MONGOLIA GROWING WITHOUT UNDUE BORROWING Enhancing efficiency of spending and revenue PUBLIC EXPENDITURE REVIEW PUBLIC EXPENDITURE REVIEW MONGOLIA GROWING WITHOUT UNDUE BORROWING Enhancing efficiency of spending and revenue PUBLIC EXPENDITURE REVIEW June 2018 1 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Currency Equivalents (Exchange Rate as of May 30, 2018) Currency Unit = Mongolian Tughrik (MNT) US$ 1=MNT 2,409 ABBREVIATIONS AND ACRONYMS BEPS Base Erosion and Profit Shifting GDP Gross Domestic Product BOM Bank of Mongolia GDSI General Department of Social Insurance BOT Build-Operate-Transfer GFMIS Government Financial Management BPIS Budget Preparation Information System Information System BT Build-Transfer GMP Good manufacturing practice CIT Corporate Income Tax GOM Government of Mongolia CMAP Comprehensive Macroeconomic Adjustment GPA Government Procurement Agency Program GTL General Taxation Law CMP Child Money Program HDF Human Development Fund COLA Cost of Living Adjustment HEI Higher Education Institute CSC Civil Service Council HMP Housing Mortgage Program CSL Civil Service Law HRMIS Human Resource Management Information CSO Civil Society Organization System CSRS Civil Service Reform Strategy IA Investment Agreement DBM Development Bank of Mongolia IAAC Independent Authority Against Corruption DMD Debt Management Division IBL Integrated Budget Law DML Debt Management Law ICT Information and Communication Technology DRC Dispute Resolution Council IFI International Financial Institution DRG Diagnosis-Related Group IMF International Monetary Fund DSA Debt Sustainability Analysis IOT Input Output Table EAP East Asia and Pacific IT Information Technology ECA Europe and Central Asia ITCCT Information Technological Centre for ECE Early Childhood Education Customs and Taxation EFF Extended Fund Facility LM Line Ministry EITC Earned Income Tax Credit LME London Metal Exchange EITI Extractive Industries Transparency Initiative LMIC Lower Middle-Income Country EM Erdenes Mongol LTU Large Taxpayers’ Unit ERP Economic Recovery Program MASAM Mainstreaming Social Accountability in FDI Foreign Direct Investment Mongolia FHC Family Health Care MCH Maternal and Child Health FHF Future Heritage Fund MDA Ministries, Departments, and Agencies FHFL Future Heritage Fund Law MDF Mongolia Development Fund FSF Fiscal Stabilization Fund MECSS Ministry of Education, Culture, Science and Sports FSL Fiscal Stability Law MED Ministry of Economic Development GAAR General Anti-Avoidance Rule MIC Middle Income Countries GAP Government Action Plan MICS Multiple Indicator Cluster Survey GBL General Budget Law MNAO Mongolian National Audit Office GDT General Department of Taxation MOF Ministry of Finance 2 ABBREVIATIONS AND ACRONYMS MOH Ministry of Health RDTC Regional Diagnostic and Treatment Center MOM Ministry of Mining SAAR Specific Anti-Avoidance Rule MP Member of Parliament SAIC State Audit and Inspection Committee MPC Monetary Policy Council SAM Social Accounting Matrix MSWL Ministry of Social Welfare and Labor SDC Swiss Agency for Development and MTEF Medium-Term Expenditure Framework Cooperation MTFF Medium-Term Fiscal Framework SDG Sustainable Development Goals NCD Noncommunicable Disease SDV Sustainable Development Vision NDA National Development Agency SES Socio Economic Survey NDC Notional Defined Contribution SF Stabilization Fund NGO Non-Governmental Organization SHI Social Health Insurance NPV Net Present Value SIF Social Insurance Fund NSO National Statistical Office SIGO Social Insurance General Office O&M Operations and Maintenance SOE State-Owned Enterprise OECD Organization for Economic Cooperation and SPC State Property Committee Development SPPCA State Property Policy Coordination Agency OT Oyu Tolgoi SSC Social Security Contribution PAYGO Pay-as-You-Go SSF Social Security Fund PEFA Public Expenditure and Financial SSIGA State Social Insurance General Authority Accountability STR Student-Teacher Ratio PER Public Expenditure Review SWF Sovereign Wealth Fund PIT Personal Income Tax SWP Social Welfare Pension PFM Public Financial Management TIN Taxpayer Identification Number PID Public Investment Department TIMSS Trends in International Mathematics and PIM Public Investment Management Science Study PIMA Public Investment Management Assessment TSA Treasury Single Account PIP Public Investment Program TT Tavan Tolgoi PIS Pension Insurance Scheme UB Ulaanbaatar PISA Program for International Student UN United Nations Assessment UNDP United Nations Development Programme PIT Personal Income Tax UNICEF United Nations Children’s Fund PM Prime Minister UPN Unique Project Number POS Point of Sale VAT Value Added Tax PPLM Public Procurement Law of Mongolia WB World Bank PPP Public Private Partnership WBG World Bank Group PSMFL Public Sector Management and Finance Law WDI World Development Indicators PSP Price Stabilization Program WGI World Governance Indicators PV Present Value WHO World Health Organization PwC PricewaterhouseCoopers 3 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue TABLE OF CONTENTS Preface 11 Acknowledgements 12 Executive Summary 13 Chapter 1. Context 28 1.1 Structure of the Mongolian Economy 30 1.2 The Critical Role of the Mining Economy 32 1.3 Sound Fiscal Policies and Poor Implementation 35 1.4 Organization of the Report 39 Chapter 2. Domestic Revenue Mobilization 40 2.1 Recent Fiscal Developments - Key Features of the Mongolian Tax System 42 2.2 Challenges of the Current Tax System 44 2.3 Priority Areas of Reform 58 Chapter 3. Aggregate Expenditure - Trends and Cross-Cutting Issues 60 3.1 Expenditure Performance 62 3.2 Public Sector Wage Bill and Employment - Trends and Issues 71 3.3 Structure and Efficiency of Public Capital Spending 85 Chapter 4. Sectoral Spending 104 4.1 Education Sector 107 4.2 Health Sector 119 4.3 Pension Reform 132 Chapter 5. Institutional Framework for Improved Fiscal Management 140 5.1 Current Institutions for Fiscal Management - Performance 142 5.2 Fiscal Management of the Mining Sector - Performance 150 5.3 An International Comparison of Fiscal Management Institutions - Lessons for Mongolia 152 References 159 4 CONTENTS TABLES Table 1.1 Key Reform Areas of the Government’s ERP 36 Table 2.1 Break-up of Mining Revenue, in Major Taxes (MNT billion) 42 Table 2.2 Structure of Government Revenue (% of GDP): 2010-2017 43 Table 2.3 Tax and Revenue Allocation between State and Local Governments 44 Table 2.4 International Comparison of Tax Rates (%) 45 Table 2.5 Tax Revenue versus Tax Expenditure (MNT billion) 47 Table 2.6 Total Tax Expenditure 47 Table 2.7 New EITC Schedule 55 Table 2.8 Adopting a Progressive PIT System - Simulation Results 56 Table 2.9 Adopting a Progressive PIT System - Alternative Scenario 56 Table 3.1 Number of Civil Servants by Region, 2007-2016 75 Table 3.2 Debt Incurred through Extra-Budget Funding of Investment Projects by 2016 (MNT billion) 86 Table 3.3 Mongolia - Capital Expenditures, 2010-2017 87 Table 3.4 Central Government Capital Spending by Sector (MNT billion) 90 Table 3.5 Sector Distribution of DBM’s Non-commercial Lending 91 Table 3.6 An Overview of Legal and Regulatory Hierarchies for PIM in Mongolia 93 Table 3.7 An Overview of Key Institutions in Mongolia’s PIM and Capital Budget Process 95 Table 4.1 Number of Education Facilities by Facility Type, 2005 -2017 110 Table 4.2 Trends in Earnings and Poverty Rates in Mongolia, 2010-2014, by Educational Attainment 115 Table 4.3 Number of Health Facilities by Facility Type, 2005 -2015 121 Table 4.4 Outpatient Visits and Inpatient Admissions by Facility Type, 2015 122 Herders’ Service Supplement: Proposed Supplementary Replacement Rate according to Eligible Table 4.5 136 Service Years (%) Mothers’ Service Supplement: Proposed Supplementary Replacement Rate according to Number of Table 4.6 136 Children (%) Table 4.7 Herders & Mothers’ Service Supplement: Stylistic Examples of Replacement Rates (%) 136 Table 4.8 Actuarially Fair Reduction Factors (%), 2019 137 Table 5.1: Revenue and Expenditures of the Future Heritage Fund (MNT billion) 151 5 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue BOXES Box 1.1 Main Features of the FSL of 2010 35 Box 2.1 Some Features of Tax Policy in Mongolia 46 Box 2.2 A Summary of Proposed changes to the Mongolia’s Tax System by the Government in 2018 57 Box 3.1 The BOOST Initiative Could Improve the Analysis of Budget Data 63 Box 3.2 Major Changes in the CSL, December 2017 73 Box 3.3 Recommendations on Government Employment and Wages, Mongolia 76 Box 3.4 Actions Required for Grading, Compensation and Payroll Administration 82 Box 3.5 The Current Legal and Regulatory Apparatus of PIM 92 Box 3.6 Purpose of Prescreening 99 Box 3.7 Example of Rationalization Criteria 101 Box 4.1 Addressing the Burden of Chronic Disease: Challenges Faced by Mongolia’s Health System 127 Box 5.1 Key Institutions for Fiscal Management in Mongolia 142 Constitutional Provisions Regarding Parliamentary Powers over Fiscal Decisions and Box 5.2 144 Provisions of the FSL on Deficit and Debt Limits Subprojects on Public Monitoring of Local Procurement of Medicine and Medical Equipment Box 5.3 as Well as of School Budget under the World Bank and Mainstreaming Social Accountability in 148 Mongolia (MASAM) Project financed by the Swiss Agency for Development and Cooperation Box 5.4 Independent Fiscal Institutions models 153 Box 5.5 Fiscal Institutions in OECD countries 153 Box 5.6 Likely Tasks of a Future Fiscal Council in Mongolia 155 Box 5.7 Stabilization Funds around the world 157 Box 5.8 Types of SWFs 157 6 CONTENTS FIGURES Figure (ES).1 Mongolia: Boom and Bust Cycles 13 Figure (ES).2 Efficiency of Public Spending Ranking 17 Figure (ES).3 Mongolia: Financial Sustainability of Pensions (% of GDP) 21 Figure (ES).4 Fiscal Councils: Aspects of Legal and Operational Independence 25 Figure 1.1 Structure of Aggregate Supply (%) 30 Figure 1.2 Structure of Aggregate Demand (%) 30 Figure 1.3 Sectoral Contribution to GDP (%) 31 Figure 1.4 Structure of Net Taxes on Production /Products (%) 31 Figure 1.5 Mineral Revenue Accounts for 5% of GDP, on Average, in 2011-2017 33 Figure 1.6 Mineral Revenue Has Been Driven Largely by Royalties and CIT 33 Figure 1.7 The Share of Mineral Exports in Total Exports Substantially Increased 34 Figure 1.8 The Mining Sector Benefited from FDI Inflow 34 Figure 1.9 Budget Balance and Government Debt (in % of GDP) 37 Figure 1.10 Spending Consolidation Performance (% of GDP), 2016 and 2017 38 Figure 1.11 General Government Debt (% of GDP): 2011-2017 39 Figure 2.1 Mongolia Tax Rate - International Comparison 45 Figure 2.2: VAT Return Filers (number of businesses during 2016-2018) 49 Figure 2.3: Lorenz Curve for Taxes and Contribution in Mongolia (2016) 54 Figure 3.1 Expenditures Increased Significantly 62 Figure 3.2 Recurrent and Capital Expenditures Mostly Driven by a Surge In 62 Figure 3.3 Interest Payments Mainly Explained the Rise in recurrent Expenditures 64 Figure 3.4 Meanwhile Off-budget Investment Spending Dominated Rise in Capital Spending in 2012-2015… 64 Figure 3.5 Mongolia Stands Out for Its High Expenditure Level by International Comparison 65 Figure 3.6 Mongolia’s Current vs Capital Expenditure by International Comparison 65 Figure 3.7 Mongolia’s Goods and Services Expenditure by International Comparison 66 Figure 3.8 Mongolia’s Wage Bill Expenditure by International Comparison 66 Figure 3.9 Mongolia’s Interest Payments by International Comparison 67 Figure 3.10 Mongolia’s Current Transfers Expenditure Remain Steady Over the Past Few Years 68 Figure 3.11 Mongolia’s Capital Expenditures by International Comparison 68 Figure 3.12 Capital Expenditures: On-budget versus Off-budget (% of GDP) 69 7 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Figure 3.13 Spending on Maintenance is Lower Relative to the Size of Economy. 69 Figure 3.14. Local Governments Took on a Growing Share of Public Spending… 70 Figure 3.15 Disbursement Rates of All Level Government Exceeded 90% in 2012-2017 71 Figure 3.16 Meanwhile, the Disbursement Rate is Particularly Low for Public Investment… 71 Figure 3.17 Civil Service Turnover has Increased Overtime Since 2001 ... 72 Figure 3.18 Mongolia’s Efficiency of Public Spending is Among the Lowest in the World. 74 Figure 3.19 Change in Civil Service Employment has Increased with Urbanization, 2005-2015 74 Figure 3.20 Number of Civil Servants, 2005-2015 75 Figure 3.21 Contribution to Growth in Personnel Spending, 2011-2015 76 Change in the Government’s Share has Remained Almost Unchanged in Figure 3.22 77 Total Employment in 2011-2015 Figure 3.23 Contribution to Growth in Personnel Spending, 2011-2015 77 Figure 3.24 The Government of Mongolia’s Personnel Spending, 2011-2017 78 Figure 3.25 Composition of Government Pay in Fiscal Accounts, 2003-2017 79 Figure 3.26 Average Wages by Sector (Thousand MNT), 2017 79 Figure 3.27 Growth in Real Wages by Sector, 2011-2017 80 Figure 3.28 Real Growth in Basic Salaries for Administrative Personnel, 2011-2015 80 Figure 3.29 Real Growth in Basic Salaries in Education Sector, 2011-2015 80 Figure 3.30 Salary Supplements in Education, 2015 81 Figure 3.31 Capital Expenditures and Net Lending (% of GDP): 2004-2017 88 Figure 3.32 Capital Expenditures (% of GDP): 2003-2017 89 Figure 3.33 Composition of Capital Spending by Key Categories: 2004-2017 89 Figure 3.34 Central Government Capital Spending by Sector 90 Figure 3.35 Local Government Capital Spending 91 Figure 3.36 PIM Institutional Arrangements: Roles and Responsibilities of Key Institutions 96 Figure 4.1 Trends of Spending in Education, 2005-2016 107 Nominal Government Spending on Education in Mongolia, 2005-2018, Figure 4.2 108 Total and by Economic Classification Figure 4.3 Number of Students and Share of Private Sector by Level of Education System, 2005-2017 111 Figure 4.4 Share of Children of Lower Secondary School Age Not in School, Mongolia, 2013 112 Figure 4.5 Share of Children of Upper Secondary School Age Not in School, Mongolia, 2013 112 Figure 4.6 Primary Education Student Teacher Ratios in Mongolia and Other Countries 113 8 CONTENTS Figure 4.7 Secondary Education Student Teacher Ratios in Mongolia and Other Countries 114 Figure 4.8 Trends of Spending in Health, 2005-2016 119 Figure 4.9 Catastrophic Health Payments, by Household Expenditure Quintile 122 Figure 4.10 Share of Outpatient Visits by Health System Tier, 2015 123 Figure 4.11 Share of Inpatient Admissions by Health System Tier, 2015 123 Figure 4.12 Infant Mortality Rate in Mongolia and Other Countries, 2014 124 Figure 4.13 Maternity Mortality in Mongolia and Other Countries, 2014 125 Figure 4.14 Maternal Mortality in Mongolia and Other Resource-Rich Countries, 2015 125 Figure 4.15 Infant Mortality Rate in Mongolia and Other Resource-Rich Countries, 2016 126 Figure 4.16 Prevalence of Leading Cases of Morbidity, 2005 -2015 126 Figure 4.17 Prevalence of Leading Cases of Mortality, 2005 -2015 127 Figure 4.18 Projected Pension Insurance Scheme Net Financing Gap 133 Figure 4.19 Population Growth, Age Distribution and Dependency Ratios 2000-2060 133 Figure 4.20 Life Expectancy at Retirement Age 2000-2060 134 Relationship between Minimum Pension, Partial Pension, Minimum Wage and Figure 4.21 137 Minimum Living Standard Figure 4.22 Transition Options for Real Value of Full Minimum Pension (Constant 2018 MNT/Month) 138 Figure 5.1 Revenues, expenditures and accumulation of the Fiscal Stabilization Fund 152 Figure 5.2 Fiscal Council: Aspects of Legal and Operational Independence 154 9 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Vice-President: Victoria Kwakwa Country Director: Bert Hofman Global Practice Director: John Panzer Country Manager: James Anderson Practice Manager: Deepak K. Mishra Task Team Leader: Jean-Pascal N. Nganou Co-Task Team Leader: Badamchimeg Dondog 10 PREFACE PREFACE A number of new developments in the management of public finances occurred in Mongolia since the last Public Expenditure Review (PER) was conducted in 2009. Mongolia adopted two major laws that introduced significant reforms in macro-fiscal and public financial management: The Integrated Budget Law (2011) and the Fiscal Stability Law (2010). The authorities also approved the Future Heritage Fund Law, in 2016, to help address macroeconomic challenges. These include mitigating Dutch disease effects, addressing the perverse effects of upside revenue volatility, and encouraging inter-generational equity, while improving the governance of mining revenues. Furthermore, in 2016, the Mongolian government also adopted the Sustainable Development Vision 2030 (SDV 2030), of which the main objective is to make Mongolia an upper-middle income country within a decade. At the height of the last economic boom, in 2011, Mongolia showed impressive economic growth and development, recording the highest GDP growth in the world at over 17%. The government adopted expansionary fiscal policies following the boom of 2011-2013, to counter the decline of global commodity prices, but this fiscal stance led to an unsustainable public debt dynamic for the country. Also, the massive increase in public expenditures, largely on the back of off-budget financing techniques, resulted in a rapid deterioration of the country’s economic and fiscal performance in 2013-2016. Following elections and change in government, the new administration adopted a major Economic Reform Program in 2016 aimed at restoring macroeconomic stability, reducing fiscal deficits, eliminating off-budget spending and reviving the confidence of private investors. Sound policies and economic diversification had indeed become essential for the country to mitigate the impact of booms and busts caused by sudden changes in global commodity prices. Against this background, a World Bank team working in close cooperation with the Mongolian officials embarked on an effort to prepare a new PER. The PER employed a two-phase approach. Two Just- in-time PER reports were delivered to the government in October of 2016 and 2017, respectively, to help the government to identify appropriate fiscal adjustment options as part of the regular budget consultation process. This report, the full PER, now aims to undertake a detailed analysis of the recent trends in public expenditures and revenue with a view to helping Mongolia’s policy makers and opinion leaders identify the main priorities for a successful implementation of its ongoing recovery program. It also intends to strengthen the fiscal foundation of rapid and inclusive growth to help achieve the objectives of the SDV 2030. Furthermore, the PER also reviews the spending needs of key sectors such as health and education, which play an essential role in the long-term development of the country and the fight against poverty. 11 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue ACKNOWLEDGEMENTS This report has been prepared by a team led by Jean-Pascal N. Nganou (Senior Country Economist, Macroeconomics, Trade and Investment) and Badamchimeg Dondog (Public Sector Specialist, Governance). The team comprised Davaadalai Batsuuri (Economist, Macroeconomics, Trade and Investment), Altantsetseg Shiilegmaa (Economist, Macroeconomics, Trade and Investment), Raul Felix Junquera-Varela (Lead Public Sector Specialist, Governance), Mohan Nagarajan (Senior Economist, Governance), Samuel Freije-Rodriguez (Lead Economist, Poverty), Anil Onal (Economist, Macroeconomics, Trade and Investment), Jay-Hyung Kim (Advisor, Governance), Rabia Ali (Senior Economist, Education), Tsolmon Tsilaajav (Consultant, Health), Luke Voncina (Consultant, Health), Pagma Genden (Operations Officer, Education and Health), Mark Dorfman (Senior Economist, Social Protection and Labor), Tungalag Chuluun (Senior Social Protection Specialist, Social Protection and Labor), Carolina Vaira (Senior Governance Specialist, Governance), Amarbayasgalan Dorj (Governance Specialist, Governance), and Marcelo Buitron (Consultant, Governance). Chapter 5 of the report draws upon a background paper on non-technical drivers of fiscal consolidation and potential implications for World Bank Group engagement (July 2017) by Verena Fritz (Senior Public Sector Specialist, Governance). Angar Enkhtur (Team Assistant, EACMF) and Otgonbayar Yadmaa (Program Assistant, EACMF) provided efficient support to the task team and mission teams on administrative and logistical matters. Sincere appreciation goes to Xavier De la Renaudiere (Consultant, Macroeconomics, Trade and Investment), Selvaraj Ranganathan (Knowledge Management Analyst, Documentation and Communication Products), and Diane Stamm (Consultant, Macroeconomics, Trade and Investment) for editing of the report. The report greatly benefitted from the very useful insights and comments provided by our peer reviewers Jens Kromann Kristensen (Head, PEFA Secretariat); Genevieve Boyreau (Program Leader, Macroeconomics, Trade and Investment); and Ralph van Doorn (Senior Country Economist, Macroeconomics, Trade and Investment). The team is grateful for the excellent collaboration in preparing the report extended from the government counterparts at the Ministries of Finance (MOF); Education, Culture, Science and Sports (MECSS); Health (MOH); Labor Protection (MLSP); Bank of Mongolia (BOM); National Statistical Office (NSO); Civil Service Council (CSC); and the Mongolian National Audit Office (MNAO). Overall guidance for the report was generously provided by James Anderson (World Bank Country Manager for Mongolia) and Deepak Mishra (Practice Manager for East Asia and Pacific, Macroeconomics, Trade and Investment). 12 EXECUTIVE SUMMARY EXECUTIVE SUMMARY I. Context 1. In 2016, the Mongolian government adopted the Sustainable Development Vision 2030 (SDV 2030), whose main objective is to make Mongolia an upper-middle income country within a decade. To achieve this objective, the SDV 2030 would have to rely on sound macroeconomic and fiscal policies and a gradual diversification of the country’s economy lessening its dependence on production and exports of minerals. 2. Sound policies and economic diversification are indeed essential to mitigate the impact of boom and bust cycles brought about by sharp fluctuations in global commodity prices. In 2010, following a sharp decline in copper and coal exports (in 2008-2009), the government adopted a Fiscal Stability Law (FSL), inspired by the Chilean experience, which aimed at dampening the economic and fiscal impact of mineral price fluctuations by stabilizing public spending through less dependence on the proceeds of mineral revenue. However, immediately after the 2008-2009 crisis, Mongolia became the fastest growing economy in the world. Real GDP growth peaked at 17% in 2011 and economic growth averaged 13.8% in 2011-2013. 3. When the last economic boom ended in 2014, the government tried to mitigate its impact through a massive increase in public expenditures. Much of this was accomplished through off- budget financing techniques, since the rules of the FSL prohibited a large increase in State budget deficits. The result was a rapid deterioration of the country’s economic and fiscal performance in 2013-2016, which led to the adoption, in 2016, of a major Economic Reform Program (ERP) aimed at reducing public finance deficits, eliminating off-budget spending and restoring the confidence of private investors. FIGURE (ES).1 Mongolia: Boom and Bust Cycles 20 120 Commodity Price Index 15 100 10 GDP growth, % 80 5 0 60 -5 40 -10 Fiscal Balance, % of GDP 20 -15 -20 0 Source: MOF data, WB staff estimates. Note: LHS – GDP growth & Fiscal balance. RHS – Commodity Price Index (2010 = 100). 13 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 4. The strong correlation between global commodity prices and economic growth in Mongolia is well established. In the past one decade, Mongolia’s economy has had two episodes of near- recession, both coinciding with sharp drop in global commodity prices (Figure (ES).1). The last economic slowdown in 2014-2016 is particularly noteworthy, as it happened despite a massive stimulation to the economy in the form of large fiscal deficits. 5. In 2017, extractive industries accounted for 22% of gross domestic product (GDP) and 21% of government revenue, but the true contribution of the mining sector to overall economic performance is much larger if one takes into account its impact on other economic sectors. Exploitation of coal, copper and other minerals is a major driver of economic growth and an important source of fiscal transfers, export earnings, and foreign direct investment (FDI). However, like many other producers of minerals, Mongolia’s mineral wealth is vulnerable to changes in the external environment. Almost all the macroeconomic crises in recent years coincided with a downturn in global commodity markets. This means that the government has a dual responsibility. First, it should stimulate investment in the mining sector, accelerate growth and maximize government revenue. Second, it should also promote economic diversification and adopt policies aimed at minimizing existing vulnerabilities, notably through sound fiscal management involving the creation of buffers against future external and domestic shocks. 6. Low savings during sunny days and high spending during rainy days remain at the heart of Mongolia’s fiscal problems. The Mongolian government did not implement the most appropriate counter-cyclical policies. Most of the country’s mineral resources were discovered recently and building strong institutions and developing effective policies to minimize vulnerabilities is a complex process which takes a long time. Consequently, Mongolia, which stands out as a country with high public spending, low tax rates and high exemptions, was unable to generate substantial savings for future years when booming mineral production and exports stimulated growth and increased government revenue. 7. Expansionary fiscal policies following the boom of 2011-2013, to counter the decline of global commodity prices, led to unsustainable public debt dynamics. Coping with low export volumes and low prices for its mineral production after 2012 was extremely difficult. Copper and coal prices dropped by 25% and 32%, respectively, between 2013 and 2015,1 and coal exports declined by 21% reflecting slower demand in China. The growth rate of the economy which averaged 15% in 2011- 2012, fell to 7.9% in 2014, 2.2% in 2015 and 1.3% in 2016. Strong political pressures and the desire to mitigate the economic impact of the commodity crisis, encouraged the government to keep public spending at a high level (about 37.5% of GDP in 2014-2016). However, total government revenue fell from 31% of GDP in 2011-2012 to 26.6% in 2014-2015. The combination of sharply reduced government revenue and expansionary fiscal policies led to high (consolidated) budget deficits (averaging 12% of GDP in 2014-2016), and unsustainable public debt and debt service ratios. 8. Mongolia has good fiscal laws, but these proved inadequate to prevent the country from borrowing 20% of GDP in foreign currencies with short maturities at the height of its economic boom. The objective of the FSL, adopted by the Parliament in 2010, was to ensure fiscal discipline by setting up a structural deficit ceiling (2% of GDP) calculated based on average commodity prices over 16 years. Strict implementation of that rule was expected to create savings during good times 1 Compared to their peak in 2011, the decline of copper and coal prices was about 45% 14 EXECUTIVE SUMMARY that would be used to mitigate the economic and fiscal impact of low export prices and sharply reduced government revenue during bad times. The law, which became effective in 2012, however, was not implemented as the government used a series of off-budget financing techniques to bypass the rules of the FSL, and fund high levels of public spending. Recurrent expenditures rose from 23.1% of GDP in 2010 to 27.4% in 2016. Capital spending, largely financed by off-budget resources through the Development Bank of Mongolia (DBM), increased from 6.1% of GDP in 2010 to 11% in 2016. High domestic and foreign borrowing led to sharp increases in public debt (up to 87.2% of GDP in 2016 from 24.1 in 2011). Interest payments rose from 0.3% of GDP in 2011 to 4% in 2016. 9. To restore a sound economic management framework, the government adopted a Comprehensive Macro Adjustment Plan (CMAP) in February 2015 with the broad-based political support of a grand coalition. The CMAP aimed to improve the macroeconomic policy framework and called for reducing the balance of payments pressure by unlocking FDI inflows for large foreign investment projects. Under the CMAP, the government revised the 2015 budget in January to include a large portion of DBM’s off-budget expenditures and amended the FSL to contain the structural budget deficit within 5% of GDP and to reduce the budget deficit to less than 2% of GDP by 2018. However, some fiscal consolidation measures proposed by the government (including reforming social transfer programs including the Child Money Program that could have saved about MNT 100 billion), failed during the parliament review of the 2015 budget. Meanwhile, the budget revenue projection (MNT 6.3 trillion) proposed by the MOF was revised upward by the parliament to MNT 6.7 trillion to allow more spending, which eventually undermined the credibility of the 2015 budget revenue projections. The monetary policy tightened, withdrawing the policy loans that were issued to the Price Stabilization Program in 2013-2014. A major development was made to revive FDI from 2016 as the prolonged negotiation on Oyu Tolgoi (OT) underground mine development was concluded. 10. Despite the commitment toward restoring sound economic management, prior to the elections of 2016, the government embarked on a series of spending measures including: (a) one-time expenses on ASEM and the dispute settlement payment to Khan Resources Ltd; (b) clearance of payment arrears from public investment projects; and (c) policy lending programs such as one-off spending Good Programs. The result was a sharp increase in the budget deficit. 11. A newly elected government in 2016 affirmed its commitment to reforms aimed at accelerating economic recovery. In November 2016, the Parliament adopted an Economic Recovery Program (ERP), which combines short-term measures aimed at restoring fiscal sustainability and long- term actions aimed not only at consolidating fiscal reforms but also at diversifying the economy, improving the business environment and attracting foreign investment for stable growth. 12. The government’s response, buoyed by positive developments in the commodities market, resulted in a swift improvement of its fiscal stance. Off-budget spending by the DBM and the Bank of Mongolia (BoM) was stopped. A pay-as-you-go (PAYGO) system limits the ability of the Parliament to increase overall spending. An independent fiscal council was created to review budget forecasts and monitor compliance with the FSL. Mining exports including coal and copper increased from about $4 billion in 2016 to nearly $5.5 billion in 2017. Copper, coal and gold production is expected to further increase as new projects enter the production phase. The combination of rising mining activities and specific tax policy and administration reforms will increase government revenue and reduce public finance deficits. Adjustment measures envisaged for 2018-2019 will include a hiring freeze and a major effort to rationalize capital expenditures (notably domestically-financed projects). 15 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 13. The latest International Monetary Fund (IMF) Debt Sustainability Analysis (DSA) (February 2018) concluded that the program adopted by the government would restore government debt to a sustainable path, provided: (a) growth picks up in 2019, (b) donor funding on favorable terms reduces the present value of government debt, and (c) external non-concessional borrowing will be successfully refinanced. Nevertheless, the Mongolian economy and public debt remains extremely vulnerable to external factors, including the global demand (mainly China’s economic growth), commodity prices, and exchange rate and interest rate shocks. II. Objective of the Public Expenditure Review 14. The objective of this Public Expenditure Review (PER) is to analyze recent trends in public expenditures and revenue with a view to identify the main priorities for strengthening the fiscal foundation of rapid and inclusive growth. The PER also reviews the special spending needs of a number of key sectors, including health and education, that play an essential role in the long-term development of the country and the fight against poverty. III. Trends in Public Expenditures and Public Investment Efficiency 15. With the onset of mineral wealth, Mongolia emerged as a big spender. The country’s public expenditure to GDP ratio averaged 37.3% in 2010-2016, much above the average in commodity dependent countries (29.5%) and other lower-middle income countries (LMICs) (31.6%). Both recurrent and capital expenditures substantially increased since 2010. The GDP ratio of total public spending increased from 31.6% in 2010 to 40.5% in 2016. Recurrent expenditures increased from 23.1% of GDP in 2010 to 27.4% in 2016. Capital expenditures, including all DBM’s spending, increased from 6.1% of GDP in 2010 to 11% in 2016. 16. The wage bill is the largest component of recurrent expenditures. Its increase from 6.1% of GDP in 2011 to 6.5% in 2017 (a decline from 7.3% in 2016) is not due to significant increases in salary levels but to the growing number of civil servants, which is high and increased from 4.4 civil servants per 100 population in 2009 to 5.7 in 2015. 17. Interest payments, which rose from 0.4% of GDP in 2010 to 4% in 2016 are by far the largest factor behind the increase in recurrent expenditures. The recent increase in interest payments is entirely due to large external and domestic borrowing used to finance an over-ambitious public investment program (PIP). Yet, in effect, recurrent spending is not the most critical issue. The high level and the recent growth of capital expenditures are the main issue. 18. Mongolia’s capital expenditure in recent years has been among the highest in the world. At an average of about 11% of GDP in 2010-2016, Mongolia’s capital expenditures far exceed the 4.8% of LMICs and the 3% of East Asia and Pacific countries. Total capital spending excluding the DBM’s commercial spending increased from 6.1% of GDP in 2010 to 12.4% in 2013, declined to 6% in 2015 but rose again to 9.9% in 2016. Capital expenditures, initially funded by the budget in 2010-2011, were increasingly financed by off-budget resources (DBM) when the rules of the FSL became effective. The decline in 2015 was the result of a first round of adjustment measures. While the decision of the government to pay off arrears on promissory notes authorized by the Parliament since 2014 was intended to make fiscal and debt statistics more accurate, it also contributed to the increase of capital spending in 2016. 16 EXECUTIVE SUMMARY 19. The composition of capital investment has been skewed towards new projects and against operations and maintenance. Most of the budget and off-budget funding for capital expenditures financed public investments in transport and electricity. Mongolia’s capital expenditures also include maintenance, equipment and feasibility studies, but a major weakness of the budget is the low percentage of funds allocated to capital maintenance. Line ministries continue to give priority to new projects over adequate funding for operations and maintenance of existing capital assets.2 20. Moreover, public expenditure efficiency in Mongolia is low as the country ranks 124th on efficiency of public spending of the Global Competitiveness Index (Figure (ES).2). The 2016 IMF Public Investment Management Assessment (PIMA) for Mongolia revealed the poor performance in terms of public investment management (PIM) practices. Mongolia’s scores are much lower than those for other world and emerging market comparators, some of which invest far less than Mongolia. Weak PIM led to poor projects selection, long project implementation delays, high cost overruns, and a rapid deterioration of existing assets (see above), thereby contributing to low spending efficiency. This report argues that successful implementation of a road map aimed at improving public investment efficiency is the highest priority. The proposal includes actions organized around six themes: (a) developing a comprehensive and consolidated approach to PIM, (b) clarifying roles and responsibilities, (c) developing a PIM database, (d) rationalizing the ongoing PIP portfolio, (e) managing the existing build-transfer (BT) activities and strengthening build-operate-transfer (BOT)/ public-private partnership (PPP) management, and (f) upgrading the allocation of maintenance expenditures. FIGURE (ES).2 Efficiency of Public Spending Ranking Malaysia 15 China 19 Indonesia 25 Botswana 37 More efficient Thailand 51 Russia 57 Kazakhstan 65 Chile 80 Peru 104 Mongolia 124 0 20 40 60 80 100 120 140 Source: Global Competitiveness Index 2016–17, World Economic Forum. 2 A recent IMF study shows that Mongolia spends less than 0.15% of GDP on road maintenance (0.3-1% in other developing countries). Road main- tenance accounts for only 20% of the total road spending against 50% in other developing countries. This is clearly an area in which Mongolia should increase its public spending in order to protect its infrastructure and reduce future investment costs. 17 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue IV. Domestic Revenue Mobilization 21. Mongolia’s revenue performance has been highly volatile, exposed to the full vagaries of global swings in commodity prices. Following the decline in commodity prices and mineral exports after 2011, total government revenue fell from 33.8% of GDP in 2011 to 24.1% in 2016. Mining-related revenue averaged only 20% of total revenue during that period, but about three quarters of the decline in government revenue was due to a sharp drop in mining- related revenue (from 11.5% of GDP in 2011 to 3.3% in 2016). In 2017, higher demand in China led to a 33% increase in coal production and a 133% increase in coal exports. As a result, total government revenue exceeded expectations, reaching 29.2% of GDP during that year. 22. Delinking the overall revenue performance from the mineral sector is a critical first step. Recent positive results should encourage the government to embark on a gradual reform of its taxation system to increase the share of non-mining revenue, thus reducing the vulnerability of total government revenue to the ups and downs of mining activities and creating fiscal space for a more effective contribution of the public sector to sound investments and high-priority social sector services. 23. Three types of measures should be envisaged. First, low statutory tax rates should be increased. Second, the number and size of tax exemptions should be revised, following a detailed assessment of the costs and benefits of the current system. Third, tax administration is weak and should be reinforced. • Mongolia’s statutory tax rates are low when compared to its peer countries. The personal income tax (PIT) rate is 10% and the added value tax (VAT), which was introduced in 1998, had its rate reduced from 15 to 10% in 2007. In April 2017, the government proposed to move to a progressive PIT,3 to increase the social insurance contribution of employers and employees and to increase excise taxes on fuel, alcohol, and tobacco. Some of these reforms met with resistance and were not be implemented. The income tax on small businesses went up to 10% and the excise tax on petroleum products and old vehicles was increased. • Too many tax exemptions have reduced the tax base. Around 50% of VAT liabilities are exempted from taxation. Direct investors and service providers in the mining sector are given tax incentives on profit and VAT taxes. A detailed evaluation of these exemptions should review the objective of each measure, the benefits expected and the feasibility of alternative approaches. As the international tax system also has potential for income tax leakage, the treatment of transfer pricing is an important issue. General Anti-Avoidance Rules (GAAR) should also improve the fairness of the tax system. • Broadening the tax base can be achieved by improving the control of compliance and bringing more taxpayers into the tax net. Electronic invoicing increases the number of active tax payers. Increased focus on the large taxpayer segment also is a key characteristic of a modern tax administration system. 24. There is room to increase revenue collection by making the current 10% flat personal income tax (PIT) more progressive. An analysis of the distributional impact of the value added tax (VAT), excise taxes and PIT reveals that the impact of the 10% flat personal income tax is only slightly progressive in relative terms although it only collects 9% of government revenues (much less than the share of VAT and Social Security Contributions (SSC), which are 20% each). Meanwhile, VAT and excise taxes are regressive as these taxes have a dis-equalizing effect because their burden is proportionately larger among the poor than among the non-poor. 3 Business entities pay income tax at the rate of 10% up to MNT 3 billion and 25% above that level. 18 EXECUTIVE SUMMARY 25. The impact of these taxes on inequality is diverse, but limited. VAT and excise taxes increase inequality as measured by the Gini coefficient by about 1 percentage point, while PIT and SSC reduce the inequality of final income by 0.7 and 0.4 percentage points, respectively. These results are compatible with the usual finding that taxation is not the main driver of fiscal policy redistribution effects. It is also consistent with the observation that PIT, SSC, VAT and excise taxes represent a small portion of fiscal revenues in Mongolia, and hence of public expenditures. This is mainly explained by the large influence of corporate and royalties’ taxation that characterize mineral commodity-based economies. The redistributive impact of fiscal policy is mostly undertaken through public expenditures. 26. The Child Money Program and the subsidies in mortgage interest are two large transfer programs (in terms of their budget shares) that have only limited redistributive impacts. A more targeted allocation of these funds towards groups in the bottom of the distribution would allow for larger transferences to these groups and, hence a more efficient poverty reduction effect of the transfers. Reallocation of these funds would enhance the poverty and inequality reduction impact of the fiscal system.  On the one hand, the subsidies on mortgage interest represent a large percentage of budget expenditures (around 1.5% of GDP) but this is mostly allocated in top income groups with a regressive impact both in absolute and relative terms. On the other hand, the Child Money Program is also a relatively large transfer program (around 1% of GDP) that is neutral in terms of absolute progressivity. 27. To conclude, revenue mobilization will improve if the government embarks on a major reform program involving a wide variety of measures, including: (a) a gradual implementation of a progressive PIT system, (b) removing distortionary tax exemptions, (c) a review of international aspects of tax laws, (d) introducing GAARs in the legislation, and (e) improvements in the ICT strategy. V. Social Sector Issues 28. Growing and high-quality education and health services and an effective social security system are essential to support the diversification of the Mongolian economy and the long-term development of the country. Mongolia’s SDV 2030 gives a high priority to these objectives. Building a Knowledge-based Society and a Skillful Mongolia – Government Spending on Education 29. Mongolia’s education spending ratios are comparable with peer countries. Total education spending peaked at 6% of GDP in 2014 and public education spending was at 18–19% of total public spending in 2014-2015. Recurrent expenditures dominate. Although wage spending per education worker declined, the wage bill is the fastest growing category of expenditures. Mongolia spends substantially on expensive school equipment but allocations to classroom learning materials are extremely low. Poor planning of education investments and human resources leads to unused capacity in some areas and severe capacity constraints in other. In addition, the efficiency of capital spending is weakened by extremely low routine maintenance expenditures. 30. Nevertheless, substantial progress has been achieved towards near universal coverage in basic education. The number of kindergartens and enrollment in early childhood education increased rapidly. The public sector absorbs most of the pupils especially at the pre-primary and general education levels. However, the number of (public) higher education institutes is declining due to privatization. 19 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 31. Outside primary education, which is nearly universal, socioeconomic gaps in coverage prevail despite rising education puts households on a higher earning path. Access to quality early childhood education is constrained, notably for herder families. Early grade literacy and numeracy data also indicate low-quality foundations that undermine the efficiency of education investments at the secondary and higher levels. 32. A comprehensive reform agenda should include five priority measures: (a) benchmark performance by participating in PISA and strengthen national assessments, as absence of standardized assessment data makes it very difficult to evaluate the effectiveness of the education system in producing learning outcomes, (b) reduce inequities in quality of learning environment across schools and kindergartens, and give schools autonomy to manage operational budgets, (c) expand access to early childhood education through community or home-based modalities for herders, (d) explore potential for increased parental contributions among the well-off (to relieve spending pressure on the public sector), and (e) strengthen planning of capital investment through more transparency and a larger role for provincial education authorities. Developing an Effective, High Quality and Accessible Health Care System - Government Spending on Health 33. Mongolia’s health spending is low. At just over 4%, the share of total health spending in GDP is lower today than that in 2003 (6.7%), and lower than that in most other middle-income countries. Since 2010, government health spending remained consistently low fluctuating between 6-8% of total government spending. Most of the public spending is on recurrent expenditures. While wage spending per health worker is growing slowly, the wage bill is the fastest growing and largest component of recurrent expenditures (before current transfers and medicines). Allocations to maintenance and repairs are extremely low. Poor sector planning that is not based on clear medium-and long-term strategies affects the quality and efficiency of investments and recurrent expenditures. There was a phenomenal expansion of the private sector in provision of both inpatient and outpatient services. 34. Maternal and child health (MCH) outcomes continue to improve, but reform is critically needed to address the burden of non-communicable diseases (NCDs). Mongolia’s strong MCH performance is due to the near-universal coverage of basic preventive services (immunization and skilled attendance at birth). However, NCDs account for 64% of the disease burden and 86% of mortality. Mongolia is not equipped to meet this growing burden as its fragmented system needs to transition to a new service delivery model that integrates preventive and curative services, modernizes role of secondary and tertiary hospitals as providers of complex care, rationalizes human resources, and deploys standardized local and national systems to measure and improve the quality of primary care, chronic disease management, and patient satisfaction. 35. The reform agenda should include three main components: (a) reorganize services to address chronic illness by focusing on high-quality primary care integrating preventive and curative services, (b) improve efficiency by basing capital investments on strategic long-term goals, and (c) align incentives to support integrated service delivery and improve management at the local and provider level. Implementing the proposed reform agenda may require a gradual increase in budget allocations to the health sector and its share of total public spending as the overall public sector reform reduces public finance deficits to more sustainable levels. 20 EXECUTIVE SUMMARY Building an Effective Social Insurance System - Pension System Reforms 36. Mongolia is at a critical time where not only its aging population is accelerating the fiscal burden of the Pension Insurance Scheme, but also the authorities have enacted several policy changes that pose considerable risks to its financial sustainability. These measures have sharply added to anticipated costs. Without increases in the contribution rate enacted in April 2017, the required State Subsidy for Pension Insurance is projected to rise from 2% of GDP in 2016 to 6% in 2030 and 11% in 2050 (Figure (ES).3). Although 2017 increases in the contribution rate will reduce the fiscal costs by about 1 percentage point of GDP, other measures enacted will more than offset this expected cost reduction. The authorities will therefore not be able to achieve the objective of the 2015 Strategy on Pension Reform, which sets out the Government’s target of a maximum State Subsidy of 2% of GDP by 2030. FIGURE (ES).3 Mongolia: Financial Sustainability of Pensions (% of GDP) Baseline: Mandatory + Voluntary Contribution Rate Increase ‐ Mandatory and Voluntary Retirement Age Increase: Mandatory + Voluntary Schemes Source: PROST Simulations of the 2018 PER Background paper on Pensions. 37. Government pension policies in recent years have not only increased costs but have lacked consistency. The authorities adjusted pension parameters in 2012, 2017 and 2018 in ways that have resulted in inconsistent policies that will materially increase the required State-Subsidy and have adversely impacted system design and incentives. These adverse policy changes include: (a) deeply discounted service buyback arrangements enacted in 2012 which enabled many workers to purchase a pension for life at retirement age at a fraction of the cost that other workers and employers have paid having contributed during their work lives, (b) supplemental benefits for herders and mothers which create substantial fiscal costs and establish incentives for workers to stay in herding into the future to qualify, (c) reductions in herders’ retirement ages, (d) requirements for employers to make social insurance contributions on behalf of mothers with small children, which create incentives for employers to avoid hiring women of child-bearing age, (e) real increases in the minimum pension since 2012, which have reduced the linkages between contributions and benefits so that about half of retirees get the minimum, and (f) substantial real increases in the partial pension, which has reduced the incentives for fully vesting for a regular pension. 21 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 38. Although the government has enacted some other measures that will have some effect in reducing expected costs, such measures will not be sufficient to offset the costs of inconsistent policies mentioned above. The following reform measures were undertaken: (a) increases in the contribution rates, which are expected to improve the financing of pension insurance by up to 1% of GDP per year, (b) increase of the reference wage period for determining benefits from 5 to 7 years, and (c) gradual increase of the pension vesting period from 20 to 25 years.4 39. Meanwhile, there are additional challenges to the sustainability and adequacy of pensions. These challenges include: (a) ad-hoc pension indexation, which leaves retirees vulnerable to erosion of the purchasing power of pensions while creating pressure for periodic benefit increases, and (b) the effective retirement age of 60 for men and 55 for women and earlier ages for eligibility for special professions and herders that is too low to sustain the Pension System and creates poor incentives for workers to stay in the labor market. 40. The distributional analysis of transfers indicates that pensions are the most important transfer in terms of both absolute and relative progressivity. These explain the largest increase of income from groups at the bottom of the distribution, thereby helping elucidate most of poverty reduction. Families with elderly are the most favored by the tax-transfer system in Mongolia. Although the distributional analysis is short-term, the share of elderly population is expected to increase in coming years, the dependence of pensions as the main redistributive policy is also bound to increase. Consequently, fiscal sustainability of the pension system in the short and long term is the most important factor to maintain the redistributive impact of fiscal policy in Mongolia. 41. The pension system needs bold and urgent reforms. A range of parametric reform measures will be needed to improve the sustainability of the Pension Insurance Scheme such as increasing the retirement age, increasing the cost of the service buyback, reducing supplementary benefits, gradually reducing the real minimum pension and the partial pension, and adopting automatic price-based indexation. VI. Other Public Expenditure Issues - Public Wage and Employment 42. The wage bill is the largest and the fastest growing component of recurrent expenditures. Wages per worker in the civil service are below the national average and the wage levels in many other sectors of the national economy. The number of civil servants in Mongolia is high as cross-country comparisons show that the number of civil servants per 100 population is significantly higher in Mongolia than its regional peers and continued to rise from 4.4 in 2009 to 5.7 in 2015. The fact that Mongolia has one of the lowest population density in the world is one of the factors behind these high numbers. Accounting for the population density requires addressing two main issues: the desirability (given such a dispersed population) and affordability (given the fiscal constraints). There is no doubt that the government is prioritizing the desirability angle over affordability, which is not a sustainable strategy. At 7.3% of GDP in 2016 and 18.3% of total public spending, Mongolia’s wage bill is between the average in middle-income and low-income countries. However, at 30% of government revenue, it is not affordable as long as the government is unable to further improve its weak revenue mobilization performance. 