authorities closed all “other accounts” and merged associated revenues all other Union Budget receipts. E. Functional Composition of Expenditure 3.22. The functional composition of spending in the Union Budget in 2009/10 was misaligned relative to the country’s development priorities. Spending for general public services and defense far exceeded the average share in middle-income countries. Conversely, expenditure shares were well below lower middle-income average for social protection, public order and safety, education, health, and housing and community amenities (Figure 3.7). 3.23. Spending on social services increased very rapidly in between 2009/10 and 2013/14, reflecting the country’s shifting priorities. For most of Myanmar’s history, the budget had been decided by the government. In 2012/13 however, the new Parliament debated its first budget. At the same time, spending on projects associated with Nay Pyi Taw were winding down. The expenditure share for general public services correspondingly declined from 45.4 percent in 2009/10 to 23.0 percent in 2013/14 (Figure 3.8), which is closer to, though higher than, the lower middle-income average. This made room for a rapid expansion in spending on social services. The expenditure share for education almost doubled, increasing from 7.2 percent in 2009/10 to 12.5 percent in 2013/14. The share for health was more than tripled, increasing from 1.7 percent to 6.4 percent over the same period. The higher expenditures financed several policy changes, among them the elimination of some household out-of-pocket costs for educational and healthcare expenses, and more teaching and healthcare personnel. These and other changes are discussed in further detail in chapters 4 and 5 of the PER. Spending on social protection also 31 increased, from 1.5 percent to 4.8 percent of total Union expenditures but still remained well below international norms (Figure 3.9). Figure 3.7: Government Function 2009/2010 Figure 3.8: Government Function 2013/14 710 Social Protection 1.5 710 Social Protection 4.8 701 General Public Service 45.4 701 General Public Service 23.0 704 Economic Affairs 704 Economic Affairs 709 Education 7.2 709 Education 12.5 707 Health 1.7 707 Health 6.4 702 Defense 20.7 702 Defense 28.7 703 Public Order and Safety 703 Public Order and Safety 706 Housing etc. 706 Housing etc. Myanmar 2013/14 Myanmar 2009/10 708 Recreation etc. 708 Recreation etc. Lower Middle Income Lower Middle Income 705 Environmental Protection 705 Environmental Protection 0 20 40 60 0 10 20 30 40 Percent of total expenditures Percent of total expenditures Sources: Ministry of Finance BOOST database and Annex 2. 3.24. Defense is an important public good. Yet, military spending remains high and necessarily limits the expansion of other public services. At around 4.2 percent of GDP planned for 2013/14 (Figure 3.10), it remains above the lower middle-income average of 2.3 percent of GDP. Average defense shares are lower still for upper middle-income countries and high income countries. Figure 3.9: Social Protection Spending Figure 3.10: Military Expenditures 25 21.4 Expenditure Shares GDP Shares GDP and expenditure shares (%) 20 35 GDP and expenditure shares (%) 28.7 30 15 Expenditure Share GDP Share 25 20.7 20 10 7.2 15 4.8 5 10 6.9 4.2 1.5 5 2.3 1.9 0.1 0.7 0 0 Lower Middle Myanmar Myanmar Lower Middle Myanmar Myanmar Income Countries 2009/10 PA 2013/14 BE Income Countries 2009/10 PA 2013/14 BE Sources: Ministry of Finance BOOST database and Annex 2. 3.25. Spending on economic affairs was around 70 percent, equivalent to 1.8 percent of GDP, for capital investments in 2013/14. Of this amount, 1.4 percent of GDP was allocatted to infrastructure and 0.4 percent of GDP was allocatted to commercial activities.23 It may be worth reviewing the latter in 23 Includes agriculture, livestock and fisheries, cooperatives, mining, industry, hotels and tourism. The number could be higher to the extent that some activities in energy, transportation, and communications could potentially be conducted by private providers as well. 32 greater detail to ensure that these are not activities better done by the private sector. Approximately 30 percent of spending on economic affairs were for recurrent needs, most likely regulatory functions. F. Economic Composition of Expenditure 3.26. The economic composition of spending was misaligned at the beginning of the review period. Capital investments in 2009/10 accounted for around 63 percent of the Union Budget compared to an average of 18.6 percent for lower middle-income countries (Figure 3.11). This crowded out other critical expenditure including employee compensation, goods and services, and social benefits. Figure 3.11: Economic Classification, 2009/10 Figure 3.12: Economic Classification, 2013/14 21 Employee Compensation 15.9 21 Employee Compensation 20.5 22 Use of Goods and Services 8.7 22 Use of Goods and Services 17.7 63.5 31 Capital Investment 38.3 31 Capital Investment 1.7 27 Social Benefits 5.1 27 Social Benefits 25 Subsidies and Transfers 25 Subsidies and Transfers Myanmar 2009/10 24 Interest Payments Myanmar 2013/14 24 Interest Payments Lower Middle Income Lower Middle Income 26 Grants 26 Grants 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Percent of total expenditures Percent of total expenditures Sources: Ministry of Finance BOOST database and Annex 2. 3.27. Capital expenditures as a share of the Union Budget declined rapidly over the review period from from 63.5 percent of the total in 2009/10 to 38.3 percent of the total in 2013/14 (Figure 3.12). This share remained above the average for LMICs (Figure 3.13). The share in other countries tends to fall with development and concurrent growth in private capital investments. The composition of capital investment shifted substantially as well, falling in general public services, while increasing in defense and economic affairs. Capital spending for social services increased substantially too, from a low base, in 2012/13 and 2013/14. 3.28. Spending on employee compensation increased rapidly over the review period, going from 15.9 percent of total Union spending in 2009/10 to 20.5 percent in 2013/14 (Figure 3.14).24 The military had the largest share of employee compensation and made the biggest gains over time. The next largest expansion occurred in the education sector, which also had the most employees in 2009. A portion of these increases can be attributed to salaries and allowances, which went up three times. Allowances were increased by MMK 20,000 per employee in 2011/12, salaries were increased by MMK 10,000 per employee in 2012/13 and allowances were pushed up again in 2013/14 by MMK 20,000. Another portion of the increase cost comes from new hires particuarly in education and health (see chapters 4 and 5). Overall, the MoF could ensure future affordability of the wage bill by linking any further increase in salaries 24 Employee compensation is defined here as salaries, honoraria and allowances including travel allowances. The share of total military spending on employee compensation is assumed to be the same as the share of total non- military spending on employee compensation. 33 or benefits to a clear strategy for pay and grading in the broader context of staffing needs, the MTFF and allocational trade-offs (Box 3.1). Figure 3.13: Capital Expenditure Figure 3.14: Employee Compensation 70 30 24.8 60 63.5 25 20.5 50 Shares (%) Shares (%) 20 38.3 15.9 40 15 30 18.6 10 7.8 20 5 3.0 10 5.9 5.8 5.6 1.4 0 0 Lower Middle Myanmar Myanmar Lower Middle Myanmar Myanmar Income Countries 2009/10 PA 2013/14 BE Income 2009/10 PA 2013/14 BE Countries Expenditure Shares GDP Shares Expenditure Shares GDP Shares Sources: Ministry of Finance BOOST database and Annex 2. 3.29. Available data suggests a misalignment between the size of the civil service and Union expenditures on capital investment. The total number of people employed by the Union government in 2009 was 883,215 compared to a population of 51.5 million people (Table 3.2). This was the equivalent of 1.7 percent of the population compared to 0.9 percent for low income countries and 2.9 percent for lower middle-income countries (Figure 3.15). More important, however, is whether the civil service was able to perform all the duties and functions needed for an ambitious program of national development. Much more research is required to answer such a question. One hint can be found by looking at the ratio of spending on capital investment to spending on employee compensation. The Myanmar ratio (1.5 to 1) looks very high relative to other country groups (Figure 3.16). The MoF and Ministry of National Planning and Economic Development (MNPED) could explore this further by reviewing Public Investment Management practices and capital expenditure efficiency, including the scope for savings through reduced spending on commercial activities in economic services. Table 3.2: Government Employment 2009 Total Employed Women Employed Women (Percent) Total 883,215 454,190 51.4 Ministry of Education 336,567 257,345 76.5 Ministry of Agriculture and Irrigation 67,768 24,478 36.1 Ministry of Health 58,107 37,184 64.0 Ministry of Energy 23,106 3,098 13.4 State Administrative Organizations 26,462 13,469 50.9 All Others 371,205 118,616 32.0 Source: Central Statistical Organization, 2010. 34 Box 3.1: Size of the Wage Bill – Lessons from ECA Countries The non-defense public sector wage bill in Myanmar increased by around 40 percent in the 2015/2016 Budget.a Changes to the wage bill will have important implications for the overall fiscal stance and the government’s ability to implement counter-cyclical policies. Upward or downward adjustments to government spending during periods of economic recession or boom respectively becomes more difficult if non-discretionary spending on wages take up a large share of total expenditure. A study by Eckardt and Mills (2014)25 of European and Central Asian (ECA) countries looking at the behavior of the aggregate public sector wage bill across business and electoral cycles, and its impact on public finances, point to relevant issues for Myanmar. The size of the wage bill in ECA varies a lot from 4 percent of GDP in Kazakhstan to 14 percent in Montenegro. Most countries spend between 20-30 percent of the budget on wages compared to around 17 percent in Myanmar. Between 2000 and 2008, the average annual growth of wage bill expenditure in ECA was 9.4 percent in real terms, compared to 48 percent in Myanmar between 2012/13 and the 2015/16 Union Budget. With the onset of the global financial crisis in 2008, many ECA countries had to drastically cut the size of their wage bills. In Latvia for example, real wage spending in 2011 was cut to 65 percent of the 2008 level. Cuts were implemented through nominal wage freeze, nominal wage cuts, hiring freeze, and retrenchment/rationalization. The study finds that the size of the wage bill in ECA countries is highly pro-cyclical in relation to both business and electoral cycles, and is more pro-cyclical than other types of government expenditure. This is consistent with developments in Myanmar where the wage bill has grown rapidly with recent economic recovery, and spiked further during an election year. Another, perhaps more surprising, finding is that the cuts implemented to the wage bill during the economic downturn was sharper than the increase during the high growth years of 2000-2008. One possible explanation for this finding is that nominal freezes lead to erosion of real wages, similar to past practice in Myanmar. The study also finds that wage bill expansions are associated with a deterioration of the overall fiscal position – a one percentage point increase in the wage bill as a share of GDP increases the fiscal deficit by half a percentage point. This is also consistent with Myanmar: a higher wage bill in 2015/16 was accommodated with cuts to other spending, but the cuts were not enough to avoid a higher overall deficit. The wage bill in Myanmar does not pose an immediate concern in terms of crowding out discretionary expenditure and constraining counter-cyclical policies. The general government wage bill (excluding defense and SEEs) is around 20 percent of general government recurrent expenditure.b There is also a clear need to compensate public servants for loss in real income. Going forward however, any increase should be linked to medium-term sustainability and to pay and grading reform/performance considerations. __________________ a. Medium-term fiscal framework. b. Ibid. 25 Eckardt and Mills (2014), “What Goes Up Must Come Down: Cyclicality in Public Wage Bill Spending,” World Bank. 35 Figure 3.15: Wage Bill and Civil Service Employment Figure 3.16: Ratio of Investment to Wage Bill 10 1.6 1.5 Civil service as a percentage of population HIC 9 1.4 8 OEC 1.2 (2002-2008 average) 7 6 UMC 1.0 0.8 5 0.8 4 0.6 0.5 3 Myanmar LMC 0.4 0.3 2 1 LIC 0.2 0 0.0 10.0 20.0 30.0 40.0 Lower Middle Upper Middle High Income Myanmar Wages and salaries as a percentage of total expense Income Income Countries (Gen. Gov.) (2002-2008 average, 2013/14 for Myanmar) Countries Countries Sources: Mills (2011) and UN Population Division (2012). Sources: MoF and World Development Indicators. 3.30. Expenditures on goods and services also expanded rapidly from 8.2 percent of total spending in 2010/11 to 14.0 percent of the total in 2012/13 and an expected 17.7 percent in 2013/14. Defense consumed the greatest portion of goods and services and also grew the fastest, particularly between 2011/12 and 2012/13. Other large gains were made in health, education and economic affairs. The Union expenditure share intended for goods and services in 2013/14 is similar to the averages in the middle- income countries, though GDP shares are lower (Figure 3.17). Figure 3.17: Goods and Services Figure 3.18: Social Benefits 25 25 20.1 20 17.7 20 Shares (%) Shares (%) 15.1 15 15 10 8.7 10 6.3 4.7 5.1 5 2.6 5 0.8 1.7 0.1 0.8 0 0 Lower Middle Myanmar Myanmar Lower Middle Myanmar Myanmar Income 2009/10 PA 2013/14 BE Income 2009/10 PA 2013/14 BE Countries Countries Expenditure Shares GDP Shares Expenditure Shares GDP Shares Sources: Ministry of Finance BOOST database and Annex 2. 3.31. Social benefits expenditure26 expanded sharply but remained below international levels. This was mainly social security and pensions rather than social assistance. Their combined expenditures increased from 0.2 percent of GDP in 2009/10 (1.7 percent of total expenditures) to a target of 0.8 percent 26 IMF GFS codes for social assistance benefits (272) and employer social benefits (273). 36 of GDP (5.1 percent of total expenditures) in 2013/14.27 This is substantial, but still far below the average for LMICs (Figure 3.18). 3.32. The number of people with pension coverage has been very limited. Most of the spending on social benefits consists of pensions and commuted pensions for government. According to MoF data, there were 916,199 government workers who had pension coverage in 2010. This was the equivalent of only 2.7 percent of the working age population. At the time of the review, there was no government operated or mandated pension scheme for the private sector and very few private sector employees have private pension coverage. As a result, the majority of the population, particularly the poor and vulnerable, has no access to predictable social protection. Instead, they have to rely on informal coping mechanisms and ad hoc assistance from donor-financed programs. 3.33. The government has been considering reforms to its civil service pension scheme but key details were undecided at the time of this review. Fundamental decisions under discussion included whether to proceed with defined contributions, or defined benefits, unfunded or partially funded, and whether to include all employees or only the new employees from 2015 onward. As a consequence, it was not possible to know the costs involved. The authorities are encouraged to ensure that the chosen pension system is prudently managed and fiscally sustainable. An expanded system will bring welcome benefits but could become a substantial government liability if not well managed. A particular risk is associated with the idea of defined contributions flowing into a central provident fund. 3.34. The establishment of a provident fund fed by defined contributions would result in a significant fiscal outlay in the short run and a rapidly growing fund that would, in a few years, become the largest single institutional investor in the country, posing significant risks. The proposal also envisions the possibility of investing in equities, corporate bonds and unspecified ‘other’ assets. At present however, capital markets in Myanmar are underdeveloped and offer no liquid securities or other assets that would normally be included in a pension fund’s portfolio. For this reason, a special government bond issued expressly for the provident fund and linked to bank deposit rates is preferable to investing in illiquid assets. This is the approach taken in Singapore, where the government issues special bonds to the Central Provident Fund.28 3.35. Social security coverage has been very limited but will expand. The 1954 Social Security Act was replaced in August 2012 by a new act that mandates a new pension scheme for the formal private sector. Under the new law, all citizens are eligible to register for social security, but enrollment is not automatic. Even so, coverage is expected to grow substantially beyond the estimated 650,000 beneficiaries in 2012 or about 1.7 percent of the labor force. 3.36. The new social security benefits program had not been costed, appeared overly generous and may strain government capacity. Most governments at Myanmar’s level of development cannot afford to provide unemployment benefits or publicly-funded housing as mandated by the 2012 Social Security Act. Administrative costs could be high: the law sets a cap of 25 percent which is far higher than the 1 percent seen in Singapore, 2 percent in Thailand, Malaysia and Korea and 10 percent in the Philippines. 27 According to notifications 106/2011 and 107/2011 by the Ministry of Finance and Revenue, retired State service personnel and political pensioners were allowed to draw increased pension rates beginning July 2011. 28 Singapore’s CPF also guarantees a nominal return of 2.5 or 4 percent per annum (depending on the type of account). This guaranteed return is lower than the actual return on investments which are made by the Singapore Investment Corporation. 37 The overall costs could be high if the new scheme is implemented as a government-funded defined benefits program. Alternatively, costs could be held down by setting up individual accounts under a contributory scheme. There is also a need for better coordination between the Social Security Board and the Ministry of Health in order to eliminate overlapping health care systems. Most fundamentally, the new law needs to tighten up its fiduciary regulations in order to ensure funds are safely invested and costs are contained. 3.37. It will be useful to finding a good balance between social security and social assistance for poor and vulnerable populations, backed by complementary policies for economic development and poverty reduction. The diversity of needs and coping mechanisms by the populations in each State/Region raises interesting challenges for shaping a national social protection approach. Recently introduced government programs (e.g., stipends) are a good way to begin expanding social assistance. Yet, some caution is justified about creating unsustainable social protection entitlements through large national programs that might exceed government capacity or resources. Sequencing and phasing will be critical. G. Conclusions and Recommendations 3.38. The government has made good progress in gradually realigning the Union Budget towards the development priorities. Efforts at mobilizing government revenue has generated early results. Reaching levels comparable to other countries will be a long-term endeavor. There are nonetheless some short- term measures that could contribute to improved revenue collections without adversely affecting equity and efficiency. On the expenditure side, the government has effectively started to address major needs in the social sectors by reallocating away from areas such as general public services in order to create fiscal space. There may be additional potential, which warrants more detailed review for example in the defense or capital expenditure budgets. The increase in the wage bill is large by most standards. While not an immediate source of concern, experience from other countries shows that downward adjustments during periods of recession can be more severe than increases during high growth periods. Issues Options Higher and more stable revenues Short-term: MoF could review and list all exemptions, assess the tax would help pay for more expenditure impact, rationalize exemptions, and consolidate process for infrastructure and social services granting any new exemption. Enhancing efficiency of tax Short-term: MoF could establish a system for rigorous analysis of tax arrears administration could enhance and prepare effective arrears management plan. revenue collection. Medium-term: The MoF could develop and implement a plan to gradually eliminate practice of “other accounts”, bringing associated receipts on - budget. Create fiscal space for priority Medium-term: The authorities could initiate an expenditure review of expenditure through further military expenditures to identify potential efficiency gains. reallocation across the Budget. Create fiscal space for priority Short-term: The MoF and MNPED could review Public Investment expenditure through efficient Management practices and capital expenditure efficiency, including the capital spending. scope for savings through reduced spending on commercial activities in economic services. Ensure sustainability of the wage Short-term: The authorities could analyze the impact of different wage bill and alignment with public reforms on the overall wage bill, and how best to link wages to sector performance. performance. 38 IMPROVING THE COVERAGE, 4 QUALITY AND EQUITY OF EDUCATION SERVICES Background: Many decades of government under-spending on education and prohibitive household financing requirements have left major gaps in education outcomes. The government has initiated several reforms since 2012/13 to improve the situation. Objectives: The objectives of this chapter are to analyze government expenditures and policies for basic education between 2009/10 and 2013/14, and to propose options for policy reforms and spending to improve coverage, quality and equity of education services. Key findings: Myanmar has suffered from low school enrolment and completion rates (one third of 1.2 million students enrolled in grade 1 made it to grade 11, and only one third of those passed the school leaving exam); poor learning outcomes (9 percent of a third grade class in Yangon cannot read a single word); and inequalities in access and quality (net primary enrolment as low as 69 percent in poorer areas compared to 85 percent average nationally). Reforms initiated since 2012/13 include elimination of primary and secondary school fees, introduction of compulsory primary education, hiring 79,000 more teachers, expansion of the stipend program for poor students, and delegation of some spending authority to district and township education offices and to school head masters through block grants. To implement these policies, the government quadrupled the budget on education between 2011/12 and 2013/14, albeit from a low base. The MoE is now grappling with pressures to deliver more despite: (i) limited administrative, budget and performance data as a basis to prioritize spending; (ii) fledgling capacity to analyze policy and budget linkages; and (iii) fiduciary risks arising out of large cash payments and weak procurement rules. In addition, pressures are emerging for other important needs including teacher training; adding more years of schooling; ensuring wider coverage to include marginalized groups; and Early Childhood Development. Another emerging issue within education is how much to decentralize and how soon. Recommended options: (i) MoE to use standardized assessments of student learning outcomes to monitor the quality of the education system. As a starting point, MoE could use the Early Grade Reading Assessments piloted in 2014 in the Yangon region; (ii) MoE and MoF to carry out thorough costing of new initiatives (e.g. more school years, more teacher training, modernized curriculum, coverage to disadvantaged groups, early childhood education) and propose resourcing needs with implementation plans; (iii) strengthen implementation through more procurement training; (iv) to empower district officers or school principals to take decisions within a framework of accountability for results to be developed under the leadership of MoE; (v) carry out a Public Expenditure Tracking Survey to ensure schools are receiving allocations in accordance with the Ministry’s guidelines (i.e. in accordance with the formula and the time table specified in those guidelines). 39 CHAPTER 4 IMPROVING THE COVERAGE, QUALITY AND EQUITY OF EDUCATION A. Background 4.1. The government has initiated profound changes in the education sector since 2011 to help redress decades of under-spending and prohibitive out-of-pocket payments that have led to low school enrolment and completion rates; poor learning outcomes; and inequalities in access to and quality of education across the country. To deal with these challenges, the education system is transitioning to a new approach. The Ministry of Education (MoE) is taking on a stronger policy role. Public spending on education is rising rapidly, having already been quadrupled from a low base between 2011/12 and 2013/14. 4.2. The PER aims to assist the government in further prioritizing spending and policy reforms to gradually improve the coverage, quality and equity of education services. It focuses on basic education, which accounts for around two thirds of government spending on education as a whole. It starts with an overview of how the sector is organized for the regulation and delivery of education services. It highlights some of the big gaps in education outcomes. It looks at the government’s recent strategy to fill education gaps, and measures that could further strengthen that strategy. It then analyzes how government spending has responded to this new strategy and suggests options for further prioritization. It ends by reviewing the current arrangements for budget preparation, implementation and reporting in the sector, with recommendations to manage the rapidly growing inflows of resources in the sector. B. The Education Sector 4.3. Almost all educational services in Myanmar are provided by the Union government, with a small role for sub-national governments. Education is clearly specified as a Union responsibility by the 2008 Constitution. Thus, the administration and funding of the education sector has been centralized within Union ministries, mainly the Ministry of Education. As such, new reforms for the delegation of functions to deconcentrated agents of the MoE (state or district education officers) have advanced further than ideas about what functions to devolve to state and region governments. Non-governmental and private sector providers are almost absent. 4.4. The Ministry of Education (MoE) accounts for more than 80 percent of Union spending in the sector. MoE has responsibility for: (i) administration of basic education services (grades 1-11) through a Department of Basic Education (DBE)29; (ii) planning and training which includes curriculum preparation, teacher training, publication of learning materials, compilation of statistical yearbooks, and costing of new policies; (iii) running the national matriculation exam; and (iv) running public universities through a Department of Higher Education. In addition, fourteen ministries provide vocational and higher education services in their own sectors with funding from their own budgets (Table 4.1). For example the Ministry of Health runs and finances health related universities. 29 Three regionally-based Departments of Basic Education were consolidated into one DBE on April 1, 2015. 40 Table 4.1: Education Expenditure by Different Ministries 2013/14 2013/14 Budget Estimates, (MMK millions) Current Capital Percent of Expenditure Expenditure Total total Education Sector 752,392 335,755 1,088,147 100.0 Education 685,376 222,534 907,910 83.4 Science and Technology 34,694 63,188 97,882 9.0 Health 10,990 24,891 35,881 3.3 Agriculture and Irrigation 2,624 13,344 15,968 1.5 Culture 7,559 4,165 11,723 1.1 Border Affairs 6,806 4,141 10,947 1.0 Cooperative 1,631 1,638 3,268 0.3 Transport 1,010 497 1,507 0.1 Livestock Breeding and Fisheries 559 739 1,299 0.1 Religious Affairs 701 596 1,298 0.1 Rail Transportation 442 22 464 0.0 Defense NA NA NA NA State Administrative Organizations NA NA NA NA Environmental Conservation and Forestry NA NA NA NA Industry NA NA NA NA Source: MoF BOOST database, with World Bank staff’s attempt at identifying departments across ministries which focus on “education.” 4.5. Sub-national authorities (States and Regions) had a very limited role in education as of November 2014, but this could change under the new education law. Some limited steps have been taken to delegate decision-making of the day-to-day running of schools. For example, since 2012 township and district education officers received increased responsibility for implementing and managing both recurrent and capital parts of the budget. Another example is that starting in 2013/14 MoE schools received part of their non-wage recurrent expenditures as a “block grant”, which they spend under the supervision of a newly established school grant committee. Nonetheless, sub-national authorities do not own or operate any pre-schools, primary schools, universities, training centers. They therefore do not receive dedicated funds for education from the Union.30 In principle they can decide to use some of the transfers they receive from the Union government (totaling around $600 million in FY2012/13) for education. However, a lack of data makes it difficult to assess this. 4.6. Non-governmental actors including monasteries and private schools play a marginal role in the education sector. The MoE is by far the largest provider. Others include the Monastic Education School System (less than 1 percent of total enrollment) and other non-state providers. In the school year 2012/13, the education sector supported by the MoE included about 35,500 primary and post-primary schools, 3,000 middle or lower secondary schools, and 2,300 high or upper secondary schools. In these schools, about 5.2 million students were enrolled in primary school (grades 1-5), 2.3 million in middle school (grades 6-9) and 0.7 million in high school (grades 10-11). Enrollment in private primary and middle school is almost non-existent, but accounts for approximately 5 percent of enrollment in high school. The MoE has no experience in sub-contracting services from the private sector. 30 There were cases where the central government has passed resources on to the office of the state or region level Chief Minister to hire “contract” teachers. 41 C. Education Challenges 4.7. Educational outcomes, for much of Myanmar’s recent history, have been below potential. Older generations completed fewer years of education (Figure 4.1). Three dimensions of educational outcomes are of special concern: (i) coverage so that more children could be enrolled in and attending school; (ii) quality to ensure more children could be gaining basic foundational skills and competencies in reading and arithmetic; and (iii) inter-regional equity so that children of every income level, ethnic group, gender, and geographical area have access to good quality education. Figure 4.1: Educational Attainment by Age Cohorts 2009/10 100% 90% 3 Higher Education 6 7 9 80% 8 11 Completed Upper-Secondary 9 Percent of age cohort 70% 13 12 Upper Secondary 60% 16 36 17 Completed Lower-Secondary 50% 42 16 40% 42 Lower Secondary 41 30% 37 30 Upper Primary 31 20% 23 17 Lower Primary 10% 10 12 7 0% No School 25-29 30-39 40-49 50-59 60-69 70+ Age cohorts Source: World Bank staff estimates based on IHLCS 2009/10. 