43. The government is taking steps to better control the growth of the wage bill yet further efforts of reform are needed. In the context of the IMF’s Extended Financial Facility (EFF), hiring freezes (except for health, education, emergency services and police) will be implemented and grade 4 The vesting period is the amount of covered service needed before an individual is entitled to a full annuitized pension. The increase in the vesting period only applies to individuals that retire at age 55 for women or 60 for men. Those who retire at an age schedule gradually increasing the eligibility age are not subject to the increase in the vesting period. 22 EXECUTIVE SUMMARY advancements will be suspended until the end of 2019. A 2009 World Bank review of the Civil Service recommended a comprehensive review of the civil service pay and grading system and development of an automated centralized payroll system that would help provide further insights and quality data analytics for formulating better informed reform options. Yet, so far little efforts were put into these actions. 44. Another key issue in the civil service is to create a professional and apolitical civil service. All the rules necessary to promote merit-and performance-based recruitment and advancement are in place but implementation remains weak. Frequent reorganizations of ministries provide opportunities for political leaders to dismiss and replace managers and staff. As a result, the aggregate turnover which did not exceed 5% in the 2000s, rose to more than 14% in 2016 after the elections. Approval by the Parliament of a new Civil Service Law (CSL) in December 2017 is expected to be an important step towards rationalizing civil service management practices. VII. Institutional Issues The legal framework 45. The two major Public Financial Management (PFM) laws include the Integrated Budget Law (IBL) of 2011 and the FSL of 2010. Prior to the IBL, fiscal management was regulated by the Public Sector Management and Finance Law (PSMFL) of 2003 and the General Budget Law of 1992. The main objective of the IBL is to regulate activities related to the state and local budgets, establish the budget principles, framework and components, define roles and responsibilities of the parties involved in the budget processes as well as regulate activities related to budget planning, ratification, spending, accounting and reporting as well as supervision and oversight. The IBL in its current form was fully enacted in 2012. 46. The FSL, on the other hand, had totally different objectives. The goal is to stabilize public spending thanks to multi-year projections of non-mining revenue and to specific ceilings on the structural deficits (2% of GDP) and public debt (40% of GDP). Unfortunately, when the law became effective (in 2012), the government could use non-concessional foreign borrowing and off-budget financing techniques (through the DBM and BoM) to expand both recurrent and capital expenditures. Growing (consolidated) public deficits led to rapidly increasing public debt ratios. The debt service ratio went up from 2.1% of total public spending in 2012 to 12.9% in 2017. However, the amendment to the FSL in 2016-2017 and the adoption of the Debt Management Law (DML) in 2015 provided a sounder legal framework for medium-term fiscal adjustment. 47. Political pressures have stymied implementation of these sound laws. The laws adopted by the Mongolian government aimed at strengthening fiscal discipline and stabilizing public spending and public deficits have in practice experienced large implementation gaps. Also, the basic institutional set- up for the management of revenue, expenditures and debt in the country was inspired by that of many other well-managed governments. So far, however, political pressures have been so strong that neither these laws nor the existing institutional frameworks have been effective in preventing the approval of budgets based on overoptimistic revenue projections, and generating unsustainable levels of recurrent and investment expenditures. The institutional setup 48. Mongolia’s public finance institutions are similar to those that have been created by many well- managed governments. Yet, the sound laws and sensible institutional and administrative structures were unable to prevent a major deterioration of the government’s fiscal performance. The Parliament (The Great Khural) is the main policy institution. It approves tax laws, annual budgets, and amendments, medium-term fiscal frameworks and audited budget reports. The MOF designs and submits for 23 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Parliament approval the appropriate fiscal policy instruments, reports on budget implementation and is responsible for public debt management. The ministry is also responsible for managing various special purpose funds like the Human Development Fund (HDF), the Stabilization Fund (SF), the Social Insurance Fund (SIF), and the Future Heritage Fund (FHF). 49. Mongolia’s institutional set-up, however, lacks a coordinated approach to public spending. For a long time, ministries, departments, and agencies (MDAs) were able to use a variety of funding techniques to bypass existing rules and spend more than agreed ceilings. Changes in policies and external pressures are the only way to restore financial discipline and the credibility of the fiscal management system. Thus, the most important reform is the emergence of a strong MOF with the power and the capacity to plan and implement sound revenue mobilization and public spending policies. 50. The Medium-Term Fiscal Frameworks would seem to be the most appropriate public finance planning instruments. The Ministry of Finance relies on departments and divisions responsible for preparing and implementing budgets collecting taxes and customs duties, auditing accounts and inspecting services. A National Development Agency, which reports to the Prime Minister, is responsible for coordinating national development policies and prioritizing development projects. The Ministry of Mining plays a major role in proposing and implementing rules governing that taxation of the mining sector which accounts for 10-20% of total government revenue depending on fluctuations of exports and commodity prices. 51. While parliament has substantial powers over budgeting, it has limited technical capacity for budget analysis. The Parliament Secretariat performs primarily administrative functions as it is not equipped to provide meaningful technical advice to MPs and standing committees. Moreover, the existing Budget Analysis Unit in the Parliament Secretariat has not fulfilled its mandate effectively, mainly due to limited capabilities. As a result, despite its key fiscal decision-making power, the Parliament’s ability to access relevant knowledge and advice, including on likely fiscal impacts, is limited. Effectiveness of parliamentary oversight institutions, namely, the Mongolian National Audit Office and the Independent Authority Against Corruption, has also been compromised by the rising trend of politicization of the public policy in Mongolia. 52. Progress has been limited in the implementation of the Glass Account Law and on the government’s commitments to the Open Government Partnership (OGP) despite initial efforts to advance fiscal transparency. On the 2017 Open Budget Index (OBI) Mongolia is among the countries that release limited budget information. In fact, Mongolia’s score decreased from 51 (out of 100 points) in 2015 to 46 (out of 100 points) in 2017. The country is also lagging the global average score regarding opening opportunities for citizen participation (7 out of 100 points). 53. There are ongoing efforts to bring more transparency to political party and election financing. The Mongolian National Audit Office now reviews all election contribution records of political parties and candidates and makes the relevant reports publicly available. However, as the need for funding required to win an election increases political parties could be looking to ‘private/secret’ funders undermining the integrity of the electoral process. 54. The independence of the newly created Fiscal Council of Mongolia has been questioned in several fora. Although effectiveness of a fiscal council is not always linked to a specific institutional set-up, the most successful fiscal councils foster transparency, promote a culture of macroeconomic stability and provide non-partisan advice regarding the budget process. Given the role that politics has often played in the budget process in Mongolia, more independence should be considered for the Fiscal Council, established in the context of the ERP. While more traditional fiscal councils around the world have relied on reputation and established practice, strict legal independence is a common feature of the new 24 EXECUTIVE SUMMARY generation of fiscal councils (Figure (ES).4). Furthermore, the ability to operate independently should also be duly considered as the adequate availability of resources would enable the Fiscal Council to fulfill effectively its technical mandate. A larger proportion of new fiscal councils around the world also have guaranteed legal access to information and benefit from safeguards on their budget. FIGURE (ES).4 Fiscal Councils: Aspects of Legal and Operational Independence Traditional New generation Legal independence Access to information Safeguards on budget Own staff attend tasks 0 20 40 60 80 100 Source: Independent Fiscal Councils: Recent Trends and Performance, IMF. March 2018 VIII. Reform Priorities 55. Experience shows that reforms initiated during difficult times are abandoned during good times. The Mongolian government seems ready to change its policies when severe economic and financial crises necessitate urgent adjustment measures. At the same time, it does not hesitate to return to old practices as soon as an improved international environment stimulates economic growth and increases government revenue. Clearly, Mongolia needs not only to implement its Fiscal Stability Law and other sound policy documents, but also strong political support for a consistent approach to economic and fiscal stabilization. 56. In the long-term, economic diversification will be the main factor that will decrease the vulnerability of Mongolia’s economy and public finances to frequent changes in commodity prices. A systematic review of Mongolia’s potential outside the mining sector, and a gradual reorientation of the public investment program, taxation and incentive systems should become a major feature of the country’s long-term development strategy. However, a number of urgent measures are also necessary to facilitate an effective implementation of the ongoing reform program. 57. Past experience shows that the institutions responsible for public finance and public sector management do not always have the political leverage and the technical capacity to implement a wide variety of drastic reforms in many sectors at the same time. To succeed, the reform process needs to be focused on the most urgent and consequential measures, those which will have maximal impact on the country’s economic and fiscal performance. 25 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue In this context, a summary of the priorities that have been identified is presented in the following table: Coverage Recommendation Timeline Develop national and sector strategies, which review not only Short to Medium public investment programs but the totality of actions necessary term to optimize public spending Leverage the recent approval of the new Civil Service Law (CSL) Short to Medium to accelerate civil service reforms. term Revisit the independence of the recently established Fiscal Council and strengthen its capacity to effectively function as a Short to Medium Setting up of the mechanism to provide non-partisan advice on fiscal policy and term right institutions performance towards bolstering fiscal discipline. Accelerate ongoing efforts to promote transparency and engage Short to Medium the public in the oversight of fiscal matters building on successful term budget savings interventions available. Harness mining revenue to accelerate Mongolia’s diversified development-ensuring a well-functioning stabilization fund and Medium term the Sovereign Wealth Fund (SWF) Promote a strong centralized management of public finances, Short to Medium fully controlled by a stronger and more competent MOF. term Equip the existing Budget Analysis Unit in the Parliament Sec- Short to Medium retariat with the capabilities needed to provide relevant knowl- term edge and advice to the Parliament. Implement gradually the planned mechanism that is, the FHF, aimed at generating savings for the future years consistent to Medium term improvement of the external environment. Improving performance of Reorganize services to address chronic illness by focusing on Short to Medium existing institutions high-quality primary care integrating preventive and curative term services. Implement a road map aimed at improving public investment efficiency articulated around developing a comprehensive and consolidated approach to PIM, developing a PIM database, Short to Medium rationalizing the ongoing PIP portfolio, managing existing term concessions and public-private partnership (PPP), and upgrading the allocation of maintenance expenditures. Facilitate greater citizens’ participation in fiscal issues and deci- Short to Medium sion making by fully implementing the existing laws and relevant term regulations. 26 EXECUTIVE SUMMARY Maintain macroeconomic and fiscal policies by continuing ongo- Short to Medium ing efforts to reduce and streamline public spending (including term stopping off-budget spending). Formulate Medium-Term Expenditure Frameworks (MTEFs) Short to Medium based on realistic projections of government revenue. term Reintroduce gradually a progressive PIT. Medium term Undertake a systematic review of ongoing tax exemptions and Short to Medium tax expenditures. term Adopt a new tax incentives strategy articulated around the find- Short to Medium ings of the systematic review. term Freeze temporarily new tax exemptions until the review has Short term been completed. Advance preparatory steps to participate in PISA and strengthen Short to Medium national assessments of education outcomes. term Reduce inequities in quality of learning environment across schools and kindergartens through effective re-allocation of Short to Medium resources from capital expenditures and increased school-level term autonomy in spending. Improve efficiency through strengthening planning of capital Short to Medium investment through more transparency and increased discretion Adopting the right term to provincial education and health authorities. set of policies Align incentives to support integrated health care service deliv- Short to Medium ery and improve management at the local and provider level. term Adjust the service buyback legislated in 2017 so that the cost of Short to Medium the buyback for beneficiaries is better aligned to the expected term pension benefits. Increase the retirement age. Medium term Establish an automatic pension price indexation of pension Short to Medium benefits. term Adjust early retirement for herders and special professions through actuarially fair benefit reductions for early retirement Medium term and supplements for late retirement. Re-establish meritocratic principles in recruitment and promo- tion to limit patronage-driven recruitment and to reduce the Medium term current turnover in the civil service. Simplify the pay structure and rationalize supplementary pay- ments following a detailed analysis of the existing pay and grad- Medium term ing scheme. Roll out a centralized personnel and payroll system to increase the control and improve data availability toward strengthening Medium term the management of the largest recurrent expenditure. 27 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue CHAPTER 1 CONTEXT OVERVIEW AND CONCLUSION Key Highlights  Mongolia’s mining sector, which accounts for 20% of gross domestic product (GDP) and 22% of government revenues in 2017, is a major source of economic growth, fiscal transfers, export earnings, and foreign direct investment (FDI).  Ironically, however, almost all the macroeconomic crises in Mongolia in recent years have coincided with a downturn in global commodity markets.  That is because the government has repeatedly failed to pursue appropriate countercyclical policies, especially saving mineral wealth during boom times to be used during periods of commodity price busts.  During the most recent commodity price downturn, the combination of sharply reduced government revenue and a pre-electoral expansion in fiscal policy led to high (consolidated) budget deficits and unsustainable public debt and debt service ratios. 28 PUBLIC EXPENDITURE REVIEW  The Fiscal Stability Law (FSL) adopted by the Parliament in 2010 was not implemented as originally intended, as the government used a series of off-budget financing techniques to fund high levels of public spending despite sharply reduced government revenue.  In November 2016, the Parliament adopted an Economic Recovery Program (ERP), which combines short-term measures aimed at restoring fiscal sustainability and long-term actions aimed not only at consolidating fiscal reforms but also at diversifying the economy, improving the business environment, and attracting foreign investment for stable growth.  Off-budget spending by the Development Bank of Mongolia (DBM) and the Bank of Mongolia (BoM) has been stopped. An independent fiscal council has been established to review budget forecasts and monitor compliance with the FSL. Elements of a Priority Reform Agenda  The short- and medium-term prospects of the Mongolian economy will depend on effective implementation of the most critical components of the ERP.  The implementation of sound macroeconomic and fiscal policies involving the creation of buffers against future external and domestic shocks is critical to economic diversification. 29 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 1.1 STRUCTURE OF THE MONGOLIAN ECONOMY 1. Mongolia’s experience with boom and bust cycles is largely explained by the structure of its economy. Dominated by the services sector, the Mongolian economy is strongly influenced by mining and mining-related activities in construction, as well as in trade and transport services. 2. To analyze the structure of Mongolia’s economy, we used the 2015 Input-Output Table of the National Statistics Office (NSO) and developed a quasi-social accounting matrix (quasi-SAM). A social accounting matrix (SAM) is a squared matrix combining intersectoral exchanges of activities and products, with factors of production, and institutions. The main findings are provided in the following paragraphs. 3. Domestic production dominates Mongolia’s aggregate supply of goods and services. Imports account for about 17% of total supply, and consumption of fixed capital and indirect taxes both claim the remaining 9%. Domestic absorption dominates total demand for goods and services. Exports account for 23% of total demand. FIGURE 1.1 Structure of Aggregate Supply (%) FIGURE 1.2 Structure of Aggregate Demand (%) Consumption of fixed capital; 5% Exports; Indirect Imports; 23% Intermediate taxes; 4% 17% inputs; 31% Investment; 11% Domestic Private production; Government consumption; consumption; 74% 27% 8% Source: Quasi-SAM developed from NSO Input-Output Table 2015. 4. Services (mostly trade, real estate, financial sector, transport, and education) account for more than 50% of the total value added. Extractive industries account for only 18% of GDP at factor cost. With 16% of GDP, agriculture (mainly livestock) continues to provide opportunity for economic diversification, which is essential to prevent and mitigate the adverse effects of boom and bust cycles. Tourism-related activities (restaurants and hotels), although viewed as a promising area for diversification, account for a negligible 1% of the value added (Figure 1.3). 5. Extractives activities also account for 18% of net taxes on production/products (indirect taxes). Manufacturing (including textiles and agroprocessing) and trade and transport services carry the heaviest burden. Agriculture, information and communication technology (ICT), and financial activities are relatively less taxed (Figure 1.4). 30 CONTEXT FIGURE 1.3 Sectoral Contribution to GDP (%) Others; 11% Mining; Public Admin; 18% 4% Construction; 5% Education; 5% Transport; Agriculture; 6% 16% Financial Sector; 6% Trade; Manufacturing; 13% 8% Real Estate; 8% Source: Quasi-SAM developed from NSO Input-Output Table 2015. FIGURE 1.4 Structure of Net Taxes on Production /Products (%) Others; 11% Mining; Public Admin; 18% 4% Construction; 5% Education; 5% Transport; Agriculture; 6% 16% Financial Sector; 6% Trade; Manufacturing; 13% 8% Real Estate; 8% Source: Quasi-SAM developed from NSO Input-Output Table 2015. 6. The sectoral contribution to GDP reveals a pattern broadly common to resource-rich developing countries, where the mining sector has a strong influence on the rest of the economy. In 2015, the top seven activities of the Mongolian economy (totaling about 84% of GDP at factor costs) included mining, agriculture, trade, real estate, manufacturing, financial sector, and transport. The services sector dominated, but the true contribution of the mining sector was much larger than 18%, if its impact is considered-through multiplier effects-on the development of other economic sectors and subsectors, including construction in the secondary sector, and commerce and other services in the tertiary sector. 31 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 1.2. THE CRITICAL ROLE OF THE MINING ECONOMY 7. Mongolia has vast mineral wealth located throughout its 1.5 million square kilometer territory. Mongolia’s potential is huge in many key commodities, and the country is one of the lowest-cost producers globally. Oyu Tolgoi (OT) (copper) and Tavan Tolgoi (TT) (coal) are among the largest mines in the world. Total proven reserves are estimated at around US$ 1 trillion, but the ultimate value could be much higher when new areas are explored and new deposits discovered. Copper and coal, the only minerals intensely produced and exported so far, make a substantial contribution to the national economy. However, diversifying the country’s mining production and exports could also be a useful component of a national economic diversification strategy. 8. Unlike other advanced mining economies with a capital-intensive mining industry, the Mongolian mining sector is underdeveloped and relatively labor intensive. This is partly explained by the slow development of the mining infrastructure, as Mongolia relies on road transportation to export large volumes of raw coal and copper concentrates. With thousands of truck drivers, the mining sector is (indirectly) as labor intensive as light and processing industries. Underdeveloped infrastructure also comes with negative externalities including on the environment (for example, dust, pollution). Developing and modernizing the country’s infrastructure should therefore be an important component of Mongolia’s public investment strategy. 9. Mining, however, is the most volatile segment of the Mongolian economy. Changes in the performance of the mining sector have had and will always have a strong influence on economic performance, government revenue, and external developments. 10. Mongolia’s economy grew at a double-digit rate (15%) in 2011-2012 as foreign investors rushed in to take advantage of the vast, partly untapped, mineral deposits. However, growth during 2014-2016 dropped suddenly following slower growth in China and a sharp decline in copper and coal prices, which dampened commodity exports. Export prices dropped by about 25% for copper and by 32% for coal between 2013 and 2015.5 Coal export volume also shrank by 21% reflecting slowing demand in China, which absorbs most of Mongolia’s coal exports. Moreover, weak exports and a precipitous drop in FDI caused a collapse in investment and reduced double-digit GDP growth to 7.9% in 2014 and 2.4% in 2015. FDI inflows dried up to less than 2% of GDP during 2014-2016 (from a peak of about 40% in 2011), as new investment projects in the mining sector were delayed. GDP growth further decreased to 1.2% in 2016 as private consumption contracted by 9%, reflecting negative real household income growth and rising unemployment. A strong rebound in coal exports in the last quarter of 2016 and higher public expenditures helped maintain a small positive growth (2%) for that year. The decline in copper and coal prices was about 45% from with their peak in 2011. 5 32 CONTEXT 11. The mining sector accounted for 33% of total government revenue and 11% of GDP in 2011, when FDI inflows peaked at US$ 4.7 billion. In 2016, mining-related government revenue dropped to 15% of total government revenue and 3.1% of GDP when the commodity market cooled and FDI inflows were sharply reduced. In 2017, following the recovery of the coal market, mining revenues went up to MNT 1.6 trillion, that is, 21% of total budget revenue and over 6% of expected GDP (Figure 1.5). FIGURE 1.5 Mineral Revenue Accounts for 5% of GDP, on Average, in 2011-2017 Percentage of GDP: 2011-2017 35% 30% 11% 5% 6% 5% 6% 25% 5% 3% 20% 15% 26% 23% 24% 24% 23% 10% 21% 21% 5% Non-Mineral 0% 2011 2012 2013 2014 2015 2016 2017e Mineral Source: General Department of Taxation (GDT), Ministry of Finance (MOF). 12. Mining revenue comes from three main sources: corporate income tax (CIT), royalties, and customs duties. In 2017, receipts from royalties and taxes/fees (including CIT, value added tax [VAT], customs taxes, and other fees) amounted to MNT 820 billion and MNT 776 billion, respectively. Taxes therefore accounted for the largest share of government receipts from the mining sector, at 90%, of which a further over 50% of mining receipts were from royalties (Figure 1.6). FIGURE 1.6 Mineral Revenue Has Been Driven Largely by Royalties and CIT Percentage of total revenue 2011-2017 35% 33% 30% 25% 11% 21% 21% 20% 19% 20% 18% 7% 7% 15% 5% 15% 5% CIT 14% 5% 7% 10% 10% 8% 10% 10% 10% Royalties 5% 5% Others 7% 5% 6% 4% 3% 4% 4% 0% Total Mineral 2011 2012 2013 2014 2015 2016 2017 Revenue Source: General Department of Taxation (GDT), Ministry of Finance (MOF). 33 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 13. The mining sector also is the key driver of external sector developments. Mining exports (including copper and coal) went up from US$ 0.9 billion in 2007 to US$ 5.5 billion in 2017, that is, nearly 90% of total merchandise exports. Mining was both the main factor behind large current account deficits (following an increase in investment-related imports) and the main beneficiary of a surge in FDI inflows (cumulative amount of nearly US$ 14 billion during 1990-2017, which largely covered current account deficits especially in 2010-2013). FIGURE 1.7 The Share of Mineral Exports in Total Exports Substantially Increased Percentage of Total exports 2000-2017 100% 80% 60% 40% 20% 0% Mineral exports 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Non-Mineral exports Source: NSO, Investment Agency for Mongolia, World Band staff estimates. FIGURE 1.8 The Mining Sector Benefited from FDI Inflow FDI stock (% of GDP): 2005-2017 50% 40% 30% 20% 10% Mining 0% Others Total -10% 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: NSO, Investment Agency for Mongolia, World Band staff estimates. 14. As international mineral prices are expected to increase, the contribution of the mining sector to the Mongolian economy should expand. Copper, coal, and gold production are projected to grow. The gold content in OT’s copper production is expected to increase and new projects will enter into the production phase, notably Gatsuurt and Tsagaan Suvarga gold mines and TT coal mine. Prospects will depend on progress in fostering an enabling environment, notably by addressing electricity and railway bottlenecks and resolving investment contract and related governance issues. Consequently, government revenue from the mining sector is expected to increase notably when the OT’s underground mine begins to produce. 34 CONTEXT 1.3 SOUND FISCAL POLICIES AND POOR IMPLEMENTATION 15. A succession of booms and busts caused by sudden changes in commodity prices is not a new experience for Mongolia. In 2008-2009, the price of copper fell by 70% from its July 2008 peak. Expecting a rebound in copper production and government revenue when the OT project would come on stream, the government decided to review its policy framework for the management of mineral resources. The FSL was approved by the Parliament in 2010. Its objective was to dampen the economic and fiscal impact of mineral price fluctuations by smoothing public spending and insulating it from changes in mineral revenue. Box 1.1 describes the new fiscal rules introduced in the context of the FSL. BOX 1.1 Main Features of the FSL of 2010 The FSL would have enacted three complementary fiscal rules to ensure fiscal discipline: (a) a ceiling on the structural deficit fixed at 2% of GDP, (b) a ceiling on expenditure growth based on the rate of growth of non-mineral GDP, and (c) a public debt ceiling of 40% of nominal GDP. The structural deficit rule is based on a definition of structural revenues as revenues to be received if prices of major minerals (copper and coal) were at a given level (a 16-year moving average)a. Expansion of expenditures would have been limited to a level consistent with the absorptive capacity of the economy. The rule does not distinguish between current and capital expenditures. The gross debt rule aims at ensuring that implementing fiscal rules is consistent with debt sustainability. The fiscal framework would be supported by a stabilization fund for mineral revenue. When mineral revenue exceeded the structured revenue, the difference would be placed in the stabilization fund maintained in an account at the BoM. Note: a.Originally a zero target was recommended to allow the buildup of stabilization buffers and to reduce debt-to-GDP ratios. A deficit target of 2% of GDP was finally approved. It was deemed appropriate if Mongolia’s borrowing strategy were to rely mostly on concessional funds. 16. Based on the Chilean experience, the rules of the FSL are generally sound and could continue to inspire the management of future mineral revenue. However, it was not implemented as intended. The fiscal position deteriorated sharply in recent years, driven by a massive increase in off-budget spending, which was not envisaged in the FSL. During 2012-2015, the fiscal situation was worse if the ‘consolidated budget deficit’ is considered, which is obtained by adding off-budget expenditures of the DBM to the budget deficit as defined by the government. The consolidated budget deficit averaged 9.3% of GDP during 2012-2015. The DBM loans were financed through noncommercial loans to finance the government’s capital expenditure (DBM-financed capital expenditure), and commercial loans to private and state-owned companies (commercial investment to bypass the budget deficit ceiling of the FSL). 17. The central bank was also heavily engaged in quasi-fiscal lending programs. In 2013-2014, loose monetary policies supported fiscal stimulus through quasi-lending activities. The Price Stabilization Program (PSP) and the Housing Mortgage Program (HMP), launched by the BoM, injected liquidity amounting to almost 17% of GDP in 2013. The PSP provided subsidized loans of more than MNT 1 trillion to construction projects and other industries, but began to be phased out by the end of 2014. The HMP-which provides subsidized mortgage loans-is politically popular and continues with the BoM’s financing. Outstanding HMP mortgage loans reached MNT 3.1 trillion at end-2016- equivalent to 13% of the 2016 GDP. 35 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 18. The government tried but failed to curb the ballooning deficit through its Comprehensive Macro Adjustment Plan (CMAP)6 adopted in February 2015 with the broad-based political support of a grand coalition. The government was not able to sustain earlier progress in deficit reduction as they could not resist pre-electoral spending pressures. By bringing DBM’s off-budget investment into the budget and reducing the consolidated deficit from 11.3% of GDP in 2014 to 8% in 2015. The fiscal adjustment efforts, however, were reversed in 2016 as on-budget expenditure was ramped up, led by large pre-electoral spending and a significant increase in capital expenditure (notably to clear arrears on promissory notes). As a result, the budget deficit widened to 17% of GDP, and government debt surged to 88% of GDP. Interest payments climbed to 4% of GDP in 2016-raising serious concern about the sustainability of the public debt position. 19. Dominated by external debt, Mongolia’s government debt increased over six-fold between 2011 and 2016. The government debt-including direct government borrowing and government- guaranteed debt-increased from MNT 3.2 trillion in 2011 to MNT 21 trillion in 2016. As a result, the debt-to-GDP ratio climbed from 24% to 87.6% over the same period. This debt-to-GDP ratio is among the highest in developing countries, far exceeding a 45.5% average in middle-income countries and a 35.6% average in low-income countries in 2015. Mongolia’s Economic Recovery Program (ERP) Program overview 20. The government reform is based on an ERP, approved by the Parliament on November 24, 2016. The objective of the ERP is to mitigate the impact of the economic crisis of 2014-2016 and to restore fiscal sustainability, while pursuing structural reform aimed at regaining the confidence of TABLE 1.1 Key Reform Areas of the Government’s ERP Economic Recovery Program Strategic Objective I Strategic Objective II Economic policy adjustment to overcome short-term Structural reforms for long-term stable growth path economic difficulties • Restore fiscal sustainability by: (i) reducing • Promote economic diversification by supporting inefficient and unnecessary expenditures; and (ii) non-mining sectors with large potential, e.g., expanding the budget revenue base. livestock products such as meat, dairy, and skin/ • Discontinue the quasi-fiscal expenditures of the leather goods. BoM and the DBM. • Expand bilateral trade and transit transport. • Assess the performance of, and the contingent • Attract foreign investments for sustainable and liabilities from quasi-fiscal expenditures. stable growth. • Prepare a plan to resolve the repayments of bonds • Strengthen the FDI environment, including the issued by the government and the DBM. establishment of the investor Protection Council. • Enhance monetary policy to: (i) increase • Strengthen the fiscal legislation to restrict the international reserves; (ii) stabilize inflation; and (ii) parliament’s power during the budget preparation. maintain flexible exchange rate. • Improve the governance, independence, and • Maintain financial stability by appropriate liquidity operational transparency of the BoM. management and proper prudential supervision of • Improve the DBM law to ensure its transparent and the banking system. efficient management. Source: Government of Mongolia (GOM), World Bank staff. 6 Under the CMAP, the government revised the 2015 budget in January to include a large portion of DBM’s off-budget expenditures, and amended the FSL to contain the structural budget deficit within 5% of GDP, and reduce the budget deficit to less than 2% of GDP by 2018. However, some fiscal consolidation measures proposed by the government (including reforming social transfer programs including the Child Money Program that could have saved about MNT 100 billion), failed during the parliament review of the 2015 budget in January. Meanwhile, the budget revenue projection (MNT 6.3 trillion) proposed by the MOF was revised upward by the parliament to MNT 6.7 trillion to allow more spending, which eventually undermined the credibility of the 2015 budget revenue projections. 36 CONTEXT foreign investors and diversifying the economy. The principal reforms to be implemented in the context of the ERP are summarized in Table 1.1. 21. Measures aimed at achieving fiscal sustainability are described in the 2017 supplementary budget adopted on April 14, 2017. Those measure include, first, off-budget spending by the DBM and the BoM to be stopped (and integrated into the budget). Second, safeguard measures to be introduced to mitigate political pressures and ensure fiscal discipline. A PAYGO system to be introduced to limit the Parliament’s ability to increase overall public spending and an independent fiscal council to be established to review budget forecasts and verify compliance with the FSL. Third, strong adjustment measures to be implemented to reduce on-budget spending and increase government revenue. The budget deficit is not to exceed to 10.6% of GDP in 2017, a reduction equivalent to 6.4% of GDP. The 2017 adjustment measures include (a) canceling a one-off dispute settlement expense and eliminating pre-electoral policy loans and (b) ending the Promissory Notes Program that led to a significant increase in capital expenditure in 2016. Adjustment measures in 2018–19 will focus on rationalizing capital expenditures (particularly domestically financed projects). It will also include a hiring freeze to reduce the public sector wage bill and a reduction of unnecessary operating expenses. Fourth, revenue mobilization measures will include raising taxes on higher income earners and on tobacco, alcohol, petroleum, and old vehicles, as well as expanding the tax base for the interest tax. Fifth, budget priorities will include social protection and basic public services (especially education and health services). The education and health budgets will be protected, notably allocations to repair and maintenance of schools and hospitals, owing to a large reduction in other capital expenditures. In 2018, the Food Stamp Program will be expanded to scale up support for households below the poverty line. Progress to date 22. Mongolia’s public finance performance improved dramatically, when the fiscal deficit fell from 16.4% of GDP in 2016 to 1.9% in 2017 (Figure 1.9). When the ERP was negotiated with the country’s development partners including the World Bank, the government assumed that the fiscal deficit would decline from 16.4% of GDP in 2016 to 10.6% the following year. By September 2017, it was already clear that the reduction of the deficit would be much faster than expected, with the year- end deficit projected at 7.6%. Finally, the pace of improvement accelerated in the last quarter, with the overall fiscal deficit turning out to be only 1.9% of GDP-the lowest since 2010. This is due to both substantial increases in revenue and substantial cuts in public spending. FIGURE 1.9 Budget Balance and Government Debt (in % of GDP) Budget Balance Trend (% of GDP): 2003-2017 4 3.0 2.6 2.4 1.4 0 -1.9 -1.7 -4 -3.7 -3.4 -4.6 -5.3 -8 -6.5 -9.1 -9.2 -8.3 -12 -10.4 On-budget deficit DBM’s social programs -16 DBM’s corporate loans -17.0 Gov’t definition of deficit -20 Consolidated deficit 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: MOF data, World Bank staff estimates. 37 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 23. Total government revenue exceeded expectations as it increased from 24.1% of GDP in 2016 to 29.2% in 2017. Mineral revenue doubled in 2017 owing to a sharp increase in royalties and CIT and customs duties on imports, following an increase in coal prices and a strong performance of the domestic coal sector. Non-mineral revenue also increased by 24% in 2017, mostly thanks to the improved VAT collection following implementation of a new VAT law, which became effective in 2015. Collection of tax arrears was also stronger in 2017 (totaling US$ 290 million, almost two- thirds of outstanding arrears of US$ 450 million). FIGURE 1.10 Spending Consolidation Performance (% of GDP), 2016 and 2017 40% 35% 30% 25% 20% 15% Recurrent Primary expenditures 10% Interest Payment Capital Exp and Net Lending 5% Off-Budget Spending (DBM commercial portfolio) 0% -5% 2016 2017 Source: MOF data, WB staff estimates 24. Improvements in public spending performance were also impressive (Figure 1.10). Total public spending declined from 40.5% of GDP in 2016 to 31.1% in 2017, mostly due to a sharp drop in capital expenditures and off-budget spending (commercial loans of the DBM) in line with new policies aimed at gradually reducing off-budget spending, notably through tight control of capital spending. Despite a year-over-year growth rate, recurrent spending was lower than projected owing to the reduction of interest payments and the impact of the hiring freeze on the wage bill. To reduce interest payments, the government also plans to halt issuance of domestic T-bills, giving priority to concessional financial support from development partners including the IMF and the World Bank. 25. All these positive developments contributed to a 3-percentage-point reduction in the government’s gross debt in 2017 (Figure 1.11). Total government direct and guaranteed debt reached MNT 23 trillion (US$ 9.4 billion) as of 2017, approximately 85% of GDP. The fiscal deficit fell from 17% of GDP in 2016 to 1.9% in 2017. The real GDP growth rate was much higher than anticipated. The exchange rate appreciated by 2.5% and the real interest rate declined from 4.3% to -2.5%. The drop in real interest rate was due to improved market confidence and a shift to concessional borrowing. However, the current debt-to-GDP ratio remains one of the highest among developing countries, far exceeding the 45.5% average of middle-income countries and the 35.6% average of low-income countries in 2015. 38 CONTEXT FIGURE 1.11 General Government Debt (% of GDP): 2011-2017 100 90 87.6 84.7 80 18.8 12.7 70 61.4 8.8 Domestic debt 60 55.4 17.2 14.4 IMF loan 50 47.5 13.3 41.0 9.6 General Government 40 6.0 13.4 External Guarantee 6.1 11.4 62.1 4.8 External bond and loans 30 24.1 51.7 3.8 Total Government Debt 20 31.8 30.6 33.6 and Guarantees 30.1 10 20.2 External debt and 0 guarantees 2011 2012 2013 2014 2015 2016 2017 Source: MOF data, World Bank staff estimates. 1.4 ORGANIZATION OF THE REPORT 26. The PER’s main objective is to provide a detailed analysis of recent trends in public expenditures and revenue. In this context, this PER hopes to helping Mongolia’s policy makers and opinion leaders identify the main priorities for successful implementation of its ongoing recovery program and strengthening the fiscal foundation of rapid and inclusive growth. 27. This report is therefore organized as follows: Chapter II discusses the recent developments in domestic revenue mobilization including volatility related to commodity prices. Chapter III examines recent trends and composition of public expenditures including in comparison with peers. It also explores two critical cross-cutting public expenditure issues including public wage and employments as well as public investment efficiency. Chapter IV analyzes government spending trends and issues in social sectors and the Social Security System. Finally, Chapter V assesses the Mongolia’s institutional framework for Public Financial Management. 39 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue CHAPTER 2 Domestic Revenue Mobilization OVERVIEW AND CONCLUSION Key Highlights  Mongolia’s fiscal revenue is volatile due to its dependence on mining revenue and lower tax base. There has been a sharp revenue decline over 2012-2016 due to falling commodity export and prices.  However, evidencing economic recovery since late 2016, the fiscal space in 2017 significantly improved mainly because of a strong mining revenue performance and tighter spending control. While this is a good sign, it does not prevent Mongolia revenue from being tied to the commodity cycle.  The current environment is therefore ideal to accelerate tax policy and administration reform to make the system more stable and robust and to improve the fiscal space further. 40 PUBLIC EXPENDITURE REVIEW  This chapter reviews the revenue mobilization  Mongolia’s tax policy environment for the mining practice in Mongolia in line with existing tax policy industry needs to be comprehensive, given its and administration regime and suggests possible leading role in entire economy and fiscal revenue solutions for the constraints to be addressed in generation. the short, medium and long terms.  Tax Administration: To diversify the sources of  Elements of an Optimal Revenue Mobilization revenue from more volatile mining revenue, the Reform Program tax base must be broadened and compliance must  Tax Policy: be improved by strengthening tax administration.  Mongolia’s tax system, notably tax rates, is too  Other Challenges and Constraints: Tax generous by international standards due to flat legislation needs to be revised toward reflecting 10% rate for major taxes. international taxation issues such as transfer pricing, general anti-avoidance rule (GAAR) and  Therefore, tax rate revisions including progressive double taxation. rates should be considered in the context of the government’s fiscal consolidation efforts.  Tax policy measures are required to reduce distortionary tax incentives and exemptions, as tax expenditures are a significant source of tax base erosion in Mongolia. 41 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 2.1 RECENT FISCAL DEVELOPMENTS - KEY FEATURES OF THE MONGOLIAN TAX SYSTEM 1. Mongolia’s fiscal revenue is volatile due to its dependence on mining revenue and lower tax base. During 2012-2016 - end of the commodity super cycle, Mongolia’s revenue share to GDP dropped continuously from 34% in 2012 down to around 22% in 2016. A decline in investment and exports has negatively affected taxes on production and consumption. Direct taxes dropped sharply, led by CIT and the associated dividend and capital gains taxes. The revenue loss was particularly severe in mining revenues (royalties, CIT, oil revenues) declining to 3.3% of GDP and 15% of total revenues in 2016, from its peak of 11.5% of GDP and one-third of total fiscal revenues in 2011. On the other hand, fiscal expenditure grew to 42% of GDP in 2016 from 32% of GDP five years ago. Consequently, Mongolia’s consolidated budget deficit expanded alarmingly to 17% of GDP in 2016 against a usual level of 9%. The high budget deficit led to a rapid accumulation of debt to 88% of GDP. 2. Mongolia’s fiscal space significantly improved in 2017 mainly because of a strong mining revenue performance and tighter spending control. The 2017 budget performance indicates that structural fiscal deficit stood at 6.3% of GDP compared to the 9.5% projected. Budget revenue receipts exceeded the plan by MNT 657 billion which is equivalent to 2.2% of GDP in 2017. Total revenue reached 29% of GDP compared to the initially planned 26%. The strong performance in the mining sector, particularly with coal production growing 29% and exports, has mainly contributed to higher CIT and VAT revenues. Royalties from mining companies were 15% higher than projected. The CIT and royalty values were doubled in 2017 relative to 2016 (Table 2.1). Aggressive recovery of tax arrears of MNT 708 billion out of a total of MNT 1,092 billion (about 65% of outstanding tax arrears) has also helped grow revenue. TABLE 2.1 Break-up of Mining Revenue, in Major Taxes (MNT billion)   2011 2012 2013 2014 2015 2016 2017 Air pollution fee 19.0 22.6 23.5 30.7 12.1 20.5 34.8 CIT 319.8 176.7 296.3 212.3 261.7 225.8 443.7 Royalty 610.1 320.8 292.2 403.2 468.8 239.6 527.9 Land fee 10.4 11.3 12.4 12.1 12.0 12.3 12.6 Domestic VAT 87.2 64.1 115.6 92.0 63.7 22.1 30.6 PIT (withholding) 34.7 56.3 75.2 73.8 70.3 75.9 107.4 Progressive royalty 34.5 196.8 213.6 214.1 132.0 28.7 292.2 Tax on immovable assets 4.2 4.5 13.9 16.2 17.2 19.9 30.6 Tax interest, fines, losses 2.9 19.3 9.2 17.4 5.6 1.7 3.2 Fee for using water and natural spirits 3.5 9.0 36.3 31.1 31.4 36.8 39.7 Other 97.0 6.6 5.6 3.4 3.2 64.1 72.6 Total 1,223.3 887.9 1,093.7 1,106.3 1,078.1 747.4 1,595.2 Source: MOF data 42 DOMESTIC REVENUE MOBILIZATION 3. Mining industry is one of the major revenue contributors to the state budget and is regulated by more than a dozen of legislations. Over the last seven years, it represented on average about 21% of total government revenue while contributing about 5-6% of GDP (Table 2.2). Mongolia’s tax policy environment for the mining industry is not comprehensive given the leading role of mining in the economy including revenue generation. No special tax regimes are in operation for the sector. Although Mongolia’s mining industry is large in terms of its impact in the overall economy, the sector itself is sensitive to cyclical fluctuations of global commodity prices and therefore, mining revenues are volatile. The introduction of ad-hoc and unpredicted legal amendments undermine the competitiveness of and investment in the mining sector and jeopardize fiscal sustainability. Moreover, mining revenue may tend to crowd out non-mining revenues in resource-rich economies.7 Therefore, the country needs competitive tax laws to boost foreign investment in the non-mineral sector while sustaining the mining revenue stream. TABLE 2.2 Structure of Government Revenue (% of GDP): 2010-2017   2010 2011 2012 2013 2014 2015 2016 2017e Total revenues and grants 33.0 33.8 29.8 30.9 27.9 25.3 24.1 29.2   Mineral revenue accumulation to FSF 0.0 1.8 0.6 0.2 0.2 0.0 0.0 2.5 Structural revenues and grants 33.0 33.8 29.2 30.7 27.7 25.3 24.1 26.6   Tax revenue 27.4 27.8 25.2 26.4 23.4 22.1 20.2 23.2     Income Tax 9.9 6.3 5.5 5.8 4.9 5.0 4.4 5.9       PIT 1.6 1.7 2.0 2.3 2.1 2.1 2.2 2.4       CIT 4.0 4.1 3.1 3.4 2.8 3.0 2.2 3.5         of which: mining companies 2.3 2.4 1.1 1.5 1.0 1.1 0.9 1.6       Windfall tax 4.3 0.4 0.3 0.0 0.0 0.0 0.0 0.0     Social security tax 5.9 8.5 7.8 7.5 6.2 4.5 4.8 6.0     Tax on immovable properties 0.1 0.1 0.1 0.2 0.2 0.4 0.4 0.5     VAT  5.9 8.5 7.8 7.5 6.2 4.5 4.8 6.0     Excise & Special Taxes 2.0 2.6 2.0 2.0 1.6 1.4 1.4 1.9     Customs duties 2.0 2.6 2.0 2.0 1.6 1.4 1.4 1.9     Mining royalties 1.8 1.5 1.9 1.5 2.1 2.6 1.2 0.7   Non-tax revenues and grants 5.6 4.3 3.9 4.3 4.5 3.2 3.9 3.5     Oil revenue 0.6 0.6 0.6 1.0 1.3 0.7 0.6 0.7 Memo  Mining-related revenues 10.4 11 5 6 5 5 3 6   (% share to total revenue and grant) 32.4 33.0 21.0 20.0 18.0 19.0 15.0 21.0 Non-mining revenues 22.6 22.7 24.5 25.2 22.8 20.7 20.9 23.4 1/ Mineral revenues from market prices exceeding structural prices accumulated in Fiscal Stabilization Fund and Future Heritage Fund 2/ 2017 figures are actuals Source: World Bank calculation based on MOF data. 4. The utilization of mining revenue is not earmarked but follows allocation principles enshrined in a series of key laws. These laws include the Fiscal Stability Law, the Integrated Budget Law, and the Future Heritage Fund Law and the Law on Government Special Funds. Additional revenues that are generated from the price difference of key commodities are collected into the Stabilization Fund and 65% of royalties are centralized into the Future Heritage Fund. 7 Belinga, V.; M. Melou, and J-P. Nganou (2016). “Does Oil Revenue Crowd Out Other Tax Revenues? Policy Lessons for Uganda”. Policy Research Working Paper 8048. World Bank. The paper estimated that a percentage increase in oil tax revenue could crowd out about 20-30% of the non-oil tax revenue. 43 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 5. There have been recent developments in Mongolia related to the lack of transparency that surrounded the OT Agreement, through which the State did not get a fair deal. While this report can’t evaluate the accusations, it is worth noting that assessments done by Otto and later Sunley (IMF) help to show the impact of the negotiated OT concessions. Instead of the tough 2006 terms, which placed Mongolia among the highest taxing mining fiscal regimes in the world (highest quartile), chiefly due to the excess profits tax (called ‘Windfall Tax’), the project proceeded under the 2009 Investment Agreement with somewhat lighter terms. Nevertheless, the IMF, at the time of the Sunley report, considered that even the OT terms “may be too demanding for other, smaller projects”8. The current standard regime differs from the 2006 regime by including a royalty with price-linked escalation in place of the Windfall Tax. In all other respects, however, it remains the same. Therefore, one could consider that the current fiscal regime is less tough than the 2006 regime but tougher than the OT terms and that new investors are very likely to seek concessions through negotiation for mine projects that are anything less than world class. Research on the terms being sought by Centerra Gold Co. Ltd for Gatsuurt gold project would help confirm this view. 2.2 CHALLENGES OF THE CURRENT TAX SYSTEM 2.2.1 Distribution of Taxing Power in Mongolia 6. The adoption of Corporate Income and Personal Income Tax Laws in 1991 and General Law of Taxation in 1993 laid the legal foundation for the establishment of a taxation system in Mongolia. Mongolia has 23 variety of taxes, of which 21 taxes are collected from legal entities and individuals. In addition to the 21 taxes, royalties are imposed on mining sector according to the Minerals Law. The distribution of taxation powers between the different levels of government is presented in Table 2.3. TABLE 2.3 Tax and Revenue Allocation between State and Local Governments State Taxes (Central Aimag and Capital Common Taxes Soum and District Taxes Government Budget) City Taxes Royalty on minerals (95% of Central budget: CIT Capital city taxes PIT other than 8.1.1 of the PIT law 5 of local budget) VAT of imported goods & Domestic VAT Land user fee Gun tax Services Immovable Stamp duties other than 11.2 of the   Excise taxes property tax stamp duty law User fee for water Custom taxes User fee for hunting on production use Wage tax (8.1.1 of License fee for the use of natural re- Gasoline taxes PIT law) sources other than minerals License fees for mining and Inheritance and User fee for herbs exploration of mineral resources gift tax Stamp duties other Water pollution fee than 11.2 of the stamp User fee for herbs duty law   Vehicle tax User fee for common minerals     User fee for drinking water and springs     Self-employment tax     Dog tax     User charges for waste Source: MOF & GDT 8 Emil M. Sunley, Jan Gottschalk, and Alistair Watson (2010). “Mongolia: The Fiscal Regime for Mining - A Way Forward.” IMF. 44 DOMESTIC REVENUE MOBILIZATION 2.2.2 Tax Policy Statutory low tax rate 7. While Mongolia’s statutory tax rates have been low by international comparison (Figure 2.1), government introduced a few rate revisions in the context of its fiscal consolidation. Business entities pay the CIT at 10% for annual taxable income of up to MNT 3.0 billion, and at 25% for income of more than MNT 3 billion. The CIT is the only tax that has differentiated rate. Since 2007 Mongolia has maintained a low statutory rate of 10% on PIT and VAT rate (10%), amongst the lowest in the world (Table 2.4). TABLE 2.4 International Comparison of Tax Rates (%)   Maximum VAT Maximum PIT PIT Types Customs duty Interest tax Mongolia 10 10 Uniform 5 0 Korea 10 38 Progressive Various 14-20 Indonesia 5-15 (10-75 for 30 Progressive 0-150 15-20 luxury goods) Malaysia 25 28 Progressive 0-50 15 China 17 45 Progressive 12.47 (0-100) 10 Thailand 7-10 35 Progressive 1-20 (60 for luxury) 10-15 Vietnam 35 35 Progressive 0-40 5 Chile 19 40 Progressive 6 35 Botswana 12 25 Progressive 0-25 10 Asia average 12 29 … … … Source: KPMG, PwC.  FIGURE 2.1 International Comparison of Tax Rates by Level of Income. Maximum PIT Rate (%) VAT Rate (%) PIT rate (%) VAT Rate 60 30 50 25 40 20 30 15 20 10 Mongolia 10 5 Mongolia 0 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000 0 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000 Per capital GNI (Atlas Method): S Per capital GNI (Atlas Method): S Source: KPMG, World Bank. 45 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 8. Mongolia has a fairly straight forward set of tax rates, plagued with several exemptions. Box 2.1 captures some features of the Mongolian tax policy before the changes introduced in 2018. BOX 2.1 Some Features of Tax Policy in Mongolia a. Business entities subject to income tax are classified as resident and nonresident. The corporate income tax law provides several exemptions. For example, income derived from sale of share of produce in the oil industry from a production sharing arrangement is tax exempt. A nonresident’s transfer of his or her share of tax exempt income from a production sharing agreement in the oil industry is exempt from withholding tax of 20%. However, tax exemptions and tax holidays for foreign investors have been phased out. b. Value added tax was introduced in 1998 and revised to 10% from 15% in 2007. VAT is levied on sales turnover that exceeds MNT 50 million. Exports are zero rated. The VAT regime, however, exempts equipment, raw materials, gasoline and diesel used in exploration, extraction and under production sharing agreement with the government in the oil industry. c. Thin capitalization interest is levied on that portion of debt when debt-equity ratio exceeds 3:1. d. Transfer pricing rule requires related business entities to charge one another the same price as unrelated entities on the sale of comparable goods and services. e. Tax loss can be carried forward to be deducted from taxable income in two future years, up to 50% of taxable income. f. Mongolia refunds sizable VAT revenue. Mongolia operates a VAT incentive scheme to promote compliance and broaden the VAT base. There are three kinds of refunds that are operated under the new VAT law introduced from January 1, 2016: (i) normal VAT refunds through input tax credit; (ii) up to 20% refund of VAT paid is made to consumers who upload e-receipts of VAT including purchases from formal sources (dealers with revenue greater than 50 million MNT); and (iii) lottery (applicable to all). In addition, Mongolia is grandfathering an incentive scheme under the old VAT law that provided VAT refunds up to 10% of investments made in installments. This could continue for up to five years to exhaust all claims. Together, VAT refunds can amount up to 30% of VAT revenue. g. Until 2017, Mongolia imposed zero tax on interest income on bank deposits. In 2017, only taxable deposits attracting a 10% tax were those larger than MNT 100 million with a term of less than 12 months, and the foregone interest tax was estimated to be over 0.2% of GDP. Effective 2018, the government has eliminated the threshold on interest income tax and introduced a flat 10% tax without reference to size of deposit or tenor. Imposing tax on interest income is consistent with the arrangement for other passive income, capital gains and dividends that are taxed at 10%. In addition to revenue implications, exempting interest income from taxation also skews the choice of financing towards debt, away from equity that is subject to tax. Substantial Tax Expenditures 9. Tax expenditures are a significant source of tax base erosion in Mongolia, mainly supported by VAT exemptions. Total tax expenditures in Mongolia are estimated as over 3% of GDP as of September 2017, which accounts for nearly 14% of end year total revenues (Table 2.5).9 9 Tax expenditure is as of September 2017 due to the end year data is not available. 46 DOMESTIC REVENUE MOBILIZATION TABLE 2.5. Tax Revenue versus Tax Expenditure (MNT billion)  Indicators 2014 2015 2016 2017e Total tax revenue 5157.0 5119.0 4996.8 6302.6 Total tax expenditure 720.8 656.8 908.3 874.0 % of Total tax revenue 14.0 12.8 18.2 13.9 Source: World Bank staff calculation based on the draft 2018 budget. Around 50% of VAT liabilities are exempted, a growing figure that will continue to compromise VAT refunds. A secondary effect of VAT exemptions is that it could further decrease VAT revenues if the exempted good is an intermediary good and breaks the VAT supply chain. Tax expenditures, due to VAT exemptions, are estimated to have decreased to 0.75% of GDP in 2017 from 1.4% of GDP in 2015 (Table 2.6). Direct investors and service providers in the mining sector are given tax incentives on profit and VAT exemptions, although this is not too unusual for a resource-rich developing economy. Additionally, a possible accelerator for revenue losses are exemptions and tax holidays-even beyond the mining sector. Exemptions from excise, customs and PIT laws represent the second and third largest sources for tax expenditures. TABLE 2.6 Total Tax Expenditure   GDP Ratio (%) Expenditure Ratio (%)   2014 2015 2016 2017 2014 2015 2016 2017e Total tax expenditure 3.24 2.84 3.79 3.22 100.0 100.0 100.0 100.0 1. CIT law 0.51 0.25 1.24 1.15 15.7 8.9 32.7 35.9 2. VAT law 1.50 1.35 1.05 0.75 46.3 47.7 27.6 23.3 3. Customs tariff and duty law 0.18 0.20 0.19 0.18 5.6 7.2 5.0 5.5 4. PIT tax law 0.39 0.44 0.51 0.44 12.0 15.6 13.4 13.8 5. Excise tax law 0.66 0.58 0.81 0.69 20.4 20.5 21.3 21.6 Source: World Bank staff calculation based on the draft 2018 budget. Note: GDP estimates are World Bank staff projections.  10. While tax expenditures are not necessarily a suboptimal policy instrument applicable in specific circumstances, an evaluation of their associated costs and benefits may be warranted. Tax expenditure should be compared with alternative policy instruments-either budgetary expenditures or regulation. However, the cost-benefit evaluation of tax expenditures is critical and should include all relevant dimensions while comparing the same to alternative policy instruments. Once this analysis is completed it will be easier to see whether there is scope to reduce tax expenditures. For instance, in income tax, profit tax holidays can also lead to profit shifting (tax planning and profit minimization) within Mongolian companies with multiple branches. On the VAT, exemptions should be revisited and even be replaced with zero rates to protect against cascading effects of breaking the VAT chain. Second, the 20% VAT refund on final consumers is an income transfer. It would be useful to estimate the amounts and determine if it makes sense both from a revenue and a policy side. Beyond the revenue forgone, tax expenditures can also have other types of costs or secondary effects. Examples include: (a) the administrative cost to tax administration, (b) compliance costs for taxpayers, (c) macroeconomic effects, (d) implications for economic efficiency, (e) horizontal and vertical equity, and (f) the effects of fraud and abuse since badly designed tax exemptions/incentives may open avenues for tax evasion/avoidance. 47 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Frequent Tax Amnesty Schemes 11. Mongolia has offered periodic tax amnesty schemes that may be counterproductive. Tax amnesty laws were passed in 2007, 2008 and 2015. Policy-makers often view amnesty programs as an efficient and equitable policy that may even become an additional source of revenue in the short term. However, the IMF has noted that the perceived benefits of tax amnesty programs could be overstated and are often unlikely to exceed the program costs, which are rarely measured.10 Also, according to the IMF, ‘successful’ tax amnesties are the exception rather than the norm as, over time, net revenue collection and compliance are often negatively affected by amnesty programs. 12. Under these periodic schemes, filers were given amnesty without paying any back taxes, penalties, or interest, or even having to declare the source of the income. The 2007 amnesty provided a one-time tax forgiveness from any liabilities, tax and social insurance contributions if individuals and legal entities voluntarily registered as taxpayers or declared their taxable income and other taxable goods, services and operations. The second Tax Amnesty Law in 2008 provided a one-time amnesty where some legal entities were exempted from certain imposed tax and social insurance liabilities, criminal and administrative sanctions. Following these laws 4,621 taxpayers sent 8,145 reports and 748 individuals and entities registered as new taxpayers. Moreover, MNT 431.6 billion of taxes from MNT 4.5 trillion of reported revenue were recovered. 13. The third tax amnesty law- the Economic Transparency Law - was effective between August 2015 and February 2016. The Law provided a one-time waiver of legal penalties and tax liabilities for individuals and legal entities, who voluntarily declared hidden/unreported immovable and movable assets or revenue by registering them and filing tax returns and social insurance reports. As of February 20, 2016, a total of MNT 33.3 trillion of revenue was reported while MNT 7.4 trillion of taxes were waived. 2.2.3 Tax Administration Leveraging Lessons of the Current VAT System 14. Since its approval in 1998, the VAT law has been revised three times and The Government of Mongolia adopted a new VAT law effective on January 1, 2016. Within the legal reform framework, the revision of the VAT Law aimed to create the regulatory and technical framework for online registration to sell goods, works and services and with holders of VAT to expand the tax base11. According to the General Taxation Law (GTL), each taxpayer is obliged to use a Point of Sale (POS) machine that is consistent with Mongolian standards, which shares basic sales information automatically with the tax administration. To encourage citizen participation, the law mandated refund of 20% of the tax paid by the consumer based on the payment receipts printed by the POS/cashier machines. Rollout of the new VAT Law was supported by an internet-based electronic receipt system called e-BARIMT. The government’s VAT promotion scheme to improve VAT compliance has two parts including introducing the above- mentioned POS machines at various retail outlets for business-to-consumer (B2C) transactions, and an e-invoice system for business-to-business (B2B) transactions.12 As a result of the revised VAT Law, the number of business entities that are registered into the E-Barimt system has been increasing (Figure 2.2) which in turn expands the tax base. 10 Tax Amnesties, Katherine Baer and Eric Le Borgne, IMF Working Paper, 2007. 11 As a result of increasing the VAT withholder threshold to MNT 50 million, 45% of total taxpayers were released from obligations such as submitting VAT report (return), maintaining records, collecting receipts and reporting etc. 12 This system generates a unique QR code and a lottery code on all customer receipts. Consumers who scan the QR code and upload the receipts into the e-BARIMT system are eligible for a 20% refund of VAT and become eligible for a lottery. This system is now installed in 36,900 of the 59,814 estab- lishments generating e-invoices. 48 DOMESTIC REVENUE MOBILIZATION FIGURE 2.2: VAT Return Filers (number of businesses during 2016-2018) 16 16 16 16 6 6 6 6 7 18 5 01 01 01 01 01 01 20 20 20 20 20 -2 -2 l-2 -2 -2 -2 n- n- n- n- n- b eb ec ec -Ju ec Ja -Ja -Ja -Ja Fe -Ju -D -D -F -D 16 7- 1 17 24 31 15 14 31 31 31 Source: MOF, GDT 15. The main VAT priority is to continue the rollout of the electronic receipt system and to cover the approximately 22 informal markets. The GDT must be vigilant to protect against sales suppression fraud, to which the POS terminals are vulnerable. Experience suggests that these devices by themselves provide limited assistance in alleviating some of the pre-existing compliance problems. There is evidence, however, that Electronic Fiscal Devices implemented as part of a comprehensive compliance improvement strategy produced positive results, both in terms of additional revenue and improved taxpayer behavior. The introduction of fiscal devices presents opportunities for the tax administration to rethink its approach to business processes, not only by automating the collection of information, but also by leveraging arrangements to improve compliance. The GDT should also use the data gathered by e-BARIMT to improve compliance in VAT and Corporate Income Tax (CIT). Limited Capacity for Enforcing Compliance of Large Tax Operations 16. The challenge is to further analyze the large taxpayer segment to strike the right balance between the number of taxpayers allocated to the Large Taxpayers’ Unit (LTU) and its share in total revenue collected. Increased focus on the large taxpayer segment13 has become a key characteristic of modern tax administration given, (a) their importance from a revenue perspective14 ; (b) their engagement in large-scale, complex, specialized, and often international operations; (c) associated high revenue compliance risks; and (d) their role as a significant source of third-party information (and withholding)15. In Mongolia, typically two main criteria are used to There are a number of different criteria to consider taxpayers in this segment, which include the size of actual taxes paid, turnover, physical assets, con- 13 solidated group, or staff employed by the company. Normally, the allocation of taxpayers to this segment is based on a combination of different criteria paying special attention to turnover. According to IMF. 2011. Revenue Mobilization in Developing Countries. On average, large taxpayers can secure 60-80% of domestic taxes (more in island 14 economies). In Mongolia, the ratio of large tax payer revenue to total revenue ratio is 90%. 15 Bird, Richard. 2010. Organizing to Tax. The importance of tightening control over this segment derives not only from the amount of revenue that is collected–directly and indirectly through withholding arrangements–but also from the significant amount of third-party information that is received and transmitted to the tax administration. This information provides the foundation of the compliance risk management system that all tax audit strategy should be based on. 49 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue assign entities to the LTU: (1) companies reporting over MNT 5 billion turnover; (2) commercial banks and businesses that entered into investment agreements or production sharing arrangements with the government (that is, mining and oil companies), and cellphone companies that are of national strategic importance. The LTU does not consolidate financial statements of subsidiaries of conglomerates. As of 2015, there were 36 registered large taxpayers that represented around 90% of total tax revenue. Companies in the LTU come from the mining, banking, construction and trading sectors. Mining sector-related companies dominate in terms of revenue contributing nearly 40% of CIT in 2015 and 81% of all revenue. The LTU has 83 staff assigned and the control or audit division also has specialist industries such as construction or mining. 17. The GDT segments taxpayers in order to control compliance and provide services to taxpayers but this arrangement is not effective. Taxpayers are segmented into large, medium, small and micro categories and compliance strategies are differentiated. An Annual Work Plan is produced that details control measures and the third-party information to be used for tax audit purposes. Roughly fifty% of the audit work is planned based on risk parameters and is system-selected. The other fifty% is based on external information. The findings are shared with the Tax Inspection Methodology Department that is responsible for tax audit. Going forward, GDT may contemplate creation of a Tax Intelligence Unit under the Tax Audit Department, which would help identify taxpayer compliance risks and tax fraud patterns. 18. The taxpayer register is the bedrock of tax administration. In Mongolia, the GDT does not register companies. The registration is carried out under Company Law assigning a unique seven-digit ID that is also used by the GDT. The GDT registers individual taxpayers, assigning a unique 10-digit alphanumeric ID. The registration of companies poses unique problems to the GDT. The quality of the data in the tax registration systems is poor. This is mainly due to non-uniqueness of the Taxpayer Identification Numbers (TINs), for instance a business can have multiple TINs, errors in the data in General Department for State Registration databases, weaknesses in the interfaces between the tax registration systems and other registries, and excessive use of manual data entry with weak controls. Large entities do not always register with the GDT, or are exempted from paying taxes under bilateral and concession arrangements. Registered entities can also become inactive. The entire registration process is manual. There is a need to improve coordination between GDT and the Registration Authority to provide for seamless sharing of information on registrants. There is also a need to ensure that applications for registration are authentic. Additionally, initiatives may be undertaken to detect unregistered businesses and individuals, especially those representing high revenue risks. 19. Problems in the quality of the taxpayer data compromise the effectiveness of tax audit, as does the limited number of tax inspectors to conduct the audit. The risk management system in the GDT forms the basis of the annual audit plan. It uses tax data, administrative and financial data on taxpayers from other sources and a software program based on 120 parameters to determine the risk ratings of taxpayers. These rates range from ‘High’ to ‘Low’. High risk taxpayers are required to be audited every other year, while low risk taxpayers are required to be audited once every five years, and the duration from the last audit is one of the 120 criteria used in the risk management system. The Department of Tax Inspection Methodology is in charge of desk and comprehensive audit. The Taxpayer Communication Department does the mass audit and returns to the Taxpayer Tax Returns with discrepancy or inconsistency. No sanctions are imposed at the level of mass audit. The LTU is in charge of audit of the large taxpayers. 50 DOMESTIC REVENUE MOBILIZATION 20. The GDT was unable to provide a full breakdown of the various categories of tax audits carried out in 2015. Audits are a key tool used by tax administrations to increase compliance. They are used to: (a) detect and redress individual cases of non-compliance; (b) promote voluntary compliance by increasing the probability of detection and penalties for non-compliant taxpayers; and (c) gather information on both the health of the tax system and the evasion techniques used by taxpayers16. Additionally, audits provide a good opportunity for the tax administration to educate taxpayers on their legal obligations or bookkeeping requirements, thereby improving future compliance. There are multiple types of audits available to tax administration ranging from registration checks, record keeping audits, and single-issue audits to comprehensive (full) audits and fraud and criminal investigations. In 2015, audits helped realize an additional tax assessment of MNT 68 million from large taxpayers and MNT 13 million from other taxpayers. While 131 tax audits that targeted the large taxpayer segment were conducted, 4,709 tax audits targeted other taxpayer segments. Issues with Tax Arrears, Audit, Refunds and Disputes 21. In the GDT’s case, outstanding tax arrears in the last three years have been about a quarter of the net tax revenue collected. Tax arrears are of two types: those that are based on the self-assessment by the taxpayers (that is, the difference between the assessed tax liabilities and those that were paid) and those resulting from tax audits. Most debt is up to 12 months overdue. Debts over two years old account for just 10-2015% of total debt, but that number is increasing. The GDT has debt collection targets of 90% for the self-assessed arrears and 85% from the arrears from audits. 22. Despite the existence of the Dispute Resolution Councils (DRC), presence of an independent dispute resolution mechanism to resolve taxpayers’ claims and to ensure a fair application of the tax system is still lacking. A dispute resolution system is a necessary corollary of a tax collection process that hinges on voluntary compliance and self-assessment. In Mongolia, Dispute Resolution Councils (DRCs) are the first point of complaint. Tax disputes are resolved within 60 days on average at the tax DRC and 60 days at the court from the date of receiving the complaints/claims, respectively. However, the DRC is not an independent body but reports to the commissioner of the GDT, which has the authority to override the council’s decisions. Second, the membership of the council is heavily skewed toward government officials. Mongolia should consider providing some autonomy to DRCs and setting up independent tax tribunals as a prior step to resolving appeals through the courts. The GDT may also consider putting in place a system of mediation for small and medium taxpayers. Unlike the normal legal system, where case law and precedents are referred to and where taxpayers in similar circumstances are, or should be treated the same, the outcome under mediation may well differ from case to case. To ensure the quality of the mediation process, rather than employing a decentralized system involving various small tax offices, one option to consider is mediation placed at the head office level. 23. A tax arrears management strategy should be developed to establish priorities for arrears collection and set performance targets. Chapter 8 of the 2008 General Taxation Law (GTL) (Articles 31 to 70) and Chapter 9 of the draft 2016 GTL (Articles 40 to 56) deal with various provisions to recover tax arrears and debt. Plans should distinguish between different categories of arrears, considering the size, nature, and age of the debts and give priority to large and more recent arrears since the latter are tax debts more likely to be collected. Specialization in arrears collection needs to be developed, including establishment of a dedicated debt collection function. Information technology (IT) applications should be developed to support debtor profiling and determine optimal collection strategies based on risk assessments. GDT can also recover tax arrears from property, salary, receivables, dividends, World Bank. 2010. Risk-Based Tax Audits: Approaches and Country Experiences. 16 51 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue owned shares and other income. Tax debts can also be settled in installments. The GDT has the power to freeze/close bank accounts and sell collateralized assets, but it has no power to write-off debt. It might be appropriate to consider providing the GDT the powers to write off uncollectible arrears. The GDT may also consider provisions that withhold government payments to tax defaulters to offset the tax due. In addition, it may use differentiated strategies for different categories of taxpayers. 24. Many countries are devising methods to detect refund fraud using intelligence, risk analysis, risk profiling, and data matching. Mongolia processes about MNT 100 billion of VAT refunds including normal VAT refunds and the fiscal lottery. In 2015, in Personal Income tax there were 73,719 refund requests for MNT 44.5 million. Of these, 63,122 requests were processed for MNT 39.6 million. No data was available for refunds in CIT. VAT refunds are administered by the Revenue Department in the Ministry of Finance after initial processing by the GDT but refunds are capped at 30% of the VAT collections annually. Audit verification action will normally follow the lines of a conventional desk audit/verification (in-office), although the activity may be restricted to the refund in question. The credibility of the business operation itself should be examined, together with the documentation used to support the claim. Procedures for refund fraud detection must be put in place to minimize refund risks. 2.2.4 Other Challenges 25. The international tax system may have potential income tax leakage. Potential channels include transfer mispricing, profit shifting through the use of interest deductions (thin capitalization), and treaty shopping by foreign multinationals operating either through subsidiaries, permanent establishments, or joint ventures in Mongolia. Two known legislative weaknesses facilitate this. First, the definition of related parties in the income tax law may not be clear, which allows companies to avoid being deemed a related party. Second, transfer pricing legislation and implementing procedures may need revision and strengthening, especially considering the recent upgrading of transfer pricing rules globally. 26. Mongolia’s GTL is currently missing a GAAR provision that is now standard in most countries’ tax legislation.17 Presently the Administration and Cooperation Department of the GDT provides guidance on GAAR-like cases and oral and written guidance on the implications is issued to the companies concerned. However, this is inadequate to provide the general and specific provisions for tax avoidance. First, tax avoidance needs to be defined and differentiated from tax evasion.18 Tax avoidance is the legal use of the tax laws to reduce tax liability while tax evasion is an illegal activity that involves underreporting or concealing sources of taxable income or taxable activities. Both tax evasion and tax avoidance are forms of noncompliance with tax laws. In this regard, globalization of economic and financial activities, the growth of multinational businesses and cross-border ownership of companies are placing increasing strain on domestic tax systems and challenging existing consensus on international tax rules. 17 GOM is now considering introducing a GAAR seeking to improve horizontal equity and fairness in implementing the tax system; ensuring compliance with BEPS standards and extending the information exchange obligations for third-parties for tax purposes. Australia has a GAAR that dates back to 1915. Recent GAARs to mention are China (2008), Indonesia (2008), Belgium (2012), and India (2017). The USA does not have a proper GAAR. In 2010, the USA codified the economic substance doctrine and one could view this as something similar to a GAAR provision. 18 GAAR rising Mapping tax enforcement´s evolution, ERNST & YOUNG, 2013. There are two major categories of anti-avoidance rules: general and spe- cific. A general GAAR is a set of general rules and principles to provide the tax authorities with mechanisms to deny tax benefits in those operations or contracts with ‘no economic substance’, which are only intended to reduce tax burden or to access tax benefits. On the other hand, the specific GAARs target schemes or transactions, which are potentially abusive. 52 DOMESTIC REVENUE MOBILIZATION 27. Double tax treaty policy, particularly considering the recent cancellation of four treaties, may also have an impact on investment and revenues. Mongolia has entered Double Taxation Avoidance and exchange of tax information treaties with 28 countries, beginning 1993. However, since 2014, treaties have been withdrawn with the United Arab Emirates, Kuwait, Luxembourg and the Netherlands. Some of the treaties were conduits for aggressive tax planning to avoid paying profit tax in Mongolia. On one hand, canceling these treaties could have contributed to weaker FDI, but it also should have resulted in higher revenues as cancelling the treaties would have resulted in higher withholding taxes for the operations still in Mongolia, especially on passive income-dividends and royalties. On double tax treaties, a thorough review of the treaty network and applicability would help identify if an abuse of such treaties allows for profit shifting to low-tax jurisdictions. Some work has begun - under the Mongolian and Japan trade agreement, to review the existing contracts to identify the benefits for Mongolia. Examining the effectiveness of related party definitions, transfer pricing instruments, thin capitalization, and double tax treaty applications (especially in view of anti-abuse legislation in domestic law) would be useful to determine whether a policy upgrade is needed to strengthen the revenue effects of related party transactions. 28. Mongolia has constituted a working group to draft a Tax Procedures Code. While a unified tax code replacing separate legislation for each type of tax is desirable, it may not be politically feasible. In such cases, the international best practice is to extract the administrative provisions from each of the tax laws and consolidate them in a single, uniform tax procedures code, separate from the substantive pieces of tax legislation. A uniform tax procedures code enhances transparency, clarity, and consistency in the application of the country’s tax laws. Recently, countries such as Indonesia, New Zealand, Montenegro, and Serbia have implemented Tax Procedures Codes (Acts) successfully; and other countries are currently developing them, including South Africa, Namibia, Ghana, Malawi, Tanzania, Botswana, and Turkey. Impact of Recent Measures and Distributional Analysis 29. A distributional analysis of selected tax and transfer policies following Lustig (2016)19 was carried out to estimate the distributive impact of taxes and transfers using Mongolia’s Household Socio- Economic Survey for year 2016 (Figure 2.3).20 The analysis assessed the distributional impact of PIT, VAT and excise taxes as well as social security contributions (SSC) (on the revenue side), monetary transfers such as social insurance-assistance programs, and in-kind transfers in public health and education (on the expenditure side). Also included were quasi-fiscal expenditures due to subsidized interest rates for housing mortgage or energy prices.21 The key findings are described below. 30. The tax and transfer system in Mongolia shows relative progressivity. When adding taxes paid and transfers received, the net effect in terms of share of pre-market income is higher the poorer the per-capita market income. Transfers received, net of taxes paid, represent more than 150% of market incomes for those in the second decile of the income distribution.22 N. Lustig (2016) “Commitment to Equity Handbook. A guide to estimating the impact of fiscal policy on Inequality and Poverty” Tulane University. 19 Rodriguez S. and J. Yang (2018). “Mongolia: Distributional impact of taxes and transfers”. Background paper of the PER. 20 21 The fiscal policies included in the simulation exercise represent about 50% in central government fiscal revenues, and 42% of central government ex- penditures. The tax-benefit incidence simulations assume that Social Security contributions are fully paid by employees, personal income tax by income recipient and VAT/excise taxes by the consumer. Assumptions about tax evasion are excluded. In addition, administrative and capital expenditures are reflected in the redistribution assumption about in-kind transfers (namely, public education and health). 22 The first decile is not included in the figure because net transfers are much too high for this group due to individuals with nearly no income before taxes and transfers. 53 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue This percentage declines to a little more than 50% for those in the third deciles and keeps declining to represent -2 and -7% for those in the top two richest deciles of the distribution because of taxes paid being larger than transfers received. Most taxes are collected from richer households. The top three deciles of the distribution are responsible for 67% of PIT and SSC, 54% of the VAT and 44% of excise tax collections. VAT subtracts a larger share of market incomes among the poor, whereas income tax does it among the richer deciles. In monetary terms, all deciles of the distribution get more in transfers than what they pay in taxes, except the two richest deciles funded partly by corporate taxes, royalties and other government revenue. The Gini coefficient of income declines from 0.4183 for market income (that is incomes before taxes or transfers) to 0.3752 for consumable income (that is monetary income that can be spent in goods and services). After including in-kind transfers (that is, public education and health), final income Gini coefficient descends to 0.3507. 31. From the tax side, impacts on inequality are diverse but limited. Direct taxes (PIT and SSC) are slightly progressive because the share of these taxes paid by high income groups is even higher than the market income they concentrate. On the other hand, VAT and Excise taxes are slightly regressive because the share paid by poorer groups is larger than their respective income share. VAT and excise taxes increase inequality measured by the Gini coefficient by 1 percentage point. On the other hand, PIT and SSC reduce the Gini coefficient of final income by 0.7 and 0.4 percentage points, respectively. These results are compatible with the usual finding that taxation is not the main driver of fiscal policy redistribution effects. In the case of Mongolia, PIT, SSC, VAT and excise taxes represent a small portion of fiscal revenues, and hence of public expenditures, due to the large influence of corporate and royalties’ taxation that characterize mineral commodity- based economies. The redistributive impact of fiscal policy is mostly undertaken through public expenditures. FIGURE 2.3: Lorenz Curve for Taxes and Contribution in Mongolia (2016) 45-degree Line Market Income PIT Deductions Income Taxes Social Security Contributions (Employee) VAT Excise Tax 100 90 80 70 60 50 40 30 20 10 0 1 11 21 31 41 51 61 71 81 91 101 Source: Rodriguez and Yang (2018). 54 DOMESTIC REVENUE MOBILIZATION 32. Considering the likelihood of regressive tax incidence in a PIT flat rate, the GOM transformed the PIT system into a progressive system from 2018 but reversed it immediately amid social and political pressures. The public consultations for the introduction of new tax incentive and exemption measures envisaged by government were completed in March 2018. Box 2.2 below presents the new tax measures. The Government has introduced a progressive earned income tax credit (EITC) system that puts more credit in the lower income brackets (Table 2.7). But without accompanying progressivity in income tax the overall impact on PIT collection of this measure is likely to be negative. The earlier taxable income threshold was set in 2006. The current annual minimum wage is MNT 2,880,000. But as the lowest income tax slab has been recast, technically those at or below the minimum wage will have to pay some PIT. The break-even wage with no PIT would be (MNT 1,600,000). Increasing the Earned Income Tax Credit (EITC) lowers the tax burden for lower income groups, and mitigates the tax burden of middle income groups to some extent that face higher marginal income tax rate. But the lowest slab design will need to be re-worked to accommodate the poor and the EITC must be linked to minimum wages. Simulation results by the World Bank indicate that the tax incidence could improve with a progressive tax system as compared to a flat PIT rate (Table 2.8 and Table 2.9). Tax burden for the lower and middle-class income groups (about 95% of salaried workers) is expected to be lower. Results could vary depending on the EITC and the tax bracket. The PIT system would need to focus on rebalancing the tax incidence, which is currently likely to put relatively higher tax burden on lower income groups. The 10% flat PIT rate is only slightly progressive in relative terms, and collects only 9% of government revenues (much less than the share of the VAT and Social Security Contributions, which are 20% each). There seems to be room for making this revenue source more effective in terms of revenue collection and more progressive in terms of redistribution. TABLE 2.7 New EITC Schedule Taxable Annual Income, MNT Tax Credit, MNT 2018 2019 onwards 0-6,000,000 160,000 240,000 6,000,000-12,000,000 140,000 220,000 12,000,000-18,000,000 120,000 200,000 18,000,000-24,000,000 100,000 180,000 24,000,000-30,000,000 80,000 160,000 30,000,000-36,000,000 60,000 140,000 36,000,000 or More - - Source: The 2018 budget. 55 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue TABLE 2.8 Adopting a Progressive PIT System - Simulation Results PIT Tax Base New PIT system Current PIT system Estimated Estimated Number % Share Income % share % share Annual Salary New Tax PIT burden Tax PIT burden of # of share of PIT of PIT MNT Bracket (MNT Bracket (MNT workers workers (%) burden burden bullion billion) 2,880,000 >= 346,195 39.5 9 8% 9 1.5 10% 25.8 4.8 2,880,001 - 6,000,000 169,897 19.4 12.4 10% 48.5 8.3 10% 61.1 11.4 6,000,001 - 12,000,000 241,430 27.5 34.6 10% 171.7 29.3 10% 189.7 35.5 12,000,001 - 18,000,000 72,178 8.2 17.1 10% 92.2 15.7 10% 97.6 18.3 18,000,001 - 24,000,000 24,042 2.7 8.2 15% 49 8.4 10% 47.6 8.9 24,000,001 - 30,000,000 9,437 1.1 4.1 15% 27.7 4.7 10% 24.3 4.6 30,000,000 < 14,184 1.6 14.7 30% 188.2 32.1 10% 87.8 16.5 Total 877,363 100 100 ... 586.3 100 ... 533.9 100 Earned income Tax Credit Earned income Tax Credit (MNT): 1,400 (MNT): 84,000 Source: MOF (PIT Tax Base Data), World Bank estimates TABLE 2.9 Adopting a Progressive PIT System - Alternative Scenario PIT Tax Base (2015) New PIT system Current PIT system Estimated Estimated Number % Share Income % share % share Annual Salary New Tax PIT burden Tax PIT burden of # of share of PIT of PIT MNT Bracket (MNT Bracket (MNT workers workers (%) burden burden bullion billion) 2,880,000 >= 346,195 39.5 9.0 8% 4.3 0.9 10% 20.3 4.3 2,880,001 - 6,000,000 169,897 19.4 12.4 10% 41.8 8.3 10% 53.6 11.3 6,000,001 - 12,000,000 241,430 27.5 34.6 10% 151.9 30.3 10% 168.7 35.7 12,000,001 - 18,000,000 72,178 8.2 17.1 10% 82.2 16.4 10% 87.2 18.4 18,000,001 - 24,000,000 24,042 2.7 8.2 15% 43.8 8.7 10% 42.6 9.0 24,000,001 - 30,000,000 9,437 1.1 4.1 20% 25.9 5.2 10% 21.8 4.6 30,000,000 < 14,184 1.6 14.7 25% 152.2 30.3 10% 78.9 16.7 Total 877,363 100.0 100.0 ... 502.1 100.0 ... 473.1 100.0 Income Tax Credit (MNT): Income Tax Credit (MNT): 101.640 84,000 Source: MOF (PIT Tax Base Data), World Bank estimates 56 DOMESTIC REVENUE MOBILIZATION BOX 2.2 A Summary of Proposed changes to the Mongolia’s Tax System by the Government in 2018 1. Personal income tax law • Setting a flat rate of 10% on employment income; presumptive tax when income is undetermined. • Annual tax credit from employment income increased to MNT 120,000 from MNT 84,000. • Interest subsidy equivalent to interest differential between commercial mortgage loan rate and Government mortgage program. • Increasing the tax credit amount on first-time housing from MNT 3 million to MNT 6 million. • Tax on capital gains from the transfer of mining license and land rights. • Broadening the definition of resident taxpayers. • Introducing the controlled foreign company rule. 2. Corporate income tax law • Increasing the threshold for upper tax rate from MNT 3 billion to MNT 6 billion. • Simplified corporate tax regime for small businesses (1% tax on gross turnover below MNT 50 million). • Decreases on nonresident withholding tax from 20% to 15%; on interest income from debt instruments from 10% to 5%; withholding tax on dividends from 20% to 10%. • Tax rate on transfer of rights decreased from 30% to 10%; tax on capital gains from the transfer of mining license and land rights. • Improving the permanent establishment related tax rules; earnings stripping rule; controlled foreign company rule and introducing ring-fencing rule on mining license holders; broadening the definition of related party for transfer pricing. • Allowing foreign tax credit; Increasing the period of loss carry forward to 5 years. • Reporting period of tax return reduced to quarterly for above MNT 6 billion entities semi- annually for below MNT 6 billion entities, and annually for entities subject to the simplified tax regime. 3. General tax law - important changes • Introducing the GAAR rule; risk management and risk-based auditing. • Ensuring compliance of standards and requirements of BEPS and Global Forum on Transparency and Information Exchange for Tax Purposes. • Extending the information exchange obligations for third-parties for tax purposes. • Requiring taxpayers to deposit 10% of payment to file a claim to the Dispute Resolution Committee. • No statute of limitation on tax fraud; strengthening tax authority’s rights on collecting tax independently; pre-collection and pre-seizure of assets. • Increasing the deferral of tax arrears from 2 months to 12 months; taxpayers to amend tax returns up to 1 year. • Introducing tax obligation rule for secondary party if assets are transferred to avoid tax obligation. 57 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 2.3 Priority Areas of Reform 33. This section presents potential areas for future reforms in the areas of tax policy, tax administration and international taxation. Tax Policy  Continue to explore the possibility of transforming the current PIT system to some form of a progressive system by introducing a graduated tax rate schedule at least for the upper echelons of tax payers and reassessing the VAT refund system.  Avoid establishing tax amnesty schemes and consider alternatives such as payment installment agreements, voluntary disclosure programs, simplification of the tax policy regimen, and strengthening of the legal framework.  Broaden the tax base by progressively removing distortionary tax exemptions/incentives by initiating a process of comprehensive and systematic evaluation (cost-benefit analysis) of tax expenditures; enhance transparency in the process of establishing tax expenditures; introduce fiscal rules and termination dates (sunset clause); and review the conditions in place to access specific benefits.  Conduct a thorough review of international aspects of tax laws including effectiveness of related party definitions, updating of transfer pricing legislation, transfer pricing instruments, thin capitalization, and double tax treaty application.  Include a GAAR in the Mongolian legislation to better fight against tax avoidance arrangements. Tax Administration VAT Implementation  Progressively phase out the 20% refund of VAT paid, which is applicable to consumers that upload their receipts into the e-BARIMT VAT system.  Adopt prevention and mitigation measures against electronic sales suppression techniques in POS machines. These measures include developing a strategy for tackling electronic sales suppression within tax administration’s overall approach to tax compliance and a communications program. Organizational Structure  The tax administration needs to align its organizational structure to a modern IT-driven organization. Large Taxpayer Unit  Consolidate financial statements of conglomerates and introduce requirement to file a consolidated tax return instead of individual tax returns for individual companies.  Establish formal criteria other than turnover to assign taxpayers to theLTU. Compliance Risk Management  Create a tax intelligence unit under the Tax Audit Department.  Improve quality of taxpayer data to facilitate a better implementation of risk assessment framework for tax audit. Tax Procedures Code  Extract the administrative provisions from each of the tax laws and consolidate them in a single, uniform tax procedures code separate from the substantive pieces of tax legislation. 58 DOMESTIC REVENUE MOBILIZATION Taxpayer Registry, Taxpayer Account, and Collection  Improve coordination between the tax administration and the registration authority to provide for seamless sharing of information on registrants.  Carry out proof of identify checks to prevent bogus entities from registering.  Keep the database clean of inactive taxpayers, invalid, and duplicate registration cases. Tax Arrears Management  Provide the tax administration with powers to write off uncollectible arrears.  Consider withholding government payments to tax defaulters to offset the tax due.  Use differentiated strategies for different categories of tax arrears and taxpayers.  Continuously monitor age of tax arrears in view of the decreasing probability of collecting debt with time. Tax Audit  The tax administration should carry out a deep assessment of its audit function’s strengths and weaknesses using international good practices. Tax Refunds  Improve the information regarding the stock of tax refunds, its evolution, and average time to resolve especially when it comes to VAT refunds. Tax Disputes  Consider setting up independent tax tribunals as a prior step to resolving appeals through the courts.  Consider setting up a mediation and dispute resolution mechanism for small taxpayers.  Put in place safeguards against frivolous appeals.  Consider providing some autonomy to DRCs.  Emphasize training on mineral taxation. 59 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue CHAPTER 3 Aggregate Expenditure Trends and Cross - Cutting Issues OVERVIEW AND CONCLUSION Expenditure Performance - Trends and Composition Key Highlights Public Sector Wage bill  Mongolia has been spending too much for Key Highlights its revenue capacity. Government spending  Mongolia’s civil service is currently facing increased threefold over the last seven years a number of challenges that could affect despite weak revenue performance. the credibility, technical competence and  Public spending peaked at 40% of GDP in 2016 effectiveness of public administration. (32% of GDP in 2013).  In Mongolia, the number of civil servants  The key drivers for the spending surge include: per capita is higher than in other East Asian (a) pro-cyclical fiscal management during the countries. mining boom in 2011-2012, (b) expansionary Elements of a Priority Reform Agenda policies during the economic downturn in 2013-2014, and (c) an abrupt and unbudgeted Re-establishing meritocratic principles in spending increase in 2016. recruitment and promotion to limit patronage- driven recruitment and reduce the current Elements of a Priority Reform Agenda turnover in the civil service and expansion in the  Reconfirm the commitment to the country’s number of civil servants. fiscal consolidation program by reducing  Simplifying the pay structure and rationalizing inefficient public spending and strengthening supplementary payments following a detailed revenue mobilization. analysis of the existing pay and grading  Improve functional classification of budget scheme. data by leveraging the BOOST platform in  Rolling out centralization of personnel and order to strengthen the analytical capacity for payroll systems to increase the control and better policy formulation (see Box 3.1). improve the management of the largest expenditure and improve data availability.  Leveraging the recent approval of the new Civil Service Law to accelerate the above reforms. 60 PUBLIC EXPENDITURE REVIEW Public Investment Efficiency Elements of a Priority Reform Agenda Key Highlights  Developing a comprehensive and consolidated approach to PIM.  Line ministries developed and executed unsustainable volumes of poorly appraised and  Clarifying roles and responsibilities. costly public investment projects through extra-  Developing a PIM database. budgetary funding.  Rationalizing the ongoing PIP portfolio and  Direct lending by ministries, departments, and prioritizing the completion of ongoing projects. agencies (MDAs) and local governments through promissory notes and a system known as Build-  Managing existing BT activity and strengthening Transfer (BT) have been prominent means of BOT/PPP management. funding new projects.  Upgrading the allocation of maintenance budget.  The Development Bank of Mongolia (DBM) has up until recently financed significant amounts of projects as well that would normally be considered More specifically, the most important issue is to create as public investment projects. institutional structures and operational procedures  Despite mounting fiscal pressures, the GOM that will enable the government to centralize the review continues to prepare large public investment of public investment projects, independently from the projects with seemingly little consideration of source of financing, and to use modern techniques of their need, appropriateness or affordability. project selection, appraisal and monitoring to not only reduce the size but also improve the quality of the  There is an unknown amount of operational and country’s capital expenditures. maintenance costs built into future budgets since recurrent costs associated with new projects are not considered in the preparation phase. 61 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 3.1 Expenditure Performance 1. Mongolia’s public spending increased sharply in 2011-2016 at the time the revenue performance was weakening.23 After taking a plunge in 2008-2010 during the global financial crisis, public spending increased from 31.6% of GDP in 2010 to an average of 39.2% of GDP in 2013-2014. Initially, that increase was fueled by double-digit growth in 2011-2012 caused by large FDI inflows and strong commodity markets, but the spending increase far exceeded the pace of revenue growth. In 2013-2014, the government revenue performance weakened but the increase in public spending was accelerated to prop up growth against external shocks caused by a sharp drop in FDI inflows and deteriorating terms of trade. In 2015, the government tightened its fiscal policy to curb expenditure in the face of weak revenue and rising debt. The fiscal adjustment effort, however, was quickly reversed in 2016, in the run-up to the election. Public spending increased from 38.3% of GDP in 2014 to 40.5% in 2016, while revenue continued to decline during three consecutive years (Figure 3.1). FIGURE 3.1 Expenditures Increased Significantly in 2011-2016 Aggregate revenue and expenditure trend in 2011-2016 (% of GDP): 2003-2017 45 Total revenues and grants 40 Total expenditures 35 30 25 20 15 10 5 0 03 004 005 006 007 008 009 010 011 012 013 014 015 016 017 20 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Source: MOF data, World Bank staff estimates. FIGURE 3.2 Recurrent and Capital Expenditures Mostly Driven by a Surge In Recurrent vs Capital Expenditure (% of GDP): 2003-2017 30.0 Recurrent Expenditures 25.0 Capital Expenditures incl DBM spending (Off-Budget) 20.0 15.0 10.0 5.0 0.0 03 004 005 006 007 008 009 010 011 012 013 014 015 016 017 20 2 2 2 2 2 2 2 2 2 2 2 2 2 2 Source: MOF data, World Bank staff estimates. uuu 23 In the absence of reliable micro-fiscal data for a functional classification of public expenditures, the analysis in this report is limited to the economic classification. 