4.8. In terms of coverage, Myanmar (along with Brunei Darussalam and Indonesia) has one of the lowest enrollment rates in the region at all levels of education. Basic education includes primary schools (G1-5) and secondary schools. Secondary schools are divided into middle schools (G6-9) and high schools (G10-11). Data for enrollments vary widely. For instance, estimates of net primary enrollment rates vary between 85 percent in 2012 (MoE) and 93 percent in 2010 (IHLCS, Figure 4.2). Even if the larger estimate is correct, it remains well below what many of Myanmar’s regional peers have achieved. Figure 4.2: Estimates of Net Primary Enrolment Rates 100 98.0 98.1 98.3 98.4 98 95.9 96 Percent 94 93.1 92.2 91.7 92 90 88 Brunei Indonesia Myanmar Lao PDR Malaysia Vietnam Thailand Cambodia Darussalam (2012) (2010) (2012) (2012) (2012) (2011) (2012) (2012) Sources: World Bank EdStat, except Malaysia, Thailand and Myanmar (IHLCS 2010/11). 42 4.9. In terms of quality, there are clear signs of weaknesses based on school leaving examination results, used both to graduate students from high school and to manage their placement in universities. Only 460,000 out of 1.2 million students starting in grade 1 eventually make it to 11th grade. Most of them leave school or drop out along the way. Since 2000, approximately 300,000 11th graders have taken the exam each year. Every year, the national average pass rate has hovered around 33-36 percent, with wide variance across the country (Figure 4.3). Others can repeat 11th grade and attempt to pass the exam the following year. Thereafter, they have to quit school and can attempt to pass the test on their own. Figure 4.3: 11th Graders Who Passed School Leaving Test in 2010 Enrolled in 11th grade Appeared for 11th grade test Pass rate (%) Numbers enrolled, appeared for test 200,000 45 50 180,000 39 40 45 37 38 160,000 35 35 36 36 36 36 40 34 34 140,000 31 35 29 Pass rate (%) 120,000 25 30 100,000 25 17 80,000 20 60,000 15 40,000 10 20,000 5 0 0 Ayeyarwa… Kayah Tanintharyi Rakhine Shan (S) Chin Magway Bago (E) Mandalay Kachin Kayin Sagaing Shan (N) Yangon Mon Shan (E) Bago (W) Sources: Central Statistical Organization, Statistical Yearbook 2009 (table 17.01 and 17.03 for state-level pass rates), Ministry of Education, Department of Planning and Training: Statistical Yearbook March 2010. 4.10. Many students are not learning the ability to read. An Early Grade Reading Assessment (EGRA) on a sample of schools in the Yangon region in 2014, where results are expected to be more favorable, revealed that as many as 9 percent of a third grade class cannot read a single word (Figure 4.4). For students who were able to read some words, 12 percent of them could not answer simple questions about the text, suggesting lack of understanding. In February 2012, UNICEF supported a government assessment of almost 23,000 students across 868 schools in the Myanmar language and mathematics. More than 75 percent of third graders could only master the most basic operations. Figure 4.4: Early Grade Reading Assessment, Yangon Region 2014 90% 77% 80% % of student who could not read a single word 70% 60% % of student who could not answer a single question correctly about a text they just read 50% 37% 40% 27% 30% 20% 12% 10% 9% 10% 0% Grade 1 Grade 2 Grade 3 Source: World Bank Myanmar Early Grade Reading Assessment (EGRA) for the Yangon Region , 2014. 43 4.11. In terms of equity, there is wide variation in education indicators across the country, with the poorer states lagging behind in virtually all dimensions (Table 4.2). For instance, Chin had one of the lowest net primary enrollment rates in 2009/10 and some of the poorest quality of education. Of the Chin students that remained in school to the 11th grade in 2010, only 17 percent passed the final examination. Within rural areas, the highest net primary enrollment rate in 2009/10 was 96 percent in Kayah and the lowest was 69 percent in Rakhine. The highest pass rate in the 2010 final examination was 45 percent in Mon and the lowest was 17 percent in Chin. Table 4.2: Education Outcomes in 2010, by States and Regions Incidence (%) Primary School Enrollment Rates (%) Pass Rate (%) Poverty Urban Rural Total 11th Grade Test Chin 73 91 85 86 17 Rakhine 44 89 69 71 31 Tanintharyi 33 87 85 85 37 Shan 33 93 86 87 36 Ayeyarwady 32 94 87 88 34 Kachin 29 94 92 93 29 Magway 27 94 92 92 35 Mandalay 27 91 91 91 36 Bago 18 88 84 85 37 Kayin 17 80 88 87 34 Mon 16 83 89 88 45 Yangon 16 94 91 93 36 Sagaing 15 91 95 94 39 Kayah 11 96 96 96 35 Total 26 92 87 88 36 Low 11 80 69 71 17 High 73 96 96 96 45 Gap 62 16 28 25 28 Sources: 2011 Integrated Living Conditions Survey and Ministry of Education Department of Planning and Training: Statistical Yearbook March 2010. 4.12. Myanmar’s children from poorer and rural households are less likely to enroll, complete, and learn in school (Table 4.3). For example, 46 percent of children from the richest income quintile attend pre-school, compared to only 7.6 percent of children from the poorest quintiles. In Myanmar, almost 95 percent of children from the richest income quintile attend primary school, compared to only 81.4 percent of children from the poorest quintiles. The comparable numbers in Vietnam are 95 percent enrollment for the richest quintile and 94 percent for the poorest quintile. Table 4.3: Education Outcomes, by Income Quintiles, Cross-Country Percentage of children aged 3-5 attending early childhood education Primary net enrollment rate 2009/10 MICS 2006 MICS 2006 MICS 2009/10 MICS 2006 MICS 2006 MICS Myanmar Lao Vietnam Myanmar Lao Vietnam Income quintiles: Poorest 1 7.6 1.4 35.7 81.4 59.0 94.1 2 16.2 2.0 48.9 90.6 70.6 95.6 3 19.5 2.8 56.7 93.7 83.7 95.7 4 32.3 7.9 61.7 94.9 93.6 96.8 Richest 5 46.0 43.7 80.7 94.9 98.2 94.9 Urban 39.1 33.6 74.7 93.0 93.0 94.6 Rural 15.9 2.3 51.4 89.2 72.8 95.6 Total 22.9 7.4 57.1 90.2 79.0 95.4 Ratios of: Poor to rich 0.2 0.0 0.4 0.9 0.6 1.0 Rural to urban 0.4 0.1 0.7 1.0 0.8 1.0 Sources: Myanmar MICS 2009/10, Lao MICS 2006, Vietnam MICS 2006. 44 4.13. Students from urban and more affluent households have a much better prospect for completing more years of education. In Myanmar, an analysis of the 2010 Integrated Household Living Conditions Survey (IHLCS) indicates that students from the highest income quintile are over three times more likely to attain a tertiary education as those from the lowest income quartile. The distribution of infrastructure is important as well. In 2012, financial constraints were frequently cited by Township Education Officers as barriers to the adequate provision of functioning latrines and clean drinking water in schools.31 Many children were unable to attend school because they had to spend time fetching water for their families. Results from the IHLCS shows that States and Regions with better rural access to safe water in 2009/10 tended to also have higher enrollment rates and higher pass rates on the 2010 school leaving exam. D. Policy Priorities in Education 4.14. The government in 2011 adopted a “10 points Education Policy” to reform education policy and address some of the above challenges. The ten points (Box 4.1) have been the motivation for several of the MoE’s initiatives in the past three years, including eliminating school fees in primary schools (in the school year 2013/14) and middle school (in 2014/15), increasing teacher salaries, and working more closely with development partners, among others. Box 4.1: 10 Points Education Policy  Implement a free, compulsory primary education system.  Increase the enrollment rate in basic education sector.  Nurture new generations as intellectuals and intelligentsias in the development of human resources.  Improve capacities of teachers in both basic and higher education sectors.  Utilize teaching aids more effectively.  Upgrade the quality and socio-economic status of educational personnel.  Provide scholarships, stipends and awards both locally and internationally.  Promulgate relevant laws for the participation and contribution of private sector in education sector.  Collaborate with international and local organizations including the United Nations and international non-government organizations (INGOs) and non-governmental organizations (NGOs).  Upgrade education standards to international levels. 4.15. It would be helpful to have a clearer focus on the quality of education and equity. Although the 10th item in the Ten Points alludes to the importance of quality, this fundamental objective seems to have received inadequate attention. For example, in the MoE’s annual reports, the sections on “quality” do not contain any information on operationally defining the quality of education (as measured by assessments of students’ learning outcomes); rather, they discuss the types of interventions that might help improve learning (e.g. child centered learning, upgrading the curriculum). To promote equity, the MoE will need to be more inclusive and proactively seek to attract and support students from socio-economically weaker backgrounds, disabled children, and children from ethnic minorities. 4.16. Achieving a clearer focus would involve choosing measurable outcomes that correspond to quality and the develop assessment tools to monitor gaps in achieving those quality outcomes (i.e. to ensure equity). For Myanmar, a “tangible target” (or “learning standard”) would be to focus on reading and numeracy skills for first, second and third graders. For instance, the measure of progress could be the number of words per minute a child can read. The MoE could set a target of ensuring that all children can 31 UNICEF (2012). Situation Analysis of Children in Myanmar. 45 read at least one word by the end of Grade 1. At the time of this review, 27 percent of children at the end of Grade 1 could not read a single word. In the early grades of schooling, the Early Grade Reading Assessments (EGRA) and Early Grade Mathematics Assessments (EGMA) could be used for this purpose.32 4.17. The government would benefit from investing early in better data and analysis on coverage, quality and equity to prioritize spending towards the 10 Points Policy. Data gaps pose challenges for effective policy-making, especially because most important decisions are made at the Union level rather than at the sub-national level where officials would be better informed of local needs and issues. Establishing a solid framework for data compilation and analysis that feeds into policy-making is a long- term endeavor (Box 4.2). In the short to medium-term however, the MoE could enhance the capacity for policy analysis by maintaining three critical databases: (i) Education Management Information System to electronically record administrative data, which is currently collected in paper format; (ii) student learning outcomes, and (iii) budget and expenditure data. Box 4.2: Data Challenges  Education Management Information System: MoE currently collects monthly and annual administrative data (e.g. number and type of schools, teachers and students), compiled in a Statistical Yearbook published with 12 months delay. An MIS would improve quality, timeliness, and accessibility of this important information.  Household surveys: These are important sources of estimates on enrollment, drop-out rates, household spending on education, and returns to education among other things. Existing surveys are dated (2009/10) and do not cover the entire country. Census data and new household survey data will become available in 2015. These data provide invaluable information on enrollments and drop outs. To be useful, though, staff in MOE need training to be able interpret and use these data for policy making.  Accounting for different education levels: Many schools that integrate multiple levels (i.e. primary, middle and high school) do not account for spending at each tier (e.g. nearly all middle schools are also primary). Going forward it will be important to separate cost of each tier to accurately assess costs per student at each level.  Assessment of student learning outcomes: There are currently no regular reviews of whether schooling translates into learning at different stages of the education cycle. The Early Grade Reading Assessments (EGRA) and an Early Grade Mathematics Assessments (EGMA) could be instituted to start the process for grades 1-3.  Government spending: Key to monitoring the effectiveness of links between policy and government spending is timely, relevant, and accurate fiscal data. This will involve review of budget classification and broader FMIS. As part of the writing of this PER, data were compiled in a consolidated BOOST database to facilitate analysis. However, this database need to be updated and actively used by budget staff in MOE – a process which will require training and experience. 4.18. The legal framework for the education sector, which has been in place for many decades is also currently being updated. The Basic Education Law and the University Education Law, in effect at the time of this public expenditure review, both date to 1973. Formal technical and vocational education (TVE) is regulated by the Agricultural, Technical, and Vocational Education Law of 1974 with amendments in 1983 and 1989 under the Ministry of Science and Technology. The Parliament and the MoE have recently begun 32 See Gove and Wetterberg (2011) for more details on EGRA. 46 updating the laws. Thus, on September 30, 2014, a new National Education Law was passed, marking an important milestone in updating the various laws governing the education system. The new law is seen as an “umbrella law” for the entire sector, with more detailed laws to follow for each of the sub-sectors. 4.19. The creation of new roles for States and Regions in the education sector is one of the more important changes under the new National Education Law. It enhances State and Regional: (i) operational responsibilities “to communicate and transfer between government and other schools”; (ii) operational responsibilities with budgetary implications, including autonomy to develop courses and language classes for different ethnic groups; and (iii) budgetary responsibility in terms of managing public finances for those education functions that are assigned to them under relevant law. 4.20. Some delegation of responsibility is warranted, particularly when it comes to the day-to-day running of schools in a large country like Myanmar, where the center may not be adequately attuned to local needs. Effectiveness of delegation will depend on many things, including local capacity, incentives, and accountabilities. There will also be asymmetry of information between central and local authorities, which can only be addressed in time through data reforms discussed above. Nonetheless, it is recommended to start delegating authority to local authorities against accountability for results (e.g. test scores, attendance, rates). Coordination of the framework could take place at the State/Region level. E. Sector Financing 4.21. The main sources of financing for education in Myanmar have been the Union government and households, with external donors contributing only a small share.33 Reflecting years of under- investment from public resources, an unusually high share of total education spending was funded by households in 2009/10. The IHLCS shows public resources accounted for around 31 percent, and household spending was around 63 percent of total education spending (Figure 4.5).34 This is triple the international average of 18.5 percent.35 Donor resources made up the remainder. 4.22. Out-of-pocket costs deterred many families from schooling. When households were asked why their children are not in schools, over half (55.5 percent) identified reasons related to economic and cost factors. Common responses included high school fees and indirect “opportunity” costs, such as lost services in agriculture, work, or care of a family member. Almost a third of the respondents (29.9 percent) indicated out-of-pocket costs were not affordable. 4.23. Beginning in 2012/13, the Union government has likely replaced private households as the main financier for education, although updated information on household spending will only become available as part of the household survey data collected in 2015. Enrolments may have been motivated by the 33 In 2009/10, external financing accounted for approximately US$40 million, and accounted for approximately 6 percent of spending in the sector (Figure 4.5). Donor support has grown in size since then to an estimated US$100 million or more per year. 34 The estimate of private spending obtained from the Integrated Household Living Conditions Survey carried out in 2009/10 (and covering some 18,000 households source) is roughly comparable to numbers reported in a 2010 UNICEF report in which 660 parents from 131 schools located in 24 townships were interviewed in details about their spending on education. UNICEF (2010): Household Education Expenditure Report. The UNICEF estimates for private spending by households (in the schools they visited) are roughly 10-15 percent higher (or around MMK 6000 per year) than the average figures obtained from IHLCS and seem mainly due to the fact that the UNICEF methodology included “pocket money to children” as an education -related expense whereas the IHLCS did not. 35 Psacharopolous and Patrinos (2004). 47 removal of primary school feels in 2013/14, the removal of secondary school fees in 2014/15 and the provision of free textbooks. Ministry of Education officials believe that these cost savings for households motivated an increase in the number of children enrolled in primary and secondary schools in 2014/15 by 300,000 to 8.5 million. It may be possible to do more. These tuition fees and textbooks are not the only out-of-pocket cost paid by households, as illustrated by Figure 4.6. Improved stipend programs for the poor and disadvantaged, which are beginning in 2015/16, could help cover some of these costs. Figure 4.5: Composition of Financing for Education Figure 4.6: Composition of Private Spending (cash and in-kind) Other Transportat education ion costs, costs, 2.7% 7.1% 6% Boarding, School fees, 21.9% 9.2% 63% 31% Private Contributio tutoring, School ns to the 41.5% stationeries school, , 10.3% 3.8% Total Gov't expenditure Text books, 3.6% Household Expenditure Externally financed spending Sources: World Bank staff calculations based on IHLCS 2009/10 and MoE data for MoE spending in 2009/10. Figure 4.7: Union Education Expenditures, 1972-2011 Figure 4.8: Government Spending by Country 7 6.3 3.5 6 5.3 3.0 5 Percent of GDP 4.1 3.7 3.6 2.5 4 3.3 3.5 3.3 Percent of GDP 3 2.7 2.0 2.1 2.3 2 1.6 1.6 1.5 1 0.7 1.0 0 Lao PDR [2009] Philippines [2009] Singapore 2012] Thailand [2009] ASEAN Vietnam [2008] Myanmar [2011/12] Myanmar [2012/13] Myanmar [2013/14] Brunei [2011] Malaysia [2009] Low income Indonesia [2009] Cambodia [2007] 0.5 0.0 1970 1975 1980 1985 1990 1995 2000 2005 2010 Sources: World Bank (1985), World Bank (1995) and Source: Ministry of Education and World Bank EdStat. ASEAN World Development Indicators, Ministry of Education. average excludes Myanmar. F. Ministry of Education Spending 4.24. In terms of overall levels of government spending, Myanmar has, for many decades, directed a much lower share of GDP to education than most other countries. In 1973/74, Union expenditures on education were estimated to be close to 3.3 percent of GDP. The GDP share subsequently and almost 48 consistently fell since then, reaching 0.7 percent of GDP by 2011/12 (Figure 4.7). The level of spending in Myanmar, even in 1973/74 and certainly in 2011/12, was far below that of regional peers (Figure 4.8). The decades of inadequate spending have, as discussed, above affected the quality of education. School infrastructure is very basic, and school supplies and teaching materials are highly inadequate. This makes learning conditions very difficult and affects the fundamental capacity to provide quality education. There are many opportunities for significant improvements even in the short-term (Box 4.3). Box 4.3: Important Gaps to Address through Higher Government Spending on Education  The basic education system consists of only 11 years of schooling compared with 12 or 13 in the rest of ASEAN  Despite evidence of high returns to early childhood education, there has been no budget for kindergartens.  The learning curriculum has not been updated on a regular basis to reflect strengths and weaknesses revealed from standardized assessments.  Most of Myanmar’s textbooks have not been modified in the past 10 years.  The budget to cover all the non-wage running costs for schools was very small, averaging less than $250 per school in 2009/10 leading to high out-of-pocket payments.  The budget for vocational education is very small and spread thinly across several ministries.  There is no budget for in-service teacher training and inadequate resources for pre-service training.  Until recently, the only targeted support for financially disadvantaged students was in the form of a stipends scheme and provided support to only 0.1 percent of all students.  There is virtually no system of support from children with disabilities  Some functions that are commonly found in ministries of education are missing in Myanmar (e.g. data collection, regular assessments of learning outcomes). 4.25. In 2012/13, the Parliament approved a substantial increase in the education budget. Estimated actual spending on education by all ministries increased to the equivalent of US$894 million or 1.6 percent of GDP. The budget for 2013/2014 was increased further to about US$1.1 billion equivalent, 2.1 percent of GDP, or about 14 percent of overall government spending (Table 4.4). The kyat value of spending was more than tripled between 2011/12 and 2013/14. Yet, even after these increases, Myanmar continues to allocate one of the smallest regional shares of GDP to education (Figure 4.8). Table 4.4: Total Education Expenditures 2009/10- 2013/14 a/ 2009/10PA 2010/11PA 2011/12PA 2012/13RE 2013/14BE In MMK million 257,380 331,150 372,314 778,215 1,088,639 In US$ million 247 355 462 909 1,135 As percent of GDP 0.7 0.9 0.8 1.6 2.1 As percent of total Union expenditures 8.0 7.5 8.3 10.1 13.8 a. Includes educational spending from all relevant ministries. Source: World Bank staff calculations based on MOF BOOST database (PA) means “preliminary actual estimate”; (RE) means “revised budget estimate” and (BE) means “budget estimate.” Exchange rates correspond to fiscal years and use market values throughout. 4.26. There were important shifts in the composition of spending on basic education associated with the overall increase in expenditure level. The biggest driver of spending was teachers’ salaries, which in 2011/12 constituted around three quarters of total spending on basic education (Table 4.5). The wage bill more than doubled between 2011/12 and 2013/14. The share of salaries and allowances however fell from 76 percent of total spending in 2011/12 to an estimated 62 percent in 2013/14. This was because 49 spending on goods and services and capital investments grew even more quickly than salaries and allowances. The share of spending on goods and services (and other recurrent items) went from 14 to an estimated 24 percent between 2011/12 and 2013/14, and that of capital from 10 to 14 percent over the same period. Table 4.5: Basic Education Expenditures, 2011/12 to 2013/14 2011/12 PA 2012/13 RE 2013/14 BE Total Expenditure (MMK millions) 270,927 521,392 732,590 Current Expenditure 244,063 467,889 628,331 Salaries and allowances 205,775 354,435 452,602 Other current expenditures 38,288 113,454 175,729 Capital Expenditure 26,864 53,503 104,259 Total Expenditure (% of total) 100 100 100 Current Expenditure 90 90 86 Salaries and allowances 76 68 62 Other current expenditures 14 22 24 Capital Expenditure 10 10 14 Source: MOF BOOST database. 4.27. Part of the wage bill increase was driven by higher teacher wage rates. Previously the government used to keep wages fixed in nominal terms over very long periods of time, leading to significant erosion of real income. During the period 1972-2010, nominal wages were only increased five times coinciding with changes in government: 1972, 1989, 1993, 2000 and 2006. In other words this implied a freeze in nominal terms of salaries for these periods: 1) 1972-88; 2) 1989-93; 3)1993-2000; 4) 2000-2006; 5) 2006-2010. This policy came to an end in 2011 and subsequent years when salaries were increased on an annual basis. It is estimated that 80-90 percent of the 2012/13 increase in teacher salaries was due to a 30,000 kyat monthly “bonus” added to the salaries of all government employees, including staff working in the education sector (Table 4.6).36 For a primary school teacher employed in an urban area, this monthly “bonus” amounted to a 64 percent salary increase. In addition, the salary bill rose as a result of introducing a “hardship allowance” (a doubling of the monthly salary) of teachers working in 87 remote locations in the country. However, in terms of its budgetary impact, the “hardship” allowance was much more modest, since it only affected 16,000 teachers (approximately 6 percent of all teachers). Table 4.6: Monthly Teacher Salaries by Rank April 2006- Jan. 2010- April 2013– April 2014 – Dec. 2009 March 2013 March 2014 March 2015 Starting Highest Starting Starting Starting Highest Starting Highest Level Salary Salary Salary Salary Salary Salary Salary Salary Primary School 27,000 32,000 47,000 67,000 67,000 72,000 87,000 92,000 Primary Head/ 33,000 38,000 53,000 73,000 73,000 78,000 93,000 98,000 Junior Middle School Middle School Head/ 39,000 44,000 59,000 120,000 120,000 130,000 140,000 150,000 Senior Teacher High School Head/ 80,000 90,000 100,000 120,000 120,000 150,000 160,000 170,000 Township Education Officer Source: Ministry of Education. 36 The bonus was added starting in April 2012 (the first month of the 2012/13 fiscal year), applies to all government staff and has been added for every month since April. 50 4.28. The education wage bill was also driven by a major drive for recruiting new teachers. At least 29,000 new contract teachers were hired in basic education in 2013/14 and at least another 43,000 in 2014/15 for a total new hiring of 72,000. Seeking to ensure that every primary school had at least one teacher per grade, the MoE has been rushing to hire more teaching staff. A key constraint in doing so, however, is that the 22 teaching colleges located across the country only produce approximately 10,000 graduates each year (and each teacher is trained for two years). To meet the ministry’s new more ambitious staffing norms, the ministry decided to let state and regions hire and then train, graduates from other fields of study, or even high school graduates, if they were willing to move to “hardship townships.” 4.29. Increased spending on goods and services and capital coupled with a policy on eliminating school fees has helped reduce out-of-pocket payments by households. The loss in revenue from fees (estimated at $63 million in 2009/10 for all levels of education including tertiary) and other household contributions (estimated at around $36 million) was more than offset by government spending on goods and services, maintenance and capital. Government expenditure on these categories increased from $29 million in 2009/10 to $188 million in 2013/14. MoE officials believe that the elimination of school fees motivated an increase in the number of children enrolled in primary and secondary schools in 2014/15 by 300,000 to 8.5 million. 4.30. Other new policy initiatives, while relatively small, were introduced to further reduce out-of- pocket payments. These included free textbooks, and scholarships and stipends (Table 4.7). Expanding stipend programs to support poor and disadvantaged students is an effective way to keep poor children from dropping out of school. Stipends can help poor families invest in schooling by supporting them with direct costs, such as uniforms, and by helping families to recover some of the opportunity costs, such as alternative income earning opportunities for the child. Experience from other countries suggests that a stipend program has the potential to lead to higher school attendance rates and lower dropout and repetition rates in the targeted grades. Table 4.7: Estimates of the Cost of New Initiatives Introduced Since 2011/12 a/ Additional Budget Initiative Targeted Allocation Type of Initiatives started Student Groups (MMK millions) FY2012/13 FY2013/14 Free Textbooks (0320) 2011/12 All Primary Students 6,876 16,000 Free Exercise Books (0309) 2011/12 All Primary Students 3,220 5,700 Cash Transfer for compulsory primary education 2013/14 1000 Kyats per primary student .. 5,597 School Operational Grants (0305 0313) 2012/13 All Schools with Primary level 12,939 14,070 Scholarship Awards (0501) 2012/13 All Middle and High School Students 100 531 Stipend Awards (0501) 2012/13 All Students (Primary/Middle/High) 650 650 Total 23,785 42,548 a. Budget codes appear in parentheses. Source: World Bank staff calculations based on BOOST database. 4.31. Block grants were another source of increased recurrent spending. This is a transfer that schools have flexibility on in terms of spending. Block grants grew from approximately US$285 per high school in 2009/10, after adjusting for inflation, to more than US$8,000 by 2013/14 (Figure 4.9). As schools did not have such flexibility in the past, there may be mistakes in recording and reporting these expenditures in the beginning. These will lessen in time and with experience, and therefore are not an obstacle for bigger transfers. 51 Figure 4.9: Block Grants Allocated Per School, for Goods and Services 9,000 8,338 8,000 2009/10 2013/14 7,000 constant 2013/14 USD) (spending per facility, 6,000 5,000 3,920 4,000 3,000 2,000 722 1,000 285 102 12 0 High Schools Middle Schools Primary Schools Source: World Bank staff calculations based on data received from Ministry of Education. 4.32. Going forward, new initiatives will require additional fiscal space and therefore warrant thorough costing. In the immediate future, adding more years of schooling, one at the beginning and one at the end of the cycle, will be the biggest driver of costs. For example, adding one year of preschool will bring 1.2 million more students into the system, which represents an almost 20 percent increase in student numbers. Assuming 25 students per class, this will require hiring 48,000 new teachers and building approximately 60,000 new kindergartens. Using the 2013/14 average recurrent cost per student (75,703), this will cost approximately 91 billion kyat a year, every year, an increase in the recurrent budget of approximately 12 percent. Keeping 322,000 students in high schools for one more year will cost approximately 180,000 kyat a year per high school student, a total of 58 billion kyat a year. Actual costs are likely to be higher if the government is to pursue higher quality of education services. 4.33. The quality of education is unlikely to improve unless Myanmar’s current teachers are retrained, and equipped with better textbooks and teacher guides. Modernizing a curriculum, writing new textbooks, re-training teachers, and rolling this out to every classroom in the country usually takes at least five years in a well-functioning ministry of education. Most likely, this could take even longer in Myanmar, where no undertaking of this scale has been attempted for at least 10 years. 