62 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES BOX 3.1 The BOOST Initiative Could Improve the Analysis of Budget Data The BOOST program is a World Bank collaborative effort that was launched in 2010 to provide quality access to budget data. The initiative aimed to make well-classified and highly disaggregated budget data available for policy makers and practitioners in government, researchers, and civil society and promote their effective use for improved budgetary decision making, analysis, transparency, and accountability. Since its launch, the program centered its activities around three work streams: (a) supporting expenditure analysis, (b) advancing fiscal transparency and public dissemination of budget data, and (c) improving public financial management (PFM) processes and systems. The program has designed and delivered more than 60 national and subnational BOOST data sets in standardized formats, whose contents are country specific. Each data set typically allows for approved, revised, and executed budgets to be cross-referenced across years with categories such as, • Government levels (central or local); • Administrative units (ministries, departments, agencies, schools, hospitals, and so forth); • Subnational authorities (districts, municipalities, and other local government units); • Economic classification categories (staff salaries, procurement of goods, and so forth); • Sources of funding (budget funds, off-budget funds, external finance, and so forth); and • Budget programs (if the country uses a program-based budgeting system). Mongolia needs to use the BOOST initiative since its micro-fiscal data are of poor quality. The lack of functional classification producing aggregate numbers concerning institutions/ministries in charge of executing the expenditure make the review of the country’s public spending a challenging endeavor. Both the functional and programmatic classifications (which are available only for the initial budgets of 2015-2018) provide incomplete descriptions that need to be improved and made comparable to other countries (some descriptions in the original datasets are incomplete and cannot be used in an analysis (unknown references). This issue needs to be addressed to allow the creation of a more comprehensive BOOST dataset. Also, the new programmatic and functional classifications need to be reviewed (due to misconceptions concerning these classifications).   Source: Adapted from World Bank Africa Pulse (May 2017). 2. Both recurrent and capital expenditures increased significantly in 2010-2016. High volatility characterized the evolution of current expenditures as a series of pro-cyclical universal transfers were launched and then stopped. The GDP share of current expenditure increased both in 2007-2008 and in 2011-2012 as increased current transfers were fueled by the commodity market boom during the two periods. Current expenditures, which averaged 24.1% of GDP in 2013-2015, rose to 27.4% of GDP in 2016 due to election spending and a substantial rise in interest payments. Capital expenditure also rose from 6.1% of GDP in 2010 to 11% in 2016.24 3. The main driver of the increase in capital expenditure was first off-budget expenditures in 2012- 2015, and then a sharp increase in on-budget spending in 2016. The combination of sluggish revenue growth and the deficit control rules of the FSL, on-budget expenditures remained tight in 2012-2014, 24 The GDP ratio of capital expenditure rose from 4.2% in 2000-2006 to 7.2% in 2007-2010, as strong revenue linked with increasing mineral exports helped finance a huge infrastructure development program. The increase in capital expenditures accelerated since 2010, driven first by the global commodity market boom and increased revenue and then by a vast public investment program financed by external borrowing in 2012-2014, with the Development Bank of Mongolia as an off-budget investment vehicle. 63 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue with its GDP share gradually declining from 36% in 2012 to 32% in 2014. At the same time, however, the government tried to stimulate growth by launching a vast program of off-budget expenditures through the DBM, using the proceeds of external sovereign or sovereign-guaranteed borrowing. The DBM’s off-budget expenditures averaged 5.9% of GDP in 2012-2014, accounting for about 15% of total public expenditure over the same period. In 2015, the government undertook to improve fiscal management by reducing total capital expenditures and including in the budget DBM-financed investment projects.25 The increase in expenditure in 2016 was largely due to a steep increase in on-budget spending while DBM-funded off-budget programs were tightened (Figure 3.3). FIGURE 3.3 Interest Payments Mainly Explained the Rise in Recurrent Expenditures 14 12 10 8 6 Wages & salaries 4 Goods & services 2 Interest payments 0 Subsidies 04 05 06 07 08 09 10 11 12 13 14 15 16 1 7e Transfers 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Source: MOF, IMF, ADB, World Bank staff estimates FIGURE 3.4 Meanwhile Off-budget Investment Spending Dominated Rise in Capital Spending in 2012-2015… On-budget Expenditures DBM’s budgetary loans DBM’s corporate loans Gov’t definition of expenditures Total Budget Expenditures Source: MOF, IMF, ADB, World Bank staff estimates However, the corporate loan programs of the DBM - funded by government-guaranteed external borrowing - remained off the budget. 25 64 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 4. In 2016, elections led to a significant increase in recurrent and capital expenditures (Figure 3.3 and Figure 3.4). Before the parliamentary election, the government launched four one-off policy loan programs (Good Herder/Student/Fence/Share), which accounted for more than 6% of on-budget spending and about 2.3% of GDP. In 2014-2016, the Parliament approved off-budget investments through the issuance of promissory notes to contractors in lieu of cash payments. In 2016, the government decided to clear payment arrears on promissory notes (equivalent to 2.8% of GDP). 5. Mongolia stands out as a country with high public spending (Figure 3.5). The expenditure to GDP ratio of Mongolia is 31.1% in 2017,26 much above the average in commodity-dependent countries (29.5%) and lower-middle income countries (LMICs) (19.3%). The GDP ratios of both current and capital spending increased substantially since 2010 and are higher than in comparator countries. Strong revenue in 2010-2011, universal cash transfers and off-budget investments (DBM) are the main factors. FIGURE 3.5 Mongolia Stands Out for Its High Expenditure Level by International Comparison General Government Expenditure Trend (% of GDP): 2003-2017 45 40 35 30 Mongolia 25 Lower middle 20 Income Countries 15 Middle income 10 countries 5 0 Low income countries 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Asia LICs Source: MOF, IMF, ADB, World Bank staff estimates. FIGURE 3.6 Mongolia’s Current vs Capital Expenditure by International Comparison27 All levels of government (% of GDP): 2017 35 30 Current expenditure Capital expenditure 25 20 15 10 5 0 Mongolia Chile Peru Malaysia Thailand Kazakhstan Indonesia EA&P (avg) Source: MOF, IMF, ADB, World Bank staff estimates. Only 33.9% of GDP in 2015 but still much higher than the country’s peers. 26 The comparator countries included in this section were selected based on EAP region, natural resource endowment as well as the level of income. 27 65 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 6. In 2017, budget expenditure substantially reduced through fiscal consolidation by the government with support of donors including the IMF and World Bank. Total expenditures declined from 40.5% of GDP in 2016 to 31.1% in 2017, mostly explained by a sharp drop in capital expenditures and off- budget spending (DBM’s commercial loans) in line with government’s policies to gradually reduce off- budget spending including tight control on capital spending. Despite showing a year-on-year growth, recurrent spending reduced compared to planned which is mainly attributed to the reduction in interest payments and the impact of hiring freeze on the wage bill. To reduce interest payments, the government also committed to halting the domestic T-bills issuance in favor of concessional financial support from development partners including the IMF and World Bank. Recurrent Spending Categories 7. Current expenditures on goods and services stayed at a level comparable to the international average but rose substantially in 2016 to 6% of GDP (from 4.7% in 2015). The large increase in goods and services expenditures was due to a series of one-off spending programs, which totaled 0.8% of GDP and 14% of total expenditures on goods and services.28 FIGURE 3.7 Mongolia’s Goods and Services Expenditure by International Comparison All levels of government (% of GDP): 2017 8.0 7.0 7.0 6.3 6.1 6.0 4.9 5.0 4.0 3.0 2.6 2.2 2.0 2.0 1.0 0.0 Mongolia Thailand Malaysia Indonesia Kazakhstan Chile Peru Source: MOF, IMF, ADB, World Bank staff estimates. FIGURE 3.8 Mongolia’s Wage Bill Expenditure by International Comparison All levels of government (% of GDP): 2016 8 7.3 6.7 7 5.9 6.0 6 5.5 5 4.4 4 3.0 3 2 1 0 Mongolia Thailand Malaysia Indonesia Kazakhstan Chile EA&P (avg) Source: MOF, IMF, ADB, World Bank staff estimates. 28 The large increase in goods and services expenditures in 2016 was particularly due to one-off spending programs. The total sum of such one-off programs reached 0.8% of GDP and accounted for 14% of total expenditure on goods and services. Such programs include: (a) dispute settlement payments to Khan Resources: MNT 143.5 billion; (b) expenses to host the ASEM-MNT 21 billion; (c) general and local election expenses-MNT 30.6 billion; and (d) Olympic expenses- MNT 5.2 billion. Note that half of the election expense is likely to be needed in 2017 for the presidential election scheduled to be held in the summer. 66 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 8. As shown in Figure 3.8, Mongolia’s wage bill is comparable with peers (see a more detailed analysis of the wage bill in section 3.2). The basic salary - which represents the major portion of the wage bill - increased from 6.1% of GDP in 2011 to 7.3% in 2016 (but declined in 2017 with hiring freeze). The main drivers of the increase were the growing number civil servants and automatic wage increases (Cost of living adjustment [COLA]). The number of civil servants per 100 population (5.7) is higher than in other East Asia countries (an average of 2.5). At 4.8%, the rate of growth of Civil service staff is nearly three times the rate of growth of the population. Civil servants move up the (basic) salary grade depending on years of services. The mechanism for calculating these increases should prudently be reviewed, and this provides an opportunity for adjustment of the wage bill. 9. Mongolia’s spending on interest payments is much higher than in comparator countries (Figure.3.9). Interest payments increased from 0.8% of GDP in 2012 to 4% of GDP in 2017 due to a rapid build- up of domestic and external debt since 2012. Accounting for less than 1% of total expenditures until 2011, interest payments rose to 13% of total spending in 2017, thus reducing fiscal space for primary expenditures necessary for social protection and economic development. Mongolia’s interest payments (4% of GDP) are far higher than the 2.1% average in LMICs. FIGURE 3.9 Mongolia’s Interest Payments by International Comparison 4.5 All levels of government (% of GDP): 2017 4.0 4.0 3.5 3.0 2.5 2.1 2.1 2.0 1.6 1.2 1.3 1.5 0.9 0.8 1.0 0.5 0.0 Mongolia Thailand Malaysia Indonesia Kazakhstan Chile LMI (avg) EA&P (avg) Source: MOF, IMF, ADB, World Bank staff estimates. 10. Current transfers rose from 9% of GDP in 2015 to 9.5% in 2016 (Figure 3.10). The main driver was an increase in social insurance payments due to expanded health services and a state contribution to health insurance equivalent to 0.3% of GDP. In addition, a large one-time compensation (0.4% of GDP) was paid to retiring state employees in 2016. Implementation of these new measures continued in 2017, keeping the pension and social insurance payment at a similar level (around 6.1-6.2% of GDP). 11. Social insurance payments increased in recent years, largely due to increased pension payments. Projections indicate that the deficit in the pension fund will increase further and will require significant additional financial support from the central government. Budget support to the pension fund, which accounted for 2.2% of GDP in 2017, will continue to grow in the coming years. The following parametric reforms are recommended to improve the financial position of the pension scheme, its equity and credibility: (a) gradually increasing the retirement age; (b) adopting automatic inflation- based indexation; (c) Reducing the minimum pension and the partial pension and linking them to a transparent index such as the Minimum Living Standard; (d) Adjusting early retirement for herders & special professions through actuarially fair benefit reductions for retiring early and supplements for late retirement. 67 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 12. The level of Mongolia’s social welfare transfers is comparable to that in other countries. It could be increased to offer better protection for the poor. Mongolia spends approximately 2.1% of GDP on social welfare transfers (Figure 3.10). This is a relatively generous contribution if compared to the 1.6% average in middle-income countries. A recent World Bank analysis shows that social welfare programs are generally pro-poor, but the poorest 20% receive a relatively smaller share (34%) of these transfers, when compared with similar programs in European and Central Asian countries (over 40%). Most of the social welfare programs are universally or categorically targeted, and some social welfare benefits are viewed as regressive, with most of the beneficiaries going to the three highest social welfare quintile groups. FIGURE 3.10 Mongolia’s Current Transfers Expenditure Remain Steady Over the Past Few Years All levels of government (% of GDP): 2010-2017 13 12 11 10 9 8 Other transfers 7 2.0 1.9 One time prize and bonus 2.1 6 2.3 2.2 0.2 1.4 Pension and social insurance 5 1.0 1.0 payment 4 5.8 6.1 Social welfare transfers 3 4.8 5.0 5.7 2 4.2 4.7 3.8 One‐time compensation to 1 govt retirees 0 2010 2011 2012 2013 2014 2015 2016 2017 Total current transfers Source: MOF, IMF, ADB, World Bank staff estimates. Capital Spending 13. Mongolia’s average capital spending in 2012-2017 was much higher than in comparator countries. The GDP ratio of capital expenditure rose from 4.2% in 2000-2006 to 7.2% in 2007-2010, when strong mineral revenue was used to finance huge infrastructure investments. The growth of capital expenditures continued after 2010, financed first by high revenues during the commodity boom and by external borrowing in 2012-2014 (off-budget financing through the DBM). As a result, capital spending averaged 11% of GDP in 2016, which far exceeds the 6.1% average of East Asian and Pacific countries in 2016. An adjustment of about 5% of GDP would therefore seem to be needed to bring Mongolia’s average in line with comparator countries. FIGURE 3.11 Mongolia’s Capital Expenditures by International Comparison 10 9.3 Capital expenditures (% of GDP): 2017 9 8 7 6.3 5.8 6 5 3.9 4 3.4 2.7 2.9 3 1.9 2 1 0 Mongolia Chile Peru Malaysia Thailand Kazakhstan Indonesia EA&P (avg) (average 2012-2017) Source: MOF, IMF, ADB, World Bank staff estimates. 68 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 14. The increase in capital spending in 2012-2015 was largely driven by investment projects funded by the DBM. When the deficit ceiling of the FSL became effective in 2012, the government began to use the newly established DBM as an off-budget investment vehicle by injecting the proceeds of sovereign and sovereign- guaranteed borrowing into the development bank. As a result, public investment projects financed by the DBM accounted for 4.6% of GDP in 2013 and 1.9% in 2015 and represented more than one-third of total capital expenditures. At the same time, public investment directly financed by the central and local government budget (budget-funded projects) continued to decline from 9.1% of GDP in 2012 to 4.2% in 2015 (Figure 3.12). FIGURE 3.12 Capital Expenditures: On-budget versus Off-budget (% of GDP) On-budget and DBM expenditures (% of GDP): 2007-2015 14 100 9 50 DBM non-commercial projects (%): LHS Government-funded capital spending (%): LHS 4 0 Capital Expenditures included in the budget: LHS % share of DBM-funded projects (%): RHS -1 -50 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017e Source: MOF, IMF, ADB, World Bank staff estimates. 15. Despite the large increase in public investment, the capital maintenance budget did not increase substantially over the last 10 years. The GDP ratio of maintenance expenditure slightly rose from an average of 0.3% in 2003-2009 to an average 0.4% in 2010-2016. The current level of maintenance expenditures is comparable to the 0.3% average in LMICs, but this level is far from adequate, considering Mongolia’s much larger investment budget. The ratio of maintenance to total capital expenditures declined to 4.8% in 2017 from 5.1% in 2010. The average in comparator countries is about 6.2% of GDP. FIGURE 3.13 Spending on Maintenance is Lower Relative to the Size of Economy Maintenance to capital budget ratio: % of total capital spending 7.9 6.2 5.9 5.2 5.5 5.1 4.8 3.9 3.5 2010 2011 2012 2013 2014 2015 2016 2017 LMIC 2010-2014 avg Source: MOF, NSO data, World Bank staff estimates SSF: Social Security Fund, HDF: Human Development Fund. 69 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 16. As shown in Figure 3.14, the local governments took on a growing share of public spending. Their share of general government spending went up from 10% in 2010 to 21% in 2017. In 2013, the IBL provided for more decentralization of public spending to local governments (21 provinces and the Ulaanbaatar [UB] city). The goal was to improve the quality of public spending and to spread the benefits of mineral wealth and mining revenue across the Mongolian territory. Local governments assumed responsibility for basic education, primary health care, urban planning, construction, social welfare services, water supply and sewerage, public transport, urban roads and bridges, and municipal services (street lighting and garbage removal). Financed by local taxes and fiscal transfers, local government spending increased from 3.7% of GDP in 2010 to 9% in 2017. FIGURE 3.14. Local Governments Took on a Growing Share of Public Spending… Composition of total spending by level of government (%). Central gov Loc gov SSF HDF 100% 2 9 4 3 3 14 10 14 12 15 15 14 16 14 15 17 17 17 15 16 14 80% 7 7 6 7 11 13 8 15 10 20 21 10 9 23 24 7 11 23 60% 40% 79 80 79 77 75 64 67 65 64 63 62 60 58 56 62 20% 0% 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017e Source: MOF, NSO data, World Bank staff estimates SSF: Social Security Fund, HDF: Human Development Fund. 17. Low disbursement rates at all levels of government - though more prominently at the local government level - are due to capacity constraints. Figure 3.15 shows that actual spending is only a fraction of amounts planned in the budget. Disbursement rates are particularly low for capital investment rather than current spending (Figure 3.16), especially in the case of local government. Low execution rates, management and infrastructure issues and institutional and administrative capacity constraints limit absorptive capacity. It should be noted that disbursement rates for capital expenditures are low not only at the local level but also for central government projects. In other words, the institutional capacity issue cuts across all levels of government. 70 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES FIGURE 3.15 Disbursement Rates of All Level FIGURE 3.16 Meanwhile, the Disbursement Government Exceeded 90% Rate is Particularly Low for in 2012-2017 Public Investment… 2012 ‐ 2017 (Average) 2012 ‐ 2017 (Average) 100 100 80 80 94.3 93.2 88.3 60 60 79.7 40 40 20 20 0 0 Central Local Central Local Government Government Government Government Source. MOF, NSO data, WB staff estimates. SSF: Social Security Fund, HDF: Human Development Fund. 3.2 Public Sector Wage Bill and Employment - Trends and Issues 3.2.1 The Civil Service Reform Agenda 18. The first post - 1990 legal framework for the civil service was adopted in 1995. It was followed by new legislations in 2002 (with major amendments in 2008) and recently in December 2017. Broadly, the law regulates the categorization of civil servants, the types of posts, recruitment and performance management principles, the management structure (the Civil Service Council (CSC) and the Cabinet Secretariat), working conditions, types of benefits, and disciplinary measures. As defined in the Civil Service Law (CSL), the four main employment categories are: (a) political service, (b) administrative service, (c) special service (that is, uniformed services, the judiciary, and staff of the oversight organizations), and (d) service area staff, including frontline service delivery staff in health, education, and other sectors. The administrative and special services are categorized as ‘core civil servants’ in the CSL. 19. The governance of the civil service is relatively weak, and the system does not function well. The main institution responsible for the management of the civil service is the CSC, together with a structure of 50 sub-councils. The CSC has a small secretariat of less than 15 staff. Members of the CSC come from all of the three branches of government (legislature, executive, and judiciary). The council is expected to set up policy directions and to monitor and enforce compliance. While rules for merit-based recruitment and performance management formally exist, implementation is very weak. 20. The Mongolian government recognizes the importance of an efficient civil service for the performance of the public sector. In 2007, the government adopted a Medium-Term Civil Service Reform Strategy and Implementation Action Plan that established a number of strategic priorities for achieving a civil service that is “oriented to meeting the needs of citizens, flexible in responding to changing circumstances, proactive, lean, capable, and outcome-oriented.” These priorities included improved selection procedures for senior civil servants, development of leadership skills, more transparent, fair, and performance-oriented personnel management, and a modern remuneration system that is also fiscally sustainable. The 2008 amendments to the CSL intended to address many of these priorities. The World Bank and the IMF also provided advice29 on these policy priorities. Government Employment and Wages, Mongolia - Selected Issues and Statistical Appendix, IMF, July 2008 and Mongolia Towards a High Performing 29 Civil Service, World Bank, June 2009. 71 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 21. Implementation of the reform strategy faced challenges due to the lack of practical mechanisms and guidelines but also political issues. Poor civil service management is clearly linked to the current structure of civil service governance, which is influenced by the three negative processes of capture, exclusion, and clientelism. Past civil service reform actions had positive aspects, but the frequent reorganizations of key ministries, which result from recurrent changes in leadership, provide opportunities for dismissing existing staff and replacing them with politically affiliated appointees. Sometimes the political capture of the civil service is through the use of temporary appointments but a deeper penetration of political appointees is also taking place at the management level, in service delivery organizations as well as in divisional and departmental heads across key national ministries. Based on CSC data, the aggregate turnover reached 14.23% in 2016 (after the elections), a significant increase from the late 2000s, when turnover was around 5% annually (Figure 3.17). FIGURE 3.17 Civil Service Turnover has Increased Overtime Since 2001 ... 16% 14.2% 14% 13.1% 13.4% 12% 11.5% 10% 9.3% 8.3% 8.4% 8% 7.6% 6% 5.2% 5.6% 4% 2% 0% 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: Civil Service Council 22. High turnover rates were particularly evident after the 2012 and 2016 general and local elections when the power shifted from one party to the other. High turnover rates undermine the stability and continuity of the civil service and affect the effectiveness of capacity development programs. The number of formal complaints to the CSC for unlawful dismissal of civil servants increased from 87 in 2009 to 144 in 2016. Many, however, do not lodge complaints for violation of their rights, because they do not trust the capacity of the state to do the right thing. In fact, leaving voluntarily the civil service because of affiliation to the opposition party has become common practice despite the protection offered by the CSL. On average, over 7,000 people leave the civil service every year on a voluntary basis.30 23. The priority goals of the new government action plan (2016) are a quick recovery of the economy, restoring growth and improving the livelihood of citizens. It is also to build the capacity of the civil service and appoint professional personnel at all levels in order to ensure an efficient and transparent delivery of public services. The action plan also aims to “improve the reputation of the civil service by putting in place the necessary disciplinary actions for civil servants who display professional inadequacy and incompetence or violate the code of conduct and ethical standards and by eliminating any type of nepotism, favoritism and preferential treatment.” 24. Before 2017, draft laws were submitted to the Parliament first in October 2012 and then in July 2015 that could not be approved. The goal of the new CSL approved in December 2017 is to limit the political capture of the civil service by instituting a system of career progression and requiring a minimum of years “Towards a Professional and Citizen-centered Civil Service in Mongolia”, Project Document, UNDP 30 72 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES and civil service experience for managerial staff (16 years of service for the position of State Secretary). In addition, the law is expected to improve accountability by introducing sanctions for patronage-driven decisions. Approval of the new CSL provides an opportunity for more reforms to professionalize the civil service and improve its institutional stability through new standards, procedures and relevant capacity building requirements31. BOX 3.2 Major Changes in the CSL, December 2017 1. For each position, heightened requirement for the number of years in the civil service is introduced. For instance, for positions at the level of a senior officer, at least 8 years of civil service experience, of which no less than 4 years at the level of specialist is required. Also, completion of a series of professional training is required. Moreover, for the position of a State Secretary, at least 16 years of civil service experience with no less than 8 years at the level of Director-General is required. The law initiators expect introduction of these re- quirements to be a major step toward making the civil service free of political influence and ensuring independence. 2. The mechanism of career progression is instituted in the law. Specifically, this means when a position becomes vacant at grade TZ39, selection for the position would be carried out from the pool of candidates at TZ38. This will allow the civil service to remain professional and sustainable. 3. Civil servants’ accountability system has been revisited. After each election, many civil ser- vants are dismissed unlawfully and new people are appointed in leadership positions. As a result, in 2008-2016, 2,100 cases related to the civil service were resolved at court, of which 1,238 or 59% were for unlawful dismissal. As per the new law, if a civil servant is found to have been unlawfully dismissed, the lost pay, social insurance, and medical insurance for the applicable period would have to be compensated by the official who made the wrongful decision. 4. Changes were made to enhance independence of the CSC. High-level civil servants’ ranking and grades of similar/equivalent positions will be determined by the Parliament as recom- mended by the CSC. Also in the law, there is a clause establishing that “All parties including the President, Speaker of the Parliament, Prime Minister, Members of the Cabinet and the Parliament, political parties, the public, staff of private entities and citizens are prohibited to engage in and influence the operations of the Civil Service Council”. 31 In February 2018, the United Nations Development Programme (UNDP), with financial assistance from the Government of Canada, launched a project involving the CSC and the Standing Committee on State Structure of the Parliament to support implementation of the civil service reform priorities outlined in the law. The objective of the project is to create a stable, impartial, professional and citizen-centered civil service as a key building block for the pursuit of medium- and long-development strategies. The four-year project will engage various stakeholders and will be focused on four main issues: (a) strengthening the legal and policy frameworks for a professional civil service; (b) improving the professional capacity of civil servants; (c) enhancing gender equality in public administration; and (d) strengthening accountability mechanisms. Key activities will include developing by laws and procedures (43 by laws and procedures) to support implementation of the new CSL, developing training curricula for leadership and administrative posts, reforming the civil service entry examinations system, improving human resource management procedures and practices (with respect to gender equality), and promoting enhanced leadership and overall management. The project also aims to engage in participatory governance initiatives to give voice to citizens commenting on their experience and to improve the quality and accountability of public service delivery. 73 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 3.2.2 Size and Cost of the Civil Service 25. The Government of Mongolia has a large workforce that is partly explained by some of the country’s unique characteristics. Cross-country comparisons show that the number of civil servants per 100 population is significantly higher in Mongolia than in regional peers and continued to rise from 4.4 in 2009 to 5.7 in 2015. Yet, efficiency of government spending (as per the Global Competitiveness Index 2016-2017) appears to be lower than its peer countries in the EAP region and other resource-rich countries with the score of 2.1 (where 1 = extremely inefficient; 7 = extremely efficient) ranks 124th among the 136 countries surveyed. This is partly because Mongolia has one of the lowest population density in the world (behind Greenland) so there are fixed costs in the provision of public services. Generally, it is not surprising that the number of civil servants per capita should be higher in countries with a low population density. FIGURE 3.18 Mongolia’s Efficiency of Public Spending is Among the Lowest in the World. Malaysia 15 China 19 Indonesia 25 Botswana 37 More efficient Thailand 51 Russia 57 Kazakhstan 65 Chile 80 Peru 104 Mongolia 124 0 20 40 60 80 100 120 140 Source: Global Competitiveness Index 2016-2017, World Economic Forum FIGURE 3.19 Change in Civil Service Employment has Increased with Urbanization, 2005-2015 7.0 50 6.5 45 6.0 5.5 40 5.0 Number of civil servants 35 per capital 4.5 Population in urban agglomerations of more 4.0 30 than 1 million (% of total population, right) 2005 2006 2007 2008 2009 2010 2011 2012 2013* 2014 2015 Source: World Bank staff calculations based on data from Civil Service Commission and World Bank’s World Development Indicators (WDI). Note: Employment series has a break in 2013. The effect of employment growth may be overstated in this year. 74 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 26. Yet, there are concerns about the adaptability of the civil service to changing economic and social conditions. It can be argued that the cost of delivering public services to a widely dispersed population should be high, but that cost should go down, as urbanization progresses. This is not what is happening in Mongolia. The share of the population living in urban areas increased from 36.9% in 2005 to 46.3% in 2015, but civil service employment has continued to rise. 27. Despite past recommendations to impose a partial freeze on new employment and allow strategic staffing in key administrative positions32, employment in the civil service continues to grow. Over the past decade, the number of civil servants increased by about 4.5% per annum (Table 3.1). In the last four years it increased by 14.5% from 162,769 in 2012 to 188,974 in 2016 (Table 3.1), and now accounts for 16.6% of the country’s workforce and 6.3% of the population (4.8% in 2005). TABLE 3.1 Number of Civil Servants by Region, 2007-2016 Regions 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Western 18,704 19,181 20,279 20,994 22,169 22,900 23,614 25,285 25,679 25,549 Khangai 22,825 23,167 24,278 24,721 25,742 26,291 27,274 28,838 29,371 29,359 Central 20,407 19,964 21,379 21,637 24,053 24,188 24,960 27,510 28,089 28,202 Eastern 9,595 9,640 10,144 10,549 11,077 11,279 11,952 12,538 12,857 12,729 Capital city 68,727 72,097 72,143 77,778 78,571 78,111 81,648 89,430 90,462 93,135 Total 140,258 144,049 148,223 155,679 161,612 162,769 169,448 183,601 186,458 188,974 FIGURE 3.20 Number of Civil Servants, 2005-2015 6.5 200 6.0 5.5 150 5.0 4.5 Number of civil servants per capita 4.0 100 Number of civil servants 2005 2006 2007 2008 2009 2010 2011 2012 2013* 2014 2015 Source: World Bank staff calculations based on data from Ministry of Finance, Civil Service Commission, and World Bank’s WDI. Note: *Employment series has a break in 2013. The effect of employment growth may be overstated in this year. 32 Selected Issues and Statistical Appendix, IMF, July 2008 75 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 3.21 Contribution to Growth in Personnel Spending, 2011-2015 2.1 2.1 1.9 1.9 1.9 1.9 1.8 1.7 1.8 1.6 1.7 7.8 7.8 7.5 9.1 9.6 9.8 10.0 9.9 10.6 10.6 10.6 18.9 18.8 18.1 17.8 19.0 18.0 18.5 17.7 19.2 19.1 19.2 in thousands 71.2 71.3 72.5 71.2 69.5 70.3 69.6 70.7 68.5 68.6 68.6 Service area Special Administrative Political 2005 2006 2007 2008 2009 2010 2011 2012 2013* 2014 2015 Source: World Bank staff calculations based on data from Civil Service Commission and World Bank’s WDI. Note: Employment series has a break in 2013. The effect of employment growth may be overstated in this year. 28. The growth in the civil service is expected to moderate through 2019. As part of its IMF Extended Financial Facility Program the government has introduced a public sector wage freeze through 2018. Through 2019; civil servants’ grade advancement is expected to be suspended and new hiring restricted, except for priority sectors (health, education, emergency services, and police), where one new employee will be hired for every two who leave).33 BOX 3.3 Recommendations on Government Employment and Wages, Mongolia Civil service spending of over 10% of GDP is high in comparison with other countries and its increase raises concern about fiscal sustainability and the crowding out of other government expenditure. The high wage bill reflects the large size of government employment and the recent increase in real wages. The wage scale, however, is highly compressed, and the composition of the civil service is highly skewed toward support staff. The CSRS adopted in 2004 needs to be fully implemented. Recently, the government took a number of important steps. It approved the Action Plan for the CSRS, completed a CSC and carried out functional reviews of three key ministries. However, more needs to be done streamline the civil service and improve the quality of services delivery. The following measures should be considered: (a) imposing a partial freeze on new employment—allowing only selected strategic staffing in key civil administrative positions; (b) making the HRMIS fully operational to improve the database on civil service employment; (c) improving coordination between the MOF, the Ministry of Social Welfare and Labor and the GSC in order to effectively implement the CSRS; (d) promulgating a law on contracting out government services (a draft has been prepared for internal discussion, and amending the Government Service Law to create a professional civil service with transparent rules for employment; and (e) formulating a fiscally affordable compensation and severance policy that is consistent with the CSRS objective of streamlining the civil service, and integrating civil service reforms into the medium-term budget frameworks. Source: Selected Issues and Statistical Appendix, IMF, July 2008. 33 IMF Mongolia Article IV, May 2017 76 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 29. The composition of the civil service has changed. The civil servants are grouped into four categories. Administrative personnel include staff of the ministries, senior staff of aimags, heads of agencies, and other ‘general managers’. Special positions include the armed forces, police, judges, staff of oversight bodies and diplomatic officers. These two categories are considered as ‘core’ civil servants. Services area employees are teachers, healthcare professionals, social welfare staff, and agricultural extension workers. The last category includes the political appointees. In 2005, there were 37.5 administrative and special personnel per 100 service area employees. In 2015, that ratio increased to 43.3. 30. The share of the civil service in total employment remained virtually unchanged since 2011. Total employment increased by 7.2% from 2011 to 2015. An expanding construction sector drove almost half of this increase. Together, trade and manufacturing contributed the same amount. Government employment also increased but at a slower rate, with significant differences depending on the sectors. Employment in education and health increased by 0.7 and 1.2% respectively, but employment in public administration including defense and social insurance grew by 18.3%. Overall, the share of the government sector in total employment remained broadly unchanged at 17% (Figure 3.22). FIGURE 3.22 The Government’s Share has Remained Almost Unchanged in Total Employment in 2011-2015 2010 Human health and social -0.2 3.5 2010-2015 work activities Education -0.5 8.2 Public administration and defence; 5.4 0.5 compulsory social insurance Share in total employment (%) Source: World Bank staff calculations based on data from National Statistical Office. FIGURE 3.23 Contribution to Growth in Personnel Spending, 2011-2015 50 40 30 20 10 Real wages 0 Employment -10 Inflation -20 2011 2012 2013* 2014 2015 Source: World Bank staff calculations based on data from MOF, Civil Service Commission, and World Bank’s WDI. Note: Employment series has a break in 2013. The effect of employment growth may be overstated in this year. 77 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 31. Over time, growing civil service employment and rising real wages increased government spending on personnel. Personnel spending more than doubled in nominal terms between 2011 and 2015 (Figure 3.23). Almost half of the increase reflected adjustment for inflation but the remaining half resulted from rising employment and higher real wages. The civil service employment increased by 15.4% during that period. The increase in real wages was 23.4%, mostly due to across-the-board increases in basic salaries in 2012. 32. Government civil service spending is not affordable under current economic conditions. At 5.3% of GDP in 2016, government spending on personnel is not particularly high (between the average in middle-income and low-income countries).34 Similarly, as a share of total spending (18.3% in 2016), personnel spending in Mongolia is comparable to what is found in other countries. The problem is that personnel spending absorbs a high percentage of total government revenue (29.9% in 2016 following a 11.9 percentage -points increase after 2011)35 (Figure 3.24). Today, as fiscal sustainability becomes a major issue and since the country’s economic recovery is expected to be gradual, current government spending on personnel may not be affordable. FIGURE 3.24 The Government of Mongolia’s Personnel Spending, 2011-2017 29.9 30.0 28.6 25.1 25.0 24.0 24.1 23.4 22.7 22.2 22.4 21.0 20.8 19.9 19.8 20.0 18.3 18.0 16.1 15.0 10.0 9.0 7.4 7.3 7.2 7.3 7.2 6.1 6.5 5.0 0.0 In % of GDP In % of Total Expenditures In % of Total Revenues 2010 2011 2012 2013* 2014 2015 2016 2017e Source: MOF and World Bank’s ‘Size of the Public Sector’ database. Note: Personnel spending includes wage bill and social contributions. 33. The composition of government spending on personnel changed from 2012 to 2017. Personnel spending includes basic salaries, supplementary compensation and payments to contractors. Unlike the share of government spending on contractors which did not change, it is the share of supplementary compensation has increased since 2012 (Figure 3.25). Personnel spending includes wage bill and social contributions. 34 The share of personnel spending declined to 22.4% in 2017 35 78 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES FIGURE 3.25 Composition of Government Pay in Fiscal Accounts, 2003-2017 Base salaries Supplementary compensation Contractors Wage bill (% of GDP): RHS 100% 10 90% 9 80% 8 70% 7 60% 6 50% 5 40% 4 30% 3 20% 2 10% 1 0% 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Source: World Bank staff calculations based on data from Ministry of Finance. 3.2.3 Pay, Grading and Compensation 34. Average civil service wages are not higher than those in the rest of the economy. A comparison of the wage levels across different sectors shows that average government wages (including defense, social insurance, education and health) are below the national average (Figure 3.26). The top-paid professions are in mining, followed by International Financial Institution (IFIs) electricity, and professional and financial services. However, it is not possible to assess the existence and the size of a public/private sector wage gap in the absence of micro-level data on professions and skill levels. FIGURE 3.26 Average Wages by Sector (Thousand MNT), 2017 2500 2000 1500 1000 500 0 IFI’s Professional industry Electricity Financial industry Transportation Manufacturing National average Construction Public admin & Defense Real estate Communication Health Whole & Retail trade Service Agriculture Education Water supply Entertainment Restaurant Mining Public admin Household business Source: National Statistical Office. 79 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 35. Wage growth in the government sector underperformed the rest of the economy from 2015 to 2017. Wage growth is generally volatile in years 2010-2014 moderating slightly since 2014. With major salary increases in 2012, the growth of real wages in public administration was well above the national average in 2012-2013 (Figure 3.27). No other major civil service salary increase was decided since 2014 and the growth of civil service wages in public administration, education and health was below the national average during that period. FIGURE 3.27 Growth in Real Wages by Sector, 2011-2017 50% National average 40% Public administration & defense 30% Health Education 20% 10% 0% -10% -20% Mar‐10 Oct‐10 May‐11 Dec‐11 Jul‐12 Feb‐13 Sep‐13 Apr‐14 Nov‐14 Jun‐15 Jan‐16 Aug‐16 Mar‐17 Oct‐17 Source: World Bank staff calculations based on data from National Statistical Office. 36. During 2011-2015, basic civil service salaries were raised twice through across-the-board increases in 2012 and 2014. The average administrative salary increased by 13% in real terms during that period-7.1% at the high-end and 30.9% at the lowest-end of the salary scale (Figure 3.28). These differential increases compressed the salary scale of administrative personnel and reduced the maximum-to-minimum salary ratio from 2.9 in 2011 to 2.4 in 2015. In the education sector, salary growth was higher with wages for administrative employees increasing by 14.4% in real terms while salary growth for teachers in pre-schools and secondary schools reached 23.3% and 19.3% respectively (Figure 3.29). FIGURE 3.28 Real Growth in Basic Salaries for FIGURE 3.29 Real Growth in Basic Salaries in Administrative Personnel, 2011-2015 Education Sector, 2011-2015 30.9 23.3 19.3 14.4 13.0 7.1 Average Minimum Maximum Secondary Pre‐schools Administrative schools Source: World Bank staff calculations based on Government Source: World Bank staff calculations based on data from Decree 239 of 2010 and Government Decree 322 of 2014 and Ministry of Education and the World Bank’s WDI. the World Bank’s WDI. 80 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 37. Across-the-board salary increases in 2012 and 2014 further compressed the salary structure. In 2009, a World Bank review found that Mongolian pay scales are highly compressed (a compression ratio of 3.3 at that time) relative to other countries. An analysis of existing pay scales for administrative personnel shows that the 2012 and 2014 salary increases further reduced the compression ratio from 2.9 in 2011 to 2.4 in 2015. However, these basic statistics for the entire system of pay grades may overstate pay progression during the career of a civil servant, as calculations are limited to a few pay grades and increments. 38. Salary supplements complicate the salary structure, as they double basic salaries for some positions. In 2015, the salaries of civil servants in education were spread over 170 different pay grades and their five levels (Figure 3.30). Excluding the outliers, the average basic salary was MNT 574,900, but the average salary supplements (including allowances, overtime and other extra payments and rewards) amounted to MNT 619,730. For some pay grades in education (preschool and secondary school teachers grade 4), salary supplements were almost twice as high as the basic salary (Figure 3.30). The main supplements were allowances for meal and transportation, overtime, and supplementary payments for those working at a laboratory school. 39. Current findings on pay scales overlap and supplementary compensation suggest that past recommendations regarding the introduction of a grading reform and simplification of pay systems by merging allowances into basic salaries were not effectively implemented. The 2009 World Bank review of the civil service noted that pay scales were very compressed and that increments in the pay scales also overlapped to a considerable degree. The current analysis reveals a similar pattern of salary compression and pay scale overlap. The government should move in a sequenced manner toward a job-based system, in which employees are paid for specific responsibilities and current inconsistencies between staff performing the same duties should be resolved. Also, allowances should be merged into basic pay to the greatest extent possible. Finally, the government should consider the introduction of a mechanism to review and upgrade salaries in the light of annual cost of living increases. FIGURE 3.30 Salary Supplements in Education, 2015 600 500 400 MNT thousands 300 200 100 0 Base pay For overtime: For performing Other: Other For PhD: Sc. Doc, PhD Allowance for meals and Special conditions: for transportation For working at a laboratory school Supplementary remuneration for working… Supplementary Supplementary remuneration for… For skills Supplementary supplementary… For ranking Supplementary additional… remuneration for being in… remuneration for being in… remuneration for leading… working at special schools For performing additional For working at a school with an international curriculum For years in public service tasks/assuming additional… Source: World Bank staff calculations based on data from Ministry of Education. 81 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue BOX 3.4 Actions Required for Grading, Compensation and Payroll Administration Pay simplification In 2009, a government resolution provided for the simplification of the pay for administrative service by merging allowances into the basic pay. The agency responsible for implementation was the Prime Minister’s Office on the basis of recommendations from the Civil Service Council, the Ministry of Finance, and the Ministry of Social Welfare and Labor In 2010 a government resolution provided for the simplification of the pay for the support service be merging allowances into the basic pay. The Prime Minister’s Office was responsible for implementation. A locality pay to encourage horizontal mobility and attract civil servants to remote locations In 2011, a government resolution introduced a locality pay for civil servants. The Prime Minister’s Office was responsible for implementation. Sequenced introduction of grading reform and benchmarking pay In 2009, the government started analytical work on pay and grading reform. The responsible agency is the CSC in consultation with the Ministry of Finance and the Ministry of Social Welfare and Labor. In 2010, a government resolution provided for new grade structures and basic pay scales for the administrative service. The responsible agency is the Prime Minister’s Office. In 2011, a government resolution provided for new grade structures and basic pay scales for the support service. The responsible agency is the Prime Minister’s Office. Development of an automated centralized payroll In 2009, a feasibility study was completed for the development of a centralized human resource database; The responsible agency is the Ministry of Finance. In 2010, a centralized human resource database was created with information on posts, personnel, and pay for all budget entities. In 2010, an Action Plan was approved for a sequenced transition to a centralized payroll. The responsible agency is the Ministry of Finance. In 2011, a Payroll Unit was established in the Treasury Department for administration of the centralized payroll. Ensuring effective coordination between stakeholders In 2009, the government created a Steering Committee for Civil Service Reform with members from the MOF, the CSC and the Ministry of Social Welfare and Labor. The responsible agency is the Prime Minister’s Office. Source. World Bank, 2009, Mongolia: Towards a High Performing Civil Service. 3.2.4 Payroll management 40. The World Bank Review of the Civil Service in 2009 recommended the development of an automated, centralized payroll. The goal of the system and the creation of a special unit was to ensure appropriate controls and support the reforms of the compensation structure, pay and grading. As part of establishing an IT system, it was recommended that the relevant business requirements for the payroll system be developed including working relationships between the individual budget entities and the dedicated payroll unit and relevant procedures for establishment registers and staff records; calculation, control and accounting; payment and remittances; payroll reports; and audit trails. Further, considerations to integration of the payroll system to the Government Financial Management Information System (GFMIS) operated by the Treasury and other relevant systems including the Budget Preparation Information System (BPIS) were also suggested. 82 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 41. As putting in place a new IT system and updating the relevant business processes is complicated, a sequenced and phased approach was recommended for the establishment of a new payroll system. The 2009 review suggested that the government could focus on an interim solution where only a centralized database was established capturing the relevant information such as personnel data and position register and allowing better planning and modeling. Also, it was recommended that the creation of the automated system should be preceded by simplification of pay and grading and relevant compensation reforms to ensure that the modern system would adopt updated processes and practices. 42. The 2015 Public Expenditure Financial Accountability (PEFA) review gave Mongolia a score of D+, which suggests that very little efforts were made to implement the recommendations for an automated and centralized payroll. Personnel records and payroll continue to be decentralized in each of the over 5000 budgetary organizations. Each of these organizations has an administration and accounting department (two separate departments in the larger organizations) that maintains the organization’s accounts and personnel records and prepares the monthly payroll. 43. The MOF exercises its control at the aggregate level for each budgetary organization. Budget ceilings for the payroll are prepared on the basis of budgeted positions at the beginning of the fiscal year and are updated annually with the budget; hence they do not take into account staff changes during the course of the year (including vacancies, retirements, new recruitments). These budgeting practices on the basis of planned positions instead of actual produce inconsistencies between payroll and personnel records. Internal controls associated with changing personnel records and payrolls data are carried out by the authorized officials at the budgetary organizations. No comprehensive payroll audit covering all central government entities has been conducted in the last three years. Instead partial audits and staff surveys were undertaken by the Mongolia National Audit Office and the Social Insurance Authority, and to a lesser extent, by the tax office. 44. Inconsistencies between the payroll and the personnel records and the lack of a centralized information database affect the availability and reliability of the data for a comprehensive analysis of the wage bill. Each budgetary organization sends its personnel records once a year to the oversight agencies, namely the MOF, the CSC, and the Ministry of Labor and Social Protection. There are significant differences in personnel data available in these three organizations, which is symptomatic of the weaknesses in the payroll controls. There is no follow-up on the information submitted and the data are not checked. Personnel records are updated only annually but there are significant irregularities in the data reported to the different central agencies. 3.2.5 Priority Areas of Reform 45. Affordability, competitiveness, and flexibility are the key aspects to consider for civil service reform in Mongolia. To deliver the desired type and level of public services, a government needs to manage its human resources properly. This, in turn, requires an efficient and effective employment and pay system that is based on three pillars: (a) a competitive pay scheme that attracts and retains the needed skills; (b) flexible employment practices to respond to evolving needs and economic conditions; and (c) overall affordability not only today but also in the future.36 46. The reform process gain momentum with the recent approval of the CSL in December 2017. The changes introduced by the new CSL are an important stepping stone for restoring meritocracy and putting in place an institutional architecture that can effectively manage the public sector reform and ensure implementation of the law. It is important to prepare subordinate rules and regulations and develop a differentiated communication strategy and training program for rolling out the new law and explaining its main provisions to different audiences and for the training of staff at the implementation 36 IMF (2016) summarizes the proper methodology to analyze the issues of wage bill management. 83 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue stage. The new CSL reaffirmed the independent status of the CSC as the main oversight body for ensuring that the personnel procedures relating to recruitment, transfers, promotions, and dismissals comply with the law and principle of meritocracy. It is also empowered to investigate grievances and act as an administrative tribunal for issues related to human resources management practices and civil service examinations with the authority to annul any related unlawful decisions. Further, the law envisages improving professional capabilities for civil servants through rolling out professionalization training by the National Academy of Governance, the specialized training institute under the Cabinet Secretariat. It is of the utmost importance for the government to carry out these reforms to the full extent. In this regard, the role of the UNDP project ‘Toward a Professional and Citizen-centered Civil Service in Mongolia’ will be vital. 47. The reforms should be supplemented by measures aimed at limiting the rapid growth of civil service staff and the wage bill. Although personnel spending in Mongolia is not particularly high at 8.2% of GDP in 2015, it is high in terms of share of revenues (31.7% after a significant 12 percentage -point increase between 2011 and 2015). In this context, potential savings could be achieved by rationalizing government employment and relying on natural attrition, relocation, and consolidation. In this regard, it would be critical to launch a detailed review of the civil service staffing increases over the past decade to understand the dynamics that are driving them, the areas in which these increases were justified, and the future demographic pressures and attrition rates. These findings would also facilitate better and broader manpower planning exercises for the future. 48. Between 2011 and 2015, about two-thirds of the increase in personnel spending (in real terms) came from rising pay levels. Even though average wage levels in the government are slightly below national averages, there may be scope for wage restraint, because past wage increases were not linked to performance. There may be scope for reducing supplementary payments. Some components of the government’s compensation scheme are classified as transfers rather than as a part of the wage bill. Making them more transparent and exploring options for elimination may provide additional savings. One such component is spending on restructurings. The amounts are not significant, 0.2% of the wage bill in 2014, but they are frequent and detrimental to the efficient working of the government. 49. To further rationalize the compensation structure in the government, it would be necessary to conduct a major pay and grading exercise. In Mongolia, the pay scales continue to be very compressed and pay increments in the pay scales overlap to a considerable degree resulting in a random pay distribution, thus significant inequalities in compensation. While in the medium-term Mongolia needs to sequentially move to a job-based system where employees are paid for the specific responsibilities they perform, in the meantime, to the greatest extent possible allowances should be merged into basic pay. In this regard, the government needs to undertake a major pay and grading exercise to determine policy actions to streamline the current pay structures and reduce anomalies within them. Moreover, consideration is needed toward developing a regular mechanism for making annual COLA to the civil service salaries. 50. A final area where progress is needed is in establishing an automated and centralized payroll management system. Past recommendations for establishing a system allowing centralized payroll should be reconfirmed. Operationalizing the Payroll Module of the GFMIS could be a pursuable option in line with the recent upgrade of the GFMIS system (a Free Balance product) to Version 7.0 and the government’s plans to operationalize additional modules of the system that were not previously used. In the medium term, two to four years, the efforts for centralized payroll could be expanded by designing and implementing an HRMIS, that provides accurate, up-to-date information on the size and composition of the civil service, including data on turnover and attrition. 84 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 3.3 Structure and Efficiency of Public Capital Spending 3.3.1 Public Investment Management Reform Agenda 51. The GOM has recently been dealing with unconventional practices and unsustainable volumes of loans for new projects that have arisen since the economic slowdown in the mid-2010s. For at least the past five years, it seems the driving motivation behind PIM activity has been the need to start as many projects as possible in the shortest period while at the same time deferring the fiscal consequences for as long as possible. In short, Mongolia has been hooked on projects. 52. Line ministries developed and executed unsustainable volumes of poorly appraised and costly public investment projects through extra-budgetary funding. This has had the effect of allowing them unconstrained access to external debt without any checks on the cumulative fiscal effects. Furthermore, the projects involved have not been subject to the normal scrutiny and quality steps that would be considered essential in good international practice. This means that it is unknown whether these projects were based on proven demand, whether they were needed at all, and whether they were of an appropriate scale, and therefore cost. The legacy of all this uncontrolled PIM activity is likely to last many years. 53. Direct lending by MDAs and local governments through promissory notes and a system known as BT have been prominent means of funding new projects. This has allowed them to short-circuit whatever controls were in place to moderate calls on capital expenditure in the state budget. Despite evidence that the repayments on promissory notes and BT projects are unsustainable, attempts by the government to constrain the use of these practices have been only partly successful. Additionally, most of the debts incurred through BT contracts have been taken out in U.S. Dollar and Chinese Yuan (CNY), leaving the government exposed to significant exchange rate risks over the short and medium term. 54. The DBM has up until recently financed significant amounts of projects as well which would normally be considered as public investment projects. Both commercially orientated projects, such as those in the mining sector, as well as non-commercial projects, which are more socially oriented, were part of their remit. Recently, legislative interventions have repositioned the DBM to focus only on the commercial projects,37 meaning that non-commercial public investment projects are no longer their responsibility. By the end of December 2016 these 348 existing non-commercial projects of the DBM have been transferred back to the responsible line ministries under Government Resolution no. 42. A total of 119 of these projects, mainly in roads, energy, and social project construction are still under implementation; the remaining projects are complete but obviously retain the debts incurred. As with all the forms of extra-budget funding of projects, it is the debt accumulated that is of concern. In the case of DBM the debt related to non- commercial investment projects is estimated at MNT 2.528 trillion (Table 3.2). This includes U.S. dollar- denominated loans of US$ 673 million,38 almost all of which remains outstanding with the commensurate exchange risk. 55. The MOF has retained responsibility for those projects financed under the state budget. However, the role of the MOF has been diminished in recent years as MDAs chose to go around it to take advantage of extra-budgetary funding through DBM or through direct borrowing through promissory notes or BT projects rather than waiting for funding to be agreed at the state budget level. Relieving the DBM of their responsibilities regarding non-commercial projects and proposed changes to the Concession Law seem likely to rebalance greater responsibility for PIM activity toward the MOF in the future. 56. The National Development Agency (NDA) has a strategic coordinating role in the PIM system as well as quality checking role for all projects developed under the concession law. It is responsible for the The ‘Law on Development Bank of Mongolia’ of February 2017 clarifies this new remit for commercially orientated projects in Article 10. 37 Source: Development Bank of Mongolia 38 85 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue TABLE 3.2 Debt Incurred through Extra-Budget Funding of Investment Projects by 2016 (MNT billion)   Debt Incurred Repaid Outstanding Promissory Notes 658 658 0 DBM 2,528 20 2,508 (non-commercial) BT39 2,200 72 2,128 Total 5,386 750 4,636 Source: MOF approval of BT projects and for other forms of concession/PPP in Mongolia. Some of these projects are significant in scale such as the US$ 1.1 billion Tevshigiin Gov 600MW Power Plant, and a proposed US$ 370 million waste water treatment plant in UB City. As a legacy of the merging of the former State Property Committee in to the NDA, they are also responsible for project appraisal checking under a 2010 Regulation known ‘Methodology for Cost Benefit Analysis’. However, there is a critical lack of capacity to do this and their role appears set to be transferred to the MOF under draft amendments to the Concession Law. This is likely to exacerbate the capacity issue further. 57. Despite mounting fiscal pressures, the GOM continues to prepare large public investment projects with seemingly little consideration of their need, appropriateness or affordability. The National Development Plan envisages major infrastructure projects that appear to lack basic demand analysis. As an example, a proposed new national airport outside UB City is under development despite the existence of the currently underutilized, recently extended Chinggis Khaan International Airport. Action is required in the short term to avoid longer term ‘white elephant’ projects burdening future budgets. 58. There is an unknown amount of operational and maintenance costs to be built in to future budgets since recurrent costs associated with new projects are not considered in the preparation phase. All public investment projects incur ongoing costs during their lifetime (For example, paying to heat new schools or maintaining new roads). Many of these costs come with a time lag due to the long lead times in construction and are therefore conveniently overlooked during this project preparation and construction. Ongoing and periodic maintenance time lags are longer still. Without knowledge of these recurrent costs there is no way of planning to deal with medium-term pressures on the budget brought about following the recent significant capital expenditure. 59. Most local governments have been creating and executing public investment projects using extra- budgetary funding in the same way as MDAs although an amendment to the Debt Management Law (DML) has now prevented local governments from entering into BT contracts. Local governments also have a significant role to play in public investment particularly UB City which dominates local government activity. The PIM activities of the UB city authority on their own would compare in scale to one of the larger line ministries. Most of local government projects concern urban infrastructure such as local roads and small-scale social protection projects. In the case of UB City, the scale of activity through BT has reduced significantly and they plan no further BT contracts. It should be noted that repayment obligations from BT projects are repaid from the UB City budget and the liabilities are not included in the total known to the MOF. 60. The result of all this fragmented public investment activity is that not one public body understands the scale and implications of it all. There is no central overview of whether all the investment fits within a cohesive development plan raising risks that projects may not fit any strategic goal, may not Central government only. An unknown additional amount has been incurred by local government. 39 86 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES be necessary, may be oversized for the demand or could even be duplicating or overlapping each other. Nor is there any view of whether this cumulative investment is affordable within a medium-term expenditure framework (MTEF) or medium term fiscal framework (MTFF). The cumulative scale of the debts incurred through this activity is not fully understood nor are the fiscal consequences and yet some of the unconventional means employed in creating these conditions still remain available. There are signs that the reduction in BT activity may be replaced by an increase in Build-Operate-Transfer (BOT) activity without any capacity in the system to prepare and negotiate them. However, identifying the consequent fiscal obligations and risks that come with signed contracts is warranted. Therefore, urgent actions are required first to stabilize and then bring the PIM system fully under control. 3.4.2. Capital Expenditures - Trends, Reforms, Sector Distribution and New Risks Trends - Financing Instruments 61. Capital expenditures increased dramatically over the past few years, from 6% of GDP in 2010 to 11.4% in 2012 and again to close to 16.8% of GDP in 2013, and then declined to a still high 9-11% of GDP in 2015-2016 before more radical cuts to 6.3% in 2017 (Table 3.3 below). TABLE 3.3 Mongolia - Capital Expenditures, 2010-2017 2010 2011 2012 2013 2014 2015 2016 2017 Total Capital Expenditures (MNT billion) 574 1,253 1,908 3,218 3,286 2,095 2,594 1,227 On-budget 574 1,253 1,421 1,591 1,842 1,389 2,347 1,702 Investment (MNT bn) 515 1,031 1,267 1,254 1,506 615 1,264 1,070 Bt financing 1/ 2 23 87 221 Maintenance (MNT bn) 30 50 53 88 92 98 147 80 Foreign funded investment (MNT bn) 29 171 100 249 244 246 696 553 Promissory notes 95 658 DBM-non-commercial projects (MNT bn) 2/ 431 240 Off-budget 487 1627 1444 706 247 -475 DBM-non-commercial projects (MNT bn) 2/ 258 885 642 DBM-commercial projects (MNT bn) 3/ 231 639 845 706 247 -475 (In % of GDP) Total Capital Expenditures 5.9 9.5 11.4 16.8 14.8 9.1 10.8 4.5 On-budget 5.9 9.5 8.5 8.3 8.3 6.0 9.8 6.3 Investment 5.3 7.8 7.6 6.5 6.8 2.7 5.3 3.9 BT financing 1/ 0.1 0.4 0.8 Maintenance 0.3 0.4 0.3 0.5 0.4 0.4 0.6 0.3 Foreign funded investment 0.3 1.3 0.6 1.3 1.1 1.1 2.9 2.0 Promissory noted 0.4 2.7 DBM-non-commercial projects 2/ 1.9 1.0 Off-budget 2.9 8.5 6.5 3.0 1.0 -1.7 DBM-non-commercial projects 2/ 1.5 4.6 2.9 DBM-commercial projects 3/ 1.4 3.3 3.8 3.0 1.0 -1.7 Source: Mongolian authorities; and WB staff estimations 1/: BT financing is included in the investment 2/: DBN’s non-commercial loans were not recorded in the budget until the 2015 budget 3/: DBM’s commercial loans remains out of budget but included in the overall budget targets for IMF and WB calculation. *: In 2017, Net commercial lending by DBM was budgeted as MNT 70 billion. However, it is recorded negative MNT475 billion by the end- Dec. Because DBM issued commercial loans in the amount of MNT 354 billion in 2017 while a loan of MNT 829 billion which was issued to “Erdenes Mongol (SOE)“ in 2012 was unexpectedly repaid to DBM in October and December. 87 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 62. Recent increases in capital expenditures were not due to increases in budget-funded investment, which remained below 10% of GDP in 2011-2012 and declined to 6% in 2015, but to a combination of: (a) foreign borrowing (1.3% of GDP in 2013), including billions of dollars in sovereign and sovereign- guaranteed bonds, and (b) systematic recourse to off-budget funding techniques, including loans from the DBM for non-commercial economic and social infrastructure projects and promissory notes. 63. To bypass the deficit ceiling in the FSL, which became effective in 2012, the government decided to use the DBM as an off-budget investment vehicle by injecting the proceeds of sovereign and sovereign- guaranteed borrowing into the bank. As a result, the DBM financed a wide range of economic and social infrastructure projects, in addition to its traditional commercial lending activity. DBM financing of public investment peaked in 2013 at 4.6% of GDP. 64. In 2014, the Parliament approved the use of promissory notes to contractors in lieu of cash payment. Public investments paid by promissory notes were not recorded under the cash-basis budget accounting system of Mongolia. The Promissory Note Program enabled line ministries to finance and implement large and small investment projects outside the budget framework and outside any central review and coordinating mechanism. 65. Figure 3.32 compares Mongolia’s capital expenditures with the performance of other LMICs. At about 9-17% of GDP in 2011-2016, Mongolia’s capital expenditures far exceed the 4.8% of LMIC and the 3% of East Asia Pacific countries in 2010-2014. An adjustment of about 11% of GDP was therefore needed in 2017 to bring Mongolia’s average in line with comparator countries. Aggregate Expenditures by Key Categories in 2003-2016 (Mongolia and other LMICs) FIGURE 3.31 Capital Expenditures and Net Lending (% of GDP): 2004-2017 14 12 10 8 6 4 2 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017e -2 -4 Maintenance to Capital Spending Ratio (%) ON-budget net-lending Capital Expenditures DBM’s corporate loans Maintenance Note: Net lending excludes DBM’s corporate loans. Source: MOF, World Bank staff estimates. 88 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES FIGURE 3.32 Capital Expenditures (% of GDP): 2003-2017 Capital expenditures Capital expenditures (LMIC) 14 12 10 8 6 4 2 0 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017e Note: LMIC means 17 lower middle income countries. Source: MOF, World Bank staff estimates. a) Distribution of Capital Expenditures 66. There are four categories of capital expenditures in the state budget: (a) Investment; (b) Capital Maintenance; (c) Equipment; and (d) Others (including feasibility studies). As shown in Figure 3.31 and Figure 3.33 investment projects dominate the budget. 67. Capital maintenance budgets are small as line ministries continue to prioritize new projects over adequate funding of operations and maintenance of existing capital assets (Figure 3.31 and Figure 3.33). A major weakness of the state budget is the low percentage of funds allocated to capital maintenance (only MNT 44 billion in 2016). A recent IMF study shows that Mongolia spends less than 0.15% of GDP on road maintenance (0.3 to 1% in other developing countries). Road maintenance accounts for only 20% of compared total road spending against 50% in other developing countries. This is obviously an area in which Mongolia should increase its public spending in order to protect its infrastructure and reduce future investment costs. FIGURE 3.33 Composition of Capital Spending by Key Categories: 2004-2017 Others Maintenance Total investment DBM-funded investment Budget-funded investment expenditures (in % of total Capex): RHS 100% 100 90% 98 80% 96 70% 60% 94 50% 92 40% 90 30% 88 20% 10% 86 0% 84 av g 03 04 05 06 07 08 09 10 11 12 13 14 15 16 7e -20 09 20 20 20 20 20 20 20 20 20 20 20 20 20 20 201 06 20 Source: MOF, World Bank staff estimates. 89 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 3.34 Central Government Capital Spending by Sector 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Infrastructure Other economic sectors Social sectors Other ministries Source: MOF, World Bank staff estimates. 68. Infrastructure projects dominate the distribution of on-budget capital expenditure by sector followed by social sectors (Table 3.4). Moreover, the share of infrastructure projects in off-budget spending is even more significant than in budget-funded capital expenditures. Table 3.5 shows the distribution of DBM’s non-commercial loans (in MNT billion) are shown. About 95% of these loans financed infrastructure projects, with road and transport projects accounting for more than half of the total. TABLE 3.4 Central Government Capital Spending by Sector (MNT billion) Sector 2000-2010 average 2012 2015 2016 2017 Allocations in Allocations in 2015 as % of 2016 as % of as % of total as % of total Actual Actual Actual Actual total total Plan Plan Plan Plan Plan   Infrastructure 91.1 69.5 735.3 653.3 191.4 229.4 780.4 1096.9 845.3 Construction and Urban 8.2 7.4 94.6 72.6 26.8 16.9 52.4 53.0 74.5 Development 45% 52% 48% 69% Road and Transport 61.9 43.2 480.4 418.8 142.8 199.9 575.2 953.1 617.1 Development Energy 20.9 18.9 160.3 161.9 21.8 12.6 152.7 90.8 153.6 Other economic sectors 11.1 4.3 143.2 100.9 55.6 11.4 37.3 26.8 68.7 Food, Agriculture and 2.78% 7.96% 2.38% 1.69% 7.1 2.7 116.2 77.0 40.7 6.4 21.5 23.0 55.6 Light Industry Environment and Tourism 4.0 1.6 27.0 23.9 14.8 5.1 15.8 3.8 13.0 Social sectors 37.5 36.5 350.5 304.6 240.5 167.8 452.8 294.4 258.5 Labor and Social Welfare 4.9 5.2 70.4 66.4 22.3 14.8 41.9 33.6 23.4 Education, Culture, Sci- 24% 24% 35% 19% 21.6 20.6 190.1 165.3 148.7 76.7 304.1 156.4 149.5 ence and Sport Health 11.0 10.8 90.0 72.9 69.5 76.4 106.8 104.4 85.6 Other ministries 29.5 33.9 21.82% 233.9 208.0 16.42% 99.9 73.0 15.15% 355.2 170.3 10.72% 275.3 Total 175.1 155.4 93% 1462.9 1266.7 100% 587.3 481.6 100% 1934.2 1588.3 100% 1447.7 Source: MOF, World Bank staff estimates 90 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES TABLE 3.5 Sector Distribution of DBM’s Non-commercial Lending Sector Total DBM lending in MNT billion % of total Infrastructure 2,399 - Construction and Urban Development 171 - Road and Transport Development 1,344 95% - UB City 302 - Energy 582 Other economic sectors 19 - Food, Agriculture and Light Industry 8 1% - Mining, Heavy Industry 11 Social sectors 2 - - Education, Culture, Science and Sport 2 Other ministries 108 4% Total 2,528 100% Source: MOF, World Bank staff estimates b) Role of Local Governments 69. Investment by and for local governments increased dramatically since 2010 (Figure 3.35). As fiscal decentralization efforts started since 2013, it peaked in 2014 and slightly declined in the following years. In recent years, there have been concerns about whether local levels of government are adequately equipped to perform such high levels of public investment, and there have been indications that existing capacity constraints mean that municipalities are being given more resources than they can handle. FIGURE 3.35 Local Government Capital Spending 50.0 2.5 40.0 2.0 30.0 1.5 20.0 1.0 10.0 0.5 0.0 0.0 03 04 05 06 07 08 09 10 11 12 13 14 15 16 1 7e 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 in % of GDP: RHS in % of Central Gov’s Capital Spending in % of Total Capital Spending Source: MOF, World Bank staff estimates 3.4.3 Diagnostic Assessment and Recent Developments on Legal and Institutional Arrangements for PIM Assessing the Current Legal and Regulatory Framework 70. The legal and regulatory framework for capital expenditure and PIM is formulated through a set of budget and investment laws and other regulatory legislation. Important laws for Mongolia PIM include the IBL (2011), the FSL (2010), the Law on DBM (2017 amended), the Law on the Concessions 91 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue (2010), and the Public Procurement Law (2005). Box 3.5. provides details on the key features relevant to PIM of these existing laws. The Public Procurement Law (2005) covers various methods of PIM procurement of goods, works, and services as well as procurement plans, procedures, and contracts. The Law establishes rules for both recurrent and investment budget purchases. The selection of which procurement procedure can be used is regulated within set threshold values assigned by the Cabinet subject to recommendations of the state administrative body for budgetary matters. BOX 3.5 The Current Legal and Regulatory Apparatus of PIM The IBL is the organic budget appropriation law that includes specific PIM legislation in articles for regulating capital expenditure and PIM. The IBL is the most important legal arrangement in the PIM system. Current IBL articles are relatively well-developed and structured to establish the key elements of PIM governance, providing clear legal mandates and establishing requirements for public investment programing and budgeting. Chapter Five of the IBL outlines in detail the Public Investment Program and Budgeting procedures. Articles 27-34 of the chapter mandate all relevant procedures and the roles and responsibilities of key stakeholders in PIM decision-making process.40 Article 27 describes how the budget proposal for capital expenditure should be prepared. Article 28 describes the PIM planning process and how proposals should be submitted and the methodology for prioritizing project proposals with required actions to be included in the PIP.41 Article 29 describes how to develop capital budgeting for PIM projects/programs. Article 30 describes how PPP projects should be managed. Articles 31-34 describe how the capital budget proposal should be approved in the State Great Khural. Government Resolution no. 123 (2010) also regulates how the public investment program should be developed and managed. It states that the projects and actions to be included in the PIP shall meet the documentation requirements as well as having demonstrated records of financial, economic, and social benefits. The FSL regulates the amount of total capital expenditure and the management of DBM projects. Article 12 of the FSL controls total capital expenditure of the MTFF Statement through an indicator including consolidated budget balance estimated with structural budget revenue, public investment, total budget revenues and expenditures and a comparison of their actual results to the plans. Articles 16 and 17 of the FSL describe how DBM projects should be identified, prepared and managed. The Law on the DBM (amended in February 2017) and the Law on Concessions (2010) have both recently been pivotal in managing PIM expenditure and procurement. With the government deficit ceiling set by the FSL effective since 2012, the DBM Law has progressively worked in creating PIM projects, to the extent that the increase in capital expenditure from 2012 to 2015 has mostly been driven by the investment projects funded by the DBM. The Law on the Concessions is another important law in the PIM system given the number of Build-Transfer (BT) projects that have widely been implemented under the Law. BT project contracts are legally based upon the definition of Article 4 of the Law. The Public Procurement Law (2005) covers various methods of PIM procurement of goods, works, and services as well as procurement plans, procedures, and contracts. The Law establishes rules for both recurrent and investment budget purchases. The selection of which procurement procedure can be used is regulated within set threshold values assigned by the Cabinet subject to recommendations of the state administrative body for budgetary matters. Chapter Five of the IBL is a chapter for ‘budget preparation, submission, and approval’ in general, where many articles are regulating the PIM planning, 40 budgeting, and approval decisions. 41 Those prioritizing methods include the following: national long-term and mid-term development program, economic efficiency and return on investment, socio-economic benefits, public service standards, industry and regional development policy, medium term budget framework, and debt management law and other requirements specified by the strategy document. 92 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 71. Global trends suggest that PIM requires an appropriate legal and regulatory framework to ensure it functions well and that common structures and standards are applied consistently.42 The legal tradition and administrative culture in any given country will determine the most suitable approach, but where possible, the guiding principle should be to avoid using an overarching legal instrument to provide detailed procedural arrangements and instructions. Procedures and methods usually have to be adapted and changed over time, particularly in the early days of PIM reform, and the more authoritative a legal instrument is, the harder it will be to initiate such change.43 This principle points to the design and adoption of a hierarchical legal and regulatory PIM framework, probably consisting of three tiers of hierarchy from top to bottom: 1) Legal authority for the PIM system as primary legislation.44 2) Procedural instructions with high-level decision criteria and designation of analytical tools.45 3) Detailed methodological manual, criteria, standardized parameter values, and procedural documentation. Generally issued directly by the PIM coordinating agency (such as MOF) in the form of manuals, templates, and circulars, under the authority granted to it by the primary legislation. 72. Mongolia has a long way to go in developing such an appropriate legal and regulatory structure for state budget-funded PIM. As summarized in 6, the Integrated Budget Law should be the primary legislation at the top tier for establishing the legal authority in the PIM system, while the existing PIM articles are not enough to explicitly regulate the separate PIM stages of project identification, preparation, appraisal, selection and capital budgeting, and project implementation. The second tier of procedural instructions and rules is also weak and inadequate. Even though there are general resolutions regulating PIM procedures, state-financed projects are weakly regulated without specific tools and criteria. At the third tier of methodological manuals, no standard technical manual for PIM activity exists, although some donor-oriented, but not state-oriented, manuals are available. TABLE 3.6 An Overview of Legal and Regulatory Hierarchies for PIM in Mongolia Hierarchy/Projects State-Budget funded projects DBM, BT, Other PIM projects DBM Law no longer works with *Integrated Budget Law works but can the 2017 amendment Tier 1: be strengthened. Concessions Law should be Legal authority *Public Procurement Law works. strengthened and upgraded or replaced Tier 2: Weak resolutions in PIM. Weak guidelines Procedural instructions Tier 3: None None Methodology manual Source: World Bank staff 42 See recent World Bank PIM technical assistance reports for Cambodia (2017), Cyprus (2016), Georgia (2016), Romania (2015), Ukraine (2015), Zambia (2016), and so on. 43 See IMF (2007), Reforming Budget Systems Laws, I. Lienert & I. Fainboim, PFM Technical Note No. 2, IMF Fiscal Affairs Department. It argues that, since laws are more difficult to change than regulations, they should not include provisions if there is a strong risk that these will be abrogated or amended one to three years later. 44 Primary legislation may take several alternative forms such as elaborated in a budget law, in a fiscal responsibility/stability law, or even in a separate PIM Law like in Lao PDR and Vietnam. 45 Usually referred to as ‘secondary’ legislation. 93 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 73. The legal and regulatory framework for DBM projects are unusual compared to other countries, while, with the recent amendment, the DBM Law will be no longer be effective in budgetary investment projects. From now on, the transferred DBM budgetary investment projects should follow the regular PIM regulations and rules that apply to state-funded projects. The Concessions Law, which is weak in creating good practice concession agreements and risk-sharing obligations in BT projects, should also be strengthened and upgraded to effectively accommodate for real PPP type contracts. This may involve the drafting of a new law. 74. While the IBL and the resolution requirements for PIM are well presented, it seems that their actual implementation has been neither strict nor consistent with what the law or resolution stipulates. To facilitate the various IBL and resolution provisions, specific practical guidance should have been developed, but only weak guidance has been provided and disseminated46. For example, there is no detailed technical guidance for (pre-) feasibility studies, no operational guidance for project implementation, and no technical guidance for DBM or PPP project management in the budget process. As such, actual project planning, appraisal, and budgeting processes appear to have been managed by internal discussions and agreements among key stakeholders including the MOF and line ministries without clear technical guidance for setting rules and decision-making. 75. Before the recent developments such as the amended law on DBM, the DBM Charter, aspects of the Companies Law and the DBM budgetary investment projects subject to weaker project appraisal and selection process compared with the state-funded projects. They were all separately managed by the DBM and without any clearance or information sharing with the MOF47. The government began to use the newly established DBM as an off-budget investment vehicle by injecting the proceeds of sovereign and sovereign-guaranteed borrowing into the development bank48. However, according to the recent amendment of the DBM Law in February 2017, no new DBM financing to budgetary investment projects is allowed. All non-commercial loans and projects are transferred and integrated into the state budget under the MOF. 76. Most of BT projects are not internationally recognizable types of concession or public private partnership (PPP) project even though such projects have private partner involvement49. While Articles 20-28 of the Concessions Law describe the details of concession agreements and the rights and obligations of the concessionaires, it is observed that most BT projects so far do not require private partners to bear real responsibilities and risk sharing arrangement as typical concession projects do. The government is legally obliged to pay the private contractors upon the transfer of the assets to the government. As such, BT projects merely postpone the payment of public investment projects until the completion of the projects. 77. The Procurement Law works for PIM, for the downstream project and budget implementation processes, but seems weak in effectively connecting procurement process with other PIM processes of project planning and budgeting more broadly.50 46 For example, in implementing Article 28.3 of the IBL, detailed project prioritization rules and criteria should have been developed. To support Article 28.4-28.6, technical guidance for pre-feasibility studies and feasibility studies should have been prepared and rolled out, but they exist only as non-standardized guidance. To support Article 28.10, operational guidance regarding how to implement the concession projects and DBM funded projects should have been developed, but only poor guidance exists. To support Article 30, a guidance note for BOT/PPP project management should have been developed but currently does not exist. 47 The DBM Law amendment request that the DBM now runs as a commercial-based development bank since 2017. 48 The DBM budgetary investment projects have been funded by the proceeds of government’s borrowing which was provided to the DBM. This cross- liability structure between the government and the DBM created significant confusion on fiscal and debt accounting. From Budget 2015, therefore, all DBM budgetary investment projects were included, but not off-budgeted, in the budget statement. 49 There is no separate PPP law. The Concession Law is applied in every form of private sector participation. In the World Bank’s PIM, there are eight must-have stages. Procurement should be effectively linked to other PIM stages in project preparation and 50 planning, project budgeting, project adjustment and ex-post evaluation. 94 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES Institutional Roles and Responsibilities 78. The PIM institutional landscape in the GOM that comprises a large number of players and stakeholders (Table 3.7 ). The State Great Khural is the highest body approving all capital investment plans, projects, and policies. The assessment of roles and responsibilities of institutions responsible of PIM follows. 79. It is less clear, however, how the MOF and NDA coordinate their responsibilities in each PIM activity and stage. According to the Article 28 of the IBL, the NDA is also responsible for preparing, coordinating, and prioritizing the public investment projects. Article 28 states that the State Administrative Body Responsible for Development Policy and Planning Matters (which is the NDA) together with the State Central Administrative Body Responsible for Finance and Budget Matters (which is the MOF) shall establish priorities and implementation orders for investment projects. TABLE 3.7 An Overview of Key Institutions in Mongolia’s PIM and Capital Budget Process Institution Role in the PIM System  Enacts the State Budget by reviewing and approving all decisions of capital invest- State Great Khural ment policies, plans, and projects The Government  Reviews and decides all capital investment policies, plans, and projects, and sub- (The Cabinet) mits the State Budget proposals to the State Great Khural  Prepares and manages the public investment program (PIP) which is a base of the NDA State Budget projects  Compiles and prepares the State Budget and manages budget execution MOF  Estimates resources available for funding budget projects Public Investment  Manages budget execution (including procurement) for domestically financed Department of MOF public investment projects (PID)  Manages single treasury account Treasury Department of MOF  Provides cash payment and settlements for implementation of state funded capital projects and programs Line Ministries (LMs)  Prepare budget requests and provide financial and progress reports to MOF and Agencies  Implement and report on public investment projects Source: World Bank staff 80. The NDA’s role remains unclear in terms of vetting all capital project proposals from LMs that should be included in the PIP. In practice, as most of the capital budget is managed and funded by PID, however, LMs tend to deal directly with PID. Budget management in MDAs is managed separately through recurrent (non-development) and capital (development) budgets in dual budgeting systems. The recurrent and capital budgets are prepared separately in different formats. Recurrent costs for operations and maintenance are incorporated into the State Budget on an annual basis when the investment project is completed and the assets require operation and maintenance funds. According to the PID, the funds available for this are usually deficient, as earmarked external funding cannot be used for this purpose. The PEFA 2015 report assessed and rated this as ‘C’ because ‘budgeting for investment and recurrent expenditure are two separate processes with the former being the responsibility of the previous Ministry of Economic Development (MED) and the latter being the responsibility of the MOF, 95 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue with little or no coordination between them’. In the new government structure since 2016 with MOF and NDA, rather than with the MED and MOF, it is still unclear how the MOF and NDA coordinate the recurrent and capital budgets with each other. The NDA oversees the PIP, which is based on various national development policies and the parameters of the state budget. 81. Procurement for externally financed projects and programs seems governed by the regulations of the respective development partners or lenders undermining national procurement systems. The Public Procurement Law establishes roles and responsibilities across thresholds for both recurrent and investment budget purchases and envisions the gradual delegation of procurement authority to line agencies. Line ministries submit procurement plans for state budget financed investment projects to receive financing. Budget execution and procurement are also managed by the PID, with delegation of procurement to line ministries and spending agencies. 82. The Treasury Department (TD) of MOF manages the public Treasury Single Account (TSA) and the provision of cash advances for implementation of capital projects and programs substantially ties up cash that could otherwise be invested. As such, this practice limits the TD’s ability to manage the cash position of the public sector and puts the TSA at risk51. 83. A review of the roles and responsibilities of the major institutions reveals some gaps in the World Bank’s eight must-have requirements of good PIM practice. Figure 3.36 shows the roles and responsibilities of major institutions engaged in the PIM system in Mongolia. Among the features of good PIM practice, independent review role/responsibility in feature 3 and ex-post evaluation role/responsibility in feature 8 are missing or non-existent in Mongolia. In the upstream decision process (features 1-4) as well as the downstream implementation process (features 5-8), the gatekeeping oversight role/responsibility of the MOF is limited and in some cases, such as BT and BOT, is almost non-existent. These weaknesses put the outcomes of projects at risk and allow fiscal risks in these projects to go unmanaged until it is too late to avoid them. In features 1-3 and 7-8, the MOF role/responsibility versus the NDA role/ responsibility is less clear and needs to be clarified. FIGURE 3.36 PIM Institutional Arrangements: Roles and Responsibilities of Key Institutions Feature 1 Feature 2 Feature 3 Feature 4 PIM Strategic Guidance Project Independent Project Selection Features and Planning Appraisal Review and Budgeting PID of MOF with Responsible MDA, LMs, MDA, LMs N/A LMs Divisions by Entity MOF Government Feature 5 Feature 6 Feature 7 Feature 8 PIM Project Project Facility Operation Ex-Post Features Implementation Adjustment and Maintenance Evaluation Responsible MDAs, PID and TD LMs MDAs N/A Entity of MOF PID of MOF Note: NDA, Line Ministries (LMs)), MOF, None Available (N/A), PID, TD, MDAs. Source: World Bank staff As the government employs the GFMIS with streamlined business processes, the need for cash advances in an automated environment is largely 51 eliminated because all budget allocations/controls and payments could be made through the system in a timely fashion. 96 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 84. Clarification of roles and responsibilities between key ministries and departments could be enhanced further to strengthen inter-ministerial coordination and collaboration, ensuring a more integrated approach to capital budget management. There has been limited coordination of roles and responsibilities in PIM and the associated decision trees within the Government. The limited number of MDAs involved in conceiving and implementing public investment projects with weak project appraisal capacity, with insufficient MOF oversight, makes coordination weak. A clear allocation of roles and responsibilities for each feature of the PIM system is essential to enhance arrangements for PIM decision making. This is particularly important where project assessments must be performed and reviewed and formal decisions made at key control points on whether to proceed further with a project proposal. 85. Furthermore, the roles and responsibilities of the key players can be differentiated for each PIM stage. Specifically, the roles can be differentiated as those of a ‘project proposer,’ ‘project appraiser,’ ‘project reviewer,’ and/or ‘decision maker.’ The system must embody clear decision steps/control points with a designated decision maker who determines whether to proceed with, or stop, a project. The allocation of roles and responsibilities could differ according to project size, sector, level of government, or all three. Similar arrangements are contained in Vietnam’s 2014 Law on Public Investment Management, which categorizes projects into a four-group hierarchy (Nationally Important Projects and Groups A-C) by investment cost, sector, and potential for significant environmental/social impacts then specifies a corresponding hierarchy of decision rights from the National Assembly down to the People’s Committees at lower levels of government52. It is desirable that the Government of Mongolia can more specifically elaborate some degree of hierarchy in the designation of roles and responsibilities. 86. In case of the former DBM non-commercial projects recently transferred to the MOF state budget, and the BT and BOT projects implemented under the Concessions Law, the roles and responsibilities of key players could be further clarified and differentiated in stages of controlling and managing the PIM processes. If the DBM budgetary projects are successfully integrated into the state budget, it is evidence of PIM advancement. For safely seating the transferred DBM projects into the state budget 2018-20, however, the role and responsibility of the PID of MOF should be strengthened in undertaking the management of combined budget projects including the formerly DBM projects. Given that the former DBM projects may not have been well prepared and appraised, the costs and scope of those projects should be tightly monitored and screened by the PID. For effectively managing the BT and BOT projects, (where there seems to be evidence of increasing numbers of BT and BOT projects) the PID’s role should also be strengthened in identifying and costing the obligations of the government in the project contract agreement. Some recent developments - New policy objectives 87. A major objective of the Economic Reform Program (ERP) was to reduce capital expenditures to around 6% of GDP and to prioritize projects financed on concessional terms by multilateral and bilateral partners. Significant results have been achieved in this regard. Investment projects financed by foreign grants and loans more than doubled from 1.1% of GDP in 2015 to 2.9% in 2016 and 2% in 2017. Budget-funded investment which had declined to 2.9% of GDP in 2015 was stabilized around 3-3.4% of GDP in 2016 and 2017. Finally, two critical reform measures were taken including stopping DBM’s non- commercial lending and terminating the Promissory Note Program. 88. A first reform was the government decision, in 2015, to include in the budget all the non-commercial loans financed by DBM. In 2017, a more radical measure was adopted as the supplementary budget stopped completely the DBM’s non-commercial lending operations. The DBM law was amended to transform the bank into an independent, self-sustainable, financial institution focused only on commercially viable projects. In 2017, 385 DBM loan accounts for 345 non-commercial projects were See the Law on Public Investment Management in Vietnam, Articles 6-10 for classification of PIM projects and the criteria. 52 97 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue transferred to the MOF. Despite some repayments, outstanding loans for these non-commercial projects now total over MNT 3.1 trillion, including foreign currency loans of US$ 667 million. DBM’s commercial loan portfolio, which has a book value of MNT 2.8 trillion for about 1,600 individual projects, will be monitored and reviewed through third party audits leading to appropriate remedial actions. 89. The first consequence of terminating the promissory note program was to increase total capital expenditures. In 2016, a new government decided to clear MNT 658 billion in payment arrears under the program thus increasing the capital spending/GDP ratio to 10.8% that year. This, however, was a one-off expenditure and total capital expenditures declined again to 4.3% of GDP in 2017. For the future, the capital spending/GDP ratio is expected to be stabilized around 6% of GDP and to decrease to less than 6% in the 2020s. 90. Stopping DBM financing of public investment and promissory notes were two important reforms essential to improve the fiscal performance of the government. However, government agencies continue to prepare a wide variety of new projects and the temptation to use non-conventional techniques has not disappeared. BT and BOT financing are two forms of so-called PPP that have been used to finance new projects. Involving the private sector in the financing and management of public infrastructure and other public investment projects is a sound decision, provided projects selected are of high priority and the distribution of risks between the public and private sectors is adequate. 91. In this context, BT financing does not seem to be the most appropriate type of PPP. BT agreements provide that the government must pay the private contractors upon the transfer of the assets financed by the project. In effect, BT financing has many common features with the now banished Promissory Note program. BT financed projects in the road sector tend to be six times costlier than budget funded road projects (World Bank 2016). As of June 2018, accumulated payment obligations for the State under BT contracts are estimated at MNT 1.5 trillion, payable by 2022. Unlike promissory notes and DBM funding, BT financing continues. However, a new resolution limits the use of BT financing to projects of high economic and social value that do not require repayments before 2020. 3.4.4. Priority Areas of Reform 92. The present means of preparing and executing public investment projects in Mongolia does not represent a PIM system that could be recognized as good practice. The arrangements are completely fragmented and there is no way of identifying the efficiency and effectiveness of development activity. A roadmap for improving PIM in Mongolia includes 17 actions organized around 6 themes, including (a) developing a comprehensive and consolidated approach to PIM, (b) clarifying roles and responsibilities, (c) developing a PIM database, (d) rationalizing the ongoing PIP portfolio, (e) managing existing BT activity and strengthening BOT/PPP management, and (f) upgrading the allocation of maintenance budget. Developing a Comprehensive, Consolidated and Cohesive Approach to a PIM System 93. All development activity should be framed by the expected results of the investment project rather than through the source of funding. The source of funding should become the secondary consideration after projects have been identified as being of value to the nation. To that extent all (regardless of funding source or type) initial project proposals in MDAs should be assessed through a standard pre- screening template that should be completed by project proposers. The template would become the single point of entry on to the PIM system in Mongolia. Implementing this would remove the ‘funding bias’ that favors projects that have identified non-budget external funding regardless of whether the project is a priority or not and regardless of the direct and contingent liabilities that may be incurred as a result. 98 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES 94. If this system is implemented successfully, it should only be possible to enter into the PIM system if the project proposal is driven by either government national or sector policies. This would provide the authorities with a useful tool that could be used, when necessary, to contain expansionary pressures on the PIP. BOX 3.6 Purpose of Prescreening Prescreening represents a common point of entry in to a formal PIM system. It prevents financial and human resources from being wasted on feasibility studies for projects that would never be funded and assists in reducing pressures on the budget by keeping the project pipeline at a manageable level. If implemented properly it is also an effective tool in reducing the possibility of unsustainable extra-budgetary finances being sourced without consideration of the wider fiscal consequences. Pre-screening is also an important first quality step that allows the proposing authorities to test the robustness of a project concept in terms of logic, risk and sustainability. It is also in the interests of the proposing authorities that their projects provide real solutions to real problems and that they are planned and implemented effectively so as to avoid constant problems throughout the oper- ational life of the project. In good practice countries, all projects, large or small, are expected to follow the same formal pre-screening process. A positive decision at this first stage of entry into the PIM system requires that the strategic policy relevance, rationale and realism of the project propos- al is demonstrated to the satisfaction of the decision-makers and a convincing case has been made to justify expenditure whether it is to consider funding to directly implement a small project or on further planning and appraisal for larger ones. 95. As a public investment planning tool, the GOM should consider the development of a rolling forecast for sustainable annual public investment expenditure within a MTFF/MTBF. This would not only provide guidance to line ministries as they plan future projects but it will help the PID of MOF to withstand unrealistic demands during the budget round each year. Given the unpredictability of government revenues due to commodity price fluctuations, the forecast should predict a range of scenarios (loosely ‘Base Case’, ‘Best Case’ and ‘Worst Case’), which may be called upon closer to the actual budgeting period when revenues are more likely to have a greater prospect of accuracy. The forecast should also identify and estimate the associated recurrent costs of projects and indicate the impact of the year in which they complete. For this reason, also, the forecast should cover commitments for other forms of finance such as BOT and Foreign Financed projects which, when they complete, often incur significant recurrent costs. 