4.34. Delivering on the promises of the new Education Law to provide education to all children, including those with disabilities is another concern.37 There is virtually no system in place to serve children with disabilities. A cross-ministerial effort—involving the Ministries of Education, Health, and Social Welfare, Rehabilitation and Resettlement—is needed to identify and remove the barriers. As a start, the MoE could consider setting up an administrative unit, tasked with developing activities (or “programs”) to be implemented in regular and special schools. These activities could be presented as part of next year’s budget proposal or submitted to development partners for financial support. 4.35. Early Childhood Development (ECD) is another issue. The new education law mentions early child development as a responsibility of “parents, villages, and appropriate departments and organizations 37 “Special education programs and services shall be established so that every school-aged child and youth, including those citizens who are disabled or who for whatever reason have not had a chance to study, can access their right to education in line with Education for All.” Myanmar National Education Law, Chapter 3, Education principles 4.c. 52 according to existing policy.” Moreover, the law makes explicit that “free education” only starts at the primary level. However, a clear budgetary commitment from the government for ECD would address the most important years for a young child’s survival, growth and development. It is critical particularly for children from poorer households to be able to start school as prepared as children from richer households. Myanmar is developing a policy on early childhood care and development (ECCD). This proposes coordination of services in education, health, nutrition, sanitation, and child protection. A recent analysis by UNICEF provides suggestions on how to strengthen this policy.38 4.36. MoE should, with MoF assistance, explore sources of fiscal space starting with efficiency gains within the education sector. One potential area of savings is MoE’s current policy on purchase of school books. The current 92 titles used throughout the school system are of such poor quality that they have to be replaced every year, compared to other countries where a textbook can easily be used for three years.39 40 million poor quality textbooks are being printed each year, for a total of 120 million over three years. 4.37. By investing slightly more, only 40 million textbooks would have to be produced over a three year period. It will also have raised the quality of the content and the reproduction to an acceptable standard in the interim period before a wholly new curriculum and its accompanying textbooks and teacher’s guides are introduced.40 Chapter 3 proposes additional options including a stronger revenue effort by the Union government and reductions in low priority and/or wasteful expenditure programs. G. Public Financial Management in Education 4.38. The budget preparation and implementation processes for basic education are much more important and expenditure levels have been substantially increased because better results are being demanded and the Parliament is involved. The budget preparation process (Figure 4.10) begins with bottom up needs of schools submitted through Township Education Offices (TEO), Regional Education Offices (REO), who then submit to the central DBE for final review and decision. 41 4.39. It would be useful if the increased focus on education policy objectives could be accompanied by greater analysis and scrutiny of budget requests within MoE before these are submitted to MoF. Previously, government spending in education was mostly wages and salaries, which were fixed in nominal terms, offering little scope or need for analysis. Now there needs to be more attention on whether budget requests align with policy priorities. This requires: (i) scrutiny and vetting of both existing and new initiatives; (ii) analysis of available data on spending and education outcomes; and (iii) improved coordination between the recurrent and capital budgets, particularly as the government increases spending on new schools. 4.40. The MoF and MNPED also need to review budget submissions more actively. During the period under review, MoF made only minor adjustments to the budget before forwarding the proposals to the Financial Commission and Parliament (See steps 5 through 10 in Figure 4.10). It would be helpful if MoF 38 UNICEF (2012). 39 Watson et al (2013). 40 More specific recommendations on this are made in Watson et al (2013). 41 This section provides a review of the budget cycle only for basic education which accounts for approximately three- quarters of all MoE expenditures. The budget processes governing higher education is, however, highly similar to what is reviewed here. 53 and MNPED could have dedicated staff to oversee, vet and challenge budget proposals coming from MoE and other ministries. Establishing this capacity will reduce Union vulnerability to unexpected cost increases. Figure 4.10: Budget Calendar for FY2012/13 Budget Preparation January Step 7: 2-3 days Cabinet level discussion September (this also involves back and forth July Step 4: MOE submits with departments) Step 1: Township prepares budget request March “proposals” (needs) and to MOF Step 10: MOF communicates submits to September - December Approved budgets to DBEs Step 5: Back and forth departments Between MOF and MOE April July October January April February April September January Step 3: MOE reviews, Step 6: MOF submits and makes adjustments budget to April Cabinet April 1: Fiscal year starts (last week of Jan) August Step 2: DBEs combine and submits to MOE February February (minister’s office) Step 8: MOF submits budget to Step 9: Parliament Parliament debates and approves budget Source: Discussions with officials at the Ministry of Education. 4.41. The budget calendar could allow more time for Cabinet review and inputs. Figure 4.10 shows that the Cabinet is allowed only two or three days to review the budget proposal. This is not enough time in circumstances when big changes are proposed or needed for the level or composition of spending. Further decompression will be needed if and when the draft Budget proposal is made available to the public. 4.42. In the medium-term, it would be helpful to make budget proposals more informative. MoE (and all ministries) could facilitate budget formation and review by arranging spending proposals by program or objective. It would also help to make a distinction between the expenditure requirements of old versus new programs and the savings made possible by the completion or elimination of some old programs. 4.43. The rapid growth in budget allocations for education could put stress on fiduciary arrangements. There is an orderly process in place for flow of funds from the Union to townships and schools. The funds flow mechanism from the Department of Basic Education to townships operates through the banking system and is secure. However, townships transfer school budgets in cash. This creates fiduciary risk, particularly when the level of transfers is growing rapidly. This warrants closer review to assess the extent to which central allocations are making it to frontline service delivery units. This assessment can be made through a Public Expenditure Tracking Survey (PETS). A PETS type exercise is planned as part of the World Bank/Australia supported school grants and stipends program. 4.44. Budget implementation has been dominated by the three DBEs42 but this is beginning to change with the introduction of school grants. Up until recently, DBEs were responsible for: hiring and transfer of 42 Now merged into one DBE, as of April 1, 2015. 54 teachers; procurement and distribution of goods through the townships to schools; contracting and execution of capital projects above $5 million. Since 2014/15 however, schools are receiving a “school grant” based on the number of students of enrolled (see paragraph 31 above). Schools decide how to spend this money, based on a list of eligible expenditures. 4.45. The current procurement system is another source of fiduciary risk in the face of rapidly rising expenditure. As noted above, budget allocations to goods and services and capital investments have grown very rapidly. There are indications that MoE has struggled with procurement, largely because of lack of experienced, trained staff. Most staff members have little or no experience in contracting services or purchasing goods from the private sector. Solving this would open up good opportunities for reforms. For example, bringing competition to the monopoly on printing textbooks held by the state-owned Printing and Publishing Enterprise would improve the quality of textbooks and value for money. Beyond MoE, there is a general need for clear and universal regulations, backed by a law, for procurement planning, implementation, and performance monitoring. 4.46. Financial management capacity within MoE is good in comparison with other less developed countries in South East Asia. The staff within the finance functions are dedicated and staff turnover is generally low. All managerial staff, apart from general administration staff, have some form of university degree.43 These strengths could be further amplified by working to ensure more staff have specialist qualifications and expertise in accounting, information technology, procurement, and internal audit. In addition, it would be helpful to ensure more staff have their own computers and know how to use them. The ability to use an electronic spreadsheet or databases such as the EMIS for analytical purposes is very rudimentary and concentrated in a few staff. Thus far, spreadsheets are used mainly used as convenient way to set up tables, rather than as a powerful tool to do analysis. 4.47. The accounting and audit systems appear to be orderly and systematic. All accounting and financial reporting is on a cash basis and manual recording and reporting processes are currently used at the DBEs. The reporting system follows a very orderly and bottom-up process based on accounts at the school level. TEOs aggregate school accounts, add expenditures of township administration and send them to REOs. REOs aggregate township reports, add expenditures of regional administration and submit these reports to DBEs periodically. The formats used for financial reporting are the same as the budget formats and they include budget versus actual variances for each budget line. The general ledger and subsidiary ledgers for inventories and fixed assets are regularly reconciled. All accounting and supporting documents are retained and are accessible. The bank reconciliation processes appear to be strong and they ensure that all receipts and payments have been correctly recorded in the accounting records. The main disadvantage is that it is paper based and it takes a long while to consolidate reports. It takes around three months to consolidate year end reports and prepare provisional actual year-end financial reports by DBE’s. The external audit arrangements appear to be adequate. Regular external audits are conducted and audit recommendations are monitored for implementation. 43 The Basic Education sector benefits from the fact that all staff members begin their careers as teacher trainees, who firstly obtain their University and teaching qualifications and then move up through the various education sector institutions management structures (school teacher, school principal, township education officer, and into DBE or Ministry of Education management). 55 H. Conclusions and Recommendations 4.48. Tackling education sector challenges is a high priority on the government’s agenda. Since 2011, it has initiated fundamental reforms to education sector policy including elimination of school fees, introduction of compulsory primary education, provision of stipends, delegation of authority to States and Regions, improvement of learning conditions, and enhancement of teacher qualifications. To implement these policies, the government’s budget on education quadrupled, albeit from a low base. A larger education wage bill has enabled higher salaries and more teacher recruitment, and increased resources for non-wage recurrent and capital investments have helped reduce the burden of education costs on households. 4.49. The MoE will be grappling with pressures to deliver more despite: (i) limited administrative, budget and performance data as a basis to prioritize spending; (ii) fledgling capacity to analyze policy and budget linkages; and (iii) fiduciary risks arising out of large cash payments and weak procurement rules. In addition, pressures are emerging for new priorities such as teacher training; more years of schooling; extending coverage to marginalized groups; and Early Childhood Development. Issues Options Prioritization of education sector policy Short-term: MoE could use standardized assessments of student and spending would benefit from greater learning outcomes to monitor the quality of the education system. focus on quality and access to key As a starting point, MoE could use the Early Grade Reading administrative, fiscal and performance Assessments piloted in 2014 in the Yangon region. information. Short-term: MoE and MoF could carry out thorough costing of new Implementation of rapidly rising budgets initiatives (e.g. more school years, more teacher training, and addressing large unmet needs will require close attention to absorptive modernized curriculum, coverage to disadvantaged groups, early childhood education) and propose resourcing needs with capacity in education. implementation plans. Short-term: MoE could develop a cadre of officers trained in procurement and contract management. Medium-term: Going beyond MoE, the Union Government could develop clear and universal regulations, backed by a law, for procurement planning, implementation, and performance monitoring. Gradual delegation of responsibility for Medium-term: The Union government could continue the gradual day-to-day management of schools could devolution of management responsibilities within a framework of ensure that spending is more closely accountability for results. Coordination of the framework could aligned to local needs, including benefit to take place at the State/Region level, where there is a natural ethnically and socially diverse households intersection between MoE staff and State/Region administrative of Myanmar. staff. Short-term: MoE could conduct a PETS to ensure schools are receiving allocations in accordance with the Ministry’s guidelines (i.e. in accordance with the formula and the time- table specified in those guidelines). Medium-term: larger schools (located near banks) could be asked to open bank accounts to allow transfers from TEOs to be made via bank transfers. The achievement of education policy Medium-term: MoE (and all ministries) could facilitate budget objectives would be more likely with formation and review by arranging spending proposals by program effective links between policy and or objective. expenditure programs. Medium-term: It would also help to make a distinction between the expenditure requirements of old versus new programs and the savings made possible by the completion or elimination of some old programs. 56 GOING FROM MORE TO 5 BETTER GOVERNMENT SPENDING ON HEALTH Background: Health services in Myanmar have been inadequate due to historically low levels of government spending, and the ensuing weaknesses in health systems and regulatory capacity. The government in 2012/13 initiated reforms meant to improve the situation. Objectives: The objectives of this chapter are to analyze government expenditures and policies for healthcare between 2009/10 and 2013/14, and to propose options for further reforms in policies and spending, in particular through more efficient and more equitable government spending. Key findings: In 2012, government spending on public health was a mere $1.6 per-capita per year, with out-of-pocket payments by households accounting for almost 80 percent of total health spending. This has created serious obstacles to healthcare access, with rural and hard to reach areas most severely affected, and some of the lowest health outcomes among ASEAN countries. The Union government introduced policies to improve access to, and affordability of, healthcare services. More medical personnel have been hired. Beginning in 2012/13, essential drugs and selected health care services are provided free of charge to children, pregnant mothers and for patients needing emergency surgery under certain circumstances in some facilities. There are early signs that these policy changes are helping to reduce prohibitive out-of-pocket payments by households. They have, however, caused a dramatic increase in expenditure on drugs and medical consumables. Further costs have been incurred due to wage increases and heavy investments in facilities and equipment. These policies have been financed through a nine- fold increase in MoH spending on health between 2009/10 and 2013/14. Ongoing challenges include constraints on policy making and monitoring due to unreliable and incomplete data, and high costs due to expenditure inefficiencies. In the medium-term, the government may also want to consider options to increase equity public service provision including more rural primary care and better financial protection policies. Recommended options: To assist in policy making and monitoring, (i) improve the coverage and comprehensiveness of data for public and external financing of the health sector and conduct household budget surveys for private financing; to reduce expenditure inefficiencies, (ii) increase the share of MoH budget resources for preventive care and health interventions with strong public goods characteristics and (iii) formulate a policy for the strategic procurement of pharmaceuticals; to make public services more equitable, (iv) formulate a policy on capital spending that better balances rural construction against acquisition of expensive tertiary care equipment; and (iv) formulate financial protection policies to ensure universal access to health services. 57 CHAPTER 5 FROM MORE TO BETTER GOVERNMENT SPENDING ON HEALTH A. Background 5.1. Historically low levels of government spending in the health sector, and the ensuing weaknesses in health systems and regulatory capacity, have translated to inadequate health services in Myanmar. In 2012, government spending on public health was a mere $1.6 per-capita per year, with out-of-pocket payments accounting for almost 80 percent of total health spending. This has created serious obstacles to healthcare access, with rural and hard to reach areas most severely affected. It also resulted in some of the lowest health outcomes among ASEAN countries. The government has introduced major policy reforms in the health sector to address these issues as a matter of high priority. Government spending in the health sector has increased nine-fold in the five years between 2009/10 and 2013/14. 5.2. The PER aims to assist the government in prioritizing spending and policy reforms to improve health outcomes in Myanmar, in particular through better targeted, more efficient and more equitable government spending. It starts with an overview of the health sector which, unlike education, has a strong private sector presence. It provides an overview of health challenges in Myanmar, how these have informed government priorities in recent years, and the implications these have had for health sector financing. It then looks more deeply at government spending on health and recommends possible changes to allocations that could improve the efficiency and equity of that spending. B. The Health Sector 5.3. The Ministry of Health is the regulator and the main health service provider in the public sector, accounting for around 73 percent of public spending on healthcare in 2011/12 which, in turn, constituted only around 15 percent of health sector spending. In that year, there were around 987 public sector hospitals and 1,565 rural health centers. MoH’s service delivery network is deconcentrated and extends to the village level. Township health departments, managing the township health system, are the backbone of Primary Health Care and provide comprehensive health services at the local level. At the regional administrative level, the Health Departments of the states and regions provide supervisory and technical support. They also manage the provision of tertiary care and referral services. 5.4. Several other government ministries and agencies incur expenditures on health,44 which together account for 27 percent of all health expenditures from public sources and 4 percent of total health spending in Myanmar in 2011/12. Most of these health expenditures were for curative care services for staff and family members through hospitals and health facilities owned and run by each ministry. The Ministry of Defense is well known for providing the most comprehensive health services and coverage to staff of the armed forces and their family members. It operates hospitals and medical schools that train health personnel, including medical doctors and nurses. Several military hospitals also provide health care services to local residents in remote areas who live near the hospitals. 44 These ministries and agencies include the Ministries of Defense, Labor, Railways, Mines, Industry, Energy, Home, and Transport, as well as Yangon City Development Committee, Mandalay City Development Committee, and Nay Pyi Taw City Development Committee. 58 5.5. The City Development Committees (CDCs) of Yangon, Mandalay and Nay Pyi Taw also provide health services, primarily health promotion and disease prevention services.45 The Yangon CDC, for instance, provides disease prevention services against dengue fever, cholera, gastroenteritis, and tuberculosis. The services include mass health education through media and health education at schools. Other health activities include food and drug control and environmental health. Yangon CDC, for instance, routinely provides food safety and control services, particularly the inspection of markets, food stores, food industry, schools, and restaurants in 32 urbanized townships of Yangon Division. 5.6. Non-governmental providers (both for profit and non-profit) are big conduits for health sector spending and employ the majority of medical personnel in Myanmar. For-profit providers typically offer ambulatory care and hospital care. Not-for-profit institutions include civil society organizations and international non-governmental organizations. According to MoH data for 2012/13, non-governmental providers employed 57 percent of Myanmar’s doctors and 73 percent of the dentists. There are over 40 private hospitals and thousands of private health clinics in Yangon city, at least five private hospitals in Nay Pyi Taw, and thousands of private health care institutions in the regional cities, and in rural villages.46 C. Health Challenges 5.7. Living standards in Myanmar are among the lowest in the region. Myanmar is a geographically and culturally diverse country, with 135 groups speaking over 100 languages and dialects. It has been one of the poorest countries of the world, and ranked a low 149 out of 168 countries on the 2012 Human Development Index. Poverty incidence is twice as high in rural areas compared to urban areas, and states and regions such as Chin, Rakkhine, Shan, Tanintharyi and Ayeyarwady are much poorer than others.47 Despite recent gains, living standards in Myanmar are among the lowest in the region and social well- being is well below those of regional peers (Figure 5.1). 5.8. Government spending on health in the past decades has been very low. In 1975/76, Union expenditures on health were approximately 0.8 percent of GDP (Figure 5.2).48 These fell to 0.4 percent of GDP by 1995/96 and just under 0.3 percent of GDP by 2010/11. The health sector, accounted for only 1.3 percent of total government expenditure in 2011/12, equivalent to about US$1.60 per person per year. As a result, private out-of-pocket payments accounted for almost 80 percent of total spending on health in 2011/12. Use of health services is disproportionately skewed in favor of those who can afford to pay for health, rather than those who most need the services. 5.9. The above has translated to very low levels of basic healthcare services in Myanmar. Approximately 30,000 doctors and 55,000 nurses and mid-wives serve over 51.4 million people. Less than half of these doctors work in the public sector, making it very difficult to staff adequately the public health facilities in rural areas.49 Medical training is poor, and does not provide midwives and nurses with full skills and knowledge to address all competencies for basic and emergency obstetric care. In-service training is mainly provided by projects and programs, and not adequately coordinated. Many health facilities require upgrading and refurbishment, including basic facilities and equipment. Transport is inadequate to ensure effective service delivery, supervision and monitoring, and referral for those in of need emergency care. 45 The three CDCs are independent public agencies covering Yangon, Mandalay, and Nay Pyi Taw. All CDCs raise their own revenues through tax collection, fees, licenses, and property development. 46 Asia Pacific Health Observatory (2014). 47 IHLCA Poverty Profile Report, June 2011. 48 World Bank (1985). 49 Myanmar Ministry of Health, Health Workforce Strategic Plan 2012. 59 Figure 5.1: Per-capita GDP and HDI Figure 5.2: Public Health Spending, 1975/76 - 11/12 6,000 0.8 GDP per capita, US$ PPP Human Development Index 0.7 1.2 5,000 5462 0.6 4,000 0.5 1.0 3635 3,000 0.4 2926 2898 0.8 Percent of GDP 0.3 2,000 2494 1883 0.2 1,000 1970 1616 1484 0.6 1417 1730 0.1 1187 935 842 500 843 0 0 0.4 Vietnam PNG Lao PDR Bangladesh Nepal Myanmar Cambodia Mongolia 0.2 (left axis) GDP per capita, US$ PPP 2000 0.0 (left axis) GDP per capita, US$ PPP 2012 (right axis) HDI (2012) 1970 1975 1980 1985 1990 1995 2000 2005 2010 Sources: HDI values are from HDRO calculations based on Sources: World Bank (1985), World Bank (1995) and World data from UNDESA (2011), Barro and Lee (2011), UNESCO Development Indicators, Ministry of Health. Institute for Statistics (2012), World Bank (2012) and IMF (2012) and World Development Indicators. 5.10. Health systems and regulation have also suffered. The supply chain system is not well developed, and there are problems with storage and distribution of supplies, especially to facilities at the township level and below. The health information system has many weaknesses, and there are gaps in data from the community level and from hard-to-reach areas. There has been no national survey of maternal and under-five mortality since 2002. Much of the growing private sector is unregulated. 5.11. Myanmar has one of the lowest health outcomes in the ASEAN region. As shown in Table 5.1, life expectancy at birth is 65 years for both sexes (71 years in Cambodia, 76 years in Vietnam). The age- standardized death rate in Myanmar is estimated at 1,112 deaths per 100,000, about 60 percent higher than that of neighboring Thailand.50 In other words, Myanmar records 259,260 deaths that could be prevented if the health system delivered outcomes similar to Thailand (Figure 5.3). If the regional relationship between GDP per-capita and the age-adjusted death rate were applied to Myanmar, there would be 17 percent lower with only 922 deaths per 100,000 people (Table 5.1). The five leading causes of death are cerebrovascular disease (related to circulation of blood to the brain), neoplasms (abnormal cell growth, possibly cancerous), HIV/AIDS and tuberculosis, chronic respiratory diseases, and lower respiratory infections.51 Some of these could be avoided with higher spending on preventive care (section G below). 50 The age-standardized mortality rate is a weighted average of the age-specific mortality rates per 100,000 persons, where the weights are the proportions of persons in the corresponding age groups. The numbers of deaths per 100,000 population are influenced by the age distribution of the population. Two populations with the same age- specific mortality rates for a cause of death will have different overall death rates if the age distributions of their populations are different. Age-standardized mortality rates adjust for differences in population age distribution by applying the observed age-specific mortality rates for each population to a standard population. 51 Global Burden of Disease Study 2010 - Myanmar Results by Cause 1990-2010, last accessed on 9-2-13; available at http://ghdx.healthmetricsandevaluation.org. 60 Figure 5.3: Age-standardized Death Rates 2012 (Per 100,000 Population) 1,400 1,152 1,112 1,200 1,009 950 916 1,000 800 722 665 612 556 600 356 400 200 0 Vietnam Indonesia Malaysia Singapore Laos Myanmar Philippines Thailand Cambodia Brunei Source: WHO Global Health Observatory Data Repository. 5.12. Maternal and child health outcomes are poor despite significant improvements since 1990. Maternal mortality fell from 520 per 100,000 live births in 1990 to 200 in 2010, while under-five mortality fell from 106.4 per 1,000 live births in 1990 to 54.2 in 2011 (Table 5.1). During the same period, infant mortality rate fell from 76.1 per 1,000 live births to 42.4 per 1,000 live births.52 This translates to 2,000 pregnant women and 50,000 children dying every year from mainly preventable causes. If the regional relationship between GDP per-capita and infant mortality rates applied to Myanmar, the infant mortality rate would be 30 percent lower or 29 deaths per 1,000 live births. Notwithstanding the gains in recent years, Myanmar is unlikely to reach its health-related Millennium Development Goals (MDGs): MDG4, to reduce child mortality (36 per 1,000 live births for under-five mortality), and MDG5, to improve maternal health (130 per 100,000 live births for maternal mortality) for 2015. 5.13. Antenatal care coverage remains poor, although there have been some improvements in recent years. Available numbers suggest that between 73 percent and 83 percent of pregnant women had at least one antenatal visit and 43 percent had at least four visits to a health care provider in 2010.53 Likewise, there are a number of different estimates for skilled birth attendance, varying between 70 and 80 percent, and for institutional deliveries, which is estimated to be 36 percent. Available statistics therefore suggest that emergency obstetric care is not adequate or affordable for all women who need it. 5.14. Malnutrition has been a serious issue in Myanmar. This remains true, even though the overall prevalence of stunting and underweight has been decreasing over the past two decades. Estimates for 2009 suggest that 1 in 7 infants were born with a low birth weight, 35 percent of children under the age of five are stunted, 23 percent are underweight, and 8 percent are wasted. Analysis of survey data suggests that only 24 percent of Myanmar children were exclusively breastfed up to the age of six months. 52 There are several discrepancies in these numbers between what is reported by national sources and by external agencies. The Central Statistical Organization (Health in Myanmar 2013, accessed at http://www.moh.gov.mm/ on 8-23-13) reports U5MR of 34.9 per 1,000 live births in 2011 (compared to 62.4 reported by WDI and 46.1 by the Myanmar 2009/10 Multiple Indicator Cluster Survey, or MICS), MMR of 142 per 100,000 live births (compared to 200 reported by WDI), and IMR of 25.6 (urban) and 27.8 (rural) per 1,000 live births (compared to 49.3 overall reported by WDI and 37.5 by MICS). 53 WHO, World Health Statistics 2013, available at www.who.int. 61 This was slightly higher in rural areas than in urban areas, and ranges from 1.3 per cent in Rakhine to 40.6 per cent in Kachin.54 Table 5.1: Health Status Indicators ASEAN Countries a/ Predicted b/ Philippines Cambodia Estimated Singapore Indonesia Myanmar Myanmar Malaysia Thailand Vietnam Laos GDP per-capita, US$ World Bank Atlas Method 1,140 1,140 880 3,420 1,260 9,820 2,960 51,090 5,250 1,560 Age-standardized death rates 1,112 922 1,152 916 1,009 722 950 356 665 612 (per 100,000 population)* Life expectancy at birth, total (years) 64.8 70 71.1 70.4 67.4 74.7 68.4 81.9 74.0 75.5 Infant mortality rate (per 1,000 live births) 42.4 29 33.9 25.8 54.0 7.3 23.5 2.3 11.4 18.4 Under-5 mortality rate (per 1,000 live births) 52.3 36 39.7 31.0 71.8 8.5 29.8 2.9 13.2 23.0 Maternal mortality rate 200 185 250 220 470 29 99 3 48 59 (modeled estimate, per 100,000 live births)** Pregnant women receiving prenatal care (percent)** 83 85 89 93 71 91 78 99 94 Births attended by skilled health staff 70.6 76 71.0 82.2 37.0 98.6 72.2 99.7 99.5 92.9 (percent of total) Malnutrition prevalence, stunted height for age 35.1 34 40.9 39.2 47.6 17.2 32.3 4.4 15.7 30.5 (percent of children under 5) Malnutrition prevalence, underweight for age 23** 20 29** 19** 31.6 12.9 20.7 3.3 7.0 20.2 (percent of children under 5) Prevalence of wasting (weight for height) 7.9 9 10.8 12.3 7.3 15.3 6.9 3.6 4.7 9.7 (percent of children under 5) Prevalence of HIV, total 0.6 0.5 0.6 0.3 0.3 0.4 0.1 0.1 1.2 0.5 (percent of population ages 15-49) Malaria incidence rate (cases per 100,000)*** 1,096 .. 778 51 Malaria death rate (deaths per 100,000)*** 1.33 .. 0.18 0.13 Prevalence of tuberculosis 506 411 817 281 540 101 484 46 161 188 (per 100,000 population)* Incidence of tuberculosis (per 100,000 population 381 243 424 187 213 81 270 37 124 151 a. Data are for 2012 or latest estimate available. b. Predicted values assumes there is a linear relationship between GDP per-capita and health status indicators. This is a gross simplification to allow the comparison of health outcomes after controlling for country GDP. Sources: World Development Indicators. *WHO Global Health Observatory Data Repository; **Millennium Development Goals, The World Bank Group; ***Health-related Millennium Development Goals 2012, WHO Regional Office for South-East Asia. Table 5.2: Selected Indicators of Health Status and Access, 2009/10 Urban Rural Immunization (Percent of 1 year olds fully immunized against measles) 91.5 79.6 Antenatal care (Percent of women who gave birth in the last five years who used skilled health 93.3 80.8 personnel for antenatal care at least once) Births Attended by Skilled Personnel (Percent) 88.6 67.9 Access to health care (percent of population within 1 hour walk of facility) 96.5 75.3 Severe malnutrition (percent of children under 5 years with low weight for height) 8.0 9.7 Source: IHLCA 2009/10. 5.15. Inequalities in health status and access are significant particularly in rural and hard-to-reach areas, and among the most vulnerable populations. There are huge gaps particularly in relation to immunization coverage, access to health care, nutritional status, antenatal care coverage and skilled attendance at birth (Table 5.2). Women and girls of reproductive age and young children are the most adversely affected by this lack of affordable, quality health services. Many rural areas are poorly served because difficult terrain, geographical remoteness, conflict, and cultural diversity hinder the delivery of essential services. Even where services are available, access is limited by financial and other barriers. Out- 54 Information obtained from the analysis of Myanmar Multiple Indicator Cluster Survey 2009-2010, available at http://www.childinfo.org/files/MICS3_Myanmar_FinalReport.pdf. 62 of-pocket expenditure is high within and outside the public health sector, and serious illness or a medical emergency can throw a family into deeper poverty. D. Policy Priorities in Health 5.16. The government’s Health Vision 2030 (2001) continues to provide the overarching policy framework to address the above challenges.55 This is backed up by a series of more recent short-term plans56 that have set a series of commitments to improve access to healthcare services, including: 80 percent antenatal care coverage; 80 percent of births attended by a skilled attendant; 70 percent access to emergency obstetric care; and 80 percent coverage for prevention of mother-to-child transmission of HIV. They also include universal childhood immunization coverage, increased coverage of newborn care, increased contraception prevalence and reduced unmet need for contraception, improvement in the midwife to population ratio from 1:5,000 to 1:4,000. The government plans to comprehensively update its overall health strategy, including the Public Health Law (1972).57 5.17. Improved accuracy, coverage and comprehensiveness of health sector spending data should underpin any revision to the overall health strategy, and will require early commitment on the part of government. As a start, there is much scope to improve the coverage and comprehensiveness of fiscal data. Box 5.1 provides a list of some of the more important gaps in the data. Secondly, critical to prioritizing government expenditure in health is a comprehensive picture of overall financing in the health sector. Public spending is the only well recorded portion of health financing data, and it explains only a small percentage of total spending on health. Spending on health from external sources, including bilateral aid and assistance from development partners, is not completely and accurately recorded. Private spending, which is the single largest financer of health, is the most poorly documented. Box 5.1: Gaps in the Coverage and Comprehensiveness of Budget Data on Health  Capital and recurrent budget/expenditure on health of non-Ministry of Health government ministries and departments, townships etc., for 2012/13 and 2013/14  Recurrent health expenditures of non-MoH government ministries and agencies in 4-digit series58 for the period 2009/10 to 2013/14  Capital health expenditures of non-MoH government ministries and agencies by investment categories the period 2009/10 to 2013/14  Capital and recurrent budget/expenditure on health at the state, regional, and township level  Capital expenditure (with the exception of construction category) that can be disaggregated into different levels of MoH’s health facilities  Recurrent expenditure on drugs that can be disaggregated into different levels of MoH’s health facilities. 55 Covers health policy and law; health promotion; health service provision; development of human resources for health; promotion of traditional medicine; development of health research; role of non-government actors; international collaboration. 56 Rural Health Development Plan 2001-06, the Hospital Upgrading Plan 2001/02-05/06, the National Health Plans for 2006-11 and 2011-16, the Five Year Strategic Plan for Child Health Development (2010–2014), the National Plan of Action on Food and Nutrition (2011–2016) and the Health Workforce Strategic Plan 2012-17. 57 At least 14 additional laws govern the sector, including the 2007 Law Relating to Private Health Care Services. 58 Data on recurrent expenditures are available in 2 levels: 2-digit and 4-digit series. The 4-digit series is a disaggregated level of the 2-digit series that allows a more clear understanding and therefore more robust analysis of expenditures in health. 63 5.18. Regular household surveys can improve estimates of household expenditures on health care. Private spending on health, as recorded in National Health Accounts (NHA) is estimated from the previous years’ numbers, which themselves are estimates from previous years. Total private spending is estimated on the basis of a formula that takes into account the country’s GDP, the private sector contribution to the GDP, national annual consumption, share of household expenditure in the private consumption, and share of medical care expenditure in total household expenditure, which is drawn from the 2006 Household Income and Expenditure Survey. NHA estimates of private spending are thus likely to be inaccurate, and could under- or over-report the actual numbers. Surveys, such as the IHLCA could facilitate more informed estimates if they could be held on a regular cycle (e.g. every three years or every five years). 5.19. Reporting on aid funded projects also has scope for improvement and stronger enforcement. In the absence of actual expenditures by development partners, the PER uses inferences based on the memorandum of understanding (MOU) that each them signs with the International Health Division of MoH. These MOUs describe multi-year planned activities. Annual spending estimates are estimated by dividing equally the amount budgeted in the MOU across the years of operation. Expenditures across different types of health interventions are also inferred from the MOUs. Hence, the budget amount stated in the MOU can be very different from actual expenditures. E. Health Sector Financing 5.20. There are three main sources of health sector financing: public, private and external. Public funding consists of general government revenues, which are channeled through the budgets of MoH, other ministries and departments, and CDCs involved in health services; they are also channeled through the government’s Social Security Scheme. Private funding is mostly out-of-pocket payments made by households. These are channeled mainly through individuals and households themselves, with a small amount going through the Social Security Scheme. External funding is mostly from Official Development Assistance, which is channeled through the government and not-for-profit institutions working in health. The relationship between financing sources and financing agents in 2011/12 is summarized in Table 5.3. Table 5.3: Financing Sources and Financing Agents, Percent of Total, 2011/12 Public Financing Agents Private Financing Agents Total Ministry of Other Social Security Not-for-Profit Households Total Health a/ Ministries Scheme Institutions In Percent Public Sources 72.6 27.3 0.1 100.0 Private Sources 0.6 99.4 100.0 External Sources 17.1 82.9 100.0 In MMK millions Public Sources 80.1 30.1 0.1 0.0 0.0 110.3 Private Sources 0.0 0.0 3.9 638.2 0.0 642.0 External Sources 9.9 0.0 0.0 0.0 47.8 57.7 Total 90.0 30.1 4.0 638.2 47.8 810.0 a. Budget estimate. Preliminary actual is MMK 86.4 million. Sources: Ministry of Health NHA and World Bank staff calculations. 5.21. Private financing, in particular out-of-pocket payments, has historically been the biggest source of health sector funding in Myanmar. The share of out-of-pocket payments has fallen from 82 percent in 2009/10 to an estimated 60 percent in 2012/13, due to increased government spending. This remains 64 high relative to Myanmar’s peers (Figure 5.4). Pharmaceuticals account for 53 percent of total out-of- pocket spending, followed by curative services at 31 percent, and diagnostic services at 15 percent. In terms of service providers, hospitals account for 76 percent of out-of-pocket spending, whereas ambulatory (out-patient) healthcare services account for around 19 percent. Figure 5.4: Out-of-Pocket Financing (Percent of Total Health Care Expenditures) 70 57 59 60 54 56 56 48 50 50 46 40 42 40 33 35 30 20 15 16 14 10 0 Viet Nam Bhutan Sri Lanka Lao PDR China Malaysia Indonesia Thailand Japan Rep. of Korea Nepal Philippines India Myanmar Cambodia Source: WHO database. All data are for 2011 except Myanmar which uses 2013/14. 5.22. The government’s share of total spending in health in 2011/12 had been among the lowest in the region (Figure 5.5). From 2012/13 onward, public financing as a share of total health care spending has been increasing while the out-of-pocket share has been declining (Figure 5.6). The relative share of public spending on health increased from 8.5 percent of total spending on health in 2009/10 to 35 percent in 2012/13 and is expected to go up to 41 percent in 2013/14 (Figure 5.7). The big increase in public spending came about in 2012/13, when it rose by 260 percent relative to the previous year and increased the share of public sources from 14 percent in 2011/12 to 35 percent in 2012/13. The share of private financing of health has fallen over time, from 82 percent in 2009/10 to 79 percent in 2011/12 and 60 percent in 2012/13. Following a further increase in planned spending from public sources, the share of private out-of-pocket is expected to further fall to 54 percent of total health spending by 2013/14. Figure 5.5: Health Financing Expenditure Shares 76 Percent of total health expenditure 80 8 70 7 60 6 Percent of GDP 49 50 46 5 40 40 34 33 31 4 30 22 3 20 2 10 1 13 0 0 Vietnam Malaysia Indonesia Singapore Thailand Laos Myanmar Cambodia Philippines Public Expenditure on Health (% of total health expenditure, 2011) Total Expenditure on Health (% of GDP, 2011) Source: WDI, Ministry of Health and IMF Country Report No.13/250. 65 5.23. External financing of health services is a relatively small share of total financing at around 5 percent, even though it has been growing in absolute terms. Actual spending may be even higher. Over 80 percent of the external financing (Table 5.3) is channeled through not-for-profit institutions that provide a range of prevention and curative health services, both in clinics and at the community level. Their main activities include health education, medical care (prevention and treatment), and distribution of drugs and other commodities (such as contraceptive pills and condoms). Their main areas of intervention are: (i) maternal, newborn and child care; (ii) reproductive health; (iii) malaria, tuberculosis, pneumonia and HIV/AIDS and other sexually transmitted infections; and (iv) nutrition. Most spending goes to preventive care services, which accounted for 81 percent in 2010/11 and 78 percent in 2011/12. As fiduciary arrangements, policy making and policy monitoring improve within MoH, the donors could be encouraged to provide their assistance directly through the budget process. Figure 5.6: Health Financing (% of GDP) Figure 5.7: Health Financing (% Total Financing) 2.5 2.5 9.2 9.4 7.1 4.6 2.4 100% 5.3 2.0 90% 2.0 1.9 1.9 80% 1.7 70% 1.5 1.5 1.5 54.2 1.4 59.6 60% 1.0 1.3 79.3 0.8 50% 82.3 77.1 1.0 40% 30% 0.5 0.3 0.3 0.2 20% 41.2 35.1 0.2 0.2 0.1 0.1 0.1 10% 0.0 8.5 13.5 13.6 2009-10 2010-11 2011-12 2012-13 2013-14 0% 2009-10 2010-11 2011-12 2012-13 2013-14 All Public Sources Private Out-of-Pocket External Sources Total Health Spending External Sources Private Out-of-Pocket All Public Sources Source: Ministry of Health 5.24. Myanmar has a Social Security Scheme established in 1956 under the Social Security Act of 1954, which provides social service benefits including healthcare. The scheme is funded out of employer and employee contributions (2 percent of salary and, after age 60, 1.5 percent contribution by the employee and 2.5 percent by the employer). The scheme covers only 650,000 persons, or less than one percent of the population. As result, the Social Security Fund is a small fraction of total health financing, ranging from 0.15 percent in 2009 to 0.48 percent in 2011. 5.25. A new law enacted in 2012 requires employees of public or private firms with three or more employees to participate in the scheme, thus broadening coverage. It will also allow formal and informal sector workers to join the scheme voluntarily. Some benefits are expanded: the sickness cash benefit will be increased from 50 to 60 percent of wages for 26 weeks and pediatric care will be provided up to the age of one year. Additional benefits are introduced: family benefit, old age pensions, disability and survivors’ pensions, unemployment insurance, and housing benefits. The law is to be implemented in phases over several years. As of April 2014, only the existing branches of social security were covered (sickness, death, maternity and work injury) but with higher cash benefits than under the 1954 law. As of April 2014, contributions under the new law include the original 4 percent plus the employer contributes an additional 1 percent of the worker’s wage which is assigned to an employment injury fund. It is estimated that full implementation of the new law will require a mandatory contribution equal to 13 66 percent of salary and up to 25 percent of salary for those who also want the new voluntary housing benefits. F. Ministry of Health Expenditures 5.26. The overall level of spending by the Ministry of Health has risen very sharply since 2011/12, from 0.2 percent of GDP to 0.9 percent in 2013/14, but continues to remain low compared to other countries in the region (Figure 5.8). MoH is responsible for three quarters of government spending on health. Its expenditures have increased over nine-fold from MMK 55.8 billion in 2009/10 (equivalent to less than $1 per person per year) to around MMK 500 billion in 2013/14 (equivalent to over $8 per person per year). As a share of total government spending, MoH expenditures on health increased from a little over 1 percent in 2011/12 to almost 4 percent in 2013/14 (Figure 5.8). Figure 5.8: MoH Expenditures Figure 5.9: MoH Capital and Current Spending $9 4 600,000 100 Percent of total government spending 3.8 $8 3.3 500,000 77 80 71 73 US$ per person per year Million Kyats $7 3 400,000 51 61 60 Percent $6 300,000 49 39 29 27 40 $5 200,000 23 2 $4 20 1.2 1.2 100,000 1.1 $3 0.9 0.8 0 0 $2 1 2009-10 2010-11 2011-12 2012-13 2013-14 $1 0.2 0.2 0.2 (left axis) Total MOH Spending on Health $0 0 2009-10 2010-11 2011-12 2012-13 2013-14 (right axis) Capital Spending, percent of Total MOH Spending (left axis) US$ per person per year (right axis) Current Spending, percent of Total MOH (right axis) Percent of total government spending Spending Source: Ministry of Health and World Bank staff calculations. 5.27. Capital expenditure, which in 2011/12 constituted around 27 percent of MoH spending, rose sharply to 51 percent 2012/13 and was one of the big drivers of the overall increase (Figure 5.9). The share of capital fell back in 2013/14 to 39 percent with a shifting focus towards primary healthcare and prevention, which requires less capital investment (see section G). With the level increase in capital spending, there was also a compositional shift away from construction (share falling from 88 percent of capital expenditure in 2009/10 to 43 percent in 2012/13) towards machinery and equipment (share of capital spending in 2012/13 was 53 percent). Preliminary findings suggest that this went to expensive machines (e.g. x-ray machines and CT scanners) in higher-level health facilities such (i.e. mostly hospitals). 5.28. Capital construction expenditure on primary care facilities increased sharply in 2012/13 (Table 5.4). Prior to this, nearly all construction expenditures went to township hospitals and general hospitals. There was no construction-related expenditure in rural health centers during this period. This changed in 2012/13, when rural health centers’ capital allocation went from zero to 14 percent of the construction budget. The number of rural health clinics increased from 1,558 in 2010/11 to 1,635 in 2012/13.59 At the same time, station hospitals, which also deliver primary care, continued to receive about 21 percent of the construction budget. Construction expenditure of teaching hospitals increased dramatically during 59 Myanmar Ministry of Health. 67 this time, up 12 times relative to 2009/10, compared to lower increases of 3.9 times and 2.4 in the budget of specialized and general hospitals. Table 5.4: Ministry of Health Capital Spending on Construction a/ (MMK millions) Minor Head 2009-2010 2010-2011 2011-2012 2012-2013 2013-2014 DOH Head Office 47.9 42.9 305.0 1,226.4 2,596.0 Central Medical Store Depot .. 50.2 81.5 .. .. Teaching Hospitals 351.4 230.0 59.7 4,100.2 6,300.0 State and Division General Hospitals 2,475.3 4,891.0 3,356.3 8,044.3 6,888.0 Specialist Hospitals 638.3 1,139.7 2,310.5 9,037.7 4,650.0 Township Hospitals 1,600.4 7,823.1 5,134.8 14,409.6 5,257.0 Station Hospitals 2,565.5 1,988.7 3,268.1 12,285.5 16,996.0 Rural Health Centers .. .. .. 8,271.4 10,852.0 Total 7,678.7 16,165.6 14,516.0 57,375.2 53,539.0 a. Department of Health only; excludes spending by Departments of Medical Science and Traditional Medicine. Source: Ministry of Health budget data. 5.29. Pay and allowances have tripled between 2009/10 and 2013/14 due to more hiring and increased wages and benefits. The largest year-on-year increase took place in 2012/13, when salaries went up by 83 percent relative to the previous year. One of the measures financed by the budget was a 23 percent salary increase to public sector workers and pensioners intended to reduce the gap between public sector salaries in Myanmar and peer countries. 5.30. The government also hired more doctors and dentists (Table 5.5). The number of doctors per 1,000 people increased from 0.5 in 2010 to 0.6 in 2012. This remains below the lower-middle-income average of 0.8 physicians per 1,000 people.60 Available data suggest that the number of nurses and mid- wives increased as well, but not as quickly as the number of doctors. Thus, the ratio of nurses and mid- wives to doctors fell from 1.8 to 1.7 between 2010 and 2012. The share of pay and allowances in the overall MoH budget however fell from above 55 percent in 2009/10 to below 20 percent in 2013/14 because other categories – namely capital and drugs and consumables – increased more sharply over this period. Table 5.5: Health Care Personnel 2009/10 - 2012/13 2009/10 2010/11 2011/12 2012/13 Number of doctors 24,536 26,435 28,077 29,832 Public 9,728 10,450 11,675 12,800 Cooperative and Private 14,808 15,985 16,402 17,032 Percent private 60 60 58 57 Number of dental surgeons 2,308 2,562 2,770 3,011 Public 703 756 774 802 Cooperative and Private 1,605 1,806 1,996 2,209 Percent private 70 70 72 73 Annual change in public employment Doctors 145 722 1,225 1,125 Dentists -74 53 18 28 Source: Ministry of Health, 2013. Data for 2012-13 are provisional. 60 WHO. 68 5.31. Drugs and medical consumables expenditure has increased most dramatically from MMK 4 billion in 2011/12 to MMK 148 billion in 2013/14. As a share of total MOH spending, drugs and medical consumables accounted for less than 5 percent in 2011/12, increasing to 30 percent in 2013/14. This sharp rise is due to three new and progressive health policies aimed at increasing access to affordable healthcare: (i) introduction of free essential drugs at primary health care facilities; (ii) free health care services for children under five years of age, pregnant mothers, and patients needing emergency surgery (first day of hospital admission); and (iii) free essential drugs to children under 5 years in township hospitals. Figure 5.10: MoH Spending on Health Resource Use 2009/10 – 2013/14 100% 23 29 27 80% 39 51 Percent of total 60% 4 3 5 40% 21 30 55 53 48 20% 21 18 0% 2009-10 2010-11 2011-12 2012-13 2013-14 Pay &Allowances Travelling Allowance Drugs & Medical Consumables Non-Clinical Goods & Services Maintenance Charges Transfer Payments Entertainment & Meals Capital Source: Ministry of Health. G. Efficiency of Government Spending on Health 5.32. The allocative efficiency of government spending has potential to improve if more resources were channeled to programs with strong public goods characteristics.61 Prevention and treatment of communicable diseases is an example of a public good in health, because it benefits the individual concerned and also reduces the risks of infection to others.62 Since the public good provider is not compensated for the positive externality of reducing risk of infection, the private sector would not provide prevention and treatment of communicable diseases in sufficient quantity in terms of societal welfare. This would result in allocative inefficiency, which the government needs to address. 5.33. MoH spending on programs with strong public goods characteristics increased, from MMK 3,401 million in 2011/12 to MMK 6,485 million in 2012/13. Examples include programs for nutrition, water and 61 Allocative efficiency refers to the production of those goods and services that are most desirable for the society, and is thus concerned with the optimal mix of health goods and services. Goods and services with public goods characteristics are those that are provided to almost everyone as a matter of policy, such as universal childhood immunization for all families. A pure public good will provide benefits that cannot be withheld from any member of the population (a positive externality), such as the reduction in infectious diseases that follows from universal immunization. 62 Note, however, prevention and treatment requires excludable inputs, such as vaccination, clean water or condoms, as well as non-excludable inputs, such as knowledge of preventive interventions. Examples of communicable diseases include HIV/AIDS, tuberculosis (especially drug-resistant TB), eradicable disease (e.g. polio), anti-microbial resistant disease, and malaria. 69 sanitation, occupational health, maternal and child health; and programs to fight leprosy, AIDS-STD, tuberculosis and water-borne diseases. As a proportion of total MoH spending on health, however, the share of these programs fell from 5.4 percent in 2011/12 to 3.7 percent in 2012/13, declining further to 2.6 percent in 2013/14 (Figure 5.11). Maternal and child health, water-borne diseases, tuberculosis and leprosy account for over two-third of total spending on these health programs. Figure 5.11: Ministry of Health Spending of Public Goods 8,000,000 8 Percent of MoH spending 6,000,000 6 MMK thousands 6.0 6.2 5.4 4,000,000 4 3.7 2,000,000 2 2.6 0 0 2009-10 2010-11 2011-12 2012-13 2013-14 Maternal and Child Health Water-borne Diseases Tuberculosis Leprosy School Health AIDS-STD Nutrition Occupational Health Water and Sanitation (right axis) % MOH Spending Source: Ministry of Health. 5.34. Assessing the adequacy of spending on public goods programs is difficult, but judgments can be based on relevant outcome indicators. For example, the fact that Myanmar is unlikely to achieve by 2015 its MDG4 target of reducing child mortality to 36 per 1,000 live births and MDG5 target of reducing maternal mortality to 130 per 100,000 live births, suggests that current levels of public goods expenditures are not enough. Thus, the government may want to consider increasing significantly public funding of programs aimed at improving maternal and child health. 5.35. Ensuring adequate preventive care can improve technical efficiency in addition to allocative efficiency.63 Preventive care can help to eliminate or delay the burden of disease and therefore a higher health bill linked to curative care. It can also prevent people with minor complaints from flooding the emergency wards of hospitals. Primary preventions (measures intended to prevent the onset of a condition) focus on administering immunizations to prevent infectious diseases and modifying unhealthy behaviors (smoking), and is the focus of study in the PER.64 Expenditures related to prevention services are recorded in the NHA based on the OECD’s International Classification for Health Accounts functional classification of health care (ICHA-HC). This classification scheme categorizes the types of goods and 63 Technical efficiency refers to the physical relationship between resources used to produce health goods and services—capital, labor, land—and the health outcome, and is achieved when the maximum quantity of an output is obtained using the minimum quantity of inputs. 64 Secondary preventions (measures intended to detect disease in clinically asymptomatic people at an early stage when it is most treatable) focus on reducing the severity of diseases, such as cancer and heart disease, through screening programs that detect the diseases or their risk factors at early stages, before they become symptomatic or disabling. Tertiary preventions (measures to slow the progression of a disease after it is clinically obvious and a diagnosis established) focus on efforts to avoid or defer the complications of diseases after they have developed. 70 services produced by health care providers and by institutions and actors engaged in related activities to health care. One such category is that of “preventive and public health services.” 5.36. Preventive care in Myanmar is funded largely through public and external sources. Even though spending on preventive care from public sources is high, the proportions of total health spending are low overall.65 Expenditures on preventive care went down from approximately MMK 60 million in 2009/10 to MMK 55 million in 2011/12, the last year for which this information is available. As a share of total spending on health, expenditure on preventive care fell from 8.3 percent in 2009/10 to 6.8 percent in 2011/12 (Figure 5.12). The bulk of the spending on preventive care comes from external sources, whose share of total spending on preventive care went up from 72 percent in 2009/10 to 82 percent in 2011/12. During this period, the share of preventive care in total public spending went down from about 30 percent in 2009/10 to 17 percent in 2011/12. Within MoH spending, the share of expenditure on preventive care declined between 2009/10 and 2011/12, while that of curative and capital expenditures increased (Figure 5.13). Figure 5.12: Health Spending by Function, Total Figure 5.13: Health Spending by Function, MoH Only 100 100% Other Other Other 90 90% Percent of total health spending 80 80% Capital Capital Capital 70 Pharma Pharma Pharma 70% Pharma 60 60% Pharma Pharma Percent 50 50% Curative Curative Curative 40% 40 Curative Curative Curative 30 30% Diagnostic Diagnostic 20% 20 Diagnostic Preventive Preventive Preventive 10% Preventive Preventive 10 Preventive 0 0% 2009-10 2010-11 2011-12 2009-10 2010-11 2011-12 Source: Ministry of Health and World Bank staff calculations. 5.37. Four out of the five top causes of premature deaths in 1990 and 2010 were due to preventable communicable diseases. This warrants more spending on preventive care. Unlike treatment, prevention faces unique challenges in demonstrating its economic value, even when the health value is apparent. The basic goal of prevention is to maximize the number of years lived in good health by deferring illness. However, the economic value of prevention is strengthened by the fact that preventive interventions, addressing even a single risk factor, can alter the prevalence and severity of a broad range of conditions. Furthermore, prevention compounds health benefits over time in ways that disease care does not. 5.38. Another potential source of technical efficiency is improvements in the procurement of pharmaceuticals. In addition, greater consumption of free drugs provided by MOH would significantly reduce out-of-pocket spending on health. Yet, as noted above, this has proven to be very expensive. The formulation of a comprehensive policy for the strategic, cost-effective and timely procurement of pharmaceuticals would allow MoH maintain its policy objective on a more affordable basis. 65 The expenditures on prevention presented here are under-estimates. This is because many of the services related to prevention are provided as part of other services and are not specifically recorded as prevention activities. Typical examples are infant and child care, vaccination services, campaigns and special reproductive health programs. 