96. The GOM should move quickly towards an entirely rules-based PIM system with new regulations and templates which are also backed with guidance. This is to ensure consistency in the system and even- handed treatment in the way that projects are identified, prepared, appraised and selected. To achieve consistency in project appraisal, project proposers should follow a standard template system, which covers all the required steps of appraising a project. 97. A project proposal should only enter the PIM system formally with a Unique Project Number (UPN) which is assigned when it has been successful in the initial pre-screening. This allows tracking of the project as it moves through the various stages of appraisal and, if ultimately selected, implementation. The UPN would be a foundation of the proposed project database and would allow all contracts and related expenditures to be tracked against the original proposal. This would make project monitoring and (eventually) project completion reports and ex-post evaluations easier to undertake, yielding valuable lessons to the GOM that can be used to improve the formulation of future projects. 99 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 98. When projects have been successfully pre-screened and allocated a UPN, they should be formally appraised to a degree that is commensurate with the scale of the project. Project appraisal should become a default position for all projects. Unlike on large projects which require full feasibility studies and cost benefit analysis, appraisal does not have to be a complex and time-consuming exercise for small projects but the necessary time and effort should be dedicated to proving the need for the project and that it is a worthwhile and efficient use of public funds. The type and scale of appraisal should be proportional to the nature of the project and the government should regulate appropriate thresholds at which they apply. Clarifying Roles and Responsibilities 99. As the entity responsible for managing the public finances, it is essential that the MOF is fully aware of and able to manage the commitments that MDAs are making on behalf of the GOM. The present unsatisfactory situation regarding unmanaged PIM activity and debts accumulated through extra-budget funding has happened whilst the role of the MOF has been restricted to somewhat limited direct budget- funded activities only. The MOF should be made aware of all financial commitments related to projects before they are entered into. This should be made mandatory through an appropriate regulation. 100. The role of some institutions engaged in PIM appears to be in flux and quickly needs to be clarified. The amended Concession Act implies a reduced role for the NDA with a rebalancing of authority back to the MOF. Clarification of these amended roles needs to be accompanied by adequate resources to fulfil them. Furthermore, the powers of line ministries to enter into extra-budgetary funding arrangements need to be curtailed. 101. The GOM should identify and empower a single entity (for convenience hereinafter ‘PIM Unit’) to develop and manage the PIM system and the PIP within in a single cohesive framework. Although MDAs and local government should retain responsibility for identifying, preparing and executing public investment projects, all this activity should come under the overview and supervision of a PIM Unit regardless of the form of procurement, the form of contract or source of funding. Additionally, the MDAs should prepare their projects within the PIM framework, which would be strengthened and monitored by the PIM Unit. The PIM Unit would also be responsible for reviewing and checking project proposals for quality assurance purposes. Internationally, PIM Units are often found in ministries of finance but not exclusively so. If a PIM Unit is not located within a MOF, it is still essential that the highest level of cooperation is undertaken with the PID in MOF. Alternatively, the PID could be converted into the PIM Unit. 102. A quality management system should be developed to check the quality of project proposals through identification, preparation, appraisal and readiness for implementation stages. This should start with the pre-screening step, which would check whether projects have been correctly identified and whether they are appropriate. Further checks should then be developed after the appraisal stage prior to the procurement procedure and again following implementation prior to the project becoming operational. The proposed PIM Unit would be responsible for the development of the quality management system, its implementation and its subsequent operation. The overall objective of the quality management system would be to ensure that only correctly identified, fully prepared project proposals with credible implementation plans are selected for funding (regardless of the source of the funding). Developing PIM Database 103. A project database that is the single source of information for all public investment projects should be developed. To create a cohesive picture of all development activity in Mongolia all projects regardless of the source of funding would be assigned UPNs on condition that a proposal passes through the first quality check at the pre-screening stage. This UPN would stay with the project throughout its entire useful life in the operational phase or until activity ceases in the case of rejected or abandoned projects. This will allow all capital estimates to be updated and expenditure to be tracked against 100 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES individual projects across multi-year implementation and across all contracts that are related to the realization of the project. It should also include for the first time a tracking item that includes and updates projected operational and maintenance costs related specifically to the project throughout its development phase. This will help the PID at MOF to forward plan budgets based on the future as well as existing commitments and planned expenditures on an aggregate basis, and thus helping them identify an available envelope for possible new projects with more confidence. This can be taken forward after the rationalization of the Public Investment Program. Rationalizing the On-going Public Investment Program (PIP) Portfolio 104. Rationalize the Public Investment Program (PIP) in order to re-structure and/or remove non- performing on-going projects from the portfolio so that the financing demands of the portfolio can be prioritized and more effectively aligned with available fiscal resources. Similar rationalization efforts have already been made in the governments of Romania in 2015 and Ukraine in 2016-2017.53 Portfolio rationalization is an urgent issue in Mongolia, because a lot of public investment projects, including more than 480 state budget projects (including 119 projects transferred from the DBM non-commercial investment projects), around 55 BT or BOT projects with contracts signed, and many other projects under preparation and/or appraisal to be adopted, are already in the PIM pipeline seeking funding. A major part of the portfolio has been seriously underfunded which has led to a huge and unaffordable overhang of partially implemented projects. BOX 3.7 Example of Rationalization Criteria A rationalization guideline could be established as a resolution of the Cabinet, providing clear guidance and criteria on how under-performing on-going projects can be effectively screened, identified, prioritized, resolved, and managed. The objective of a portfolio rationalization guideline is to improve the effectiveness and efficiency of public investment expenditure under deteriorated fiscal space conditions by prioritizing well performing on-going projects or restructuring under-performing ones from the PIP. For the scope and coverage of the rationalization guideline, the guideline will cover the portfolio of all on-going/uncompleted state level public capital investment projects, whether funded by State Budget (former DBM included), BT, BOT, or foreign-financed. The proposed rationalization procedure will focus on: (a)  identifying under-performing projects from within the overall PIP; and (b) removing or suspending these projects from the PIP and subjecting them to a rationalization procedure to resolve their status. This would involve re-costing, re-structuring, postponing, closure and liquidation, or curtailed completion and handover. In the absence of a comprehensive PIP database, it will be necessary to carry out a detailed survey/review of uncompleted projects led by MOF and LMs. Possible examples of the Rationalization Criteria include checking the strategic relevance/importance, project implementation progress, and project feasibility. The ‘strategic relevance/importance’ criterion is to ensure the relevance/importance of the project in medium-term and long-term national and sector plans. ‘Implementation progress’ criterion checks whether the physical construction is over 50% or other percentage , and whether the remaining funding requirement is less than the percentage of total estimated cost. The ‘project feasibility’ criterion is to assess the socio-economic feasibility with information on demand estimates. This analysis, however, cannot include economic feasibility in the assessment criteria, due to the lack of sufficient cost and time to collect and assess individual feasibility studies (feasibility studies shall be done again for some selected ones only, if necessary). As an auxiliary criterion, contractual obligation is checked additionally. Projects financed by foreign loans or grants have often contractual obligation, and these projects should be additionally prioritized. Similarly, projects that already signed concessions also received an additional consideration. 53 For country examples of rationalization efforts, see the World Bank technical assistance reports as follows. For Romania, “Report on Rationalizing the Existing Public Investment Portfolio (2015)”, and, for Ukraine, “Development and Introduction of a Methodology for Public Investment Portfolio Cleaning (2016)”. 101 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue Managing BT Activity 105. The system of BT financing for projects should be closed down permanently by removing loopholes created by the amendment to Resolution no. 37. BT represents a form of borrowing that deliberately aims to defer the cost of payment for projects to future years. It is an expensive form of finance and represents a huge distortion to the present PIM system. The public investment plan therefore needs to account for the future repayment schedule of the BT projects that are currently under implementation. There is a risk that these repayments will eventually crowd out budgetary investment projects under tight revenue conditions. Its continuing presence is also a major constraint to any reform program aimed at moving PIM practices in Mongolia to what might be considered good international practice. 106. BT repayments should be prioritized to clear the arrears as soon as possible with the contracts incurring the most punitive penalties and interest charges being cleared first. BT schemes represent an expensive form of finance that is already accruing interest and penalty charges for late payment - making it even more expensive. The current slow rate of repayment adds significantly to the arrears and runs the risk that some of the investments made under BT will be no longer in operation or become obsolete before they are paid for. Strengthening BOT/PPP Management 107. The increasing scale and volume of BOT/PPP projects in Mongolia demands urgent actions to be taken to strengthen implementing and monitoring capacity. Unlike BT procurement, BOT is a more credible form of Public Private Partnership (PPP) as it usually involves some genuine risk taking by the private partner. Already the value of investments through BOT have reached MNT 3,767 billon and appear to be set to increase further as project proposers see other non-budget funding options being closed. If done correctly BOT/PPP projects can bring efficiencies in to public investment projects and deliver effective outcomes. Furthermore, maintenance provision is the responsibility of the private provider. However, if they are done incorrectly the fiscal consequences could last for decades. Given the stark contrast between success and failure, it is imperative to dedicate sufficient capacity to the subject area and ensuring that the regulatory and institutional arrangements are appropriate. 108. All further activity related to PPP should be focused on identifying and preparing opportunities for implementing projects with BOT/PPP within a single cohesive PIM system. PPP projects should not be generated through separate and parallel legal and institutional arrangements but as a part of the integrated PIM system that recognizes projects of value to the nation regardless of the form of financing and implementation. Therefore, PPP should not be used as a creative accounting technique; instead it should be employed when it is more efficient and effective to do so and it can be assessed qualitatively that a PPP implementation is likely to deliver a better outcome than more conventional procurement. The GOM will need to focus on the following additional subject areas in addition to other normal considerations in preparing an investment project: (a) Is the project feasible or interesting to investors as a PPP? Soft market test results are required to make this assessment and judgement; (b) Value for Money assessment compared with conventional implementation; (c) Fiscal Risks – analysis required. 109. Fiscal risks that come with all BOT/PPP contracts should be understood, quantified, recorded in accordance with the DML and monitored. The current stock of BOT/PPP projects, estimated at MNT 3,767 billion, contain unknown contingent liabilities and there is an existing legal requirement to report them. An urgent exercise should be undertaken to evaluate existing BOT/PPP contracts and assess the contingent liabilities contained within. These should then be reported to the DMD. 110. New BOT/PPP projects should not be signed until the fiscal consequences have been identified and quantified with MOF. This includes an assessment of both explicit and contingent liabilities within the final draft contract prior to signature. An assessment of the cumulative effects of new liabilities on top of existing ones should also be undertaken. The GOM may also wish to consider placing a ceiling on the total amount of PPP liabilities it incurs. 102 AGGREGATE EXPENDITURE - TRENDS AND CROSS-CUTTING ISSUES Upgrading the Allocation of Maintenance Budget – Investing in Existing Assets 111. Maintenance costs and all other operational costs should be estimated at an early stage of project preparation. Projects should not be selected without sight of these costs. Without adequate maintenance, assets resulting from investments quickly deteriorate, leading to underperformance, increasingly disappointing outcomes and ultimately increased pressure to replace them earlier than expected. It is also impossible to adequately appraise an investment project without an appraisal of the operating costs. There is no ‘golden rule’ for the provision of maintenance budget levels, which in any case are usually provided for in the operational recurrent budgets of MDAs. In the case of BOT/PPP projects the provision, and therefore the cost of maintenance, is almost always the responsibility of the private company54. This means that the future maintenance costs are estimated by the private company in its business plan prior to submitting its tender for the project. Similar behavior is therefore required of public entities in preparing their project proposals for appraisal. The costs of future maintenance should therefore be estimated as a matter of course along with all other future operational costs associated with the project. In a school building proposal, for example, the operational costs should estimate annual teacher and other staff costs, the costs of heating and other utility costs, and the costs of providing other goods and services that might be expected in a school. In this way the MOF is better able to plan its future recurrent provisions for completed projects. 103 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue CHAPTER 4 Sectoral Spending OVERVIEW AND CONCLUSION Education and Health sectors Key Highlights  Mongolia’s education spending ratios are average when compared with similar countries, but health spending is low. Total education spending peaked at 6% of GDP in 2014 with public education spending in 2014-2015 at 18-19% of total government spending, comparable to that in Mongolia’s regional and economic peers. However, at just over 4%, the share of total health spending in GDP is lower today than that in 2003 (6.7%), and lower than that in most other middle- income countries. Since 2010, government health spending has fluctuated between 6-8% of total  Reorganize services to address chronic illness by government spending - also consistently low. focusing on high-quality primary care integrating preventive and curative services;  Recurrent expenditures dominate in both sectors. Although wage spending per education worker  Improve efficiency by basing capital investments declined somewhat in recent years, while that in on strategic long-term goals; and the health sector has been growing slowly, the  Align incentives to support integrated service wage bill remains the fastest growing category of delivery and improve management at the local expenditures in both sectors. Sector planning is and provider level. often not based on clear medium- or long-term strategies. For example, Mongolia spends a lot  Projected increases in required state subsidies can on expensive school equipment but allocations be attributed to growing pension system old age to essential classroom learning materials are dependency ratios, and a number of legal changes extremely low. Capacity is unused in some areas introduced in recent years that dramatically add while severe constraints exist in others. And to the pension costs. efficiency of capital spending is weakened by extremely low routine maintenance expenditures  Increases enacted in contribution rates marginally in both sectors. improve the financing of pension insurance but are insufficient to offset the effects of other measures  Pension enacted in 2017 and 2018 (including a service buyback arrangement, reduction in herders’ Key Highlights retirement ages, and supplemental benefits  Managing the growing fiscal burden resulting from for herders and mothers, thereby affecting the accelerating state subsidies to pensions is one of authorities’ objective of limiting the State Subsidy key medium-term challenges of any Mongolian to 2% of GDP by 2030. Furthermore, those changes government. Under current conditions, state negatively affect the principal design, incentives subsidies to pensions are projected to rise from 2.0% and fairness of the pension system. of GDP in 2016 to 6.0% in 2030 and 11.0% in 2050. 104 PUBLIC EXPENDITURE REVIEW Elements of a Priority Reform Agenda Pension: Education: maintain overall expenditures but improve  Adjusting the service buyback legislated in 2017 efficiency. so that the cost of the buyback for beneficiaries is better aligned to the expected benefits.  Benchmark performance by participating in PISA and strengthening national assessments, as  Increasing the retirement age. absence of standardized assessment data makes  Reducing the minimum pension and the partial it impossible to evaluate the effectiveness of the pension and linking them to a transparent index education system in producing learning outcomes such as the Minimum Living Standard.  Reduce inequities in quality of learning  Establishing an automatic pension price indexation environment across schools and kindergartens of pension benefits. and give schools autonomy to manage operational budgets;  Adjusting early retirement for herders & special professions through actuarially fair benefit  Expand access to early education through reductions for retiring early and supplements for community or home - based options for herders; late retirement.  Explore potential for increased parental contributions among the well - off; and  Strengthen planning of capital investment including through increased transparency. Health: Implementing the proposed reform agenda may require a gradual increase in budget allocations to health and its share in total public spending as overall public sector reform reduces public finance deficits to more sustainable levels. 105 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 106 SECTORAL SPENDING 4.1 Education Sector 4.1.1 Recent trends in expenditure 1. Enrollment level of 100% in primary and vocational education and a lifelong training system is one of the ten goals of Mongolia’s SDV approved by the Parliament in 2016. Substantial progress was made, over the past decade, towards near universal coverage in basic education. However, upon graduating from the education system, many young Mongolians are not adequately equipped with skills in high demand and one-fifth report that the biggest obstacle to securing a job is education-related. 2. Breakdowns at multiple levels of the system contribute to the problem. First, and most critically, access to quality early childhood education (ECE) services is low, especially among the most disadvantaged, leading to large differences in school readiness. Second, absence of standardized assessment data makes it difficult to evaluate the effectiveness of the system in producing learning outcomes. Third, the large socioeconomic gaps revealed by early grade literacy and numeracy data indicate low-quality foundations, undermining the efficiency of education investments at the secondary and higher levels. These shortcomings in the foundations of skill-building compound over time as skill formation is a lifelong process and skills beget other more complex skills in a complementary and dynamic way. Despite a desire for a ‘whole-child’ approach in education, national curricula and classroom practices remain focused on academic skills, so that when students transition from school to work, they lack the so-called ‘21st century’ socio-behavioral skills for which employers report high unmet need. These include problem-solving skills, critical thinking, working in teams, and so on. Finally, higher education and technical and locational education and training policies remain divorced from labor market needs, leading to poor quality of technical or job-specific skills, low relevance, and mismatches. 3. Mongolia’s total and public education spending is viewed as average as compared with other countries at similar levels of GDP per capita. Overall, total education expenditure relative to GDP exhibited a slightly upward trend over the last decade as it increased from 4.6% of GDP in 2005-2006 to 6% in 2012- 2014 but declined to less than 5% in 2015-2016 (Figure 4.1). Public spending on education increased from 14-16% of total government spending in 2006-2010 to 18-19% in 2013-2014 but declined to 13% in 2016. FIGURE 4.1 Trends of Spending in Education, 2005-2016 6.5 6.0 5.5 5.0 4.5 4.0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: BOOST Database, MOF. 107 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 4.2 Nominal Government Spending on Education in Mongolia, 2005-2018, Total and by Economic Classification 1,600 1,400 1,200 billion MNT 1,000 800 600 400 200 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Total Current Capital Source. MOF (2017). Note: The base year for GDP-deflated real expenditures in Figure 4.1 is 2010. 4. Recurrent spending accounts for the bulk of the sector’s total budget, averaging about 85% in 2015- 2016, while capital spending claims the remainder. In real terms, per capita recurrent expenditures grew by an average of 8% per year between 2005 and 2016, but this average masks considerable variations over the years. Recurrent education spending declined in real terms in 2015 and 2016. 5. Wages, salaries and bonuses constitute the largest component of recurrent education expenditures. In 2017, wages, salaries and bonuses accounted for 53% of total recurrent education spending. In real terms, wage expenditure grew at an average of 21% per year between 2005 and 2016, and was the fastest growing recurrent expense category in this period. In some years, however, wage expenditure showed no change in real terms (2008-2010), and failed to adjust for inflationary pressures in 2015-2016. 6. The other important spending categories include ‘fixed’ operational costs, government transfers, ‘standard’ costs covering health and nutritional needs among some students, and services contracted to others. Fixed expenses related to office buildings (electricity, heating, clean water sewerage, and office rent) constitute the second largest component (11.2% in 2018), followed by current transfers55 (excluding the government domestic transfer - 9.6%), standard costs (medicines, meals, and clothing and bedding - 6.7%), services by others56 (6.6%), and social insurance (6.2%). 7. Wage expenditure per education worker declined in recent years, and further analysis is needed to determine how the wage level compares with general labor market trends in Mongolia. In fact, it is hard to make the case that the wage level is high in absolute terms, based on these data alone. In real terms, growth in wage expenditure per worker was modest until 2015, and wage expenditure The 9.6% current transfer exclude the domestic government transfer to local governments, but include government foreign transfers and ‘other’ 55 current transfers (other bonus and benefits by employers; benefits and allowance by the state; home leave travel allowance; one off benefit, allowances, and bonus; hardship benefit to civil servants working in rural areas; and compensation to public employees who got laid off due to restructuring). 56 Work, service payment, and fees for services performed by others include other common payments for contracted-out services and works; audit, credit rating and certification fee; Insurance services; vehicle taxes; vehicle inspections; IT services; land fees; bank and financial institution’s service fees; and printing of state forms. 108 SECTORAL SPENDING per worker declined by 5% in 2015 and 2016. Studies based on 2005 data show that teachers’ wages in Mongolia were high relative to other countries that experienced the transition from Soviet to market systems (World Bank, 2016). However, a comparison with estimates based on OECD data for 34 countries indicate that teachers’ wages are about one-fifth lower than the values suggested by the regression model. The National Statistical Office (NSO) estimated that monthly wages in the education sector averaged MNT 802,000 as of 2017Q4 compared to MNT 947,000 for Public administration and Defense, MNT 865,000 for Health sector and MNT 1,039,000 for the national average. 8. Expenditure on classroom learning materials is extremely low, accounting for under 1% of recurrent spending on education. With limited government spending on learning materials and supplies, kindergartens and schools are forced to largely finance these items including through contributions from parents, fund raising, or transfers from local governments. Theoretically, we expect this capacity to mobilize local or community resources to vary greatly in different parts of the country, leading to differences in the material learning environment across schools and kindergartens – for instance, in terms of availability and quality of learning materials, their variety, and the range of classrooms activities offered. In rural and UB ger-area kindergartens, the learning environment and provisions for learning materials and toys are highly inadequate, especially in relation to numbers and cognitive skills development (World Bank, 2017). Likewise, in schools, reading and math books are either too few or in bad condition and in need of replacement. 9. Although volatile in recent years, real capital expenditures peaked in 2014 as they accounted for 22% of total education spending, mainly supported by buildings and facilities, and equipment. The lowest share recorded was in 2005, when capital spending constituted 5.6% of total spending. 10. Repair expenditure, whether as part of capital or recurrent expenditures, is very low. Within recurrent spending, ‘renovation expenditures,’ the only expense category capturing repair expenditure, never exceeded 0.02% of recurrent spending between 2005 and 2018. Within capital spending, expenditure on repairs stood at 5.3% in 2018. Experience shows that deferred routine maintenance ultimately leads to higher overhaul maintenance costs and indicates inefficient expenditure planning. Many buildings constructed in the 1960s and 1970s are now in need of rehabilitation. It is also unclear how this situation will evolve under decentralization of authority over budget decisions pertaining to capital expenditures. In this context, monitoring local capital investments should be a priority for central government authorities, and a rapid assessment is needed to determine the rehabilitation needs of education facilities in rural areas and improve provisions for routine operations and maintenance in school budgets. 11. In sharp contrast to maintenance, spending on expensive school equipment is very large. As a share of total education expenditure incurred by the government, spending on expensive school equipment varied between 0.3% in 2005 and 3.4% in 2013. Spending on expensive equipment may not be large, but has been consistently sustained over the last decade. Interviews with Ministry of Education and Culture and Science reveal that this equipment was intended to improve schools’ learning environment (for example computers and laptops for teachers). 12. Expenditures for demand-side financing offered at the higher education level declined in recent years. Demand-side financing data were available for 2015 and onward. These declined from MNT 78.4 trillion to MNT 43.5 trillion between 2015 and 2018, representing 8% and 3% of the recurrent spending by the government on education in these years.57 All these expenditures were incurred at the higher education level. National scholarships constituted the largest share of this spending, and these were not offered in 2017 and 2018. These include the sum of scholarships, grants, travel allowances, and loans. 57 109 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 13. Actual expenditures incurred departed significantly from allocated budgets, higher in some years and lower in others. World Bank (2015) found that actual expenditures were lower than the adjusted budget by MNT 117.7 trillion (15% of budget) in 2011, exceeded the budget by MNT 59.7 trillion (6%) in 2012, and then again were lower than the budget by MNT 278.2 trillion (21%) in 2013. 14. The limited decision space granted to local government and school managers undermines micro- efficiency, a concern raised repeatedly regarding spending at the local level. Since the passage of the IBL, between 59% and 75% of the total education expenditure has been executed at the local government level under the special purpose transfer allocation of the education budget (64% of the budget in 2018). Thanks to the regularity of the annual budget process, and the functional TSA and GFMIS, schools receive resources on time and according to the budget plan, and this information is accurately accounted for and audited, and made publicly available, representing a significant achievement of Mongolia’s PFM reforms over the past decade. The key issue regarding local government spending is the limited decision space granted to local government official under the IBL. For example, local officials could be granted more flexibility in allocating resources across facilities within their jurisdiction to strengthen weaker schools or respond to unanticipated changes in demand. The system leaves almost no scope for school managers to plan and budget as they see fit for the specific circumstances of their school. Even reallocations across line items require cumbersome approvals. 15. Among public education facilities, the number of kindergartens expanded most rapidly in recent years, while that of higher education institutes (HEIs) fell due to privatization. Private kindergartens expanded rapidly in the last decade, from 86 in 2005 to 538 in 2017 – an increase of over 500%. In 2017, 38% of kindergartens in the country were privately owned. Growth in privately owned schools was not nearly as steep, with the number increasing only slightly from 139 to 146 during the same period and representing 18% of schools in the country. The higher education services are largely provided by the private sector, with 81% of HEIs privately owned. TABLE 4.1 Number of Education Facilities by Facility Type, 2005-2017 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Kindergartens 729 742 768 783 814 839 879 945 1,067 1,171 1,288 1,354 1,416 Public 643 648 666 696 712 722 734 750 764 777 826 854 878 Private 86 94 102 87 102 117 145 195 303 394 462 500 538 Share of private 11.8 12.7 13.3 11.1 12.5 13.9 16.5 20.6 28.4 33.6 35.9 36.9 38.0 Schools 724 742 754 748 755 751 752 755 756 762 768 778 798 Public 585 590 597 594 605 609 614 621 628 628 636 645 652 Private 139 152 157 154 150 142 138 134 128 134 132 133 146 Share of private 19.2 20.5 20.8 20.6 19.9 18.9 18.4 17.7 16.9 17.6 17.2 17.1 18.3 HEI 180 170 162 154 146 113 101 99 100 101 100 95 96 Public 49 48 47 48 42 16 15 15 16 16 17 17 18 Private 131 122 115 106 104 97 86 84 84 85 83 78 78 Share of private 72.8 71.8 71.0 68.8 71.2 85.8 85.1 84.8 84.0 84.2 83.0 82.1 81.3 Source: MECSS (2018). 16. Enrollments in ECE grew rapidly, while those in basic education were sustained at near-universal levels. Preprimary enrollments reflect the rapid growth in the provision of kindergarten services. Between 2005 and 2017, the number of children enrolled in ECE services more than doubled from 122,104 to 256,720. With basic education enrollments close to universal for a few decades now, the changes in enrollments in schools were sustained and overall rose from 556,876 to 572,752 in this period. HEIs enrollments increased modestly by 12% from 138,019 to 155,248. 110 SECTORAL SPENDING 17. The public sector absorbs most of Mongolian pupils, especially in preprimary and general education. The private sector held only 14% of all ECE enrollments in 2017. The share of students in private schools is even smaller, constituting only 6% of all enrollments in the country. Forty two% of HEI enrollments were in the private sector, and while this is the highest private sector share across the three subsectors, it is still somewhat low given that over 80% of HEIs are privately owned. Public HEIs, while small in number, have much larger capacity than private ones. FIGURE 4.3 Number of Students and Share of Private Sector by Level of Education System, 2005-2017 700 50 Number of students (thousands) 600 40 500 30 Share private (%) 400 300 20 200 10 100 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Kindergartens ‐ % private Schools ‐ % private HEI ‐ % private Kindergartens ‐ students Schools HEI ‐ students Source: Author’s calculations based on MECSS (2018). Note: Number of students on LHS and share of private sector on RHS. 4.1.2 Outcomes and Key Challenges 18. While gender gaps in access to education persist only at the higher education level, where men lag behind women, socioeconomic differences also matter. With almost universal access to basic education, Mongolia never exhibited a gender gap in school enrollments: girls’ share in total enrollments stood at 50% in 2017. The same is true of ECE enrollments - 49% of those enrolled were girls in 2017. At the higher education level, men lag behind women: in 2005, 61% of those enrolled were women, and this share has not changed much, standing at 58% in 2017. Other social and economic inequities are harder to overcome. At the ECE level, the number of children enrolled from herders’ families rose steadily in the last decade, but at 11%, the share of these children in the enrolled population is significantly lower than the share of nomadic herders in the country population (around 25%). At the school level, the shares are much more equitable at 19% of all children. 19. Socioeconomic gaps in coverage are large, except at the primary education level. Despite impressive gains in coverage, Mongolian children in greatest need of ECE are still not covered (World Bank, 2017). Geographic and ethnic gaps remain large, and household wealth and nomadic status are key determinants of access. Figure 4.4 and Figure 4.5 shed light on the characteristics of children at risk of being out of school at the lower and upper secondary levels, respectively. In both cases, children from the poorest households are at the highest risk of not being in school. Based on the latest available Multiple Indicator Cluster Survey (MICS)58 of 2013, one in five children of upper-secondary school age from households classified as the poorest 20% of households were not in school (and one in 10 children of lower secondary age). Overall, children of secondary school age children living in soums were at a much higher risk of not being enrolled in school than children 58 MICS is an international household survey designed to collect statistically sound, internationally comparable estimates of about 130 indicators to assess the situation of children, women and men in the area of health, education, and child protection. 111 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue in urban areas and boys lagged behind girls. Among children of secondary school age, the share of boys out of school was almost three times that of girls: 5.1% of girls compared to 14.2% of upper secondary school age boys, and 2% of girls compared to 6% of lower secondary school age boys. At the higher education level, the gender gap is compounded with 61% of women between the ages of 18 and 24 enrolled in a college or university in 2013, compared to 44% of men. Only 10% of individuals in this age group from households in the poorest quintile were enrolled, compared to an astounding 88% of those from the richest quintile. FIGURE 4.4 Share of Children of Lower Secondary School Age Not in School, Mongolia, 2013 25 20 15 (%) 10 5 0 Aimag center UB Soum center Poorest Other Rural Second Middle Fourth Richest Khalkh Kazakh Location Wealth Ethnicity Source: UNICEF (2016). FIGURE 4.5 Share of Children of Upper Secondary School Age Not in School, Mongolia, 2013 25 20 15 (%) 10 5 0 UB Aimag center Soum center Other Rural Poorest Richest Kazakh Second Middle Fourth Khalkh Location Wealth Ethnicity Source: UNICEF (2016). 20. The education sector labor force (that is, the number of teachers and other education sector personnel) rose most rapidly in preprimary services, followed by general education. Trends in the education sector human resource outputs reflect the prioritization by the government of ECE service delivery. The number of teachers employed in the subsector more than doubled between 2005 and 2017. In schools, the number of teachers rose by 29% to 29,242 during this period; and the total number of employees by 31% to 47,991. In higher education, in sharp contrast, the changes were much more muted, registering an increase of 207 teachers, bringing the total number of teachers to 6,724 in 2017. The number of employees in the higher education subsector increased to 12,705. 112 SECTORAL SPENDING 21. However, student-teacher ratios (STRs) in kindergartens are significantly higher than recommended by international best practices. Kindergarten-level STRs remained very high throughout our period of observation, and the average STR nationally declined only slightly from 37.3 in 2005 to 33.1 in 2017. When examined in conjunction with the massive expansion in enrollments in this subsector, these numbers indicate that the emphasis has been to expand services to enroll more children with less regard to quality of services, with little concern expressed for the fact that overcrowded classrooms may contribute little or nothing to child development. 22. At the national level, the average STR in general education schools declined in recent years but remains comparable to Mongolia’s middle-income peers. The average STR at the national level across schools declined from 24.6 in 2005 to 19.6 in 2017. However, the STR at the primary level (28 students per teacher) is among the highest among selected comparators including South Africa, Indonesia, and the Philippines. At the secondary level (Figure 4.7), the situation is somewhat different with a significantly lower average STR of 15:1, but still, by international standards, it is higher in Mongolia than in many of its middle income peers in Eastern Europe and Central Asia. FIGURE 4.6 Primary Education Student Teacher Ratios in Mongolia and Other Countries 35 30 Primary pupil-teacher ratio 25 20 15 10 5 0 10,000 20,000 30,000 GDP per capital (constant 2005 US$) Source: WDI (2017). 23. The efficiency of Mongolia’s large and sustained capital investments in education is undermined by the consistently low and declining routine maintenance budget at the school level. Mongolia’s progress in expanding coverage of basic education services is underpinned by its sustained capital investments (11-13% of total education expenditure during 2012-2014). In the coming years, there is a critical need to optimize the value achieved from these large investments through adequate maintenance of school infrastructure routine facility maintenance and repair and replacement of furniture and equipment. However, as seen earlier, school-level operations and maintenance spending is very low, under 1% of the total recurrent budget by some accounts. The deferred routine maintenance implied by this low spending leads in the long run to higher overhaul maintenance costs. To maintain the efficiency of Mongolia’s capital investments in education, complementary investments in routine maintenance of school infrastructure are vital. 113 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 4.7 Secondary Education Student Teacher Ratios in Mongolia and Other Countries 35 30 Secondary pupil-teacher ratio 25 20 15 10 5 0 10,000 20,000 30,000 GDP per capital (constant 2005 US$) Source: WDI (2017). 24. Overall, the system exhibits efficiency losses due to shortcomings in the planning of capital investments and deployment of human resources. The system continues to be plagued by concerns about unused capacity in some schools in rural Mongolia and even in some parts of UB, contrasted with capacity constraints in some urban schools. While no systematic studies are publicly available, anecdotal reports suggest that some soum schools have ‘inefficiently few’ students, in soums with high rates of outmigration. Further, even in urban areas, the location of new schools and kindergartens is often not aligned with actual needs, leading to excess capacity in some areas while overcrowding persists in others, especially in the fringe ger areas of UB city where enrollment in central city schools exceeds capacity in several khoroos. By contrast, khoroos in Bayanzurkh, Songinokhairkhan and Nalaikh tend to have an oversupply of space. A related concern is the location of schools relative to where residents live in UB, schools in the fringe areas are located far from the residential areas they serve. On average, ger area children must walk more than twice as long as the apartment area children to get to school. Informal reports also suggest uneven quality of infrastructure in schools and kindergartens, with too much being spent on new construction in some places with little regard to maintenance or upkeep in others, especially in rural areas, and poor quality of construction reported in other places, leading to high unmet maintenance needs. Each of these undermines the efficiency of capital spending. 25. In an environment of high returns to education, the rising levels of education in Mongolia put households on a higher earning path in recent years. Estimated returns of education in 2014 stood at around 9% for each additional year of schooling, with much higher returns for completing a university degree (World Bank, 2015). Despite increases in educational attainment between 2010 and 2014, the better-educated continued to earn more on average and were less likely to be poor. With rising educational attainments during this period, the proportion of people earning higher incomes rose, contributing to growth in household labor income and boosting income-generating capacity. 114 SECTORAL SPENDING 26. Coverage of ECE services expanded rapidly in recent years, while that of other levels of the system remains high. In 2013, 79% of the children enrolled in grade 1 of primary school had attended some form of preschool in the preceding year. Within the 36-59 months age group, the preschool enrollment rate stood at 68%, up from 58% in 2010, a sizable increase over a three-year period. This increase has placed Mongolia’s preprimary enrollment rate closer to that of significantly more productive economies, in per capita terms, than to its middle income peers. The adjusted net attendance ratio for primary education stood at 98% of children of primary school age, 93% of children of lower secondary school age, and 85% among children of upper secondary school age. This finding is consistent with recent empirical literature (for instance, Del Boca et al. 2018) which indicate that both government spending in education and attendance to early childhood education lead to better later school outcome. Thus, access to basic education, which is compulsory in Mongolia, is nearly universal at 98.1% of children of basic education age enrolled in school, but drops significantly at the upper-secondary level. These rates have not changed much – in 2010, 97% of children of basic education age were enrolled in basic education or higher, and the number stood at 93% for lower secondary age children. Finally, 52% of young adults between the ages of 18 and 24 were enrolled in a college or university in 2013. TABLE 4.2 Trends in Earnings and Poverty Rates in Mongolia, 2010-2014, by Educational Attainment Median Earnings (in 2010 Prices) Growth in Median Earnings (%) Education completed 2010 2012 2014 2010-2012 2012-2014 2010-2014 Less than primary 981,359 1,198,125 1,661,732 22.1 38.7 69.3 Completed primary 1,052,878 1,473,090 2,035,682 39.9 38.2 93.3 Lower secondary 1,402,380 1,829,629 2,217,323 30.5 21.2 58.1 Upper secondary 1,855,039 2,388,375 2,603,809 28.8 9 40.4 Vocational training 2,728,905 2,649,960 3,037,247 -2.9 14.6 11.3 Degree or diploma 3,151,287 3,751,129 4,174,596 19 11.3 32.5 All 1,943,214 2,520,917 2,924,625 29.7 16 50.5 Poverty Headcount Rate (%) Change Distribution of the Poor 2010 2012 2014 2010-2014 2010 2012 2014 Less than primary 65.5 55.1 45 -20.5 5.2 6.7 7.9 Completed primary 58.4 44.7 31.1 -27.3 15.2 14.4 12.6 Lower secondary 52.7 40 32.5 -20.2 29.5 31.5 28.4 Upper secondary 39 26.3 24.8 -14.2 33.9 33.1 31.5 Vocational training 23.1 20 16.8 -6.2 12.2 8.8 14.8 Degree or Diploma 14.6 7.6 4.6 -10 4 5.4 4.8 Source: World Bank (2015). Note: The major source of income is defined as a source of income constituting more than half of the household’s total income. 27. Benchmarking system performance is not possible as Mongolia does not participate in any large-scale international assessments. Mongolia does not participate in international assessments like Program of International Student Assessment (PISA) or TIMSS59, which makes it impossible to benchmark the system’s performance on learning outcomes or improvements in outcomes over time. Nationally, while students take standardized assessments at the end of grades 5, 9 and 12, the micro data for the assessments are not published. The Educational Evaluation Center publishes an annual report on key findings, for example, average scores disaggregated for each province by subject, but the analyses are not informative for policymaking and not relevant for any type of evaluative purpose. Data quality has also never been systematically assessed, but is thought to be low. However, the assessment of early development outcomes and early grade literacy and numeracy skills, both revealed large socioeconomic gaps in school readiness. Mongolia participated in the 2007 TIMSS; however, data collected were deemed by the government to be of poor quality and were never published. 59 115 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 28. While in principle Mongolia’s school financing is set up to allow resources to follow the student, low school autonomy constrains efficiency, leaving much room for improvement. The Mongolian government provides primary and secondary schools with school budgets comprising three components: (a) a normative amount based on a per student allocation, whereby the amount per student varies according to grade and location of the school; (b) a so-called ‘fixed budget’ to cover utility costs; and (c) funding for targeted social assistance to support children of low-income families. In principle, this system of financing provides resources where they are needed, as ‘funds follow students.’ However, without school autonomy, as is the case in Mongolia, per capita financing alone has limited impacts in improving efficiency and school accountability. 29. The preparation of the recurrent and investment budgets takes place in silos, with little transparency around allocations for capital expenditures. The recurrent and investment budgets for primary and secondary education are prepared and managed in silos, with responsibility and oversight vested in two separate departments of MECSS that rarely coordinate with one another. This is problematic because the investment budget includes elements that come close to operational maintenance, such as the provision for furniture and the provision for routine maintenance. Relatedly, while the normative allocation of budget to cover recurrent costs is flawed in some areas, the existence of norms imposes some amount of transparency on the process of arriving at allocations. This is not the case for capital investments, which are not informed by publicly available documentation of how unmet needs are addressed by allocation each year. The decision of which locations or schools will be rehabilitated, and where new schools will be built, takes place through a consultative process between MECSS and MOF, under the influence of provincial governments and State Khural Members of Parliament (MPs). Only the decisions made as a result of these discussions are made public. 30. A small but significant share of school income is derived from parental contributions. While officially education is free, voluntary contributions are allowed (World Bank, 2016). In all schools, it is common practice for parents to make ‘voluntary’ contributions for refurbishment of classrooms at the beginning of the school year. Parents may even contribute to cleaning materials that are kept by the classroom teacher. When compared to the normative budgets allocated for schools’ operational expenditures, parental contributions are negligible for rural bagh schools but substantial for the larger aimag and UB schools. In kindergartens, too, private contributions cover learning materials and supplies, and a share of meal costs (World Bank, 2017). According to PFM regulations, all revenues received by schools should be deposited in the school’s bank account within the TSA system, which would hence capture all sources of income. However, most-if not all-of the contributions made by parents are directly managed by those parents and are not captured in school accounts. 4.1.3 Priority Areas of Reform 31. Benchmark education system performance by participating in the PISA and strengthening national assessments. This will allow regular monitoring of efficiency and equity of education investments. As Mongolia spends a significant share of its government expenditure on education, it is therefore essential to assess the system’s effectiveness in generating learning. Mongolia is well past the stage when monitoring enrollments alone was the only feasible or relevant option – enrollment rates in general education have been sustained at near-universal levels for decades. Benchmarking of learning is critically needed if Mongolia wants to join the ranks of high-performing EAP countries, which adopted a systemic approach to measuring and monitoring learning outcomes and successfully used assessments to spur reforms improving instruction in the classroom. Beyond participation in the PISA, learning assessments, especially at the early grades, need to be strengthened and data made publicly available to allow for reliable and regular monitoring of efficiency and equity of investments. 