71 5.39. Yet another potential source of technical efficiency is improved linkage between recurrent and capital spending. This can be assessed through very indicative benchmarking. Heller (1991) proposes r- coefficients, which approximate recurrent costs as proportional to capital cost. For the health sector, the value of r ranges from 0.27 to 0.71 for rural health centers, 0.17 for urban health centers and 0.11 to 0.30 for hospitals. Another rule of thumb is that building maintenance costs should be 2 to 4 percent of the replacement cost of the building.66 These are indeed broad indicators, but in practice, some method of approximation is necessary to budget for project-specific recurrent cost implications.67 5.40. Public spending in Myanmar on health infrastructure and capital goods increased nine-fold from MMK 23,840 million in 2011/12 to MMK 185,569 million in 2012/13. Maintenance allocations also went up during this period, doubling from MMK 2,737 million in 2011/12 to MMK 5,518 million in 2012/13 and then increasing six-fold to MMK 35,800 million in 2013/14. Details of these allocations are not available, but the 23 percent set aside of maintenance charges in 2012/13 relative to capital spending in 2011/12 (appropriately lagged by one year) suggests that budgetary implications of increases in capital spending are being taken into account. Maintenance charges in 2011/12 were only 12 percent of capital spending in the previous year. This would need to be analyzed in greater depth. H. Equity in Government Spending on Health 5.41. The gap between spending on urban and rural facilities increased between 2009/10 and 2013/14. This outcome risks exacerbating inequalities in health status and access discussed in section C above. Urban centers accounted for about two-thirds of total public spending on health in 2011/12. Teaching hospitals, state and division hospitals, specialist hospitals, and township hospitals are located in urban areas. Station hospitals, secondary health centers, and rural health centers are located in rural areas (Table 5.6).68 Despite increased attention to rural facilities discussed above, the share of expenditures going to urban health facilities went from 67 percent in 2009/10 to 70 percent in 2013/14, with a concurrent decrease for rural health facilities. 5.42. Healthcare seeking behavior and expenditure on healthcare across quintiles suggests that wealthier households seek more healthcare than the poor (Table 5.7). On average 8.5 percent of the population reports any illness or injury, with fewer among the poorest quintile reporting illness. On average, 78 percent of those reporting an illness or injury seek care, but there are huge differences in health care seeking between the poorest (67.3 percent) and the richest (87.4 percent) quintiles. Out-of- pocket spending varies significantly across income quintiles, with the richest 20 percent spending over six 66 Building Research Board (1990). 67 Some research suggests that a dollar of investment may frequently yield as little as 50 cents worth of capital. The reasons are various, and include poor project selection and implementation, diminishing returns arising from bottlenecks during investment surges, as well as outright corruption. Thus, public investment ceases to be productive if inefficiency, waste, or corruption leave behind a trail of poorly executed and ineffective projects and distort the impact of public spending on capital accumulation,. For an index of public investment efficiency, together with selected country-specific scores, see: Dabla-Norris, E., J. Brumby, A. Kyobe, Z. Mills, and C. Papageorgiu (2011) “Investing in Public Investment: An Index of Public Investment Efficiency” IMF Working Paper, WP/11/37. 68 Note that only current expenditures (Form 2, collected by the Ministry of Finance) can be disaggregated to the health facility level. Capital expenditures (collected by the Ministry of National Planning and Economic Development) cannot be disaggregated to the health facility level. Most drug-related expenditures are also not disaggregated to the facility level, since most drugs are procured centrally through the Central Medical Store Depot. 72 times compared to the poorest. The difference is most pronounced in expenditure on medicines (nine times), followed by inpatient care (eight times) and outpatient care (four and a half times). Table 5.6: Urban-Rural Public Spending on Health at Facility Level (MMK Millions) 2009/10 2010/11 2011/12 2012/13 2013/14 Expenditure Category Urban Rural Urban Rural Urban Rural Urban Rural Urban Rural Pay and Allowances 21,675 7,824 29,408 11,017 30,100 11,180 57,507 19,800 66,394 23,417 Travelling Allowance 262 95 316 115 350 122 661 418 729 427 Goods and Services 2,003 222 2,534 272 2,913 357 4,055 462 11,182 784 Maintenance Charges 1,321 126 2,862 420 1,261 40 3,614 407 32,548 8,407 Transfer Payments 10 - 12 - 23 - 22 29 1 Entertainment/Meals 2 - 3 - 3 - 3 4 - Total Current 25,271 8,267 35,134 11,825 34,648 11,699 65,861 21,088 110,885 33,036 Source: Ministry of Health budget data. Table 5.7: Illness and Health Care Seeking Behavior 2009/10 Poorest Richest All Q1 Q2 Q3 Q4 Q5 In percent Any illness or injury 8.5 7.5 7.8 8.3 9.6 9.9 Any acute illness 7.7 6.6 7.0 7.4 8.7 9.2 Any chronic illness 2.6 2.6 2.4 2.3 2.8 2.9 Seeking care when ill 78.2 67.3 76.6 77.8 82.1 87.4 Seeking care in public facilities 27.6 30.7 32.2 25.9 27.5 22.7 Seeking care in private facilities 43.4 32.8 32.8 45.6 48.1 54.1 In kyat Total health out-of-pocket spending 287,819 88,734 163,973 180,693 333,242 597,711 Expenditure on medicines 132,004 36,372 61,550 75,079 101,478 325,727 Expenditure on inpatient care 88,472 22,183 60,874 47,915 128,698 160,374 Expenditure on outpatient care 51,501 19,859 29,433 41,457 66,645 89,170 Expenditure on other care 15,842 10,319 12,116 16,242 16,421 22,441 Sources: IHLCA and World Bank staff calculations. 5.43. Benefit incidence analysis suggests that government spending on health is weakly progressive in that it helps to close the gap between the rich and the poor. Table 5.8 compares the distribution of public expenditures with population shares and presents cumulative quintile shares of the service-specific subsidies and for the total public spending across all services. The data shows that the poorest quintile’s share of public expenditures is less than 20 percent for general, specialized and station hospitals but very close to 20 percent for ambulatory care. At higher quintiles, the cumulative public expenditures shares are lower than the respective population shares for general and specialized hospitals, but higher for station hospitals and facilities providing ambulatory care. 5.44. Households face a real risk of incurring large medical care expenditures if any member falls ill due to the absence of any prepayment mechanisms, such as health insurance. Not only does this disrupt the material living standards of the household, it is also potentially impoverishing and catastrophic. For many households in Myanmar, most of the budget resources are allocated to food and other items essential for sustenance, leaving very little to spend for health care. A common approach to measure the size of this problem is to define catastrophic expenditures in relation to expenditures net of basic 73 necessities, such as non-food expenditure, which better distinguishes across different consumption quintiles than total expenditure (Table 5.9).69 Table 5.8: Distribution of Health Expenditures, Public 2009/10 a/ Cumulative Shares Household General Specialized Station Ambulatory Total Expenditures Hospital Hospital Hospital Care (%) Expenditures (%) (%) (%) (%) (%) Poorest 20% 4.8613 14.8725 13.4051 16.6078 19.9534 16.4892 (std. error) (.1088) (1.7706) (2.5986) (2.7224) (1.3183) (1.1781) Poorest 40% 11.9805 30.9743 27.1545 45.8446 40.5219 37.9472 (std. error) (.2671) (2.2991) (3.3918) (3.6440) (1.6195) (1.5728) Poorest 60% 21.0021 45.1964 47.7081 60.0365 59.8867 53.8416 (std. error) (.4670) (2.4735) (3.8083) (3.5826) (1.6168) (1.5899) Poorest 80% 32.6383 68.9998 62.8888 80.5433 81.5791 75.3067 (std. error) (.7228) (2.2970) (3.6811) (2.8955) (1.2801) (1.3523) Tests of Dominance b/ against 45 degree line - - + - against Lorenz curve + + + + + Concentration Index .165458 .2194398 -.0076327 -.0116394 .0703147 (robust std. error) (.0632778) (.0679418) (.0522549) (.0427346) (.0352413) Kakwani Index (robust std. er) -.4530304 -.3990486 -.6261211 -.6301278 -.5481737 (.1049665) (.1151551) (.0989894) (.0966091) (.0915197) a. Benefit Incidence Analysis describes the distribution of public spending across individuals ranked by their living standards. The methodology and approach used for this analysis is derived from O'Donnell et al. (2008). Public spending data is obtained from NHA 2009/10. Utilization of public facilities is derived from IHLCA survey. The survey distinguishes between public and private care for all provider categories. Unit costs are obtained by dividing aggregate expenditures by the weighted sum of utilization reported in the IHLCA survey data, where weights are expansion factors indicating how many individuals in the population are represented by each sample observation. b. Dominance tests: – indicates the 45° line/Lorenz curve dominates the concentration curve; + indicates concentration curve dominates 45° line/Lorenz curve; blank indicates non-dominance. Sources: NHA 2009/10, IHLCA 2009/10, and World Bank staff calculations. 5.45. Catastrophic payments are defined for health payments as a share of both total household expenditure and non-food expenditure, using various threshold budget shares. As the threshold is dropped from 25 percent to 5 percent of total expenditure, the estimate of the number of people who suffer adversely increases from 7.6 percent to 35.6 percent of the population. At the same time, the average degree by which payments (as a proportion of total expenditure) exceed the threshold as a proportion of total expenditure – or intensity – increases from 3.5 percent to 6.7 percent of expenditure. Those spending more than 5 percent of total expenditure on health care spent, on average, 23.9 percent of their total expenditure on health. Likewise, those spending more than 25 percent of the household budget on health care spent, on average, 70.7 percent of their total expenditure on health. 5.46. Catastrophic health care costs are a high share of household spending, regardless of how they are measured. Table 5.9 also reports the headcount (incidence) and intensity of catastrophic payments for health defined as a share of non-food expenditure. As expected, both the headcount and intensity are higher for a given threshold when catastrophic payments are defined with respect to health payment relative to non-food expenditure. For instance, for about 20 percent of households, health spending was 40 percent of non-food expenditure, but only 10 percent of total expenditure. At the threshold level of 25 percent, these households spend 94 percent of their non-food expenditure on health care. For the 18 69 The methodology and approach used for this analysis is derived from O'Donnell et al, (2008). 74 percent of households that spend more than 40 percent of their non-food expenditure on health, the average spending on health is 25 percent higher than their non-food expenditure – which is indicative of significant borrowing or dis-saving to finance health care. Table 5.9: Incidence and Intensity of Catastrophic Health Payments 2009/10 Catastrophic payments measures Thresholds: Catastrophic spending as percent of household expenditures Out-of-pocket spending (%) 5 10 15 25 40 As share of total expenditure (%) Incidence (Head count) 35.5 20.6 14.1 7.6 .. Standard error 0.9 0.7 0.6 0.5 .. Intensity (Overshoot) 6.7 5.4 4.5 3.5 .. Standard error 0.7 0.7 0.7 0.7 .. Mean positive overshoot 18.9 26.1 31.9 45.7 .. Share of Household Expenditure Spent on Health Care 54.4 46.7 46.0 53.3 .. As share of non-food expenditure (%) Incidence (Head count) .. .. 42.3 29.1 18.2 Standard error .. .. 0.9 0.8 0.7 Intensity (Overshoot) .. .. 22.4 18.9 15.4 Standard error .. .. 2.3 2.3 2.3 Mean positive overshoot .. .. 52.9 65.0 84.5 Share of Household Expenditure Spent on Health Care .. .. 95.1 94.0 102.7 Sources: IHLCA and World Bank staff calculations. I. Conclusions and Recommendations 5.47. The government’s policies and budget on health went through a major transition in 2012/13. The Union government introduced policies to improve access to and affordability of healthcare services. More medical personnel have been hired. Beginning in 2012/13, essential drugs and selected health care services are provided free of charge to children, pregnant mothers and for patients needing emergency surgery under certain circumstances in some facilities. There are early signs that this is beginning to help reduce prohibitive out-of-pocket payments by households. This policy, however, has caused a dramatic increase in expenditure on drugs and medical consumables. Further costs have been incurred due to wage increases and heavy investments in facilities and equipment. These policies have been financed through a nine-fold increase in MoH spending on health between 2009/10 and 2013/14. 5.48. Ongoing challenges include constraints on policy making and monitoring due to unreliable and incomplete data, and high costs due to expenditure inefficiencies. In the medium-term, the government may also want to consider options to increase equity public service provision including more rural primary care and better financial protection policies. This chapter recommended a number of options to help strengthen policy making and monitoring, improve the efficiency of MoH expenditures and make the provision of health care more equitable. These are summarized below. 75 Issues Options The ability to make effective Ongoing: MoH could conduct periodic household budget surveys for policy and monitor policy reliable data on use and out-of-pocket spending on health. outcomes would be helped by Short-term: MoH could require development partners, private and stronger analytical capacity and not-for-profit institutions financing or delivering health services to access to better information. provide annual data on health spending by function and use. Medium-term: MoH could further develop its analytical capacity. Medium-term: As fiduciary arrangements, policy making and policy monitoring improve within MoH, the donors could be encouraged to provide their assistance directly through the budget process. Improvements in health Short-term: MoH could increase the share of budget resources to outcomes could be gained more preventive care and health interventions with strong public goods efficiently and effectively with characteristics (e.g. nutrition, water and sanitation, occupational changes in the composition of health, maternal and child health, leprosy, AIDS-STD, tuberculosis and expenditures and in procurement water-borne diseases). policy. Short-term: To make free health services less expensive to the Union, MoH could formulate a comprehensive policy on pharmaceuticals that would allow MoH to strategically procure and distribute drugs and consumables to targeted populations. Short-term: To avoid future costs for rehabilitating facilities and equipment, MoH could establish a framework to monitor the extent to which new capital investments are supported by adequate operations and maintenance costs. Public service provision would be Short-term: MoH could formulate a comprehensive policy on capital made more equitable if capital spending that would allow the government to plan and prioritize investments could be better investments on construction and equipment to address geographical prioritized. disparities and urgent needs. Household inequality and Medium-term: MoH could formulate financial protection policies, vulnerability could be reduced funded entirely by the Union, or through social health insurance through better financial mechanisms, that would ensure that nobody faces financial barriers protection policies. when seeking care at point of service. 76 A SOUND FISCAL FRAMEWORK FOR 6 SUB-NATIONAL SERVICE DELIVERY Background: Myanmar has embarked on a transition from a centralized to a decentralized system of government based on its 2008 Constitution. New institutions of sub-national governance are slowly being established but, as yet, with limited authority. Objectives: This chapter provides an assessment of intergovernmental fiscal relations over the period 2009/10 through 2014/15 and suggests options for prioritizing administrative and fiscal reforms. Key findings: The 2008 Constitution provides the overarching framework for sub-national administration in Myanmar. Initial steps have been taken to establish legislative and executive bodies at State and Region level. There are functional tensions between newly established sub-national institutions and traditional arrangements of central control. This could lead to confusion over management of, and accountability for, public services. Accountability for service delivery will improve when results monitoring is systematic and publicly reported. States/Regions have little impact on public sector borrowing because their expenditures are budgeted to match the sum of expected revenues and Union transfers. Most functional responsibilities in the 2008 Constitution are assigned to the Union government, with smaller and more localized responsibilities assigned to States/Regions. The sub-national spending share in 2013/14 was therefore small (7 percent of public sector budget). A relatively large portion of State/Region spending is used for capital investment. Per-capita spending by States/Regions is weakly correlated with poverty incidence. Revenue assignments to States/Regions are limited (6 percent of public sector revenue). Own-source revenues collected by States/Regions financed 36 percent of total expenditures in 2013/14, the remaining 64 percent was financed from Union transfers. The distribution of own-source revenues is uneven: 68 percent of own-source revenues in 2013/14 were collected in Yangon and Mandalay. The burden of State/Region revenue is moderately progressive when comparing revenue collection per-capita to poverty incidence. Intergovernmental fiscal transfers have grown rapidly in the past three years but the lack of a rules-based system for determining general-purpose grants has made them ad hoc and unpredictable on a year-to-year basis. Recommended options: (i) strengthen accountability of public administration to new institutions of sub-national governance by clarifying lines of authority of GAD officials to locally elected State/Region authorities; (ii) improve accountability for service delivery by introducing a systematic performance monitoring and evaluation framework to be designed and developed under the leadership of GAD in partnership with sector ministries ; (iii) develop and adopt a policy on functional and expenditure assignments; (iv) develop and adopt a policy on State/Region revenues with options for potential sub-national revenue autonomy; and (v) adopt a simple formula-based system of intergovernmental fiscal transfers. 77 CHAPTER 6 FISCAL FRAMEWORK FOR SUB- NATIONAL SERVICE DELIVERY A. Background 6.1. Myanmar has embarked on a transition from a centralized to a decentralized system of government based on its 2008 Constitution. The objective is to improve the responsiveness, quality and efficiency of service delivery. Steps have been taken to establish sub-national legislatures, engage local authorities in budget allocations, and increase fiscal transfers to States and Regions. The government is engaging in reforms to strengthen policies and institutions of intergovernmental relations. This is a complex policy agenda involving many stakeholders and requiring careful coordination and sequencing of reforms. 6.2. The PER looks at options for prioritizing administrative and fiscal reforms. It summarizes sub- national administration in Myanmar and analyzes the main pillars of fiscal decentralization, namely: what government functions and expenditures are assigned to whom; what local receipts help to fund those assignments; and on what basis are Union transfers helping to close fiscal gaps. The PER proposes options to strengthen current policies and institutions, and discusses their coordination and sequencing. The chapter ends with a brief review of sub-national budget preparation and implementation. B. Sub-national Administration in Myanmar 6.3. The 2008 Constitution provides the overarching framework for sub-national administration and decentralization in Myanmar. Prior to this, central authorities undertook all government functions and managed all public resources across the country. The 2008 Constitution sets a vision of decentralized self- government within a strong national framework. This is elaborated in the 2012 Framework for Economic and Social Reforms (FESR), which calls for a comprehensive decentralization policy, underpinned by new laws and regulations to support increasing autonomy of State and Regions.70 New institutions of self- government at State/Region level have been active for approximately two years. The decentralization experience to date therefore is at its infancy. The future shape and direction of relations between sub- national units of government and the Union level have been core parts of Myanmar’s ongoing peace process, aimed at finding an end to the multiple conflicts with ethnic armed groups that have affected Myanmar since independence. 6.4. Four different types of sub-national government under the Union are recognized in the 2008 Constitution (Figure 6.1). These include: 7 States, 7 Regions,71 six self-administered areas (SAA),72 and 70 Government of Myanmar, 2012, Framework for Economic and Social Reform. 71 States and Regions are constitutionally equivalent. Regions were formally known as divisions and have been historically populated by an ethnic Bamar majority. They include: Ayeyarwady, Bago, Magway, Mandalay, Sagaing, Tanintharyi and Yangon. States, which were historically recognized and named for their ethnic minority dominated populations, include Chin, Kachin, Kayah, Kayin, Mon, Rakhine and Shan. 72 There are 6 SAAs including 4 self-administered zones and 1 self-administered division in Shan state, and one self- administered zone in Sagaing region. They are areas that have been engaged in civil conflict and are equivalent to States and Regions under the Constitution, in that they may form their own assembly comprising elected and 78 one Union territory (Nay Pyi Taw). Levels below States/Regions serve primarily as deconcentrated units of central government and are not regarded as independent sub-national authorities. They include 67 Districts that form States and Regions, and 325 Townships that form districts. Townships are further divided into towns, villages and urban wards, but with limited governance structure. Figure 6.1: Overview of Government Hierarchy in Myanmar Union Government Nay Pyi Taw Self Admin. Areas States/Regions Union Territory Districts Districts Townships Townships 6.5. The 14 States and Regions each have their own legislatures (hluttaws). These include a mix of elected representatives from particular groups (e.g. 2 from each township, representatives from each ethnic groups with more than 0.1 percent of the Union population), and appointed representatives (e.g. one third representation from defense services). The State/Region executive is headed by a Chief Minister (appointed by the Union President) who appoints a Cabinet. The latter lead up to 12 sub-national ministries that, in turn, have up to 21 departments and State Economic Enterprises (SEEs). Sub-national ministries are part of the new structure and therefore not yet fully functional. The land mass and population of each State and Region varies significantly from around 10,000 Km2 to 155,000 km2 and from 0.5 million to just over 7 million people. Population density is relatively low across most States and Regions (14 to 214 people per km2) except for Yangon (638 km2). See Figure 6.2. Figure 6.2: State/Region Structure and Population Density a/ 14 Shan Blue are Regions. 12 Orange are States. Number of districts 10 Size of bubbles reflect Mandalay 8 population density. Sagaing Ayeyarwady Yangon 6 Tanintharyi Rakhine Magway Chin 4 Kayin Kachin 2 Kayah Bago Mon 0 0 10 20 30 40 50 60 Number of Townships a. Population data are best estimates and subject to significant margins of error. Sources: Data taken from State and Region Governments in Myanmar, Asia Foundation 2013. appointed officials. They are headed by a chairperson appointed by the Union President. They may propose annual budgets financed from funds allocated to them by region and state governments. 79 6.6. Accountability for service delivery would be more effective with greater clarity about subsidiarity between the various levels of government. For example, although the Union could design routes for municipal bus services, municipalities could do the job as efficiently and with greater accountability to the local populations being served. The 2008 Constitution recognizes the three branches of sovereign power (executive, legislative and judicial) as being shared between the Union, Regions, States and SAAs.73 6.7. Central government however retains a strong role, which is consistent with a unitary system. At the same time, unitary systems typically have relatively strong local governments (i.e. township and district level) in terms of raising revenue and delivering services; regional governments (i.e. states and regions) focus mostly on coordination and intra-regional issues. Myanmar seems to be moving in the direction of a stronger regional role. Under a unitary system, this arrangement could pose risks that would need to be managed. These include: centralization within States/Regions (i.e. bloated regional administration, lack of delegation to local authorities), and lack of progress on redistributive or other development policies of central government. There has been mention of potentially moving to a federalist structure, though this would require even stronger sub-national legislative and executive capacity. 6.8. Public administration at the sub-national level retains strong centralized features from the pre- 2008 period. Prime among these is the role played by the General Administration Department (GAD) of the Ministry of Home Affairs (MoHA). Centrally appointed senior GAD officials serve as Executive Secretary of the Office of the State or Region, and are responsible for coordinating between State/Region ministries and departments. This Executive Secretary is appointed by the MoHA, and reports upwards to the Union level ministry, although in theory this position is also meant to report to the Chief Minister. The GAD is responsible for administration at the township level, coordinating efforts of central line agencies such as the ministries of health and education as well as directly providing services through deconcentrated central line agencies (e.g. land registration, tax collection). 6.9. The creation of new sub-national institutions alongside older, centralized agencies is contributing to challenges in establishing management and accountability for services at sub-national level. The establishment of State/Region level legislative and executive bodies represents a significant move towards greater autonomy and accountability to local citizens. By contrast, centrally appointed officials are not directly accountable to regional or local elected representatives for their performance. Although there are State/Region level cabinet positions with specific responsibilities for service delivery, the capacity to develop and execute policies is at a nascent stage. GAD performs the secretariat function to these new Ministers but GAD officers are upwardly accountable to the MoHA rather than to the State/Region Minister. This significantly impairs the ability of sub-national governments to manage staff, execute policies and hold officials accountable for their performance. 6.10. Options for addressing management and accountability issues by effectively managing this transition should be considered early. As a start, line agencies should coordinate with and be accountable to sub-national government. Currently there are no formal mechanisms allowing State/Region executive and legislative branches to have meaningful input on the central provision of key public services in their area. There are two possible solutions but each involves making some trade-offs. One option is to reassign all GAD staff from the Office of the States and Regions to the direct supervision and control of the state or region executive (with checks and balances held by the state or region parliament). Under this option, the position of Executive Secretary of the Office of the States and Regions 73 2008 Constitution Article 11.b. 80 would be directly accountable to the Chief Minister, who would have power to hire, fire, discipline, or reward as necessary. The existing Government Office and Parliament Office would continue to provide support to the executive and the parliament respectively; however, staff under the General Administration Office would be reassigned to the state and region ministerial portfolios. This would maintain existing capacity but it could also require a potentially difficult transfer of loyalties. The second option would be to refocus existing GAD staff on coordination between the Union level government and state and region governments, instead of day-to-day operations, and to permit states and regions to hire and manage their own individual administrations. The existing cadre of Municipal Officers could serve as a foundation for this new sub-national public administration. This option would entail a significant capacity development effort and time. 6.11. If deconcentrated line agencies remain the primary service providers at sub-national level over the medium-term, then the government could also adopt a system of dual subordination. This would mean upward accountability to the central line agency, as well horizontally to the State/Region. The President has proposed regulatory changes that would move the system in this direction. The proposed changes would allow States/Regions to compile performance evaluation forms for line agency officials at the State/Region level, and submit these evaluations to the Union government. Such a system has been used effectively in Pakistan following the promulgation of their local government law, and is being considered in Thailand. C. Fiscal Developments 6.12. Public sector borrowing was not affected much by the fiscal performance of the states and regions. This is because the overall balances of each state and region had been kept close to zero. The overall balance for all states and regions as a share of GDP was a surplus of 0.06 percent in 2012/13 PA and deficit of 0.003 percent of GDP in 2013/14 RE. The authorities achieved this by ensuring that budgeted expenditures would closely match expected revenues, as illustrated in Table 6.1 below. Total budgeted expenditures in 2012/13 BE and 2013/14 BE were very close to 100 percent of budgeted revenues and transfers. The only exceptions were Kayin and Yangon in 2012/13. Table 6.1: Total Expenditures as Shares of Total Revenues and Transfers a/ 2012/13 BE 2012/13 PA 2013/14 BE 2013/14 RE Ayeyawaddy 100 95 101 100 Bago 99 93 100 100 Chin 93 96 99 101 Kachin 99 94 101 100 Kayin 54 94 100 100 Magway 99 94 99 100 Mandalay 100 98 100 98 Rakhine 104 101 100 100 Sagaing 101 96 101 99 Shan 100 94 100 100 Taninthayi 100 88 101 100 Yangon 80 97 99 100 Total 96 96 100 100 Standard deviation 14 3 1 1 Overall Balance, Millions of Kyat 27,024 -1,668 Overall Balance, Percent of GDP 0.06 0.00 a. Excludes SEEs. Data were unavailable for Kayah and Mon. Sources: Ministry of Finance and World Bank staff calculations. 81 6.13. The authorities ensured that actual expenditures were also kept very close to actual revenues and transfers, as shown in Table 6.2 for 2012/13 PA and 2013/14 RE. Total spending by all states and regions as a share of total revenues and transfers was 96 percent in 2012/13 PA and 100 percent in 2013/14 RE. This was achieved by adjusting Union transfers to the states and regions when actual revenues plus budgeted transfers were insufficient to cover expenditures. This is illustrated in Table 6.2 below. For example, in 2012/13, actual revenues collected by Kayin plus budgeted transfers were only 53 percent of budgeted revenues and transfers. The authorities therefore increased transfers to compensate for the missing 47 percent. This implies that actual Union transfers will often deviate from budgeted amounts. Table 6.2: Budget Components as Shares of Total Revenues and Transfers (Percent) a/ Own Revenue Transfers Sub- Transfers Net Other Expenditures Budgeted + Deviation + Budgeted = total + Incremental + Actual = Actual 2012/13 PA Ayeyawaddy 22 5 72 99 1 -5 95 Bago 36 2 56 95 5 -7 93 Chin 33 -1 62 94 6 -4 96 Kachin 27 -1 73 99 1 -6 94 Kayin 55 -39 37 53 47 -6 94 Magway 13 2 80 95 5 -6 94 Mandalay 65 7 28 100 0 -2 98 Rakhine 13 -5 85 93 7 1 101 Sagaing 16 2 50 68 32 -4 96 Shan 71 -25 48 93 7 -6 94 Taninthayi 12 8 59 80 20 -12 88 Yangon 45 12 7 64 36 -3 97 Total 40 0 44 84 16 -4 96 2013/14 RE Ayeyawaddy 24 10 67 101 -1 0 100 Bago 25 -1 57 81 19 0 100 Chin 10 0 88 99 1 1 101 Kachin 21 1 71 93 7 0 100 Kayin 17 0 67 84 16 0 100 Magway 22 -13 85 95 5 0 100 Mandalay 56 13 25 94 6 -2 98 Rakhine 9 0 62 70 30 0 100 Sagaing 18 -1 73 90 10 -1 99 Shan 48 2 51 100 0 0 100 Taninthayi 9 1 80 90 10 0 100 Yangon 70 14 13 97 3 0 100 Total 37 4 52 93 7 0 100 a. Excludes SEEs. Data were unavailable for Kayah and Mon. Sources: Ministry of Finance and World Bank staff calculations. 6.14. The sections that follow will review the level and composition of expenditures, revenues and transfers. The data will be compared to other countries around the world and analyzed also for issues of equity across states and regions. D. Expenditure Decentralization 6.15. Functional and sub-functional responsibilities of different levels of government are laid out in Schedule One (Union), Schedule Two (States/Regions) and Schedule Three (SAAs) of the 2008 Constitution. Major public services are either exclusively or primarily the responsibility of the Union 82 government. These include education, public health, infrastructure, energy production and distribution, resource extraction and processing, and social welfare. This suggests that, as per the Constitution, the Union government could deconcentrate responsibilities, though it may be difficult to devolve them completely. State/Region responsibilities are mostly confined to localized, small-scale services. (See Annex 3 for a summary.) The newly established ministerial portfolios at the State/Region level do not align perfectly with those of central ministries. This has created overlaps that need to be addressed, especially if the authorities are intending to develop a system of dual subordination. 6.16. State/Region expenditure as a share of overall public sector expenditure is unsurprisingly low relative to international comparators (Figure 6.3). Since Myanmar is still in the very early stages of decentralization, States/Regions accounted for approximately 7.1 percent and 7.4 percent of public sector expenditure in 2012/13 and 2013/14 respectively. Budgeted State/Region expenditure for 2014/15 grew to approximately 11 percent, inclusive of spending out of transfers from the Union government. The average across other countries in the 1972 – 2000 period is closer to 25 percent. In 2013/14, 64 percent of sub-national government spending (excluding SEEs) was financed out of transfers from the Union due to low local receipts.74 The level of dependence on transfers across other countries in the 1972-2000 period is closer to 35 percent (Figure 6.4). Figure 6.3: Spending and Revenue Decentralization Figure 6.4: Decentralization and Transfers 60 80 Sub-national revenue/general govt rev. (%) 70 50 Transfers/sub-national revenue 60 Myanmar 40 50 30 40 Myanmar 30 20 20 10 10 0 0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Sub-national spending/general govt spending (%) Sub-national spending/public sector spending (%) Sources: World Bank fiscal decentralization indicators and staff estimates. 6.17. The economic composition of sub-national expenditure varies significantly across States/Regions.75 In general, however, capital spending as a ratio of State/Region budgets is large, with 12 of 14 States/Regions spending at least 40 percent of their budgets on capital (Figure 6.5). Interestingly, the two States with highest poverty incidence, Rakhine and Chin, have very different shares of capital and recurrent expenditure. This may reflect the composition of Union expenditure in those States, but there is likely little coordination between levels, so this would need further separate analysis to explain the differences. 74 Excludes financial receipts. 75 There was no data available on the functional breakdown of sub-national expenditure. 83 Figure 6.5: Expenditure Shares, by Economic Categories, 2013/14 RE 100 11 17 16 16 16 12 17 18 12 11 90 20 22 80 Salary 19 23 20 26 70 35 36 45 34 40 60 61 48 50 54 Other 40 70 30 60 62 62 45 49 52 20 38 44 29 35 10 26 Capital 0 Rakhine Chin Kayin Mandalay Yangon Shan Bago Magway Sagaing Kachin Ayeyawaddy Taninthayi Source: Data from Ministry of Finance. 6.18. The share of wages and salaries in State/Region budgets is low, at an unweighted average of 14 percent. This outcome is almost certainly explained by the Union’s dominance in service delivery. This is likely to change as sub-national legislative and executive bodies get fully established, particularly if more functional responsibilities are assigned to sub-national authorities. Going forward, it will be important to balance the need for more staff for service delivery against the need for more staff for general administration. 6.19. Budget execution rates for expenditures were close to 100 percent for most states and regions. A few States/Regions overspent, notably Kayin, Sagaing and Yangon (Figure 6.6). This was made possibly mainly because of higher than expected revenues. Very few States/Regions underspent: Shan deeply underspent in 2012/13 while there was some minor under-spending by Kachin in 2012/13 and Magway in 2013/14. This was generally the consequence of less than expected revenues. Figure 6.6: Budget Execution Rates for Expenditures 250% Actual as % of budget 200% 2012/13 PA 2013/14 RE 150% 100% 50% 0% Chin Shan Ayeyawaddy Mandalay Rakhine Magway Bago Kayin Yangon Kachin Sagaing Taninthayi Source: Data from Ministry of Finance. 84 6.20. The level of per-capita spending across States/Regions seems weakly correlated with poverty levels (Figure 6.7). The correlation improved between 2012/13 and 2013/14, moving from 0.36 to 0.47. Expenditures in poorer States/Regions with low population density are expected to play a bigger role due to higher service needs and costs of delivering those services. Chin for example, has high poverty incidence and low population density, which has translated into high per-capita spending. This was thanks to higher transfers, the overall redistribution effect of which is discussed in section F. Two States/Regions with lower incidence of poverty (Kayah and Yangon) appear to be outliers with large per-capita spending. A fuller picture of per-capita spending and poverty links would be required to determine whether expenditures and transfers for these outliers should be modified. Moreover, any analysis is incomplete without full knowledge of what the Union government spends in each state and region. Such spending could usefully augment expenditures in States/Regions with particularly low capacity for public service delivery. Figure 6.7: Poverty and Expenditures Per-capita a/ 80,000 70,000 Chin Expenditures per-capita 60,000 Kayah 50,000 40,000 Yangon 30,000 20,000 10,000 0 0% 10% 20% 30% 40% 50% 60% 70% 80% Poverty incidence P(0) a.Expenditures exclude SEEs. Population estimates pre-date the 2014 census. Sources: Integrated Household Living Conditions Survey, Ministry of Finance and World Bank staff calculations. 6.21. Greater decentralization of functional and spending assignments could potentially improve prioritization and outcomes of service delivery, but this needs to be systematic and based on clear principles. Assigning functional responsibility to different levels of government can be seen at two levels. The first is the assignment of the actual government function (e.g. education) or sub-function (e.g. primary education). Secondly, for each function or sub-function, there should be clarity on who is assigned responsibility for specific tasks such as: (i) policies, standards, oversight (e.g. which level of government is responsible in primary education for policies on education curriculum, standards on student/teacher ratio, oversight of policy implementation); (ii) provision and administration (e.g. which level of government is responsible for recruitment and performance management of teachers); (iii) production and distribution (e.g. which level of government is responsible for contracting the services of the private sector to provide school meals). 6.22. Some functions should exclusively be the responsibility of central government (e.g. defense or foreign policy) as these are national in scope. Other government functions and sub-functions can be shared or concurrent across different levels of government (e.g. both central and sub-national authorities 85 responsible for tertiary education). Generally, functions should be decentralized to the lowest level of government that is fully capable of performing those functions and ensuring responsiveness to citizen needs. For shared/concurrent functions, however, only one level of government should be responsible for each of the tasks noted above within a particular jurisdiction (e.g. in one district, only one level of government should be involved in provision of primary education) to avoid overlap and coordination costs. Usually, regulation of service delivery (i.e. policies, standards, and oversight) is assigned centrally to ensure adequate access and quality across the country. The provision of services and administration is often decentralized. 6.23. The decision on what functions and tasks to assign to different levels can be based on several considerations. These include: economies of scale (whether more centralized provision is more efficient); externalities (if an activity in one area affects another area, a higher level may be best suited); equity (if equal access is a priority, centralized provision may be better); and variation in needs (if people have different requirements and preferences, lower levels could better address these variations).76 6.24. A policy on functional and expenditure assignments based on the above principles and provisions in the 2008 Constitution would be beneficial to all stakeholders. When developing such a policy, it would be helpful to be clear about any exclusive responsibilities assigned to a particular level of government. For exclusive assignments, one order of government is responsible for regulation, financing and service delivery. Secondly, the policy can elaborate on which functions and sub-functions can be shared between the Union and States/Regions. The 2008 Constitution would suggest that although the central government retains responsibility for most major functions, it still has the option to deconcentrate and therefore share some responsibilities to States/Regions. Actual decentralization can be asymmetric (i.e. States/Regions with more capacity can have greater decentralization). Thirdly for each of the shared functions, the policy should elaborate on responsibilities for regulation, standards, oversight; provision and administration; and production and distribution. Fourthly the policy can lay out options for adopting, implementing and monitoring some of these measures. E. Fiscal Transfers 6.25. Intergovernmental fiscal transfers increased from 0.6 percent of GDP in 2011/12 to 1.2 percent of GDP in 2011/12, and 2.3 percent of GDP in 2014/15.77 These transfers account for the majority of States/Regions revenues. In 2013/14, transfers accounted for 64 percent of total receipts collected by the states and regions (excluding financial receipts). Transfers from the Union government can be grouped under two broad categories: (i) grants for State/Region authorities, which are similar to general purpose transfers; and (ii) development funds (e.g. poverty, township development), which are similar to conditional transfers. 6.26. Fiscal transfers do more than provide extra resources; they have also helped equalize the distribution of per-capita receipts across States/Regions. Equal per-capita revenues per jurisdiction can be visualized as a 45-degree diagonal line, as depicted in the Lorenz curves in Figure 6.8.78 Inequality can 76 Raghunandan (2013). 77 Loans to States/Regions should be recorded as financial receipts. 78 A fuller measure of equalization can be defined in terms of the ability of each State/Region to provide reasonably comparable levels of public service at reasonably comparable tax burdens. While there is insufficient data to support 86 be visualized as curves below the diagonal line. Per-capita own-source current receipts are unequally distributed across States/Regions, as discussed earlier. This inequality is exacerbated with the addition of large capital receipts collected mainly by wealthy Yangon and Mandalay. The bottom four quintiles collected only 44 percent of all revenues in 2013/14 while the top quintile collected 56 percent (Figure 6.8). The addition of intergovernmental transfers increases the share of the bottom four quintiles to 55 percent and reduces the share of total per-capita revenues held by the top quintile to 45 percent. This analysis should be considered preliminary: a more complete analysis should also include direct Union financing of services across States/Regions. Figure 6.8: Degree of Equity in Distribution of State/Region Receipts Per-capita 100% 90% Cumulative share of resources Equal shares 80% 70% 60% 55% 50% Revenues and transfers 40% 44% 30% 20% 10% Own total revenues 0% 0% 20% 40% 60% 80% 100% Cumulative Share of State/Region Resources Each quintile = 20% Sources: Data from Ministry of Finance and World Bank staff calculations. 6.27. The process for determining general-purpose grants begins with States/Regions’ budget proposals. These are reviewed and approved by the State/Region parliament before being forwarded to the Finance Commission, which recommends the level of transfers. These are consolidated into the Union budget that is submitted to the Union Parliament, which makes the final decision. 6.28. The lack of a rules-based system for determining general-purpose grants has made them ad hoc and unpredictable on a year-to-year basis. The rapid growth in transfers between 2011/12 and 2014/15 was not supported by an analysis of costs in providing services, nor by any formal rules. The multiple reviews and approvals needed at Union level creating delays in communicating the approved level of transfers to States/Regions. In some instances, transfers were only confirmed one month after the start of the new fiscal year. This jeopardizes any attempt at bottom-up planning and budget preparation, even for a limited set of functional and expenditure assignments. 6.29. The states and regions would benefit from the development of a simple, transparent and formula-based system of general-purpose grants to encourage sound budget management by the nascent institutions of States/Regions. Ordinarily, a system of general-purpose transfers to sub-national authorities would aim to enhance the ability of each State/Region to provide reasonably comparable levels of public service. This could be based on a formula that helps estimate fiscal capacity on the one hand (i.e. based on estimations of own-source revenue collections) and spending needs on the other (e.g. such an analysis for Myanmar, anecdotal evidence suggests major disparities in service delivery quality and access across states and regions. 87 based on criteria that help assess the level of development and performance in delivery of services assigned to the State/Region). The gap between the two would be the level of transfers. 6.30. It is strongly recommended to keep the approach as simple as possible, given States/Regions’ evolving and very limited revenue and expenditure responsibilities at this stage. Since the majority of services in States/Regions are still provided directly by the Union authorities, the transfer system at the moment would not in itself have a major impact on overall level of public service delivery in the States/Regions. It would nonetheless impact significantly on the key economic services, and some social services. While these may still be marginal and supplementary to Union level provision of services at State/Region level, it is critical to embed early the principles of transparency, equalization, and predictability in the transfer system. 6.31. The government could consider two possible options on formulae for general-purpose grants. The first would aim to fill the horizontal gap between fiscal capacity and spending needs across States/Provinces. This would be different to the current arrangement because spending needs would be estimated ex ante across all State/Regions on a consistent basis drawing on specified criteria (i.e. rather than simply using the State/Region budget estimates). Since most of the State/Region expenditure assignments are for economic services, the criteria for estimating spending needs could be simple indicators of: (i) economic development (e.g. spending needs are inversely proportional per-capita GDP); equalization (e.g. spending needs are directly proportional to population); and (iii) efficiency (e.g. spending needs are inversely proportional to population density or share of State/Region that is rural). 6.32. A second option for general-purpose grants would be for the authorities to apply the formula on a predetermined pool of Union resources. Under this option, instead of having to determine the fiscal gap for each State/Region, the government may decide to increase the overall pool of transfers each year by a certain percentage, then apply the above criteria for each State/Region to determine their share of that pool. One possible advantage of the second option is that it may create less incentive to underestimate State/Region revenue and thereby inflate the size of the needed transfers. As discussed earlier, underestimation of revenue seems to be an ongoing problem, e.g. actual revenues have often been over-realized in comparison to original estimates. 6.33. A clear policy for what happens to over-realized revenues would be useful. If States/Regions are allowed to retain the over-realized revenue to spend at their discretion, this could lead to consistent underestimation. Conversely, if they are not allowed to retain the over-realized revenue, then it could discourage local revenue effort. Balance is needed. These issues would need to be monitored closely as the transfer system evolves. 6.34. Development funds are another type of transfers. These are aimed at promoting national objectives such as poverty reduction fund or township development. For FY2014/15, development funds were around 5 percent of total transfers. The poverty reduction fund totaled around some 50,000 million kyats, and was divided on the basis of poverty incidence rates, with 15,000 million each for Rakhine and Kachin States, 5,000 million for Shan state, and the remaining 15,000 million divided equally amongst states and regions. The constituency/township development fund grant represented some 33,000 million kyats, and was allocated equally across townships. 6.35. As capacity grows, Myanmar could consider turning the development funds into output- oriented grants to link transfers to results. These grants aim to promote sub-national autonomy by giving managers the freedom to make spending decisions consistent with local conditions and priorities, while 88 also preserving accountability by making transfers conditional on delivery of results – typically in the form of some national minimum standards for service delivery in the particular area. Good access to reliable and timely information on service delivery performance is required – as such, introduction of such grants may be a medium term consideration for Myanmar, and may exist as part of a suite of reforms to improve public sector accountability. 6.36. Myanmar could also consider capital transfers to help address infrastructure deficiencies across the country. These could be awarded based on centrally planned criteria – for example, proximity of villages or townships to certain categories of roads, schools within defined distance from residents. Currently, capital needs are articulated on a project-by-project basis through bottom up planning. However, there is a risk more developed States/Regions end up attracting a larger share of the total pool of capital resources because they have more capacity to develop and articulate proposals and capital needs. In such cases, a formula based capital transfer program may prove more effective in promoting relatively comparable access to capital infrastructure both within and across different States/Regions. F. Sub-National Planning and Public Finance Management 6.37. Sub-national planning and public financial management capacity will be central to effective implementation of the above policies. The preparation of State/Region plans and budgets is conducted in parallel. Planning Departments in the States/Regions are meant to coordinate the preparation of 5-year plans and the investment budget, while State/Region Budget Departments prepare the recurrent budget and coordinate the overall budget. In reality, institutional responsibilities are more blurred and depend in big part on capacity across States/Regions. 6.38. Annual plans in Myanmar are based on production targets in different sectors of the economy. These are determined by iteration between the Planning Departments and Union level ministries. There is often considerable divergence between the ‘bottom up‘ estimates of states and regions and the ’top down’ estimates of the Union ministries. The key formal mechanism for reconciling these differences is through the Planning Commission, which includes both Union and State/Region representation. Annual plans however have scope for improvement if they are to better inform budget allocations. The plans should cover the policies and activities that require financing through State/Region budget and should explain how these policies and activities can help achieve the State/Region’s development objectives that cannot be performed by the private sector. 6.39. Budget preparation is led by the MoF Budget Department, which starts the process in August through issuance of budget circulars, including to States/Regions. State/Region Budget Departments coordinate the process at sub-national level. Each sub-national ministry/department submits budget estimates to the Budget Department based on standard guidelines and forms. Budgets are prepared on an incremental basis rather than informed by plans and policies. Allocations down to districts and township offices are managed at the budget execution stage with no budget ceilings agreed prior to budget approval. Ministries/departments are also at liberty to make changes to allocations during the course of the financial year (e.g. moving budgeted funds from one township to another). There is also no parliamentary oversight of how funds will be allocated down to district and township level. 6.40. Budgets are first scrutinized by State/Region Finance Ministers before going to budget vetting committees chaired by Vice President 2. Budgets are then submitted to the Union Financial Commission chaired by the President with the Minister of Finance acting as Secretary. Chief Ministers from all the states 89 and regions are members, as well as the Vice Presidents, the Auditor General, the Attorney and the Chairperson of Nay Pyi Taw Council. The next stage is submission of State/Region budgets for approval to the Union level parliament. Only after being approved by the Union hluttaw are budgets finalized and submitted to state and region hluttaws, which play a very limited role in influencing budget decisions. 6.41. States/Regions are calling for more influence over sub-national budgeting. At present, States/Regions lack authority to influence Union-level service provision in their jurisdiction. Even for functional responsibilities assigned to States/Regions, budget allocations are agreed at the Union level and there is limited scope for States/Regions to reallocate resources between their ministries/departments. Responsibility for key decisions that might influence budgetary allocations has also not been decentralized: for example, the recruitment of additional staff would require approval from the Union. Capital and recurrent budget preparation are managed through separate processes and as such there is limited scope to reallocate resources from one to the other. Authorities therefore have little control over ensuring an appropriate balance between capital investment and the necessary recurrent expenditures to use those assets productively. 6.42. Budget execution procedures are similar to those in place at the Union level. All expenditure financed through State/Region budgets is meant to take place through accounts held at Myanmar Economic Bank (MEB). Once the budget has been approved, responsibility for authorization for spending (within approved budgetary allocations) passes to MEB. Expenditures are controlled using the same system of quarterly ‘drawing limits’ that is in place at the Union level.79 State/Region offices of MEB are responsible for administering these controls. Each spending unit80 that receives funds through state and regional budgets holds an individual account at their local MEB branch for making payments. Payments from accounts held at the MEB are made using bank cheques or cash. It is then the responsibility of the spending units to make payments and to record those transactions. 6.43. Accounting and financial reporting are based on Union financial regulations. Financial statements are prepared on a monthly basis and show cash payments and receipts as well as movements in cash balances. Responsibility for preparing financial reports is highly decentralized. Each spending unit is responsible for submitting monthly accounts to respective departments. The finance sections in each department compile information and prepare a summary report to be submitted to the State/Region Budget Department. State/Region offices of MEB compile a consolidated monthly bank statement for each department in their state or region. These statements are then submitted to State/Region budget departments for reconciliation with the financial reports that are submitted by the finance sections of each state and regional departments. Once financial reports and bank statements have been reconciled, State/Region Budget Departments forward the information to the Union Budget Department. 6.44. State/Region accounts are audited by the deconcentrated Offices of Auditor General (OAG). The OAG submits the audited reports to the respective hluttaw, where the findings and recommendations are discussed and reports are approved. Recommendations are sent back to each ministry with guidance on improvements. An example from the FY12-13 Audit Report recommends that civil servants taking annual leave of more than 10 days should not be paid for extra days. Internal audit mechanisms are established in some state and region departments (e.g. Mandalay region Construction Department), but this is not uniform in across the country. The scope of internal audit is limited to basic financial audit. 79 See World Bank (2013), Public Financial Management Performance Report for further details. 80 Within each department, there may be a number of spending units (also called primary accounting units). Certain departments have offices at district and township level, each of which would constitute a spending unit. 90 6.45. The current lack of transparency surrounding State/Region budgets inhibits linkages between the local priorities of citizens and the allocations that are approved by State/Region parliaments. There have been some progress on transparency of the Union budget in recent years. The draft budget law presented to parliament is now routinely published in the national press,81 which shows the size of transfers (both grants and loans) that are made to States/Regions. At present, however, it is more difficult to access information in the media about how State/Region budgets are being allocated and actually used. Further, the comprehensiveness of budgets presented to state and region hluttaws is also unclear. 6.46. Giving state and regional governments greater influence over the use of public funds will also require an effective reporting and oversight framework between the center and States/Regions. Existing procedures for financial reporting ensure that there are numerous checks and balances during preparation of financial reports and that detailed information is available on each and every spending unit. These reports do not lend themselves easily to analysis. As such, they are rarely used to inform dialogue between the Union and States/Regions. 6.47. The decentralized approach to reporting involves over 5,000 spending units that must submit accounts on a monthly basis. This means there is enormous potential for evidence-driven expenditure and policy analysis. However, a lack of a standardized classification and manual systems mean that the information available is unsuited to deeper analysis. For example, it is not possible to analyze how funds are being allocated and spent by deconcentrated units at the township level. Putting in place the information and technology systems and processes to analyze this rich data will enable improved understanding of how expenditures are being distributed to the population. 6.48. The existing system of practice-based regulation of public financial management should move towards one which clearly defines the boundaries in which states and regions can operate. The Union government has previously relied upon a practice-based system of regulation. As states and regions gain new powers and recruit new staff, new risks will arise as well. On the one hand, states and regions will not feel sufficiently empowered to innovate, as it is unclear as to where they can follow their own procedures and where they must fit in with existing Union practices. On the other hand, in an absence of regulation, there is also a risk that states and regions may deviate from certain core practices that the Union government would view as non-negotiable. These risks can be addressed by better defining the regulatory environment for public financial management. 6.49. A lack of regulation also makes it difficult to exercise proper oversight of state and regional spending. It is difficult to sanction officials, if there is no clear guidance on what is, or is not permitted. Oversight is also currently limited to a fairly basic financial audit of their budgetary transactions. However, often the greatest risks to fiscal stability are the transactions undertaken by regional and local entities that take place off-budget. G. Revenue Decentralization 6.50. The different types of own-source revenue for States/Regions are set out in Schedule 5 of the 2008 Constitution. These include municipal taxes, transportation taxes, taxes on a limited range of forestry products, salt tax and an entertainment tax. Non-tax revenues include land revenues, registration fees, and service fees including tolls, fines and profits from investment. These are broadly in line with 81 See for example the New Light of Myanmar. 91 sound principles of revenue assignment, which includes assignment of taxes with immobile bases (e.g. land and property) to local authorities because there is no scope to shift location to avoid taxation. It is also helpful to provide the States/Regions authorities with autonomy on fees and charges for local services. 