116 SECTORAL SPENDING 32. Reduce inequities in quality of learning environment across schools and kindergartens by increasing non-salary operational budgets and giving schools autonomy to manage these budgets. First, public spending on critical aspects of school operations, including teacher development, learning materials, and repairs and routine maintenance, is either extremely low or almost negligible. Part of the needed increase could be financed by reallocation from the capital investment budget, which includes provisions for school maintenance. Such a reallocation would also improve the efficiency of capital investments. Second, there is absolutely no flexibility at the school level to reallocate budgets across line items. The system could realize efficiency gains through increased school-level autonomy, accountability, and transparency around the development and management of budgets for operational costs. Finally, to ensure that schools meet the same minimum standards, the budget allocation formula needs to include additional factors such as special needs, distance to the aimag capital and to UB as a proxy for price levels, and poverty as a proxy for the capability of parents to contribute. 33. Improve equity of education investments by expanding access to ECE in rural areas through community- or home-based modalities for hard-to-reach populations, for example, nomadic herders. Investing in high-quality ECE is a cost-effective strategy, which addresses early inequalities in opportunity, and improves efficiency by strengthening academic and nonacademic performance through all levels of the system. But it is impossible for formal ECE to reach Mongolia’s remote or nomadic populations. For these groups, home-based ECE modalities are the only feasible option. Increasing the intensity of family engagement with children on activities at home, such as reading, singing, and playing together, is associated in Mongolia with significantly higher child development outcomes in numeracy and language. Children participating in such interventions exhibited ECE outcomes considerably superior to those of comparable children in ger kindergartens. There is an urgent need to test and deploy such services. 34. To relieve spending pressure due to increased access and quality improvements, explore potential for increased parental contributions among the well-off, and partner with ECE private providers. Parental contributions are a significant albeit small source of income for schools. Further potential for parental contributions should be explored to supplement public investments made to improve quality. At the same time, the welfare implications of out-of-pocket (OOP)expenditure should be monitored. At the preschool level, those currently out of the system are the most economically disadvantaged, so monitoring is needed to ensure that private expenditures are not prohibitively high or do not pose a barrier to entry. Contributions could be linked to socioeconomic status. Another way to relieve the pressure on ECE public spending could be to partner with the private sector to exploit the tremendous growth in the private sector, especially in UB. When considering the potential for an expanded role for the private kindergartens, the government could focus on developing quality standards and/or providing technical assistance to ensure service provision of adequate quality and minimize quality differences. Effective monitoring will be an important aspect of this initiative. 35. Strengthen planning of capital investments through increased transparency and a bigger role for provincial education authorities. Keeping the entire investment budget at the central level is neither practical nor responsive to the spirit of the IBL, and contributes to a myriad planning and quality problems. The capital investment budget should be unpacked to determine which elements could be feasibly decentralized to schools, aimag, and capital-city governments. Increased local oversight of construction projects and procurement processes will improve the quality and transparency of Mongolia’s capital investments. Also, assessments are needed to ascertain the effectiveness of Mongolia’s investments in new technology and equipment. Spending on expensive school equipment has been consistently sustained over the last decade in sharp contrast to other critical categories like maintenance. However, no assessments have been conducted on how effective or useful this equipment was from the point of view of the schools. In fact, anecdotal evidence suggests that very little methodological assistance or targeted training was provided to schools in how such equipment could be used to improve teaching within the classroom or how to integrate these technologies into classroom practices. 117 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 118 SECTORAL SPENDING 4.2 Health Sector 4.2.1. Recent Trends and Structure of Expenditure 36. Mongolia’s commitment to the Sustainable Development Goals is reflected in the strides the country has made in maternal and child health (MCH). With changing lifestyles and rising life expectancies, however, a growing burden of NCDs brings new challenges. Weak service delivery integration leads to fragmented care for NCDs, which, together with the low capacity at the primary health care level and population preference for hospital care, compromises the core functions of prevention and early treatment, increasing the need for expensive, acute care, especially in rural areas. 37. Overall, the health system is characterized by weaknesses in primary care; hospital centrism; lack of integration between primary, secondary, and tertiary health care; and low quality. In addition, strategic purchasing of services by the General Authority of Social Insurance remains curtailed. A key constraint is the fragmentation of the health payment system between the state budget and social health insurance fund (SHIF). While output-based contracts with health providers are implemented by the insurance organization, they do not incentivize quality improvements. Together, these factors act as impediments to addressing the growing burden of NCDs, and achieving better health outcomes and higher returns to investments in health. 38. While public spending on health increased over the last decade, it did so at a slower pace than overall government spending and was not always adjusted adequately for inflation. Government health expenditure grew steadily in nominal terms between 2005 and 2016, except in 2009. In fact, both per capita government health expenditure in nominal and real terms increased on average by 7% and 9% per year, respectively over the 10-year period between 2005 and 2015. However, this growth is low when compared to that in total government expenditure, which increased at an average rate of 14% per year during the same period. Further, health expenditure is not always adjusted adequately for inflation. For example, in 2010 and then again in 2013, real spending on health declined significantly from one year to the next. FIGURE 4.8 Trends of Spending in Health, 2005-2016 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: BOOST Database, Mongolia Ministry of Finance. 119 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 39. Mongolia’s health expenditures are low when compared with regional and economic peers. At just over 4%, the share of total health spending in GDP and the share of government health spending (at 2%) are both lower than in most other LMICs. This is surprising for two reasons. First, one would expect that Mongolia’s status as one of the least densely populated countries in the world would exert an upward pressure on the cost of service delivery. Second, this finding is in contrast with previous observations. The 2008 World Bank report ranked Mongolia as one of the highest spenders on health. In 2003, the share of total health spending in GDP was 6.7% and the share of government spending on health stood at 4.5% of GDP. 40. Most of the government health spending is for recurrent expenditures, mainly wages. Capital expenditures fluctuated significantly over the last decade. The number of state health facilities did not change much during the period. Allocations to repairs and maintenance are extremely low. 41. Recurrent expenditures represent the bulk of total health expenditure, averaging about 85% in 2015-2016. At 38% in 2016, wages, salaries, and bonuses constitute the largest share of recurrent health expenditures incurred by the Mongolian government. Since 2012, current transfers increased significantly and in 2015 constituted the second highest share in recurrent spending (24%), followed by medicines and drugs (16%). 42. Despite fluctuations from year to year, expenditure on wages, salaries and bonuses is the fastest growing expense category, growing at an annual rate of roughly 14% in real terms between 2005 and 2015. Growth in real expenditure on medicines has been slower, at an average of 9% per year between 2005 and 2015. The high and growing share of wages in recurrent expenditures is due to a rapid increase in the number of health workers, notably doctors and nurses, but growth in wage expenditure per worker has been slow and very uneven suggesting the need to increase health sector wages. However, a more focused analysis is needed based on wage data for each type of health worker to explore how well the basic wage level in the health sector aligns with general labor market trends in Mongolia. 43. Capital expenditures, which accounted for 14.3% of total health expenditure in 2016 fluctuated significantly over the last decade, mainly reflecting the absence of an effective sector spending planning. In 2011, the capital share of health expenditure reached a peak at 24%. Buildings and facilities accounted for 63% of total capital health expenditure, while health equipment claimed 23% in 2016. Existing sector analyses confirm that current planning standards lead to purchases of unnecessary and expensive medical equipment. For instance, the capacity associated with medical equipment purchased for diagnostic centers in 11 aimags was 10 times higher than actual needs, increasing unit costs tenfold than was necessary due to higher maintenance and associated recurrent costs.60 44. Poor sector spending planning also led to a low budget for maintenance and repair. Within recurrent spending, expenditure on repairs stood at a mere 0.3% in 2016 and, between 2005 and 2015, declined in real terms by 60%. Within capital spending, expenditure on repairs stood at 8% in 2015. Capital repair spending increased slightly but steadily in real terms up to 2012 and since then has fluctuated quite a lot from year to year. 45. The number of state health facilities has not changed much in the last decade, although there has been a small increase at the secondary level and a small decline at the tertiary level. Among state- owned health institutions in 2015, Mongolia had 13 tertiary-level hospitals and centers, 33 aimag/ Regional Diagnostic and Treatment Centers (RDTCs) and district general hospitals/health centers, 6 rural general hospitals, 39 inter-soum hospitals and 291 soum and village health centers (Table 4.3). Authors’ estimates based on Selenge and Khovsgol aimag diagnostic center data. 60 120 SECTORAL SPENDING TABLE 4.3 Number of Health Facilities by Facility Type, 2005-2015 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 State Health Institutions Tertiary level hospitals and centers 17 15 15 15 16 16 16 16 16 13 13 Aimag/RDTC and district general 33 33 33 33 33 33 33 33 33 33 33 hospitals/health centers Rural general hospitals 4 4 4 6 6 6 6 6 6 6 6 Inter-soum hospitals 31 34 35 35 35 37 37 39 39 39 39 Soum and village health centers 311 310 310 305 295 291 293 290 290 290 291 Private Health Institutions Private hospitals 160 168 159 159 160 166 171 179 197 202 224 Private clinics 523 612 698 904 922 947 1013 851 822 969 1006 Family health centers 228 224 229 228 226 218 219 221 228 218 218 Drug supplier 155 133 66 140 144 158 158 155 128 160 210 Drug manufacturer 37 37 5 46 57 41 42 42 42 32 33 Pharmacies 735 671 706 741 636 666 703 855 789 936 967 Other 114 116 116 115 115 118 169 202 173 206 207 Total 2348 2357 2376 2727 2645 2697 2860 2889 2763 3104 3247 Source: MOH, World Bank staff estimates Note: Other include policy and health administration institutions, educational institutions, maternity homes, sanatoria, borderline hospital, defense hospital, and railway hospital. 46. While health spending at the local level remains low, the limited spending discretion granted to local governments and managers of health facilities undermines micro-efficiency, including the financing of health providers. Since the passage of the IBL, just under a fifth of recurrent expenditures were incurred at the local government level under the special purpose transfer allocation of the health budget. Owing to the regularity of the annual budget process, and the functional TSA and GFMIS, basic health units receive resources on time and according to the budget plan, and this information is accurately accounted for and audited, and made publicly available, representing a significant achievement of Mongolia’s PFM reforms over the past decade. That said, financing of health providers needs to be improved in many ways and the overall level of financing is too low for effective service delivery, for a variety of reasons presented in this Public Expenditure Review (PER) and World Bank (2016). Barring these problems, the key issue raised repeatedly regarding the execution of expenditure at the local level is the limited decision space granted to local government officials and service delivery providers. For example, local officials could be granted more flexibility in allocating resources across facilities within their jurisdiction to strengthen weaker health facilities or respond to unanticipated changes in demand. 47. The private sector has seen a phenomenal expansion in its capacity to offer both inpatient and outpatient services. In the private sector, between 2005 and 2015, the number of hospitals grew from 160 to 224, and clinics from 523 to 1,006, while family health centers (FHCs) declined from 228 to 218 (Table 4.4 ). The number of beds in private sector hospital beds grew during this period from 1,982 to 5,262, and in 2015 constituted about 24% of hospital beds available in the country. 121 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue TABLE 4.4 Outpatient Visits and Inpatient Admissions by Facility Type, 2015 Outpatient Inpatient   Total Number Visits Admissions   facilities beds (%) (%) Primary level 48 14 Soum health centers 291 2546 10 11.2 Inter-soum hospitals 39 667 3 2.5 FHCs 218 - 36 - Secondary level 29 38 Rural general hospitals 6 353 1 2 Aimag general hospitals 16 3378 10 17 District general hospital and health centers 12 2112 15 12 RDTCs 5 1484 3 7 Tertiary level 7 18 Single specialty centers 9 3 11 3837 Multi-specialty general hospitals 4 4 7 Private hospitals 224 5262 - 18 Private clinics 1006 - 12 N/A Other hospitals 141 2081 5 12.5 Total 1971 21720 100 100 Source: MOH, World Bank staff estimates Note: Other includes hospitals in other ministries, maternity homes, and sanatorias. 48. While OOP are high at 40% of Mongolia’s total health expenditure, catastrophic expenditure and impoverishment due to health spending is low. The share of OOP health payments in total household expenditure declined from 3.2 in 2009 to 2.8% in 2012. Catastrophic health expenditure occurs when households spend 40% or more of their capacity to pay on the use of health services. The share of households incurring catastrophic health payments in Mongolia on average stood at 1.5% in 2010, 1.8% in 2011, and 0.9% in 2012. However, richer households were likelier to incur catastrophic expenditures than poorer ones (Figure 4.9). FIGURE 4.9 Catastrophic Health Payments, by Household Expenditure Quintile 8 7 6 5 4 (%) 3 2 1 0 2009 2010 2011 2012 2009 2010 2011 2012 2009 2010 2011 2012 2009 2010 2011 2012 2009 2010 2011 2012 Poorest 2 3 4 Richest quintile quintile Source: Tsilaajav et al (2015). 122 SECTORAL SPENDING 49. In recent years, the share of outpatient visits held by state-owned secondary institutions increased at the expense of tertiary and primary institutions, while the private sector’s share remained unchanged. It is instructive to also look at how the utilization of facilities across the levels of the health system has varied during this period. Between 2004 and 2015, the share of public tertiary-level institutions in outpatient visits declined from 10% to 7%, and the share of primary also declined from 56% to 49%, while that of secondary rose from 24% to 29%. Surprisingly, despite the explosion in the number of private clinics, their share in outpatient visits did not change much over this decade. Thus, the only change appears to be an increase in concentration of service utilization at the secondary level in public sector health facilities. FIGURE 4.10 Share of Outpatient Visits by Health System Tier, 2015 Tertiary, 7% Other, 5% FHCs, 36% Clinics, airy 12% Tert e at Priv Rural, Primary 4% ary nd o Aimag Sec general, 10% District Soum/intersoum, general, 13% 15% Source: MOH, 2016 FIGURE 4.11 Share of Inpatient Admissions by Health System Tier, 2015 Soum/intersoum, Aimag general, 14% 17% vate Tertiary, Pri S 18% ec ond Tertairy District general, ary 12% Primary Rural, 9% Other, 13% Hospital, 18% Source: MOH, 2016 123 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 50. The share of inpatient visits at state-owned secondary institutions increased much more dramatically, at the expense of both tertiary and primary institutions, as did the private sector’s share. Between 2004 and 2015, the share of public sector tertiary-level institutions in inpatient visits declined from 30% to 18% that of public secondary-level facilities increased from 25% to 38% and that for the primary level declined from 27% to 14%. The share of inpatient visits in the private sector health institutions increased from 10% to 18%, while that of others also increased from 8% to 13%. For inpatient care, thus, the situation changed much more dramatically than for outpatient care, with a clear transition toward concentration of utilization of services at the secondary level. 4.2.2. Outcome and Key Challenges 51. MCH outcomes continue to improve, but some middle-income peers now perform better than Mongolia. MCH is often used as a broad marker of whether a health system is successful in improving the health status of a population. On this metric, Mongolia continues to perform well. Between 2005 and 2016, infant and under-five mortality rates declined from 33.4 to 15.4 and 49 to 23.5 per 1,000 respectively, while the maternal mortality ratio declined to 44 per 100,000 live births. The strong performance on MCH outcomes is likely due to near-universal coverage of basic preventive services, that is, the near-universal coverage of immunization and skilled attendance at birth. One difference however is that while in 2005 these mortality rates were quite low relative to other countries at Mongolia’s level of economic development, in 2014, several middle-income economic peers were performing better than Mongolia (Figure 4.12). Among resource-rich countries only, however, Mongolia performs significantly better than others at similar levels of GDP per capita. Mongolia can transform its low health expenditures into strong outcomes much more effectively than its resource-rich economic peers. FIGURE 4.12 Infant Mortality Rate in Mongolia and Other Countries, 2014 120 100 Infant deaths per 1,000 live births 80 60 40 20 0 GDP p.c. (current international US$) Source: WDI (2017). 124 SECTORAL SPENDING FIGURE 4.13 Maternity Mortality in Mongolia and Other Countries, 2014 700 600 500 Maternal Mortality Ratio 400 300 200 100 0 - 5,000 10,000 15,00 20,000 25,000 30,000 GDP p.c.(current international US$) Source: WDI (2017). FIGURE 4.14 Maternal Mortality in Mongolia and Other Resource-Rich Countries, 2015 900 800 Maternal mortality ratio (modeled estimate, 700 per 100,000 live births) 600 500 400 300 200 100 0 0 20,000 40,000 60,000 80,000 100,000 GDP per capital (constant 2010 US$) Source: WDI (2017). 125 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 4.15 Infant Mortality Rate in Mongolia and Other Resource-Rich Countries, 2016 80 70 Mortality rate, infant (per 1 ,000 births) 60 50 40 30 20 10 0 0 20,000 40,000 60,000 80,000 100,000 GDP per capital (constant 2010 US$) Source: WDI (2017). 52. NCDs constitute a growing share of morbidity and continue to be the leading cause of mortality. NCDs such as heart disease, stroke, cancers, hypertension, diabetes, and chronic lung disease, as well as injuries, have been the leading causes of mortality and morbidity for at least the last two decades (Figure 4.16 and 2008 PER). NCDs accounted for 64.3% of the disease burden and 85.9% of deaths in 2015. Underlying this are trends in risky behaviors including high smoking prevalence; high intake of fats, sugars and calories in the diet; high levels of alcohol consumption; and low physical activity. There is a clear need to strengthen primary and secondary prevention of chronic disease. Rapid aging of the population over the next few decades renders the need for action even more urgent: the share of elderly (those of age 65 and older) in Mongolia’s population is projected to rise from 7% in around 2030 to 14% over a subsequent period of 25 years (Figure 4.17). FIGURE 4.16 Prevalence of Leading Cases of Morbidity, 2005-2015 1,550 1,350 per 1 0,000 population 1,150 950 Diseases of the respiratory system 750 Diseases of the digestive system Diseases of the genito-urinary system 550 Diseases of the circulatory system 350 Injuity, poisoning and certain other consequences of external causes 150 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: MOH, 2016 126 PUBLIC EXPENDITURE REVIEW FIGURE 4.17 Prevalence of Leading Cases of Mortality, 2005-2015 25 20 15 Injuiry, poisoning and certain other 10 consequences of external causes Neoplasms 5 Diseases of the respiratory system 0 Diseases of the digestive system 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: MOH, 2016 53. The health system is not equipped to meet the growing burden of NCDs. As the prevalence of NCDs expands, Mongolia will need to address their underlying causes as well as increase early detection and chronic disease management. For this, Mongolia’s fragmented health system will need to transition from one biased toward treatment to one centered around improving population health outcomes. Box 4.1 summarizes the key obstacles that stand in the way for this transition. BOX 4.1 Addressing the Burden of Chronic Disease: Challenges Faced by Mongolia’s Health System The quality of health services remains low, despite being highlighted as a priority by the Health Sector Stra- tegic Master Plan for 2006-2015 and other policies. Differentials in service quality cause patients to bypass primary care services in favor of higher-quality secondary or tertiary services and, for those who can afford to do so, opt for overseas treatment over health care within the country. A high share of admissions in tertiary and district hospitals is medically unnecessary or inappropriate. The high ratio of hospital beds to population described earlier also suggests that planning of health inputs continues to be hospital-centric. There is no standardized performance monitoring system to collect information on quality of care and facilitate quality improvement. Most importantly, there are no incentives for health providers to improve quality of care. Fragmented care, combined with low capacity at the primary care level, compromises the core functions of prevention, case detection, early treatment and care integration. This leads to under-diagnosis, delays in treatment, and a higher need for more expensive, acute care, especially in rural areas. Providers at var- ious levels have no incentive to manage population health in a coordinated way. Links between hospitals and primary health care providers, including structured referral systems, patient discharge and handover mechanisms, and patient outreach, are generally not in place. Further, there is no horizontal integration across type of care, namely promotion, preventive, curative, and palliative. Information constraints underpin these shortcomings, hampering efforts to hold providers accountable for per- formance. There is no functional and effective electronic health (eHealth) system. eHealth can greatly enhance the functionality and effectiveness of primary care systems by connecting providers to achieve horizontal and vertical integration, coordination, and continuity of information over time. eHealth not only lays the founda- tion for successful communication between facilities but also provides health professionals and patients with the tools to more fully engage with the care process and improve care management and decision making. The integration of health data and eHealth solutions has been piloted in selected pilot sites in Mongolia under the eHealth project supported by the World Bank; however, there is an urgent need to accelerate implementation and scale up the eHealth system to the entire country. 127 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 54. Fragmentation of the payment system continues to hamper strategic purchasing of services. The 2015 Health Insurance Law mandated gradual channeling of government budget resources to social health insurance (SHI) so that the latter could serve as a pooled, independent purchaser with strengthened capacities and operating resources. At present, however, the payment system remains fragmented between two financing sources, that is, the central government budget and SHI contributions. In 2014, the central government spending (excluding SHI) and SHI spending accounted for 38.5% and 16.9%, respectively, of total health expenditure. Health services funded from the government budget are provided to all citizens free of charge and regardless of SHI status, and include public health, primary care, MCH, infectious diseases, cancer, and mental health. The SHI on the other hand mainly covers inpatient care, but in recent years has also covered some outpatient services and diagnostic tests. 55. While Mongolia’s provider payment system exhibits little evidence of adverse outcomes observed elsewhere, there is room for improvement. Mongolia is implementing a mixed payment system with a global and provider-level cap on all revenues. Four payment methods are employed: (a) capitation payment is used by Ministry of Health (MOH) to purchase all services provided by FHCs and soum health centers, and some government-funded services provided by soum and inter-soum hospitals; (b) case- based payment using Diagnosis Related Groups (DRGs) is used by the Social Insurance General Office (SIGO) to purchase services covered by SHI; (c) fee-for-service payment is used for direct payments by clients; and (d) line-item budget is used by the MOH to pay the residual of the revenue cap after other revenue sources are deducted. 56. Efficiency of capital investments is undermined by very low and fluctuating expenditure on repairs and maintenance of facilities. As seen earlier, repair expenditure, whether as part of capital or recurrent expenditure, is very low. Most of Mongolia’s government expenditure on capital investments is absorbed by construction and building. The deferred routine maintenance implied by the low spending on repairs leads in the long run to higher overhaul maintenance costs. To maintain the efficiency of Mongolia’s capital investments in health, complementary investments in routine maintenance of health infrastructure are vital. Procurement procedures undermine the efficiency of spending on drugs. Medicines stand out as the third largest category in recurrent spending. World Bank (2017) found that tertiary hospitals procure medicines individually. Hospitals do not collaborate and no information on parallel hospital procurement processes is routinely shared through formal channels to inform procurement practice. Tendering procedures for identical goods undertaken by the MOH administered hospitals are replicated by each hospital. The MOH analyses point to substantial differences between tendered prices across hospitals. Such fragmented procurement of hospital medicines in tertiary hospitals leads to inefficiencies in public spending. 57. Market competition is constrained by existing procurement regulations that favor domestic producers, leading to inflated prices. Several large companies, domestic producers and importers of medicines, dominate the market. Government Resolution No.336, issued in August 2015, requires health facilities to favor domestic producers. Meanwhile, the size of the country requires substantial investments in logistics infrastructure to distribute medicines to remote areas that few companies can afford, and hospital tendering combines procuring medicines as a good and distribution as a service. In addition, the fact that different companies deliver fragmented shipments also contributes to the high costs of distribution. According to the World Health Organization, the contracted prices of some pharmaceutical products in Mongolia are substantially higher than average prices for the same drugs in the Western Pacific region.61 58. Absence of quality assurance mechanisms for drugs also undermines efficiency. Drug regulatory functions are fragmented and are undertaken by a number of stakeholders. Medicines are WHO 2011, Report on Drug regulatory system assessment of Mongolia. 61 128 SECTORAL SPENDING administratively registered with the national center for Health Development, requiring no good manufacturing practice (GMP) certification. Quality control of medicines is in major part undertaken using a risk-based approach by the State Professional Inspection Agency that hosts the national drug control laboratory. The domestic GMP inspection process is carried out on a voluntary basis by the MOH and is internationally not recognized as implemented according to best practice standards. The trends observed in the utilization of inpatient services suggest efficiency gains in the delivery of inpatient care but not for outpatient care. Not only did the share of inpatient visits held by the various levels of the health system move overall in the same direction as that of distribution of beds, with declines at the tertiary and primary level, and increases at the secondary level, the increase in utilization at the secondary level was significantly larger in magnitude (13 percentage points) than the increase in the tier’s share in beds nationally (9 percentage points). The increase in the private sector was somewhat smaller (8-percentage-point increase in the share of visits compared to a 14-percentage-point increase in share of beds, nationally). Thus, the increase in use of inpatient services at the secondary level seems to have outpaced the increase in the share of beds at the secondary level, and this could be interpreted as a positive sign in terms of efficiency in the provision of inpatient care. For outpatient care, the picture is less positive. While the decline in the share of outpatient visits held by tertiary-level state- owned facilities and the increase in that held by secondary-level facilities seems to be a step in the right direction, the 6-percentage-point decline at the primary level is worrying. 59. At the subnational level, health resources do not seem to be optimally allocated and the efficiency implications of existing allocation rules need to be studied further. By international standards, Mongolia continues to have the highest number of hospital beds and physicians per 1,000 population, a situation that has not changed since the 2008 PER exercise. One of the explanations offered for these high population-adjusted health outputs is that Mongolia is one of the least densely-populated countries in the world and that this lowers the efficiency with which services can be delivered. Roughly half the country’s population lives in UB. If population density was the key driver responsible for the high population-adjusted outputs, we would expect to see significantly lower population- adjusted outputs in the most densely populated part of the country, that is UB. However, this is not the case. The hospitalization rate in UB was comparable to those in a handful of aimags; yet, UB still has the highest ratio of medical personnel to population. From an efficiency perspective, thus, the investment decisions underlying the subnational allocation of health resources needs to be investigated further. 4.2.3. Proposed Reform Actions 60. Reorganize services to address chronic illness, focus on high-quality primary care, and increase health spending to support this deep system reform. This PER outlines directions for the deep reorientation of health services needed to address existing health challenges and to meet future needs. The system needs to transition to a new service delivery model that integrates preventive and curative services, modernizes the role of secondary and tertiary hospitals as providers of complex care, rationalizes human resources, and deploys standardized local and national systems to measure and improve the quality of primary care, chronic disease management, and patient satisfaction. Mongolia needs a continued commitment to hospital governance reform through improving management capacity and accountabilities both within hospitals and the newly created governing boards, and setting clear roles, functions, and incentives for performance, to prepare for increased decision-making autonomy over time. This reorientation cannot be achieved without an increase in Mongolia’s budget allocation for health, which, as seen in this PER, in certain years is not even adjusted to keep up with inflation. One-off investments are needed to introduce the new model, translating into short- to medium-term increases in both capital and recurrent spending. Mongolia essentially faces the choice between an efficient and high-value path to public spending, with medium-term increases in health investments to deepen 129 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue ongoing reforms and achieve the service delivery model needed, or to continue business as usual, with the current shortcomings ultimately leading, given the trends in demographic and epidemiological profile, to persistently low value and quality and, in the long term, increasing cost of care associated with unmet needs for chronic disease treatment, long-term care, and hospitalizations. The expenditure analysis in this PER also suggests that there may be a need to improve compensation levels of health workers, but further assessment is needed to determine how health sector wages align with general labor market trends in Mongolia. In the long run, wages, and provider payments in general, should be aligned with incentives linking income with performance and long-term system goals for high-quality integrated care. 61. Improve efficiency of capital investments by aligning these with the system’s strategic long-term goals. The traditional input-based planning of capital investments employed in Mongolia lowers the value for money that can be achieved through each Tughrik invested, leading to purchases of unnecessary and expensive medical equipment. Capital investment planning needs to shift from the current input-based planning system, in which decisions are not based on actual demand but are driven by high-level macro standards, toward an approach that considers Mongolia’s changing epidemiological and demographic profiles and emphasizes effective regionalization and integration of care with new technologies. Second, the decision-making process for arriving at how the capital investments are allocated each year, and to which facilities and where, needs to be made transparent. Rather than being clearly informed by system- level strategic priorities, decisions are currently reached following negotiations between the MOH, MOF, and provincial and national government representatives. Third, the current level of expenditures on repairs and routine maintenance is extremely low, and undermines efficiency of Mongolia’s capital investments in health, and needs to be increased and sustained at a higher level. 62. Strengthen purchasing and align incentives to support integrated service delivery with due attention to measures needed for improved management at the local and provider levels. The government’s recent decision to pool health funds, channeling the state health budget through the insurance system, is a critical step forward in this area. In the short run, efforts to pool health funds and channel the government budget through the health insurance system need to be continued, together with the strengthening of the capacity of the new insurance organization (that is, the ‘purchaser’), ensuring adequate human resources, and introduction of more rational, evidence-based processes for benefits package revision. Deepening of provider payment reform should prioritize shifting the basis for government budget payment from inputs to parameters reflecting population needs, refine payments and capitation to adjust for real costs of service delivery, and increase flexibility to move expenditures across line items. Long-term measures include the enhancement of selective contracting to enable the purchaser to monitor performance and quality more effectively and the design and introduction of financial and other incentives within a network-wide design in support of population health, quality and cost containment, encouraging cooperation between purchaser and provider to improve performance rather than punitive measures that lead to misreporting. Centralized procurement of medicines would increase efficiency of public spending on drugs. Finally, local officials could be granted more flexibility in allocating resources across facilities within their jurisdiction to strengthen weaker health facilities and respond to unanticipated changes in demand. 130 SECTORAL SPENDING 131 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 4.3. Pension Reform62 4.3.1. Some Key Features of the Mongolia’s Current Pensions System 63. Workers covered under the Pension Insurance Scheme increased substantially since 2009 even though the structure of the labor force did not change as rapidly. The Pension Insurance Scheme is mandatory for workers with labor contracts who contribute to receive old age, disability and survivorship benefits. The scheme was established in 1994 with a contributory, pay-as- you-go, defined-benefit design based on the provisions applicable during the socialist period. The self-employed and other workers without labor contracts can voluntarily participate. A Notional Defined Contribution (NDC) scheme was established in 1999 to reduce the fiscal burden and better align contributions and benefits. The scheme was to apply to cohorts born after 1960 but the 2015 State Policy on Pension Reform pushed back the scheme to apply to those born in 1979 or later and cohorts born between 1960 and 1978 receive the better of either the defined-benefit pension or the NDC pension. 64. Total coverage of the labor force under mandatory and voluntary schemes which was about half in 2009 grew to 81% in 2016. Part of this increase may have resulted from legislation enacted in 2012 (taking effect in 2014) which provided a highly subsidized service buy-back arrangement while part of the increase can be attributed to some growth in payroll-based employment. This increase in coverage over such a short amount of time is dramatic when compared with labor force coverage growth in most other countries. 65. Mandatory and voluntary worker coverage is roughly consistent with the composition of the labor force. About half of workers are employers or have labor contracts and, consistently, a little more than half are mandatory contributors to the Pension Insurance Scheme. Between 15% and 20% of workers are self-employed and about 30% are herders. 66. Coverage of the elderly by pension insurance is effectively universal for men aged 60 and women 55 reflecting the legacy of full coverage from government employment during the socialist era, the 2012 buyback, and early retirement for those in special professions. 4.3.2. The Cost of the Pension System and Sustainability Concerns 67. Fiscal transfers from the State Budget for the Pension Insurance Scheme have been running about 2% of GDP (2.2% for 2017) yet are projected to substantially increase to about 6% in 2030 and almost 11% of GDP by 2050 (Figure 4.18).63 These compare with total fiscal revenues in 2017 forecasted to be about 23% of GDP so the budget for pensions may go from about 9% of revenues today to almost half of revenues in 40 years unless its parameters are modified. The estimated financing gap would be about 78% of 2015 GDP for this baseline scenario64. 62 This section summarizes the report, Mongolia Policy Options for Pension Reform, World Bank, May 1, 2018. 63 The baseline scenario does not include the effects of the April 2017 increase in the contribution rate and retirement age or the February 2017 changes in qualifying conditions for herders, the buyback arrangement and increases in benefits for retired mothers. 64 The financing gap is defined as the present value of the cashflows of the scheme for the coming 75 years. The financing gap is another indicator for determining the impact of a change in parameters. 132 SECTORAL SPENDING FIGURE 4.18 Projected Pension Insurance Scheme Net Financing Gap 1.0% 0.0% -1.0% -2.0% Baseline: Mandatory + Voluntary -3.0% Contribution Rate Increase - -4.0% Mandatory and Voluntary -5.0% Retirement Age Increase: Mandatory + Voluntary -6.0% Schemes -7.0% Price indexation: -8.0% Mandatory + Voluntary -9.0% Schemes -10.0% 2% Linear Accrual Rate -11.0% Combined Scenarios 1-4 - Mandatory and Voluntary -12.0% Source: PROST Projections, 2017. 68. The causes of the projected increases in state subsidies include an aging of the population, relatively low retirement ages particularly for women, a low contribution rate when compared to the benefits promised, and the effects of multiple amendments including the subsidized service buyback. Growing old age dependency ratios result from rapidly growing numbers of retirees and a leveling of the number of contributors. Declines in fertility rates and increases in life expectancy are the demographic factors that have led to growing old age dependency ratios (the ratio of those aged 60+ to those aged 15-59) which are expected to accelerate (Figure 4.19). Fertility rates fell from 7.5 children per woman in the 1970s to 2.1 children per woman in the early 2000s while life expectancy at retirement age increased substantially and is projected to continue to rise (Figure 4.20).65 As a result, the old-age dependency ratio is projected to increase fourfold from about 10% in 2015 to 40% in 2055. This projected rapid increase in the relative size of the retiree population while the covered labor force remains relatively flat profoundly challenges the financial sustainability of the Pension Insurance Scheme. FIGURE 4.19 Population Growth, Age Distribution and Dependency Ratios 2000-2060 Population (thousands) (%) 60+ 14‐59 0‐14 Old Age Dependency Ratio (%) 60 and over (% of total) Source: World Population Prospects, 2015 revision. 65 The total fertility rate rose to about 3.0 in 2016 although it will take many years to realize the effects. 133 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue FIGURE 4.20 Life Expectancy at Retirement Age 2000-2060 Life Expectancy at Retirement Age Female ‐ Age 55 Female ‐ Age 60 Male ‐ Age 60 Source: UN Population Projections, 2015 révision. 4.3.3. Recent Changes to the Pensions System - Impact on Current and Projected Costs 69. The authorities have adjusted pension parameters in 2012, 2017 and 2018 that have provided inconsistent policies, materially increased the required State-Subsidy and adversely impacted system design, fairness and sustainability. a) Deeply discounted service buyback arrangements enacted in 2012 and in 2017 enable many workers to purchase a pension for life at retirement age at a fraction of the cost that other workers and employers have paid having contributed during their work lives.66 This not only is highly inequitable; it also threatens the contributory character of the design and will materially increase the fiscal costs. By enabling affected workers to receive service credit at a fraction of the cost that workers have contributed in the years since 1995, the 2017 buyback undermines the payroll tax financing design of the scheme for many workers.67 b) Supplementing each year of herding by ½ year of service credit creates substantial fiscal costs and establish incentives for workers to report as herders to qualify. c) Reductions in herders’ retirement ages further segments the labor market and introduces costs that adversely affect the system’s sustainability. With life expectancy for men at age 60 of about 15.3 years and women at age 55 about 23.2 years, this provision will materially increase the years in retirement that the State will need to support herders. d) Supplemental service years for women with children68 further adds to the unsustainability of the pension system. e) Requirements for employers to make social insurance contributions on behalf of mothers with small children creates incentives for employers to avoid hiring women of child-bearing age. f) Real increases in the minimum pension and partial minimum pensions since 2012 have reduced the linkages between contributions and benefits. As the partial pension has grown to about 83% of the minimum pension, the incentives for working and contributing an additional 10+ years to fully vest in the pension insurance scheme have been greatly reduced. 66 The buyback enacted in 2017 and to be effective from January 2020 enables herders, the self-employed and informal workers to buyback service between 1995 and the present. The buyback would cost the worker 10% * the minimum wage in the years from 1995- to the present. This buyback is in addition to the 2012 buyback for the period 1991-2000. The effect is to increase the fiscal burden and further erode the differentiation in benefits based on service years and wages. 67 Total contribution rates were 19% of covered wages from 1995-2008 and 14% of covered wages from 2008-2017 and 16% of covered wages in 2018. 68 This provision was legislated in 2017 and is to be effective from January 2020. 134 SECTORAL SPENDING 70. The authorities have enacted select measures that will have some effect in reducing expected costs, although such measures will not be sufficient to offset the costs of the measures mentioned above which increase costs and will not achieve the objective set in 2015 State Policy on pension reform of reducing the required State-Subsidy to 2% of GDP by 2030. a) Increases in contribution rates enacted in 2017 improve the financing of pension insurance by up to 1% of GDP per year. The legislation mandates gradual increases in pension contribution rates returning them to 19% where they were before the 2008 financial crisis.69 b) Increasing the reference wage period for determining benefits from 5 to 7 years will have some effect on reducing costs.70 c) Gradually increasing the pension vesting period from 20 to 25 years for those retiring before age 65 will also somewhat reduce costs. However, the relative ly (relativey) high level of the partial pension with only 10 years vesting may discourage some from fully vesting. 71. Additional weaknesses to sustainability and adequacy of pensions have been identified which include: (a) ad-hoc pension indexation which leaves retirees vulnerable to erosion of the purchasing power of pensions while also creating pressure for periodic benefit increases; (b) the retirement age of 60 for men and 55 for women and early retirement for special professions which also creates incentives for workers to leave the labor market; and (c) other weaknesses which if remedied could improve incentives to participate, moderate fiscal costs and improve the equitability between workers. 4.3.4. Priority areas of reform 72. A package of reforms in parameters and qualifying conditions can improve sustainability and contain the required State Subsidy. This includes modifications to changes legislated in 2017-2018 and other changes needed to parameters and qualifying conditions. 73. The costs and incentives of the 2017 Service Buyback can be improved by increasing the worker cost of purchasing a year of contributory service.71 This reduces the distributional consequences and mitigates some of the dramatic costs of the buyback. It is suggested to make the cost of purchasing service years equal to [the current voluntary contribution rate * the currently applicable minimum wage72] Further, it is suggested to limit the years which can be purchased to only those which can fulfill the vesting requirement and, in this way, prevent individuals from being able to purchase high levels of State subsidized pension annuities. Finally, it is suggested to set a timebound window for the buyback such as 18 months so that it doesn’t become a permanent fixture of the pension system. 74. The costs and incentives of the 2017 herders’ service supplement can be improved by reducing the accrual rate for supplemental service to 0.5% per year.73 The table below illustrates the supplemental benefit expressed as a replacement rate associated with different years of eligible service as a herder: 69 Contribution rates will go from 7% for employers and 7% for employees to 9.5% each over a three-year period of 2018-2020. Contribution rates for the self-employed will go from 10% to 12.5 % for the same period. 70 The legislation increased the reference wage period from 5 to 7 years at a rate of 1 year per year beginning in 2018. The reference wages are not adjusted or ‘valorized’ to neutralize the effects of real wage growth 71 The change was legislated in 2017 and is slated to take effect in 2019. The buyback provides the option for herders, informal workers and the self-employed to purchase service from 1995 to the time of the purchase. The precise cost of the buyback to the worker was not specified. 72 In February 2018, the minimum wage was MNT 240,000 per month. For example, suppose that the minimum wage was MNT 240,000 per month in January 2019, the cost of purchasing a year of service would be [11.0% * 240,000 * 12] or MNT 316,800. 73 Under the changes legislated in 2017 and slated to take effect in 2020, a retiring herder receives a supplement of 0.5 years’ service credit for each year spent working in herding. The implicit accrual rate for such service credit after the vesting period would be 1.5%/year. This proposal is for the supplementary benefits instead have an accrual rate of 0.5%/year. This means for example, that a herder with 30 years of herding service would receive a supplemental benefit of [15 years * 0.5% accrual rate] or an additional 7.5% replacement rate. 135 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue TABLE 4.5 Herders’ Service Supplement: Proposed Supplementary Replacement Rate according to Eligible Service Years (%) Year of eligible service 15 20 25 30 35 Supplementary replacement rate 3.75 5.00 6.25 7.50 8.75 75. Mothers’ service supplement - the applicable accrual rate for supplemental service years for mothers should be 0.5%/year.74 This benefit supplement needs to be considered in the context of other benefits for mothers with children, including the early retirement age for Mother Heroes and service buybacks (legislated in 2012 and 2017), as well as non-pension maternity and motherhood benefits. The table below illustrates the replacement rate attributed to the proposed supplemental benefit according to different numbers of children: TABLE 4.6 Mothers’ Service Supplement: Proposed Supplementary Replacement Rate according to Number of Children (%) Children 1 2 3 4 5 6 7 8 9 Supplementary replacement rate 0.75 1.50 2.25 3.00 3.75 4.50 5.25 6.00 6.75 76. The combined effects of supplementary service credit for mothers and supplementary service credit for herders are illustrated in the table below. Limiting the accrual rate for supplemental benefits both limits system costs and prevents dramatic differences between the replacement rates of different kinds of workers. TABLE 4.7 Herders & Mothers’ Service Supplement: Stylistic Examples of Replacement Rates (%) (Non- (Non-herding) Herder Herder herding) Woman women Men Men Years of contribution 20.25 20.25 20.25 20.25 20.25 20.25 Number of children 1 8 1 8 Years of herding 12.5 12.5 30 Herders’ supplement (years) 6.25 6.25 15 Mothers’ supplement (years) 1.5 12 1.5 12 Total supplement year 1.5 12 7.75 18.25 0 15 Years of service with Supplement 21.75 32.25 28 38.5 20.25 35.25 Pension replacement rate for regular 45% 45% 45% 45% 45% 45% contribution years Pension replacement rate for supplement years 0.8% 6.0% 3.9% 9.1% 0.0% 7.5% Pension Replacement rate (TOTAL) 45.8% 51.0% 48.9% 54.1% 45.0% 52.5% 77. The reduction of five years in the age of pension eligibility for herders should be accompanied by an actuarially fair reduction in the benefit calculation for those herders that choose to retire early. Given the authorities’ commitment to retain early eligibility for pensions for herders, an actuarially fair reduction would fulfill the commitment while also ensuring incentives to work longer. Such actuarially fair reduction factors could be adjusted in line with any future increases in the age for normal service retirement benefits. Table 4.8 shows what percentage reductions in benefits would be applicable at different ages in 2019. 74 The provision requires that employers of women with labor contracts pay the social insurance contributions on their behalf during pregnancy and during the time that any of their children are age 3 or under. 