6.51. The overall level of revenue of States/Regions is unsurprisingly low, in line with limited spending assignments. All major sources of general government revenue in Myanmar (i.e. commercial tax, personal and corporate income taxes, customs duties, stamp duties, various fees and fines, and the majority of revenues derived from natural resources) accrue to the Union government. Myanmar has a centralized and harmonized tax regime. Own-source revenue of States/Regions account for around 6 percent of public sector revenue in Myanmar, which is low compared to other countries (Figure 6.9). Figure 6.9: Sub-national Revenues (Percent of Total Public Sector Revenues) 50 45 40 35 Regional average 30 25 20 15 10 5 0 Vietnam Tajikistan Germany Portugal Mexico China Malaysia Indonesia Russia Latvia Belarus Slovenia Denmark Moldova Romania Sweden Thailand Norway Argentina Philippines Poland Spain France Lithuania Costa Rica Myanmar Hungary Switzerland Finland Brazil Chile Mongolia Estonia Kyrgyz Rep. Bulgaria Italy Bolivia Kazakhstan Azerbaijan East Asia East Europe and Central Asia Europe Latin America and Caribbean Source: World Bank Fiscal Decentralization Indicators and Ministry of Finance. 6.52. The composition of total own-source revenue collections in 2013/14 shows that 54.2 percent came from non-tax sources, 30.4 percent came from capital receipts (i.e. sale of government assets), and 15.4 percent came from taxes (Figure 6.10). It is helpful to treat capital revenue separately to tax and non-tax revenue because these are receipts from sale of fixed assets: channeling them to recurrent activities is not sustainable. Tax collections are likely to remain low in the years to come given a historically narrow tax base and limited tax assignments. Non-tax receipts, on the other hand, are more transaction- based and in exchange for delivery of particular services, therefore easier to collect. One issue to monitor is that non-tax fees and charges do not become overly regressive, such that there is disproportionate burden on poorer households. 6.53. The level and composition of own-source revenue across States/Regions vary significantly. In 2013/14, 68 percent of total own-source revenues were collected in Yangon and Mandalay (Figure 6.11). They are among the top three most populated States/Regions after Ayeyarwady. Both Yangon and Mandalay are also major economic hubs in Myanmar, which provide the local authorities the scope for earning important non-tax receipts. It is also important to note that Yangon and Mandalay collected nearly all of the capital receipts, most likely through sale of land given the property development boom in both cities in the past three years. 92 6.54. One source of non-tax receipts are contributions from SEEs operating within the State/Region. Twenty percent of SEEs’ after tax profits must be paid to States/Regions. Available data suggests that these contributions are less than 10 percent of revenue for most State/Regions (Figure 6.12). This may be because many of these SEEs are small or are not very profitable in the first the place. The exceptions are Chin and Rakhine which have very low own-source revenues. In all other States/Regions, most tax and non-tax revenues are paid by the private sector. Figure 6.10: Own-source Revenue, Composition Figure 6.11: Own-source Revenue, States/Regions Percent, 2013/14 Kyat million, 2013/14 160,000 140,000 Tax Non-tax Capital 120,000 15.4 100,000 80,000 30.4 60,000 40,000 20,000 54.2 0 Mon Rakhine Chin Yangon Mandalay Kayah Bago Kayin Ayeyawaddy Shan Magway Sagaing Taninthayi Kachin Tax Non-tax Capital Source: Ministry of Finance. Figure 6.12: SEE Non-tax Contributions (Share of State/Region Own-Revenue) 14% 12% 10% 8% 6% 4% 2% 0% Chin Rakhine Kayin Shan Magway Bago Mandalay Yangon Taninthayi Kachin Ayeyawaddy Sagaing Source: Ministry of Finance. 6.55. The burden of State/Region tax revenue is moderately progressive when comparing recurrent revenue collection per-capita to poverty incidence (Figure 6.13).82 The poorest States/Regions show the lowest tax payments per-capita. For example, Chin shows the lowest tax receipts per-capita and Yangon 82 Capital receipts are omitted because they tend to originate from isolated incidents such as the sale of property rather than a steady stream of revenues paid by a large share of the population within a state or region. Moreover, capital receipts were collected almost exclusively in Yangon and Mandalay. 93 shows the highest tax receipts per-capita. As a next step, it will be useful to extend this analysis to examine the progressivity of each specific tax within each state and region. For example, property taxes are likely to affect a different category of taxpayers than those who pay transportation taxes. Figure 6.13: Poverty and Own Current Revenues Per-capita 16,000 14,000 Yangon 12,000 Revenues per-capita 10,000 8,000 Kayah 6,000 Chin 4,000 2,000 0 0% 10% 20% 30% 40% 50% 60% 70% 80% Poverty incidence P(0) Sources: Integrated Household Living Conditions Survey, Ministry of Finance and World Bank staff calculations. 6.56. Own-source revenue autonomy for States/Regions in terms of policy and administration is very limited and warrants early review to see whether there may be options to grant some independence. Some degree of revenue autonomy is important for accountability. It helps to better link local citizens’ tax payments to local service delivery compared to financing service out of central transfers. Evidence from other countries also suggests that funding local services out of local taxes can promote more responsible spending. This however is long-term endeavor for Myanmar. Developing countries seldom have large taxing powers devolved to sub-national governments, especially at the onset of decentralization. This can dilute central authorities’ capacity to conduct its macro-stabilization and fiscal redistribution functions. Moreover, subnational governments may be reluctant to take on the responsibility of making politically unpopular decisions to raise their own taxes. Nonetheless, some preliminary steps could be considered. 6.57. The authorities may find it helpful to develop a policy on State/Region revenues. This policy could be based on provisions in the 2008 Constitution and an analysis of possible options for providing some revenue autonomy. It would be useful to align the design of this policy with the policy on functional and expenditure assignments. Misalignment between spending and revenue assignments can lead to unfunded mandates (i.e. insufficient local revenue to finance local responsibilities), or wasteful spending (i.e. local revenue in excess of local responsibilities). In terms of revenue autonomy, the policy paper could: (i) Clarify the degree of autonomy on existing revenue assignments set out in the 2008 Constitution. This includes for example the State/Region authority over setting rates on the tax and non-tax revenue sources set out in Schedule 5. (ii) Propose additional sources of revenue, which are typically assigned to sub-national authorities but currently not covered in Schedule 5 (e.g. environmental and property taxes). An important consideration is whether introduction of such new taxes is possible through legislation given that they are not included in the 2008 Constitution. 94 (iii) Propose sharing arrangements for major sources of revenue that are assigned to central government. This would involve decisions on what revenue pool to share (e.g. commercial tax, corporate income taxes) and how to allocate that pool across States/Regions. A common practice in other countries is to share taxes at the same rates across States/Regions. (iv) Explore the option of States/Regions being able to impose a surtax on existing Union level taxes. This maintains a harmonized system while allowing sub-national authorities some independence on the tax rate. 6.58. The tax administration function can remain centralized, even with enhanced revenue autonomy. In fact there is scope to further consolidation at the Union level because responsibility for tax administration is quite fragmented. Several Union agencies (e.g. Internal, General Administration, Roads, Transport, Fisheries) and their deconcentrated units in States/Regions are involved in revenue collection and administration. Given this, and existing weaknesses in tax administration at Union level, it would not make sense to further split this function now. Additionally, separation of tax administration between Union and State/Region level potentially increases administrative costs – particularly when sub-national revenue assignments are so limited. H. Accountability for Results 6.59. Good sub-national governance requires both autonomy and accountability. Giving sub-national governments the freedom to respond to the needs of their populations within their own unique social, economic, and environmental conditions can be very helpful. The probability that they will succeed will increase if they are also held to account for their performance. Sub-national accountability depends critically on the ability of both the central government and the population to monitor sub-national government performance. Sub-national autonomy is therefore best framed in a governance structure which includes central regulation and oversight, as well as both upwards and downwards accountability mechanisms. 6.60. The current framework of oversight has focused primarily on central controls rather than monitoring how well public services are provided by states and regions and the municipalities within them. Reporting and recording systems in Myanmar have not focused on whether government programs are performing well and getting good results. Instead, they have been concerned with compliance with procedures. There have been some positive exceptions. For example, the establishment of school grants committees and the requirement that school budgets are visible at each school (see chapter 4) have made it possible for communities to become more involved in monitoring the performance of their public schools. 6.61. Myanmar’s framework for performance monitoring and local accountability is still evolving. Given that the process of decentralization has only recently begun, authorities are still considering which institutional arrangements and mechanisms will best support prudent financial management and accountable and responsive local service delivery. 6.62. Accountability for results in sub-national public service delivery easier when sub-national fiscal data and service delivery outcomes are reported publically. Systems for collecting and reporting sub- national fiscal data are being developed and modernized; this data could be compared against service delivery performance and be made publically available in a simple and appropriate format. The GAD is well positioned to undertake this role in coordination with sub-national governments and Union ministries. 95 6.63. Simplicity is important, especially in the short-run. As a counter-example of what not to do, the state of Minas Gerais in Brazil attempts to keep track of 85,000 performance indicators. This far exceeded the capacity of the Brazilian government to monitor and analyze. A medium-term step could be the introduction of national minimum service delivery standards based on the data previously collected from sub-national governments. I. Conclusions and Recommendations 6.64. The government has taken a sound approach to very gradual decentralization of administrative and fiscal responsibilities to States/Regions as per the 2008 Constitution. However, the transition necessarily involves tensions between new institutions of sub-national governance and centralized structures of public administration. This can be managed by developing and adopting new policies that help spell out how to implement the provisions on decentralization in the 2008 Constitution. First among these is the establishment of clear lines of accountability for centrally administered public servants to the newly appointed Chief Ministers in the States/Regions. This can be followed by new policies on functional and expenditure assignments, revenue assignments (including more autonomy for States/Regions), and a rules-based systems for intergovernmental transfers. 6.65. The above transition is a very broad and complex agenda of institutional reforms involving many stakeholders. Those involved include the Office of the President, Ministry of Finance, Ministry of National Planning and Economic Development, Ministry of Home Affairs, Chief Minister of States/Regions, and several line agencies. To oversee and coordinate the decentralization process envisaged above, the authorities have established an inter-ministerial Committee on Union-State/Region Relations (CUSRR) chaired by the President. Similar committees have been established in other countries to coordinate across stakeholders; set common objectives; develop and consider policy options; and take decisions as appropriate. The objectives of the CUSSR would be to promote a common vision and shared set of objectives; establish a sound policy and institutional framework; and regularly monitor and evaluate implementation. 6.66. The chapter closed with some important observations about accountability for good results. Existing systems of accountability have, thus far, been more focused on procedural matters. In fact, accountability for results in sub-national public service delivery will be more effective when the authorities publicly report sub-national financial data and service delivery outcomes. The PER therefore proposes options for gather and sharing fiscal data and service delivery outcomes. 96 Issues Options Confusion over the management of, Medium-term: To improve conflicts over accountability, the and accountability for, public authorities could either (i) reassign all GAD staff from the Office of services could be reduced by the states and regions to the direct supervision and control of the clarifying lines of authority between state or region executive (with checks and balances held by the Union ministries and new sub- state or region parliament); or (ii) transition GAD officers to a national institutions. supporting role and allow states and regions to hire and manage their own administrations, building on the existing cadre of Municipal Officers. To improve coordination between sub-national administration and deconcentrated central line agencies, the authorities could give sub-national administrators input into the operations and performance of Union ministries acting at the sub-national level. Union ministries and States/Regions Medium-term: The authorities could develop and adopt a policy on would benefit from more clarity on functional and expenditure assignments that outlines exclusive expenditure assignments, including responsibilities; shared responsibilities; and responsibility for the separation of individual policies, standards, oversight, provision and administration, and activities within broader functions, production and distribution. and the alignment of State/Region cabinet portfolios with the structure of Union ministries. State/region responsibilities will be Short-term: The authorities could develop and adopt a policy on easier to finance if their revenue State/Region revenues that sets out options for revenue autonomy base and autonomy can be including: (i) clarity on autonomy of existing assignments; (ii) expanded. additional revenues typically assigned to sub-nationals; and (iii) sharing of major revenues assigned to central government through the imposition of a surtax on existing Union taxes. It will be easier for States/Regions Short-term: The MoF could consider two options for formula-based to plan and execute their grants: (i) estimate future revenue and spending needs based on expenditure programs if transfers clear and consistent criteria (using economic development, from the Union can be made more equalization, and efficiency indicators) to determine fiscal gap and predictable. level of transfer; or (ii) apply consistent criteria to a predetermined pool of resources. Accountability for results in sub- Short-term: National and sub-national authorities can (i) continue national public service delivery will developing and modernizing systems for collecting and reporting be more effective when the sub-national fiscal data; (ii) compare the fiscal data against service authorities publicly report sub- delivery performance; and (iii) make these data publically available national financial data and service in a simple and appropriate format. The GAD is well poised to play delivery outcomes. this role in coordination with line agencies. Medium-term: Union ministries in collaboration with sub-national authorities could establish national minimum service delivery standards based on the data previously collected from sub-national governments. 97 REFERENCES Chapter 1 Department of Population. 2014. Population and Housing Census of Myanmar, 2014, Provisional Results, Census Report Volume 1, Nay Pyi Taw: Ministry of Immigration and Population. Hori, M., and Ching Wong Y. 2008. “Efficiency Costs of Myanmar's Multiple Exchange Rate Regime.” IMF Working Papers, 1-30. Integrated Household Living Conditions Assessment. 2007. Integrated Household Living Conditions Survey in Myanmar: Poverty Profile. Yangon: United Nations Development Programme. Integrated Household Living Conditions Assessment. 2011. Integrated Household Living Conditions Survey in Myanmar: Poverty Profile (2009-2010). Yangon: United Nations Development Programme. International Monetary Fund. 2001. Government Finance Statistics Manual. Washington D.C. International Monetary Fund. 2014. “Assessment of Data Adequacy for Surveillance.” IMF Country Report No. 14/307. Kheyfets, I., Mastruzzi, M., Merotto, D., and Sondergaard L. 2011. “A New Data Tool to BOOST Public Spending Efficiency.” Open Knowledge Repository, The World Bank. Chapter 2 Baffes, J., and Cosic D. 2015. Global Economic Prospects: Commodity Markets Outlook, April. Washington D.C.: The World Bank. International Monetary Fund. 2013. Article IV Consultation and First Review under the Staff-Monitored Program, IMF Country Report No. 13/250. International Monetary Fund. 2014. Second Review under the Staff-Monitored Program, IMF Country Report No. 14/91. International Monetary Fund. 2014. Article IV Consultation—Staff Report, IMF Country Report No. 14/307. International Monetary Fund. 2014. Debt Sustainability Analysis, IMF Country Report No. 14/307. International Monetary Fund and World Bank. 2009. Developing a Medium-Term Debt Management Strategy (MTDS)—Guidance Note for Country Authorities. Norwegian Agency for Development Cooperation. 2015. Oil and Gas Revenue Management: Selected Issues, Roundtable on Institutional Development and Modernization of State Economic Enterprises in the Oil and Gas Sector. Yangon: Oil for Development Program. 98 Republic of the Union of Myanmar. 1989. State Economic Enterprise Law. Republic of the Union of Myanmar. 2014. Union Budget Law 2014/15. Republic of the Union of Myanmar. 2015. Union Budget Law 2015/16. World Bank. 1995. Myanmar: Policies for Sustaining Economic Reform, Report No. 14062-BA. World Bank. World Development Indicators. Chapter 3 Central Statistical Organization. 2010. Statistical Profile of Children and Women in Myanmar, 2009. Nay Pyi Taw: Ministry of National Planning and Economic Development. Department of Economic and Social Affairs. 2012. World Population Prospects: The 2012 Revision. New York: United Nations. Eckardt, S., and Mills Z. 2014. “What Goes Up Must Come Down--Cyclicality in Public Wage Bill Spending.” World Bank Policy Research Working Paper 6760. Internal Revenue Department. 2013. Strategic Reform Plan for the Internal Revenue Department – 2013 to 2018. Nay Pyi Taw: Ministry of Finance. Mills, Z. 2011. “Wage Bill and Pay Compression Summary Note.” Internal memo, PREM Public Sector and Governance, Washington, D.C.: World Bank. Republic of the Union of Myanmar. 1990. Commercial Tax Law. Republic of the Union of Myanmar. 1976. Goods and Services Tax Law. Republic of the Union of Myanmar. 1974. Income Tax Law. Republic of the Union of Myanmar. 1976. Profit Tax Law. Republic of the Union of Myanmar. 1992. Tariff Law. World Bank. 2014. Logistics Performance Index. World Bank. World Development Indicators. Chapter 4 Central Statistical Organization. 2011. Statistical Yearbook, 2009. Nay Pyi Taw: Ministry of National Planning and Economic Development. Department of Planning and Training. 2010. Statistical Yearbook. Nay Pyi Taw: Ministry of Education. 99 Department of Planning and Training. 2012. “Access to and Quality of Education: Education for All in Myanmar.” Conference on Development Policy Options with Special Reference to Education and Health in Myanmar. Nay Pyi Taw: Ministry of Education. General Statistics Office. 2006. Vietnam: Multiple Indicator Cluster Survey. United Nations Children’s Fund. http://www.childinfo.org/files/MICS3_Myanmar_FinalReport.pdf Gove, A. and Wetterberg A. 2011. The Early Grade Reading Assessment: Applications and Interventions to Improve Basic Literacy. Washington D.C.: RTI International. Hygiene and Prevention Department of Ministry of Health. 2008. LAO PDR: Monitoring the Situation of Children and Women: Multiple Indicator Cluster Survey 2006. Vientiane: Ministry of Planning and Investment. Integrated Household Living Conditions Assessment. 2011. Integrated Household Living Conditions Survey in Myanmar: Poverty Profile (2009-2010). Yangon: United Nations Development Programme. Ministry of National Planning and Economic Development. 2011. Myanmar: Monitoring the Situation of Children and women: Multiple Indicator Cluster Survey Myanmar 2009-2010. Nay Pyi Taw: Ministry of Health. http://www.childinfo.org/files/MICS3_Myanmar_FinalReport.pdf Psacharopoulos, G., & Patrinos, H. A. 2004. “Returns to Investment in Education: a Further Update.” Education economics, 12(2), 111-134. Republic of the Union of Myanmar. 1973. Basic Education Law. Republic of the Union of Myanmar. 1973. University Education Law. Republic of the Union of Myanmar. 1974. Agricultural, Technical and Vocational Education Law. Republic of the Union of Myanmar. 2014. National Education Law. United Nations International Children's Emergency Fund. 2011. Annual Report 2010. UNICEF Executive Director’s Annual Report to the Executive Board. United Nations International Children's Emergency Fund. 2012. Situation Analysis of Early Childhood Care and Development in Myanmar. Nay Pyi Taw: Ministry of National Planning and Economic Development and United Nations Children’s Fund. Watson, J., Khine, E.E., and Myint T. 2012. “Textbook Printing and Distribution for Basic Education in Myanmar.” Phase 1 for the Comprehensive Education Sector Review. Nay Pyi Taw: Ministry of Education. World Bank. 2014. Early Grade Reading Assessment for the Yangon Region. World Bank. Educational Statistics. http://datatopics.worldbank.org/education/ 100 Chapter 5 Asia Pacific Observatory on Health Systems and Policies. 2014. “The Republic of the Union of Myanmar Health System Review.” Health Systems in Transition 4(3). http://www.wpro.who.int/asia_pacific_observatory/hits/series/myanmar_health_systems_review.pdf. Barro, R. J., & Lee, J. W. 2013. “A New Data Set of Educational Attainment in the World, 1950– 2010.” Journal of development economics, 104, 184-198. Building Research Board. 1990. Committing to the Cost of Ownership—Maintenance and Repair of Public Buildings, Committee on Advanced Maintenance Concepts for Buildings, National Research Council, National Academy Press, Washington, D.C. Central Statistical Organization, 2013. Health in Myanmar. Nay Pyi Taw: Ministry of National Planning and Economic Development. http://www.moh.gov.mm/ Dabla-Norris, E., Brumby, J. Kyobe, A., Mills, Z. and Papageorgiu. C. 2011. “Investing in Public Investment: An Index of Public Investment Efficiency.” IMF Working Paper WP/11/37. Washington DC: International Monetary Fund. https://www.imf.org/external/pubs/ft/wp/2011/wp1137.pdf. Department of Economic and Social Affairs. 2011. World Population Prospects: The 2010 Revision. New York: United Nations. Department of Health Planning, 2001. Hospital Upgrading Plan (2001/02-05/06), Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2001. Rural Health Development Plan (2001-06), Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2006. National Health Plan (2006-11) , Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2010. Five Year Strategic Plan for Child Health Development (2010–2014), Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2011. National Health Accounts, Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2011. National Health Plan (2011-16) , Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2011. National Plan of Action on Food and Nutrition (2011–2016) , Nay Pyi Taw: Ministry of Health. Department of Health Planning, 2012. Health Workforce Strategic Plan (2012-17) , Nay Pyi Taw: Ministry of Health. Global Health Data Exchange, 2010. Global Burden of Disease Study 2010 - Myanmar Results by Cause 1990-2010. http://ghdx.healthmetricsandevaluation.org 101 Heller, P. 1991. “Operations and Maintenance.” In Ke-young Chu and Richard Hemming (Eds.). Public Expenditures Handbook: A Guide to Public Expenditure Policy in Developing Countries. Washington DC: International Monetary Fund. 52-59. file:///C:/Users/wb192048/Downloads/1991_public_expenditure_handbook_chu_fmi%20(1).pdf. International Monetary Fund. 2013. Article IV Consultation and First Review under the Staff-Monitored Program, IMF Country Report No. 13/250. Integrated Household Living Conditions Assessment. 2011. Integrated Household Living Conditions Survey in Myanmar: Poverty Profile (2009-2010). Yangon: United Nations Development Programme. Ministry of National Planning and Economic Development. 2011. Myanmar: Monitoring the Situation of Children and women: Multiple Indicator Cluster Survey Myanmar 2009-2010. Nay Pyi Taw: Ministry of Health. http://www.childinfo.org/files/MICS3_Myanmar_FinalReport.pdf O'Donnell, O., van Doorslaer, E., Wagstaff, A. and Lindelow M. 2008. Analyzing Health Equity Using Household Survey Data. Washington DC: The World Bank. http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.430.3297&rep=rep1&type=pdf Republic of the Union of Myanmar, 2007. Law Relating to Private Health Care Services. World Bank. 1985. Burma: Policies and Prospects for Economic Adjustment and Growth, Report No. 4814-BA. Washington D.C. World Bank. 1995. Myanmar: Policies for Sustaining Economic Reform, Report No. 14062-BA. Washington D.C. World Bank. World Development Indicators. World Health Organization. 2013. World Health Statistics. www.who.int Chapter 6 Asia Foundation. 2013. State and Region Governments in Myanmar. Washington, D.C. Integrated Household Living Conditions Assessment. 2011. Integrated Household Living Conditions Survey in Myanmar: Poverty Profile (2009-2010). Yangon: United Nations Development Programme. New Light of Myanmar Newspaper. 2013. Speech by President Thein Sein at Meeting with Union Ministers, Region/State Chief Ministers, and Deputy Ministers. August 9. Nay Pyi Taw. Raghunandan, T.R. 2013. Expenditure Assignments – Theory and Practice. Unpublished, World Bank Workshop. Nay Pyi Taw: World Bank. Republic of Union of Myanmar. 2008. Constitution of the Republic of Union of Myanmar. Nay Pyi Taw: Ministry of Information. 102 Republic of Union of Myanmar. 2012. Framework for Economic and Social Reforms. Nay Pyi Taw. Republic of Union of Myanmar. Various Years. Gazettes of the Union of Myanmar. Nay Pyi Taw: Ministry of Information. World Bank. 2013. Public Financial Management Performance Report, Report No. 75897-MM. Washington D.C. World Bank, Fiscal Decentralization Indicators. http://www1.worldbank.org/publicsector/decentralization/fiscalindicators.htm 103 ANNEX 1: DATA AND ACCOUNTING ISSUES The time period under review includes five years. It includes budget estimates (BE) for 2009/10 through 2013/14 and provisional actual (PA) outcomes for 2009/10 through 2011/12. The provisional actual outcomes for 2012/13 were unavailable at the time of the review, so intermediate results, the revised estimates (RE), were used for that year. Neither the revised estimates nor provisional actuals were available from the government authorities for 2013/14 at the time the review was underway. Estimated outcomes for 2012/13 and 2013/14 from the IMF are used, however, in Chapter 2. The data coverage is almost comprehensive. It includes the Union state administrative organizations, the Union ministries, Union transfers to the states and regions, the Union state economic enterprises, the cantons and the district development committees. The data do not include the off-budget “other accounts” used by Union ministries for their retained revenues collected by states and regions, revenues collected by the states and regions, or economic enterprises owned or controlled by state or regional governments. The data do not include the two military holding companies. The government maintains a cash-based accounting system. This means that expenditures measure what has actually been paid for, regardless if some revenues due have not been collected (tax arrears) or if some payments due to vendors or creditors have not been fully paid (payments in arrears). Non-cash concepts are not recorded. An example found in the IMF GFS Manual (2001) is the consumption of fixed capital, meaning the decline in the value of the stock of fixed assets during the accounting period as a result of physical deterioration, normal obsolescence, and normal accidental damage. The data do not include final estimates provided by the Office of the Auditor-General. These are not published and are not shared outside of the government. The budget estimates are the original budgetary allocations approved just before the start of each new fiscal year. The revised estimates are modified allocations based on various ad-hoc policy changes and unanticipated needs. These are approved at the end of each fiscal year. The preliminary actual estimates are generated by the Accounting Department approximately 12 months after the close of each fiscal year. Estimates for GDP, population and poverty may not be accurate. Historical data show high real growth rates that were inconsistent with economic fundamentals and available data from key sectors. Public sector activities were recorded at deeply over-valued exchange rates. Growth rates from recent years appear to be more realistic and consistent with other indicators but there remains considerable uncertainty about the actual level of GDP. In particular, if the historical growth rates were incorrect, then all subsequent estimates based on those growth rates will also be incorrect. Population and poverty estimates are also subject to scrutiny. Both are expected to benefit from the recent census. The new census data were considered preliminary at the time of this review and have not been utilized. The IMF Assessment of Data Adequacy for Surveillance observes several additional short-comings but also notes good progress in improving the quality of new estimates. To minimize the probability of errors, this review uses estimates published in various IMF country reports. The review uses a mix of data sources. Chapter 2 utilizes aggregated data in various IMF reports published through 2014, augmented by World Bank staff calculations. The analysis in Chapters 3, 4 and 5 examines Union and SEE operations individually using the BOOST database. The BOOST database produces results highly similar to, but not the same as, those provided by the IMF. Importantly, for the years 2012/13 and 2013/14, the BOOST database numbers are revised estimates and budget estimates while the IMF data are preliminary outcomes and revised estimates respectively. Annex Tables 1 and 2 provide a comparison. Chapter 6 relies on state and region data provided by the MoF. 104 Annex 1 Table 1: Union Fiscal Accounts, MMK Billions, 2009/10-2013/14 Prelim. Estimated 2009/10 2010/11 2011/12 2012/13 2013/14 GENERAL (UNION) GOVERNMENT a/ 1 Revenues and Grants = rows 2 + 10 2,034 2,529 2,825 4,592 6,337 2 A. Revenues = rows 3 + 7 2,033 2,529 2,824 4,566 6,226 3 1. Tax Receipts 1,077 1,318 1,678 3,375 3,916 4 Income Tax Receipts from SEEs = row 23 295 287 387 748 839 5 Commercial Tax Receipts from SEEs = row 24 211 177 167 563 616 6 Other Tax Receipts 571 854 1,124 2,063 2,461 7 2. Non-tax 956 1,211 1,146 1,192 2,310 8 Contribution Receipts from SEEs = row 22 793 783 988 774 811 9 Paid by Private Sector 163 427 158 418 1,499 10 B. Grants, Recurrent and Capital 0 0 0 26 111 11 Expenditures = rows 12 + 15 3,175 4,379 4,517 6,461 7,833 12 A. Recurrent 1,159 1,548 1,945 3,521 4,767 13 Interest Payments 263 359 474 683 830 14 Other Recurrent 896 1,189 1,471 2,838 3,937 15 B. Capital 2,016 2,831 2,572 2,940 3,066 16 Overall Balance = rows 1 - 11 -1,141 -1,850 -1,692 -1,869 -1,496 STATE ECONOMIC ENTERPRISES b/ 17 Revenues = rows 18 + 19 2,500 2,805 3,364 7,334 8,056 18 A. Contributions = row 22 793 783 988 774 811 19 B. Receipts excluding Contributions 1,706 2,022 2,376 6,560 7,245 20 Expenditures = rows 21 + 25 3,081 3,126 3,679 6,268 7,462 21 A. Recurrent = rows 22 + 23 + 24 + 25 2,669 2,754 3,301 5,315 5,998 22 1. SEE Contribution Payments to Union 793 783 988 774 811 23 2. Income Tax Payments to Union d/ 295 287 387 748 839 24 3. Commercial Tax Payment to Union d/ 211 177 167 563 616 25 4. Other Recurrent c/ 1,370 1,507 1,760 3,230 3,732 26 B. Capital 412 372 377 953 1,464 27 Overall Balance = rows 17 - 20 -581 -321 -315 1,066 594 CONSOLIDATED PUBLIC SECTOR b/ 28 Revenues and Grants = rows 29 + 38 3,740 4,551 5,201 11,152 13,582 29 A. Revenues = 30 + 34 or rows 6 + (17 - 18) 3,740 4,551 5,200 11,126 13,471 30 1. Tax Receipts d/ 1,077 1,318 1,678 3,375 3,916 31 Income Tax Receipts from SEEs = row 23 295 287 387 748 839 32 Commercial Tax Receipts from SEEs = row 24 211 177 167 563 616 33 Other Tax Receipts = row 6 571 854 1,124 2,063 2,461 34 2. Non-tax Receipts 2,663 3,232 3,522 7,752 9,555 35 SEE Receipts excluding Contributions = row 19 1,706 2,022 2,376 6,560 7,245 36 Contribution Receipts from SEEs = row 22 793 783 988 774 811 37 Paid by Private Sector = row 9 163 427 158 418 1,499 38 B. Grants, Recurrent and Capital = row 10 0 0 0 26 111 39 Expenditures = rows 40 + 43 5,463 6,722 7,208 11,955 14,485 40 A. Recurrent = rows 41 + 42 3,035 3,519 4,259 8,062 9,955 41 1. Interest Payments = row 13 c/ 263 359 474 683 830 42 2. Primary Recurrent = rows 14 + (21 - 22) 2,771 3,160 3,785 7,379 9,125 43 B. Capital = rows 15 + 26 2,428 3,203 2,949 3,893 4,530 44 Overall Balance = 27 – 32 -1,723 -2,171 -2,007 -803 -903 45 Discrepancy 841 46 Net Financing = rows 47 + 48 1,723 2,171 2,007 1,644 903 47 A. Foreign financing 4 3 6 583 555 48 B. Domestic financing = rows 49 + 50 + 51 1,719 2,167 2,002 1,061 348 49 1. Central bank credit 1,189 1,407 790 645 411 50 2. Bank financing 467 437 643 429 -147 51 3. Other 63 323 568 -13 84 Comparison data from BOOST database: e/ Union Revenues and Grants 2,093 2,566 2,895 3,316 4,050 Union Expenditures 3,181 4,402 4,505 7,571 7,827 Consolidated Public Sector Revenues and Grants 3,842 4,760 5,446 10,009 11,370 Consolidated Public Sector Expenditures 5,468 6,669 7,164 12,348 14,500 a. Includes the central government and its support for the states, regions, cantons, Nay Pyi Taw Development Committee and Nay Pyi Taw Council. b. Includes the Central Bank of Myanmar, the Myanmar Economic Bank, the Myanmar Foreign Trade Bank and three other banking institutions. c. IMF does not record interest payments made by the SEEs. d. Tax payments by SEEs should be, but are not, consolidated. e. BOOST data for 2009/10-2011/12 are preliminary actuals, 2012/13 are revised estimates and 2013/14 are budget estimates. Sources: IMF Country Reports No. 13/250, 14/91 and 14/307 and World Bank staff calculations. 105 Annex 1 Table 2: Union Fiscal Accounts from BOOST Database, MMK Billions, 2009/10-2013/14 Revised Budget Provisional Actuals Estimate Estimate 2009/10 2010/11 2011/12 2012/13 2013/14 GENERAL (UNION) GOVERNMENT a/ 1 Revenues and Grants = rows 2 + 10 2,093 2,566 2,895 3,316 4,050 2 A. Revenues = rows 3 + 7 2,092 2,565 2,893 3,177 3,882 3 1. Tax Receipts 1,077 1,282 1,683 1,898 2,609 4 Income Tax Receipts from SEEs = row 23 289 278 388 748 839 5 Commercial Tax Receipts from SEEs = row 24 211 177 167 563 616 6 Other Tax Receipts 577 827 1,129 587 1,154 7 2. Non-tax 1,016 1,284 1,210 1,278 1,273 8 Contribution Receipts from SEEs = row 22 793 783 988 912 949 9 Paid by Private Sector 222 500 222 366 324 10 B. Grants, Recurrent and Capital 1 1 2 139 168 11 Expenditures = rows 12 + 15 3,181 4,402 4,505 7,571 7,827 12 A. Recurrent 1,162 1,552 1,950 3,938 4,832 13 Interest Payments 263 358 473 689 728 14 Other Recurrent 899 1,193 1,476 3,249 4,104 15 B. Capital 2,019 2,850 2,555 3,633 2,995 16 Overall Balance = rows 1 - 11 -1,088 -1,836 -1,610 -4,256 -3,777 STATE ECONOMIC ENTERPRISES d/ 17 Revenues and Grants = 16+17 2,542 2,977 3,539 7,605 8,268 18 A. Contributions = row 22 794 859 1,043 925 949 19 B. Receipts excluding Contributions 1,748 2,118 2,496 6,680 7,319 20 Expenditures = rows 21 + 25 3,081 3,127 3,702 5,701 7,622 21 A. Recurrent = rows 22 + 23 + 24 + 25 2,669 2,755 3,325 4,206 6,112 22 1. SEE Contribution Payments to Union 794 859 1,043 925 949 23 2. Income Tax Payments to Union d/ 289 278 388 748 839 24 3. Commercial Tax Payment to Union d/ 211 177 167 563 616 25 4. Interest c/ 137 182 205 198 209 26 5. Other Recurrent 1,238 1,259 1,522 1,772 3,498 27 B. Capital 412 372 377 1,495 1,511 28 Overall Balance = rows 17 - 20 -539 -149 -163 1,904 646 CONSOLIDATED PUBLIC SECTOR b/ 29 Revenues and Grants = rows 29 + 38 3,842 4,760 5,446 10,009 11,370 30 A. Revenues = 30 + 34 or rows 6 + (17 - 18) 3,841 4,759 5,445 9,870 11,202 31 1. Tax Receipts d/ 1,077 1,282 1,683 1,898 2,609 32 Income Tax Receipts from SEEs = row 23 289 278 388 748 839 33 Commercial Tax Receipts from SEEs = row 24 211 177 167 563 616 34 Other Tax Receipts = row 6 577 827 1,129 587 1,154 35 2. Non-tax Receipts 2,764 3,477 3,761 7,971 8,592 36 SEE Receipts excluding Contributions = row 19 1,748 2,118 2,496 6,680 7,319 37 Contribution Receipts from SEEs = row 22 794 859 1,043 925 949 38 Paid by Private Sector = row 9 222 500 222 366 324 39 B. Grants, Recurrent and Capital = row 10 1 1 2 139 168 40 Expenditures = rows 40 + 43 5,468 6,669 7,164 12,348 14,500 41 A. Recurrent = rows 41 + 42 3,037 3,447 4,232 7,219 9,995 42 1. Interest Payments = row 13 + 25 c/ 400 541 679 887 938 43 2. Primary Recurrent = rows 14 + (21 - 22 - 25) 2,637 2,906 3,553 6,332 9,057 44 B. Capital = rows 15 + 27 2,431 3,222 2,933 5,128 4,506 45 Overall Balance = rows 29 – 40 -1,626 -1,909 -1,718 -2,339 -3,130 46 Net Financing = rows 47 - 48 + 49 1,626 1,909 1,718 2,339 3,130 47 A. Financial Receipts 13 32 14 863 779 48 B. Financial Payments 5 1 4 266 699 49 C. Net Other 1,618 1,878 1,707 1,742 3,051 Comparison Data from IMF Estimates: Union Revenues and Grants 2,034 2,529 2,825 4,592 6,337 Union Expenditures 3,175 4,379 4,517 6,461 7,833 Consolidated Public Sector Revenues and Grants 3,740 4,551 5,201 11,152 13,582 Consolidated Public Sector Expenditures 5,463 6,722 7,208 11,955 14,485 a. Includes the central government and its support for the states, regions, cantons, Nay Pyi Taw Development Committee and Nay Pyi Taw Council. b. Includes the Central Bank of Myanmar, the Myanmar Economic Bank, the Myanmar Foreign Trade Bank and three other banking institutions. c. IMF does not record interest payments made by the SEEs. d. Tax payments by SEEs should be, but are not, consolidated. Source: MoF BOOST database. 106 ANNEX 2: CROSS-COUNTRY DATA COMPARISONS Comparing Myanmar’s level and composition of expenditures with other countries can be a useful way to begin thinking about whether the level or composition of spending could be further improved. With this in mind, Annex 2 Table 1 was constructed to make such comparisons easy. The data come from the IMF Government Financial Statistics which are standardized across a wide range of countries. This makes it easier to compare one country with another. The sample of countries used in the table below was selected on the basis of three key criteria: each country had to have data for at least two years between 2000 and 2007, the accounting system is cash-based, and the level of government is general government. The specified time period allows one to avoid the impact of the global food and fuel crisis and the subsequent global financial crisis and recession. (Myanmar was largely unaffected by these shocks due to its low level of trade and financial integration.) The resulting sample shown in Annex 2 Table 2 includes 40 countries: 2 low income, 12 lower middle-income, 15 middle-income and 11 high income. Of these, 39 have data for the economic classification of government (ECOG) but only 25 have data for the classification of government functions (COFOG). Average GDP shares and expenditure shares have been calculated for each ECOG and COFOG category within each of the income groups. The results for the low income group are not included in the table since the sample size (2) is too small to be representative. Annex 2 Table 1: Shares of General Government Expenditures a/ Lower Middle-income Upper Middle-income ---- High Income ---- In Percent GDP Spending GDP Spending GDP Spending By Function of Government Total b/ 33.7 100.0 32.5 100.0 39.8 100.0 701 General Public Service 5.8 17.3 7.5 23.1 6.0 15.2 702 Defense 2.3 6.9 1.6 4.9 1.4 3.6 703 Public Order and Safety 2.2 6.5 1.9 5.9 1.8 4.6 704 Economic Affairs 5.7 16.9 5.2 16.1 5.2 13.3 705 Environmental Protection 0.2 0.5 0.4 1.2 0.6 1.4 706 Housing and Community Amenities 1.6 4.6 1.7 5.5 1.4 3.4 707 Health 2.8 8.2 2.8 8.5 5.1 12.7 708 Recreation, Culture and Religion 0.8 2.5 0.7 2.3 1.0 2.4 709 Education 5.1 15.2 4.0 12.5 5.0 12.6 710 Social Protection 7.2 21.4 6.6 20.2 12.3 31.0 Sample Size (6) (6) (10) (10) (8) (8) By Economic Classification of Government Total b/ 31.5 100.0 31.5 100.0 38.7 100.0 21 Compensation of Employees 7.8 24.8 8.3 26.4 9.7 25.1 22 Use of Goods and Services 6.3 20.1 7.1 22.5 5.9 15.4 24 Interest Payments 2.2 6.9 3.2 10.1 2.7 7.1 25 Subsidies and Transfers 2.5 8.0 1.3 4.1 2.5 6.6 26 Grants 0.2 0.6 0.0 0.1 0.2 0.6 27 Social Benefits 4.7 15.1 5.7 18.1 11.5 29.6 28 Other Recurrent 1.8 5.8 1.9 6.1 2.7 7.1 31 Net Acquisition of Non-financial Assets 5.9 18.6 4.0 12.6 3.3 8.5 Sample Size (12) (12) (14) (14) (11) (11) Memorandum Item: Ratio of Investment to Employee Comp. 0.8 0.8 0.5 0.5 0.3 0.3 a. Averages are simple averages. The use of weighted averages does not significantly change the results. b. Totals from the COFOG and ECOG data will not match due to the different compositions of the associated data samples. Totals within columns may not equal the sum of components due to rounding. Sources: IMF Government Financial Statistics and World Bank staff estimates. 107 Annex 2 Table 2: List of Countries in Sample Economic Classification of Government Classification of Functions of Government Low Income (2) (2) (1) Afghanistan X X Tajikistan X Lower Middle-income (12) (12) (6) Armenia X Bhutan X X Congo, Rep. X Egypt, Arab Rep. X X Georgia X X India X Lesotho X Moldova X X Mongolia X Paraguay X Ukraine X X Yemen, Rep. X X Upper Middle-income (15) (14) (10) Albania X X Belarus X X Brazil X Bulgaria X X China X Costa Rica X Iran, Islamic Rep. X X Jamaica X X Kazakhstan X X Macedonia, FYR X Maldives X X Mauritius X X Peru X Seychelles X X Thailand X High Income (11) (11) (8) Barbados X Canada X X Croatia X Czech Republic X X Latvia X X Malta X X Russian Federation X X Slovak Republic X X Slovenia X X St. Kitts and Nevis X Switzerland X X The comparison of Myanmar with the three income group averages is only a first step in what should be a more systematic process. The results should be interpreted with care. Income per-capita is not the only driver of expenditure levels or composition. Recent research shows that the old and often cited relationship between GDP per-capita and the size of government (Wagner’s Law) no longer exists in many countries. Brady and Lee (2013), for example, found that government spending in many OECD countries has been falling as a share of GDP since the 1990s despite continued growth in GDP per-capita. Shelton 108 (2007) finds that GDP per-capita becomes insignificant as an explanatory variable when demographic variables are introduced: more aged (healthier) societies are wealthier and spend more on social security. The GDP shares of most other categories of spending tend to fall as GDP per-capita grows. Shelton finds several other explanatory variables are relevant as well. For example, more populous and/or ethnically diverse countries tend to be more decentralized. Policy and external shocks matter too. For example, defense expenditure shares have been falling almost everywhere in the world since the collapse the former Soviet Union and the end of the “cold war.” Governments that favor universal health care will spend more than those that rely heavily on household “out-of-pocket” contributions. Governments facing falling commodity prices for its producers may choose to boost spending temporarily on social safety net programs. 109 ANNEX 3: ROLES AND RESPONSIBILITIES OF VARIOUS ORDERS OF GOVERNMENT This table broadly draws on the 2008 constitution to broadly define categories of public sector activities and examine how various responsibilities within these broad categories are assigned across different tiers of government. For some activities, responsibilities may be largely exclusive to a single level of government (such as Defence and Security, which are exclusively undertaken by the Union level); whereas for others, responsibilities may be shared (such as in the Finance and Planning sectors). An N.A. implies no role for a level of government in a specific area of public sector activity. Union The region or state Self-Administered Division/Zone leading bodies a) Defence of the Republic of the Union N.A. N.A. of Myanmar and every part thereof and preparation for such defence; b) Defence and Security industries; Defense and Security Sector c) Arms, ammunition and explosives including biological and chemical weapons; d) Atomic energy, nuclear fuel and radiation and mineral resources essential to its production; e) Declaration of war and conclusion of peace; f) Stability, peace and tranquility of the Union and prevalence of law and order; and g) Police force. a) Representatives of the diplomatic, N.A. N.A. consular and other affairs; b) United Nations; c) Participation in international, regional and bilateral conferences, seminars, meetings, associations and other organizations and implementation of Foreign Affairs Sector resolutions thereof; d) Conclusion and implementation of international and regional treaties, agreements, conventions and bilateral agreements and treaties; e) Passports and identification certificates; f) Visas, admission into the Republic of the Union of Myanmar, stay, departure, immigration and deportation; and g) (g) Extradition and request for extradition. 110 Union The region or state Self-Administered Division/Zone leading bodies a) The Union Budget; a) The region or state budget; N.A. b) The Union Fund; b) The region or state fund; c) Currency and coinage; c) Land revenue; d) The Central Bank of Myanmar and d) Excise duty (not including financial institutions; narcotic drugs and e) Foreign exchange control; psychotropic substances); f) Capital and money markets; e) Municipal taxes such as taxes g) Insurance; on buildings and lands, water, Finance and Planning Sector h) Income tax; street lightings and wheels; i) Commercial tax; f) Services of the region or state; j) Stamp duty; g) Sale, lease and other means of k) Customs duty; execution of property of the l) Union lottery; region or state; m) Tax appeal; h) Disbursement of loans in the n) Services of the Union; country from the region or o) Sale, lease and other means of state funds; execution of property of the Union; i) Investment in the country from p) Disbursement of loans from the Union the region or state funds; Funds; j) Local plan; and q) Investment of the Union Funds; k) Small loans business. r) Domestic and foreign loans; s) Acquisition of property for the Union; and t) Foreign aid and financial assistance. a) Economy; a) Economic matters undertaken N.A. b) Commerce; in the region or state in accord c) Co-operatives; with law enacted by the Union; Economic Sector d) Corporations, boards, enterprises, b) Commercial matters companies and partnerships; undertaken in the region or e) Imports, exports and quality control state in accord with law thereon; enacted by the Union; and f) Hotels and lodging houses; and c) Co-operative matters g) Tourism. undertaken in the region or state in accord with law enacted by the Union. a) Land administration; a) Agriculture; a) Maintenance of pasture b) Reclamation of vacant, fallow and b) Protection against and control b) Conservation and Agriculture and Livestock Breeding Sector virgin lands; of plants and crop pests and preservation of forest c) Settlements and land records; diseases; c) Preservation of natural d) Land survey; c) Systematic use of chemical environment in accord e) Dams, embankments and irrigation fertilizers and systematic with law promulgated by works managed by the Union; production and use of natural the Union f) Meteorology, hydrology and seismic fertilizers; survey; d) Agricultural loans and savings; g) Registration of documents; e) Dams, embankments, lakes, h) Mechanized agriculture; drains and irrigation works i) Agricultural research; having the right to be managed j) Production of chemical fertilizers and by the region or state; insecticides; f) Fresh water fisheries; and k) Marine fisheries; and g) Livestock breeding and l) Livestock proliferation, prevention and systematic herding in accord treatment of diseases and research with the law enacted by the works. Union. 111 Union The region or state Self-Administered Division/Zone leading bodies a) Petroleum, natural gas, other liquids a) Medium and small scale d) Water and electricity and substances declared by the Union electric power production and matters in towns and Energy, Electricity, Mining and Forestry Sector b) Law to be dangerously inflammable; distribution that have the right villages c) Production and distribution of to be managed by the region or electricity of the Union; state not having any link with d) Minerals, mines, safety of mine national power grid, except workers, and environmental large scale electric power conservation and restoration; production and distribution e) Gems; having the right to be managed f) Pearls; by the Union; g) Forests; and b) Salt and salt products; h) Environmental protection and c) Cutting and polishing of conservation including wildlife, natural gemstones within the region or plants and natural areas state; d) Village firewood plantation; and e) Recreation centers, zoological garden and botanical garden. a) Industries to be undertaken by the a) Industries other than those e) Urban and rural projects Union level; prescribed to be undertaken by b) Industrial zones; the Union level; and c) Basic standardization and specification b) Cottage industries. Industrial Sector for manufactured products; d) Science and technology and research thereon; e) Standardization of weights and measures; and f) (f) Intellectual property such as copyrights, patents, trademarks and industrial designs. a) Inland water transport; a) Ports, jetties and pontoons f) Construction and b) Maintenance of waterways; having the right to be managed Maintenance of Roads c) Development of water resources and by the region or state; and Bridges rivers and streams; b) Roads and bridges having the Transport, Communication and Construction Sector d) Carriage by sea; right to be managed by the e) Major ports; region or state; and f) Lighthouses, lightships and lighting c) Systematic running of private plans; vehicles within the region or g) Shipbuilding, repair and maintenance; state. h) Air transport; i) Air navigation, control and airfields construction; j) Land transport; k) Railways; l) Major highways and bridges managed by the Union; m) Posts, telegraphs, telephones, fax, e- mail, internet, intranet and similar n) means of communication; and o) (n) Television, satellite communication, transmission and reception, and similar means of communication and housing and buildings. 112 Union The region or state Self-Administered Division/Zone leading bodies a) Educational curricula, syllabus, a) Matters on traditional g) Public Health teaching methodology, research, medicine not contrary to h) Prevention of Fire plans, projects and standards; traditional medicine policies Hazard b) Universities, degree colleges, institutes prescribed by the Union; and other institutions of higher b) Social welfare works within the education; region or state; c) Examinations prescribed by the Union; c) Preventive and precautionary d) Private schools and training; measures against fire and e) National sports; natural disasters; f) National health; d) Stevedoring; g) Development of traditional medicinal e) Having the right of science and traditional medicine; management by the region or h) Charitable hospitals and clinics and state, the following: private hospitals and clinics; f) preservation of cultural i) Maternal and child welfare; heritage; j) Red cross society; g) museums and libraries. k) Prevention from adulteration, h) Theatres, cinemas and video manufacture and sale of foodstuffs, houses; and drugs, medicines and cosmetics; i) Exhibitions such as Social Sector l) Welfare of children, youths, women, photographs, paintings and the disabled, the aged and the sculptures. homeless; m) Relief and rehabilitation; n) Fire Brigade; o) Working hours, resting-hours, holidays and occupational safety; p) Trade disputes; q) Social security; r) Labour organizations; s) Managements by the Union, the following: (i) Ancient culture or historical sites, buildings, monuments, records, stone inscriptions, ink inscriptions on stucco, palm-leaf parabaiks, handwritings, handiworks, inanimate objects and archaeological works; (ii) Museums and libraries. t) Literature, dramatic arts, music, traditional arts and crafts, cinematographic films and videos; and u) Registration of births and deaths. 113 Union The region or state Self-Administered Division/Zone leading bodies General administration; a) a) Development matters; i) Development affairs Administration of town and village b) b) Town and housing j) Market matters of towns land; development; and and villages c) Tenants; c) Honorary certificates and d) Narcotic drugs and psychotropic awards Management Sector substances; e) Union secrets; f) Associations; g) Prisons; h) Development of border areas; i) Census; j) Citizenship, naturalization, termination and revocation of citizenship, citizenship scrutiny and registration; and k) Titles and honours. a) Judiciary; N.A. N.A. b) Lawyers; c) Criminal Laws and procedures; d) Civil Laws and procedures including contract, arbitration, actionable wrong, insolvency, trust and trustees, administrator and receiver, family laws, guardians and wards, transfer of Judicial Sector property and inheritance; e) Law of Evidence; f) Limitation; g) Suit valuation; h) Specific relief; i) Foreign jurisdiction; j) Admiralty jurisdiction; and k) Piracies, crimes committed in international waters or in outer space and offences against the international law on land or in international waters or in outer space. Source: 2008 Constitution, Schedules One, Two and Three 114 ANNEX 4: BUDGET OUTCOMES OF THE STATES AND REGIONS Total Millions of Kyat 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 691,386 978,566 A. Own Revenue 254,348 352,232 1. Tax 54,601 54,076 2. Non-tax b/ 182,012 191,001 3. Capital 17,736 107,155 B. Union Support, excl. grant for rural poverty alleviation 437,038 626,333 Expenditure 664,362 980,234 A. Interest 19 218 B. Primary Expenditures 664,343 980,016 1. State Administrative Organizations 18,607 136,461 2. Ministry Recurrent 372,469 437,091 3. Ministry Capital 273,267 406,463 Overall Balance 27,024 -1,668 State and Regions SEE Operations a/ Revenue 211,474 301,610 Expenditure c/ 200,665 308,084 Overall Balance 10,809 -6,474 Unallocable Spending from Supplementary Allocation 180,174 140,024 Combined Balance a/ -142,340 -148,166 Own Revenue as % Total Revenue and Grants 37 36 Own Revenue as % Total Expenditure 38 36 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. Continued … 115 Kachin Kayah Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 25,642 40,761 6,259 19,097 A. Own Revenue 6,700 9,056 816 2,221 1. Tax 936 1,175 120 226 2. Non-tax b/ 5,764 7,881 696 1,996 3. Capital 0 0 0 0 B. Union Support, excl. grant for rural poverty alleviation 18,942 31,704 5,443 16,876 Expenditure 24,229 40,631 5,908 19,237 A. Interest 0 0 0 0 B. Primary Expenditures 24,229 40,631 5,908 19,237 1. State Administrative Organizations 956 4,237 501 3,329 2. Ministry Recurrent 14,248 17,236 3,691 7,319 3. Ministry Capital 9,025 19,157 1,716 8,589 Overall Balance 1,413 130 351 -140 State and Regions SEE Operations a/ Revenue 12,617 19,082 2,748 8,941 Expenditure c/ 12,085 18,701 2,571 8,811 Overall Balance 533 381 177 131 Unallocable Spending from Supplementary Allocation 1,172 4,888 940 3,852 Combined Balance a/ 774 -4,377 -412 -3,862 Own Revenue as % Total Revenue and Grants 26 22 13 12 Own Revenue as % Total Expenditure 28 22 14 12 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 116 Kayin Chin Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 14,525 24,589 13,246 33,410 A. Own Revenue 3,320 4,164 4,230 1,834 1. Tax 570 640 120 132 2. Non-tax b/ 2,748 3,523 4,111 1,702 3. Capital 2 0 0 0 B. Union Support, excl. grant for rural poverty alleviation 11,205 20,425 9,016 31,575 Expenditure 13,638 24,481 12,690 35,212 A. Interest 0 0 0 19 B. Primary Expenditures 13,638 24,481 12,690 35,192 1. State Administrative Organizations 538 4,654 741 10,666 2. Ministry Recurrent 11,749 13,119 9,873 10,879 3. Ministry Capital 1,352 6,708 2,076 13,647 Overall Balance 886 108 556 -1,802 State and Regions SEE Operations a/ Revenue 7,104 11,862 4,419 11,731 Expenditure c/ 6,268 11,274 4,749 11,364 Overall Balance 836 588 -330 367 Unallocable Spending from Supplementary Allocation 6,781 5,929 1,752 1,123 Combined Balance a/ -5,059 -5,232 -1,526 -2,558 Own Revenue as % Total Revenue and Grants 23 17 32 5 Own Revenue as % Total Expenditure 24 17 33 5 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 117 Sagaing Taninthayi Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 52,884 80,659 24,902 36,927 A. Own Revenue 9,454 13,141 5,143 5,668 1. Tax 2,474 2,369 1,517 729 2. Non-tax b/ 6,980 10,772 3,626 4,939 3. Capital 0 0 0 0 B. Union Support, excl. grant for rural poverty alleviation 43,429 67,518 19,759 31,258 Expenditure 50,574 81,457 21,806 36,948 A. Interest 0 0 0 0 B. Primary Expenditures 50,574 81,457 21,806 36,948 1. State Administrative Organizations 1,650 12,301 847 8,975 2. Ministry Recurrent 41,838 54,817 9,623 7,472 3. Ministry Capital 7,086 14,339 11,335 20,502 Overall Balance 2,310 -798 3,097 -22 State and Regions SEE Operations a/ Revenue 23,864 31,595 12,246 16,672 Expenditure c/ 21,891 32,416 11,809 16,613 Overall Balance 1,973 -822 437 60 Unallocable Spending from Supplementary Allocation 17,928 9,641 11,186 7,128 Combined Balance a/ -13,645 -11,260 -7,652 -7,090 Own Revenue as % Total Revenue and Grants 18 16 21 15 Own Revenue as % Total Expenditure 19 16 24 15 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 118 Bago Magway Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 30,521 57,155 61,264 107,227 A. Own Revenue 12,685 14,541 9,179 9,721 1. Tax 2,854 2,667 1,181 1,154 2. Non-tax b/ 9,829 11,870 7,995 8,563 3. Capital 3 4 3 5 B. Union Support, excl. grant for rural poverty alleviation 17,835 42,614 52,085 97,506 Expenditure 28,475 57,506 57,798 107,226 A. Interest 0 0 0 0 B. Primary Expenditures 28,475 57,506 57,798 107,226 1. State Administrative Organizations 1,569 8,380 1,339 7,429 2. Ministry Recurrent 22,586 25,554 24,127 44,454 3. Ministry Capital 4,320 23,572 32,332 55,343 Overall Balance 2,046 -351 3,466 1 State and Regions SEE Operations a/ Revenue 6,934 18,400 24,551 25,425 Expenditure c/ 6,629 18,115 23,804 25,922 Overall Balance 305 285 747 -497 Unallocable Spending from Supplementary Allocation 1,130 10,385 2,172 0 Combined Balance a/ 1,221 -10,451 2,041 -496 Own Revenue as % Total Revenue and Grants 42 25 15 9 Own Revenue as % Total Expenditure 45 25 16 9 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 119 Mandalay Mon Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 73,231 106,131 18,731 25,570 A. Own Revenue 51,406 72,630 7,235 8,938 1. Tax 13,290 13,190 1,482 1,661 2. Non-tax b/ 28,290 31,213 5,735 7,084 3. Capital 9,827 28,227 18 192 B. Union Support, excl. grant for rural poverty alleviation 21,824 33,502 11,496 16,632 Expenditure 73,084 103,227 17,618 25,947 A. Interest 0 149 0 0 B. Primary Expenditures 73,084 103,078 17,618 25,947 1. State Administrative Organizations 2,409 12,131 644 4,178 2. Ministry Recurrent 38,388 54,436 11,837 11,190 3. Ministry Capital 32,288 36,510 5,136 10,579 Overall Balance 146 2,904 1,114 -377 State and Regions SEE Operations a/ Revenue 8,403 11,902 7,565 11,169 Expenditure c/ 8,996 11,878 6,856 11,089 Overall Balance -593 23 709 80 Unallocable Spending from Supplementary Allocation 11,667 19,305 5,179 5,268 Combined Balance a/ -12,114 -16,377 -3,356 -5,565 Own Revenue as % Total Revenue and Grants 70 68 39 35 Own Revenue as % Total Expenditure 70 70 41 34 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 120 Rakhine Yangon Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 48,456 75,106 197,091 198,097 A. Own Revenue 3,810 5,475 112,166 166,132 1. Tax 818 715 20,171 17,363 2. Non-tax b/ 2,992 4,760 84,137 70,070 3. Capital 0 0 7,858 78,699 B. Union Support, excl. grant for rural poverty alleviation 44,646 69,631 84,926 31,965 Expenditure 49,131 74,929 191,254 199,601 A. Interest 0 25 19 25 B. Primary Expenditures 49,131 74,904 191,234 199,575 1. State Administrative Organizations 1,195 23,727 2,895 22,710 2. Ministry Recurrent 44,738 43,406 64,453 75,032 3. Ministry Capital 3,198 7,771 123,887 101,833 Overall Balance -676 177 5,838 -1,504 State and Regions SEE Operations a/ Revenue 15,354 33,979 24,121 5,115 Expenditure c/ 14,806 33,591 23,608 5,591 Overall Balance 548 388 514 -475 Unallocable Spending from Supplementary Allocation 379 30,242 112,101 37,379 Combined Balance a/ -506 -29,676 -105,749 -39,358 Own Revenue as % Total Revenue and Grants 8 7 57 84 Own Revenue as % Total Expenditure 8 7 59 83 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 121 Shan Ayeyawaddy Millions of Kyat 2012/13 PA 2013/14 RE 2012/13 PA 2013/14 RE State and Regions Own Operations a/ Revenue and Grants 68,004 105,702 56,629 68,135 A. Own Revenue 12,923 18,265 15,281 20,446 1. Tax 1,988 1,822 7,081 10,233 2. Non-tax b/ 10,909 16,415 8,200 10,212 3. Capital 26 28 0 0 B. Union Support, excl. grant for rural poverty alleviation 55,082 87,437 41,349 47,690 Expenditure 64,408 105,701 53,748 68,133 A. Interest 0 0 0 0 B. Primary Expenditures 64,408 105,701 53,748 68,133 1. State Administrative Organizations 1,912 8,303 1,411 5,441 2. Ministry Recurrent 54,196 51,986 21,122 20,191 3. Ministry Capital 8,299 45,412 31,215 42,501 Overall Balance 3,597 1 2,881 3 State and Regions SEE Operations a/ Revenue 35,475 62,054 26,072 33,682 Expenditure c/ 32,917 69,020 23,678 33,700 Overall Balance 2,558 -6,965 2,395 -18 Unallocable Spending from Supplementary Allocation 5,249 2,516 2,537 2,368 Combined Balance a/ 906 -9,480 2,738 -2,384 Own Revenue as % Total Revenue and Grants 19 17 27 30 Own Revenue as % Total Expenditure 20 17 28 30 a. Omits financial receipts and expenditures. b. Includes receipts collected by SAOs, contributions from the SEEs of the states and regions and poverty alleviation fund transfers. c. Includes payment of taxes and contributions to state and region governments. Sources: Ministry of Finance and World Bank staff calculations. 122 123