136 SECTORAL SPENDING 78. The provision which requires employers’ payment of pension contributions for mothers of young children should be replaced with an explicit budget transfer for mothers which serves as a matching grant for voluntarily contributions. Matching contributions should also apply to workers without formal labor contracts. TABLE 4.8 Actuarially Fair Reduction Factors (%), 2019 Age Male Female 50 6.3 51 5.1 52 3.9 53 2.6 54 1.3 55 10.7 56 8.7 57 67.0 58 4.6 59 2.4 Source: UN Population Projections, 2015 revision 79. The partial pension level which is currently 88% of the minimum pension needs to be gradually reduced and the years of contributions required for eligibility gradually increased so that the partial pension is gradually eliminated (See Figure 4.22). It is suggested that changes in qualifying conditions -- 10 years of contributory service -- could be gradually increased until the partial minimum pension no longer exists. To better align the partial and full minimum pension, it is suggested that the benefit level is prorated according to the following formula: [Individual years of contribution/ applicable vesting period] * the full minimum pension = the individual benefit. FIGURE 4.21 Relationship between Minimum Pension, Partial Pension, Minimum Wage and Minimum Living Standard 160% 140% Minimum Full 120% Pension / Minimum Wage Minimum Partial 100% Pension / Minimum Wage 80% Minimum Partial Pension / Minimum Full Pension 60% Minimum Full 40% Pension / Minimum Living Standard 20% Minimum Full 2008 2009 2010 2011 2012 2013 2014 2015 2016 2018.2 Pension/Average Wage Source: World Bank estimates based on data from the National Statistical Office, Ministry of Labor and Social Protection. 137 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 80. A transition program is needed to gradually increase the number of years of eligibility for a partial pension while at the same time pro-rating the benefit as a proportion of the full minimum pension for vested workers. Parameters for the level and indexation of the partial pension need to be designed to gradually align with the desired goal of elimination, partially protecting old age income security for those receiving the benefit, and differentiate more adequately between the partial and full minimum benefit. To both protect the purchasing power of those with the partial pension while also reducing the benefit in real terms, it is suggested to: (a) index the partial pension for existing retirees to CPI growth; and (b) to reduce the level of the benefit for eligible new partial pension retirees to the proportional amount of the applicable minimum pension. In addition, it is proposed to gradually increase the required years of service for eligibility as indicated above. 81. A target minimum pension such as 100% of the Minimum Living Standard (MLS) should be established as well as a transition program over several years to achieve it. With about half of retirees receiving the minimum pension and the minimum representing almost 90% of the average benefit, the level and basis are important parameters for pension system design. Broadly speaking, linking the minimum pension to a basic consumption basket such as the MLS is a means of assuring a minimum level of consumption. A second important issue is how to transition towards the desired minimum while not reducing the nominal benefit. This is particularly challenging in Mongolia where the minimum pension in February 2018 was about 140% of the MLS (see graph below). It is challenging to smoothly move the level of the Minimum Pension towards the level of the MLS while also not reducing the benefit level for existing retirees. FIGURE 4.22 Transition Options for Real Value of Full Minimum Pension (Constant 2018 MNT/Month) Full Minimum Pension (Post 2018 Retirees) Minimum Living Standard 300,000 280,000 260,000 Monthly Benefit 240,000 220,000 200,000 180,000 Source: World Bank estimates. Note: Assumes no real growth in the Minimum Living Standard. Full minimum pension is assumed to grow at a rate of [CPI growth – 2.5%] but will not be reduced if CPI growth is less than 2.5%. 82. One option to consider is to set as a target of the minimum pension to be 100% of the MLS and to increase the minimum pension at a rate of [CPI growth - 2.5%] however never reducing the nominal benefit of the minimum pension (See Figure 4.22 below).75 Under the assumption that inflation exceeds 2.5% for each year, this transition could be completed in about 13 years. 83. Pension benefits should be automatically indexed to ensure the security of retirees and the indexation should be a clear, transparent and highly regarded index as the growth in the Consumer Price Index (CPI). The existing benefit formula provides no automatic pension benefit indexation thereby exposing retirees to the risk and vulnerability of a loss of purchasing power of their pension benefits. Moreover, under the current framework, all adjustments 75 It is important to gradually achieve the objective though not at the expense of reducing the nominal value of the minimum pension. 138 SECTORAL SPENDING are subject to government approval which invites delay and uncertainty and, in recent years, volatility. Automatic indexation of pensions according to the growth in the CPI will improve the security of workers and retirees and reduce the pressure for periodic across-the-board benefit increases. 84. A small actuarially fair reduction in the benefit according to age at retirement could improve system finances and create incentives to work longer. An actuarially fair reduction should be calculated at a reduction associated with a 3 month increase in the retirement age for each for the first 3 years and 6 months per year thereafter until a reference point of age 65 is reached prior to the calculation of the reduction or supplement. 85. Reducing the ceiling on covered wages for contributions and adopting a different metric would better align the parameters with the core functions of a mandatory contributory pension insurance scheme. The current ceiling on covered wages of 10 times the minimum wage is both very high relative to the distribution and may not be the appropriate metric. At end-2017, although the average covered wage was over four times the minimum wage, about a quarter of contributors contribute at or close to the minimum wage. It is important to recognize that State- sponsored pension insurance schemes generally seek to protect a basic level of income replacement and therefore have a cap on covered wages with the expectation that workers with incomes above the cap can voluntarily save in other savings or pensions instruments. It therefore is worth considering a lower cap on covered wages which will also prevent some workers with very high wages from taking advantage of very high pension benefits. In addition, many countries have moved the metric they use as the basis for the cap from minimum wage to average covered wage, essentially because minimum wage policy aims at an objective of a working wage which is entirely separate from the concept of a cap on covered wages for pensions purposes. In Mongolia, the authorities therefore might want to consider changing the metric for limiting covered wages to average covered wages of the prior year and setting a limit of perhaps 300% to 500% of average covered wages as the cap. 86. Establishing a ceiling on pension benefits could prevent excessive replacement rates, particularly in view of multiple supplementary benefits recently enacted. One means of limiting the replacement rate resulting from recent service supplements and buyback arrangements is to introduce a cap on benefit replacement rates, perhaps at 80%, which is a common feature of pension insurance schemes. 87. Changes to the administration of disability pensions are needed including: (a) measures to re-assess partial and full disability; and (b) measures to proactively work with employers to retain and rehire disabled employees. The costs of disability pensions have grown substantially and may likely keep growing in the future. Disability costs can be contained by changes in benefits such as changing the minimum benefit formula applied to disability as well as refining the vesting rules for disability benefits. 88. Additional parametric reforms are recommended in the medium-term. Those are presented in the following paragraphs and would help improve sustainability as well as improve the equity and incentives in the scheme 89. Increasing the retirement age would both help financial sustainability and strengthen incentives for workers to remain in the labor force. Increasing the retirement age at a rate of 6 months per year to bring men and women up to a retirement age of 65 is projected to reduce the projected State Subsidy in 2050 by about 2.0% of GDP. 90. Unifying the retirement age for all workers would also be beneficial. Applying the same retirement age to all workers, including those in special and hazardous professions, and Mother Heroes with four or more children would further reduce the required state subsidy.76 Importantly, it would also equalize the incentives in the mandatory scheme for older workers to stay in the labor force across professions. 91. Further increasing the length of the wage reference period and indexing or ‘valorizing’ the wage base will improve the fairness of the benefits relative to contributions. Increasing the wage reference period as was legislated in 2017 will somewhat reduce pension costs for new retirees because individual wages are typically rising during the wage reference period. It is therefore suggested to extend the number of years in the wage base from the current 6-7 years and indexing or valorizing the wage base used to calculate benefits. 76 According to the MLSP, 38% of all old age retirees were classified as Mother Heroes (the overwhelming majority of women retirees), and 10% were classified as from special or hazardous professions. The proportion of Mother Heroes will undoubtedly decline substantially over the coming years as a result of reductions in fertility rates so that there are far fewer mothers of 4 children. 139 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue CHAPTER 5 Institutional Framework for Improved Fiscal Management OVERVIEW AND CONCLUSION Key Highlights  The roles and responsibilities of the key Elements of a Priority Reform Agenda institutions in fiscal matters and the The reform strategy will mainly include: relationships and coordination between them are well defined in the legal framework,  Revisit the independence of the recently specifically in the Integrated Budget Law and established Fiscal Council. the Fiscal Stability Law.  Strengthen capabilities of the existing Budget  The performance and effectiveness of these Analysis Unit in the Parliament Secretariat. institutions are varied following persisting  Facilitate greater transparency and citizens’ challenges including: participation in fiscal issues and decision o MDAs don’t observe hard budget making by fully implementing the existing constraints, laws and relevant regulations. o The Parliament has substantial power  Accelerate ongoing efforts to engage the over budgeting while lacking on technical public in the oversight of fiscal matters building capacity, on available budget savings successful stories. o Accountability institutions like the MNAO  Harness mining revenue management and the IAAC have limited effectiveness through using revenues to accelerate due to their weak independence, Mongolia’s diversified development - Ensuring a well-functioning stabilization fund and o The recent progress in public procurement the Sovereign Wealth Fund is an important reform was reversed as the dedicated element this effort. Government Procurement Agency was integrated into the State Property Policy Coordination Agency, and o Despite the efforts in advancing the fiscal transparency agenda, the country has run short in implementing the relevant laws and in fulfilling the OGP commitments. 140 PUBLIC EXPENDITURE REVIEW 141 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 5.1. Current Institutions for Fiscal Management - Performance77 1. The two major Public Financial Management (PFM) laws include the Integrated Budget Law (IBL) of 2011 and the FSL of 2010. Prior to the IBL, fiscal management was regulated by the Public Sector Management and Finance Law (PSMFL) of 2003 and the General Budget Law of 1992. The laws adopted by the Mongolian government aimed at strengthening fiscal discipline and stabilizing public spending and public deficits have in practice experienced large implementation gaps. Also, the basic institutional set-up for the management of revenue, expenditures and debt in the country was inspired by that of many other well-managed governments. So far, however, political pressures have been so strong that neither these laws nor the existing institutional frameworks have been effective in preventing the approval of budgets based on overoptimistic revenue projections, and generating unsustainable levels of recurrent and investment expenditures. 2. Overall, the budget process has been orderly, but trade-offs between expenditures have not been made as envisaged in the law, and no hard budget constraint was observed. The three main stages of the process include (a) initial circular, submission by MDAs and compilation by MOF, (b) decision-making by the Cabinet, and (c) discussion and approval process in Parliament. Annual budget preparation is preceded by the preparation and two stage approval of the Medium Term Fiscal Framework by the Government and by the Parliament (IBL, Para 8) (preparation by April 15, Government approval by May 1 and Parliament approval by June 1). Specific budget ceilings for MDAs are prepared between early June and early July (IBL, Para 8.3). MDA submissions to MOF are made between July 25 and August 15, jointly with submissions of half-yearly budget execution reports and financial statement (8.9.5 and 8.9.6). Revenue estimates for the coming fiscal year and the medium term are carried out by the Fiscal Revenue Division under the Department of Fiscal Policy and Planning in MOF. For estimating commodity prices, the division draws on international forecasts (London Metal Exchange). BOX 5.1 Key Institutions for Fiscal Management in Mongolia In fiscal management in Mongolia, the following institutions play a major role: the State Great Khural (the Parliament), the MOF, and the agencies reporting to MOF, that is, the General Department of Taxation and the General Department of Customs, the MNAO, and so on. • The State Great Khural (Parliament) is the main policy institution in terms of approving major pieces of legislations such as annual budget, MTFF, annual audited budget performance, medium- term and annual strategic programs, and amendments to the annual budget and tax laws. • The MOF plays a major role in formulating fiscal policy documents for submission to the Parliament and carry out, implement, and report to the Parliament on the execution of fiscal policies. The MOF is responsible for drafting annual budget and MTFF, reporting on budget execution, managing the government debt, and proposing amendments to the tax laws. Moreover, state investment programs and its financing are formulated and the related transactions are processed by the MOF. The MOF is also responsible for managing various funds created for special purposes such as the HDF, the Stabilization Fund, and the SIF. • The General Department of Taxation is a major arm of the MOF for collecting tax revenues. Through its local offices located at each administrative districts, this general executing agency collects tax payments to the department’s account, which is swept into the TSA daily. uuu The analysis in the greater part of this section was carried out as part of a background note by Verena Fritz and Sue Harding on Mongolia: non-technical 77 drivers of fiscal consolidation and potential implications for WBG engagement, July 2017, (not published) 142 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT • The General Customs Office is a government executing agency for collecting customs tax and duties on exports and imports. The department directly reports to the MOF. • Other Institutions: The following institutions also play a critical role in managing and formulating policies in the mining sector, coordinating policies across ministries and agencies, and auditing the budget expenditures. • The Ministry of Mining is responsible for formulating and implementing policies in the mining sector at the state as well as at local levels that are detrimental to the fiscal management in the country. The Ministry of Mining is responsible for proposing mining policies that would affect livelihoods of local citizens through changes in royalties, taxes, and tariffs in the Minerals Law. • The NDA is a government regulatory agency primarily responsible for coordinating development policies and prioritizing development projects. This agency directly reports to the Prime Minister’s Office and oversees coordinating development and investment policies across ministries, formulating regional development policy as well as policies in PPP. • The National Audit Office is a body responsible for carrying out audits of public financial transactions and reporting to the Parliament on its findings and recommendations. • Local Governments get their local government budget ratified by the local Citizens’ Representatives Khural, which also oversee the budget implementation. Aimag Governor’s Office is responsible for submitting budget proposals to the Khural and is responsible for its implementation at the aimag level. Similarly, each soum has democratically elected Khural of Representatives, which approves the budget and oversees its implementation. Soum Governor’s Office is responsible for budgeting and proposing the budget to the Representatives’ Khural. • Citizens and Civil Society Organizations (CSOs) have a strong legal backing for holding government accountable for public expenditures, but in practice there is a lack of capacity in the CSOs and inefficiency or lack of coordination or functioning of the accountable mechanism. Currently, there are around 40,000 CSOs registered in Mongolia. • Approximately 28,000 of these are officially registered nongovernmental organizations (NGOs), but they are not active. According to the definition of civil society adopted by World Bank, the term ‘CSO’ can refer to a wide range of organizations including NGOs, community groups, research and policy institutes, advocacy groups, trade unions, some parts of the media, professional associations, and faith-based institutions. 3. When making their budget submissions, MDAs do not observe the ceiling set out by MOF, but typically make submissions that are substantially higher78. The Department of Fiscal Policy and Planning then makes initial reductions to these submissions, questioning any new policy initiatives and similar items. It then prepares a list of trade-offs to be made by the Cabinet during its deliberations at the end of September. The complete budget is then submitted to Parliament on October 1, where the budget undergoes three readings and votes, plus one final overall vote. 4. During the parliamentary stage of the budget, there are several important stakeholders: the various standing committees (and powerful members within these), but also party caucuses,79 the Standing Committee on the Budget, and the Speaker. Key changes are discussed and voted on between the first and the second reading. Following the second reading, the budget standing committee submits the proposed revisions back to MOF to formalize the adjustments. Since the mid-2000s, parliament has begun interpreting constitutional provisions and existing legislation in ways that allow it to change the composition of the budget as well as the overall envelope. Thus, some investment projects are removed from the list, and others are added in during the parliamentary stage. 78 For example, in 2008, line ministries submitted requests for capital expenditures exceeding the final MOF proposal by MNT 700 billion – submitting requests of MNT 1.2 trillion versus MNT 482 billion proposed by MOF to parliament and approved. Party caucuses are formed by any party represented in the State Great Khural with more than 8 members. 79 143 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 5. A key political rationale for MPs during the deliberation is to seek to channel funds both towards party- wide as well as individual election promises. The latter type of election promises is focused on specific election districts - the size of which has varied between elections, as in some elections, multi-member and in others, single-member districts have been used80. For party-wide pledges, recurrent expenditures such as transfers or civil service salary increases, play a more significant role than investments. 6. Furthermore, while parliament has substantial powers over budgeting, it has only limited technical support staff for budget analysis, and hence limited capacity. While there is a Parliamentary Secretariat, it primarily has administrative functions rather than a capability to provide substantial, technical advice to MPs and standing committees. One advantage is that in a context of frequent rotations in offices, and MPs moving in and out of being cabinet members, some members and chairs of the standing committee on the budget have been previously ministers of finance. Thus, some members have substantial own technical expertise and experience in understanding key budget provisions. However, this expertise is not permanent and not evenly distributed among members, or available to other committees. Thus, while parliament holds key fiscal decision-making power, its ability to access relevant knowledge and advice, for instance on likely fiscal impacts, is limited. BOX 5.2 Constitutional Provisions Regarding Parliamentary Powers over Fiscal Decisions and Provisions of the FSL on Deficit and Debt Limits • Constitution, Article 20: “The National Parliament is the highest organ of state power, and the supreme legislative power is vested only in the National Parliament.” • Article 25 (7) defining competencies of the Parliament: “to define the State’s financial, credit, tax, and monetary policies, to lay down the guidelines for the country’s economic and social development, to approve the Government’s program of action, the State budget, and the report on its execution.” • FSL provisions (2010, amended). • Article 6.1: The structural balance of consolidated budget shall either be not more than the deficit of 2% of the GDP of the particular fiscal year or be in surplus; Net present value of Government debt outstanding shall not exceed 40% of nominal gross domestic product of the particular fiscal year. • Article 7: “It is prohibited to prepare and approve medium term fiscal framework statement violating the special fiscal requirements and prepare and approve annual budgets and make any amendments and changes to the approved budget if they are inconsistent with the medium term fiscal framework statement except in the circumstances as defined in this law.” • Article 18: Any violation of Article 7 of this law shall be the rationale for making the authority and officers liable under the law. […] The case where the State Great Khural did not meet the special fiscal requirements in the approval of the Medium term fiscal framework statement and annual budget and making amendments to them shall be the rationale for the President of Mongolia to exercise his power to veto as stated in article. • 33.1.1 of the Constitution. • According to the original version of the FSL, the 2% deficit limit and a limit to the growth of public expenditure were to come into force from 2013 and the debt limit to 40% (net present value [NPV] relative to nominal GDP) was to be in force from 2014, that is,. both following the 2012 elections. 80 In addition, in the 2012, a share of seats was allocated through a proportional vote. This was reversed for the 2016 elections which went back to a system of single-member electoral districts. 144 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT 7. Even more problematic than parliamentary influence on the budget has been the creation of large scale off-budget expenditure programs since 2010. These have included (a) commercial and public investments through the Development Bank of Mongolia (DBM), (b) Public Private Partnerships, and specifically ‘Build-Transfer’ projects (PPP/BT), and (c) lending programs by the Bank of Mongolia (BoM) at subsidized interest rates. In addition, there was at least one instance of forward sales of mining output. In particular, the DBM operations and the PPP/BT investments were facilitated by the opening up or growth of commercial borrowing opportunities, with the former using mainly externally, and the latter mainly domestically borrowed funds. These off-budget programs fragmented fiscal decision making and reduced transparency both about the overall fiscal stance as well as concerning individual approvals for specific projects. 8. The creation of these off-budget programs followed the adoption of a Fiscal Stability Law in 2010, that is, a legal act designed specifically to constrain government fiscal decisions. According to key staff who worked in the Ministry of Finance at the time, MOF was concerned about the recent experience of fiscal volatility, pro-cyclical expenditure decisions, and a rapid expansion of social programs (such as the introduction of a Universal Child Benefit in 2006) for political reasons; seeing these as risks to longer term good fiscal management and for using the unprecedented boom in commodity prices for large- scale investments that would help to boost the economy over time81. MOF’s ‘plan A’ at the time was to achieve a Constitutional amendment that would have limited the powers of Parliament to expand the overall revenue and expenditure envelope82. However, efforts towards such an amendment failed, and hence MOF pursued a Fiscal Stability Law with very stringent sanctions for breaching fiscal discipline. Efforts at developing the law and advocating for it were supported by the WBG and the IMF. 9. Looking at the public administration system more broadly, several institutions of constraint have been established, with some, but still limited effectiveness83. Key institutions include the Mongolia National Audit Office (including aimag-level sub-offices), and the Independent Authority Against Corruption (IAAC). The MNAO was established in 2003, based on a previous SAI that was set up in 1995, and the IAAC in 2006, following the adoption of an Anti-Corruption Law in Parliament. 10. Through various twinning efforts and other forms of external support, the MNAO has achieved relative organizational effectiveness, as also reflected in the 2015 PEFA assessment. The pertinent indicator PI-26, ‘Scope, Nature and Follow up of External Audit’, received a B+ rating. However, as a pivotal organization, the MNAO has not been protected from the trend towards politicization. Auditor Generals have been changed following electoral changes; with the most recent AG being appointed in early 2017 following the request from the previous AG to be released from his duties despite not having served his full mandate84. 11. The Mongolian National Audit Office is responsible for auditing operations of all levels of government from ministries down to agencies and entities that received financing from the budget including state owned enterprises. Also, it is responsible for auditing usage of government loans, grants, donations, operations of various government funds and its results. Since the scope of their activity can become too wide, auditing is conducted by private audit firms. MNAO directly reports to the Parliament every year and the General Auditor is appointed by the Parliament for 6 years. The operations of the MNAO is overseen by the Budget Standing Committee of the Parliament. At the implementation level, if the 81 Interviews, UB, March 2017. 82 Mongolia has a single-chamber Parliament with 76 seats. Election modalities hav e changed several times since the first elections held in 1992, but have been predominantly majoritarian. One report on the 2010 discussions about Constitutional Amendments is the following: http://www.loc.gov/law/ foreign-news/article/mongolia-constitutional-amendment-procedures-to-be-discussed-by-parliament/. 83 This set of issues can only be briefly touched on here. The OECD (2015) report provides a good overview from an anti-corruption perspective. 84 http://en.montsame.mn/economics/8273. 145 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue entity does not agree with the audit report, it can appeal to the court for the resolution. The MNAO’s annual report states that in 2016 86.7% of cases had no conflict, whereas MNAO imposed Togrog 14 billion of fines of which 40% was reflected in the financial report. At the Parliament level there is no indication that there is a backlog in reviewing audit reports and it became customary for the Parliament to approve the report without recommendations. 12. Relative to the MNAO, the IAAC is still a younger organization, and whether and how to prosecute corruption has been highly contested. According to the 2015 OECD assessment on Anti-Corruption, the IAAC has a broad mandate but too little capacity to actually fulfill this, and hence recommended focusing on key mandates and in particular the investigation and where appropriate prosecution of high level corruption cases. An initial National Anti-Corruption Program (2002-2010) was followed by a hiatus, as different stakeholders could not agree on a new program. Eventually, a new Anti-Corruption Program was adopted by Great Khural representatives, following the 2016 elections (OECD 2016)85. An Action Plan for implementation was subsequently also adopted. The IAAC also agreed sector specific action plans for health, education, construction, and mining, that is, two service delivery sectors, and two sectors that are globally prone to high-level corruption and rent-seeking. 13. In addition to these two institutions, there have been several institutions and ad hoc councils created with a view to constraining opportunities in various areas of economic governance. These include oversight over procurement (most recently re-centralized in a Government Procurement office), the creation of a Financial Regulatory Commission, a Monetary Policy Council established at the Bank of Mongolia, with a view to broadening the decision making of the BoM, as well as EITI for the mining sector. One of the key demands of the IMF EFF is the creation of a Fiscal Council, which is intended to support constraining fiscal decision-making, discussed further below. A key challenge for these institutions meant to constrain decision-making as well as to make it more transparent, is to remain sufficiently independent as well as effective. For example, the Monetary Policy Council appears to have become more politicized over time86. 14. A parliamentary resolution was passed to establish a Fiscal Council under the Budgetary Standing Committee of the Parliament. Given this arrangement does not guarantee political independence of the Council as advised and evidenced from international good practices, the IMF considered this structural benchmark not met during its third review of the program in March 2018. 15. Substantial staff turnover, frequent restructuring of institutions, and bringing politically affiliated staff into positions combined has a substantial cost in terms of public sector capacity and integrity. While Mongolia readily outperforms LMIC averages for WGIs on Voice and Accountability and for Political Stability, performance on ‘Government Effectiveness’ has been close to the LMIC average and dropped below this in 2012, recovering somewhat since then. For Control of Corruption, Mongolia performs at the average level (38 compared to a 37 average) on a scale from 0 (worst) to 100 (best), recovering from a drop during the boom years of 2010 and 2011. 16. In terms of public perceptions, in 2017 the ‘National Government’, that is, the overall executive was ranked among the top five corrupt institutions in the country for the third consecutive year since first appearing on the list of top five in 2015. While previously, the top five had tended to be implementing agencies - land administration, customs, judges, and then increasingly political parties - since 2015, three core institutions: political parties, national government, and parliament/legislature were named among the top five corrupt institutions, jointly with Land administration and mining87. 85 https://www.oecd.org/corruption/acn/OECD-Mongolia-Progress-Update-Oct-2016-ENG.pdf. 86 Interviews, UB, March 2017. Current membership in the MPC is comprised of 9 BoM senior officials, and 7 independent members. The latter include the economic advisers to the PM and to the Speaker of Parliament. The MPC publishes statements on a quarterly basis. 87 Survey on Perceptions and Knowledge of Corruption, The Asia Foundation (March 2017). 146 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT 17. By 2017, political parties moved up to the second place from the fifth in 2015 on the list of the most corrupt entities as per the Asia Foundation’s Corruption Perception Survey. Attempts to improve the transparency and oversight on political party financing have been introduced in the Law on Parliamentary Elections (2015) and the Law on State Auditing. Accordingly, the Mongolian National Audit Office is mandated to audit the contributions received by the political parties and election candidates as well the expenses incurred by each candidate. The audit of the 2016 Parliamentary election revealed a total of MNT 8.36 billion was received in revenues (of which MNT 4.55 billion from individual contributions, MNT 853.1 million from business entities, and MNT 2.96 billion from assets of the parties and the candidates) and a total of MNT 8.35 billion in election campaign expenses. The audit concluded that all candidates’ campaign expenses were retained within the approved thresholds for their respective constituencies. As required by the law, the MNAO had made the list of individuals and business entities that made a contribution of above MNT 1 million and MNT 2 million, respectively, while the maximum allowed contributions are MNT 3 million and MNT 15 million for individuals and firms, respectively. Despite these efforts of transparency and oversight, the results of the corruption perception survey still signal a worrying trend. As the need for funding required to win an election increases political parties could be looking to ‘private/secret’ funders undermining the integrity of the electoral process.88 18. The Public Procurement Law of Mongolia (PPLM) was developed based on international best practices, requires competitive bidding as the main procurement method, and defines the conditions under which other procurement methods can be applied. The current PPLM still includes, however, significant exemptions, notably for the Development Bank of Mongolia (DBM) and for maintenance of roads, an activity entrusted directly to SOEs. Since its establishment in 2011, the DBM has organized a large number of procurement. Given projects financed by the DBM are exempted from following the PPLM, DBM has mainly used direct contracting, affecting the effective functioning of the entire procurement system. Consequently, Mongolia scored a mere D+ in the most recent PEFA report against the indicator for “competition, value for money and controls in procurements”. 19. The revisions in the PPLM adopted in 2011 introduced significant changes in the public procurement system. Many responsibilities for procurement were transferred to the Government Procurement Agency (GPA), a newly established procurement professional agency at the time. At the same time, local governments were given more responsibility for procurement. Thus, the procurement function underwent simultaneous centralization and de-centralization. The GPA was made responsible for all procurements of national and regional nature as well as for establishing framework agreements for common use items (including office supplies) to be purchased by line ministries. Local governments were made responsible for all procurements of works, goods, and services to be financed from the local budget, as well as for local projects (including schools and hospitals) financed from the national budget. Another significant change in the PPLM was to strengthen the role of civil society and the private sector in public procurement. For this purpose, the revised law required bid evaluation committees to include at least three representatives from outside of the government. The revision also required the need for all bid evaluation committee members to be specialized in procurement, thus required special certification. 20. Achievements introduced by the 2011 amendments have been partially reversed after the change of government in 2016. The GPA became part of the State Property Policy Coordination Agency (SPPCA), while the Procurement Policy Department at MOF was integrated as Procurement Division into the Ministry of Finance’s Legal Department. Under the PPLM, MOF must play the central role in public procurement. This change can be interpreted, along with the simultaneous restructuring of the GPA, to signal a diminishing commitment of the Government in procurement development. 88 https://asiafoundation.org/2017/10/18/political-party-financing-mongolia-road-grand-corruption/ 147 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 21. In order to help ensure efficient and proper use of state and local government funds, the transparency of decisions and actions concerning budget management and public oversight of the same, Mongolia passed the Law on Glass Accounts. The Law was approved by the Parliament in July 2014 and entered into force on January 1, 2015 with the aim to enable transparent budget management decisions and processes and create a glass account portal allowing public monitoring of the spending of state and local budget as well as state and local assets. 22. The national efforts have been reinforced by global commitments towards ensuring more transparency and accountability to the Mongolian people. Mongolia joined in 2013 the Open Government Partnership (OGP), a multi-country initiative aimed at supporting governments to become more open, accountable, and responsive to their citizens. Following the OGP membership rules, Mongolia has so far adopted two national action plans (NAP). As part of the second action plan, commitment #9 relates to the effective implementation of the Glass Account Law which created high expectations in Mongolia’s performance among members of the OGP initiative and the commitments obtained a ‘star’ for its potential transformational impact in the country. BOX 5.3 Subprojects on Public Monitoring of Local Procurement of Medicine and Medical Equipment as Well as of School Budget under the World Bank and Mainstreaming Social Accountability in Mongolia (MASAM) Project financed by the Swiss Agency for Development and Cooperation Under the MASAM project, an independent monitoring initiative of Uvs aimag’s local authority and CSOs demonstrated that not only the procurement law implementation is feasible but also citizens’ monitoring has positive implications on access to and delivery of quality medicine, participatory procurement practices, empowerment of the people, and better resource utilization. In the past, the procurement process by the Aimag Health Department, as undertaken by the Bid Evaluation Committee, led to purchase and delivery of low-quality medicines and supplies, which, when dispensed to patients and clients, has contributed to growing dissatisfaction over the services, medicines, facilities, and equipment of the hospital. In the local CSOs’ analysis, the procuring entity is unable to exercise accountability on the supplier in cases of delayed deliveries and delivery of drugs that are not compliant to specifications (either low quality or totally inappropriate drug replacements). To resolve this issue, the local CSO network, coached by Transparency International Mongolia, initiated the practice of preparing tender and contract documents together with the Bid Evaluation Committee and monitoring the procured items at the Aimag General Hospital since 2017. Consequently, the associated budget savings totaled MNT 112 million in 2017 due to the use of the revised bid documents as a clearer delivery schedule and technical specifications were introduced. Delays and mismatch product delivery were reduced for execution of 20 contracts with a total value of MNT 4 billion. In 2018, even after the sub-grant project ended, the participatory procurement practice continued and a total of MNT 55 million was saved from the planned budget. Moreover, the local CSOs’ procurement knowledge and monitoring capacity have been strengthened and patients’ satisfaction toward regular medicine supply and quality of medicine was increased according to citizens’ satisfaction survey conducted by the local CSO together with the hospital staff. These efforts are expected to be continued in the future. 148 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT In Govi-Altai aimag, the local CSOs led by Altai Swiss NGO carried out school budget monitoring for the first time at four schools (two aimag center schools and two rural soum schools) as part of a subproject funded by the MASAM project in late 2017. During the monitoring, the local CSOs developed and employed relevant questionnaires and interviews as survey methods despite their lack of experience and knowledge on public expenditure monitoring. A total of 252 respondents shared their views regarding the school problems. As a follow-up, the local CSOs in cooperation with Aimag Governor’s Office and Aimag Education Department held a public hearing and openly discussed the school budget and human resource issues, which were raised from the monitoring exercise. It concluded that there is a lack of accountability on education budget expenditures, which is approximately MNT 25–27 billion per year. The public hearing is worth highlighting because it was the first-time ever event held on local budget transparency in the country since the adoption of the Public Hearing Law (2015). As a result of the sub grant project, the local stakeholders including the local CSOs, school administration, parents, and local authorities became more knowledgeable and aware of transparency of school governance and budget matters along with the general human resource problems (for example, insufficient level of participation by parents and teachers in school budget planning and its expenditure, lack of budget information disclosure by school administration, teachers and parents’ inadequate financial knowledge, as well as weak capabilities of school professionals, and so on), which were hardly acknowledged and monitored by the relevant stakeholders before. Moreover, the policy suggestions and immediate follow-up actions directed to school principal, Aimag Education Department, and the Aimag Governor’s Office were formulated and presented at the public hearing for further resolution. 23. Despite the initial efforts in advancing the fiscal transparency agenda the country has run short in both implementing the law and in fulfilling the OGP commitments. On the 2017 Open Budget Index (OBI) Mongolia is among the countries that release limited budget information. In fact, Mongolia’s score decreased from 51 (out of 100 points) in 2015 to 46 (out of 100 points) in 2017. The country is also lagging the global average score regarding opening opportunities for citizen participation (7 out of 100 points). With regards to OGP, the completion rate of Mongolia’s commitments was ranked only at 24% – significantly below the global average of 51% with major drawbacks on advancing those activities related to citizen participation and public accountability. 24. The uneven results for advancing the transparency and public participation agenda on the account of both national and international commitments have been confirmed by the clear evidences of low level of citizens’ involvement in decision making and citizens’ feedback. According to a baseline survey conducted in 2016, in total, 72 per cent of respondents had not attended a local khoroo or bag meetings in 2015. Lack of information was cited by nearly 40 per cent of respondents who were not able to participate because they were not properly informed. Also, 8 per cent of respondents did not participate because they felt that they it was not their place to contribute or that no one would listen to them. Only 63 out of 1082 respondents (six per cent) had lodged a form of formal complaint over the past year. Only 13 per cent of those citizens who had made a formal complaint stated that they were either ‘very’ or ‘somewhat’ satisfied with the outcome. 149 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 5.2. Fiscal Management of the Mining Sector - Performance 25. For resource rich countries like Mongolia, resource revenue management is the most important part in the fiscal management. In an attempt to establish a mechanism similar to a SWF for accumulating revenues from the mining sector for the economic and human development of the country, the government created the statutory Mongolia Development Fund (MDF) in 2007 and subsequently the Human Development Fund (HDF) in 2009 as a replacement of the MDF. Unfortunately, despite the good intentions, these funds that contributed as major domestic factors to the 2008-2009 economic crisis, were used to fulfill political promises such as universal cash transfers, and untargeted social welfare measures during the parliamentary elections of 2004 and 2008. Subsequently, in 2011 political parties agreed on a memorandum that stipulates non-competition among each other through political promises of cash handouts to the citizens. However, the idea of a more robust fiscal and mineral revenue management remains at the center of economic policy debates in Mongolia and further attempts were made to meet the goal. 26. The FSL attempted to impose budget discipline and delink fiscal policy from mineral revenue fluctuations and was meant to go into force in 2013. The FSL introduced the following three key special requirements: (a) A ceiling on the structural deficit of 2% of GDP. Budget revenue has to be estimated on a structural basis and structural revenues are defined as revenues that would be received if the prices of major minerals were at a particular level, defined as a 16 year moving average of mineral prices (past 12 years moving average plus 3 years’ projection). (b) A cap on expenditure growth based on the non-mineral GDP growth rate. The total budget expenditure growth rate shall not be greater than the greatest of the non-mineral GDP growth rate of that particular year and the average of the non-mineral GDP growth rate for 12 consecutive years (Expenditure Growth =< Greatest of [Non mineral GDP] or [Average of Non-mineral GDP growth of past 12 years]). (c) Net present value of public debt cannot exceed 40% of GDP (effective from 2018). 27. However, due to practical challenges of implementing the special requirements stipulated in the original version of the law, the Parliament amended the law. The amendments entailed postponing the effective dates of the special requirements for them to be progressively met in the years leading up to 2018. Key changes introduced by the amendments were as follows: (a) A new structural budget deficit ceiling set at 5% of GDP for 2015 that is mandated to be gradually tightened to 2% of GDP by 2018, including the budgetary projects financed by the Development Bank of Mongolia. (b) The ceiling on the present value of the government debt – formerly 40% of GDP - relaxed to 58.3% of GDP for 2015 and mandated to be gradually reduced to 40% of GDP by 2018. The government debt is not included government guarantees, central bank foreign liabilities, or state-owned enterprises borrowing. 28. In addition to numerical rules, the FSL also incorporates procedural and transparency reforms to strengthen budget discipline. For example, the law specifies that decisions on fiscal policy must be taken first within a medium-term fiscal framework. The budget is then decided on within the approved framework. The fact that amendments to the FSL require a two-thirds parliamentary majority signals that the authorities are fully committed to the rules-based approach. The government also manages funds such as Human Development Fund (HDF) and Social Insurance Fund (SIF) whose budget is ratified by the Parliament along with the state budget. These finds are main programs of the social welfare programs in the country. SIF is operated in accordance with the Social Insurance Law and Public Health Insurance Law and implemented by Government executing agency the General Department of Social 150 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT Insurance (GDSI). The GDSI is responsible for reporting implementation of the social insurance policy to the National Committee on Social Insurance and the Ministry of Human Development and Social Protection, which in turn accountable to the Parliament for implementation of social insurance policies. In contrast, the main purpose of the HDF was to allow citizens to benefit from the mining sector boom. This fund was operational in accordance with the Law on the Human Development Fund, which was repealed in 2016. The repeal of this law is partially related to the fact that the commodity prices were depressed, but also the Parliament enacted the Future Heritage Fund Law. 29. The Fiscal Sustainability Fund (FSF) and Future Heritage Fund (FHF) are funds designed to first stabilize the government expenditure as well as create a wealth fund for future generations. The Parliament ratifies annual budget for each of these funds and the MOF is responsible for the accumulation and overseeing of funds. These funds are operated under the Fiscal Stability Law and the Law on Future Heritage Fund, respectively, and have the following characteristics: 30. The FHF is expected to be an intergenerational savings fund with $100 million–$200 million being deposited into it annually from mineral revenues and royalties. According to the FHF Law passed in 2016, the FHF will be effective upon clearing all outstanding debts of the former HDF, which were estimated 1070 billion as of December 31, 2016. In addition to the retiring HDF’s debts in years 2017- 20, FHF will make transfers to state budget in pre-determined amounts totaling to MNT 210 billion. This is to facilitate smooth transfer of HDF expenditures, that is, child money and servicing of HDF debts. 31. Given current applicable gross revenue collection of approximately MNT 500 billion per annum, the HDF’s outstanding debts should be cleared by 2020, with FHF assets becoming positive from that time onwards. The revenues collected and accumulated in the FHF consist of the following: (a) about 65% of all royalty payments on minerals collected by the government (in accordance with Article 47 of the Law on Mineral Resources, after deductions for the Stabilization Fund established under the Fiscal Stability Law), plus an additional 20% of government mineral revenues over and above the level collected in the base year of 2017; (b) all dividends received attributable to state-owned shares of both state-owned and privately managed mining companies; and (c) half of any revenue raised from new or incremental taxes on exploitation of mineral resources. TABLE 5.1 : Revenue and Expenditures of the Future Heritage Fund (MNT billion)   2017 Amendment-II 2018 plan Revenue 333.3 525.7 Transfers To the state budget 160.0 100.0    To pay off payables of the HDF 173.3 425.7 Source: The 2018 budget document, MOF 32. The Fiscal stability law (FSL) already has countercyclical function for stabilization and but it’s very limited due to continued budget deficit and the intensifying fiscal pressure in recent years. As commodity prices has been falling since 2011, the budget mineral revenues has been declining. The budget spending has not been adjusted much, disconnecting from the rule and the FSF acted as a buffer in 2014 and 2017. 151 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 33. Development needs / investment [Quality of expenditure]: There is a rule in the FSF that seeks to focus withdrawals on key development needs. Once the fund reaches 10% of GDP, the excess above that amount is to be invested in foreign markets and with the Development Bank of Mongolia for investments in railroads, oil processing, power stations, or high quality export goods and services. There are not clear rules in the FHF yet. 34. To ensure long-term sustainability, the FSL requires FSF to be greater than 5% of GDP with the excess above 10% of GDP to be invested with a longer-term perspective. However, this provision could create inconsistencies among rules if a gap between the actual and structural revenues is not high enough to meet the 5% of GDP floor. FIGURE 5.1 Revenues, expenditures and accumulation of the Fiscal Stabilization Fund 600 572 500 384 400 336 342 307 318 300 241 242 200 118 Expenditures 100 48 95 40 Accumulation 2 11 0 Revenue 2011 2012 2013 2014 2015 2016 2017 Source: The 2018 budget document, MOF 5.3. An International Comparison of Fiscal Management Institutions - Lessons for Mongolia 5.3.1. Independent fiscal institutions 35. In resource rich countries like Mongolia independent fiscal institutions can be very helpful to curtail large deficits caused by the meddling of politics as commodity cycles and large short-term spending pressures can lead to insufficient savings. Moreover, resource rich countries also struggle when it comes to the allocation between regions and inter-generational distribution of of the revenues coming from non-renewable resources. An independent fiscal institution can shed some light and provide technical advice on issues that are idiosyncratic to resource-rich countries. 36. In the last ten years, the number of independent fiscal institutions or agencies around the world have tripled. Although not a novel concept, as the first independent fiscal institution was established in Belgium in the 1930s, the Great Recession of the late 2000s has created momentum to revisit and bolster fiscal discipline as a mean to gain public trust. This has resulted in the proliferation of fiscal institutions. In the mid-90s less that 5 countries had active fiscal institutions, by 2017 there are 39 active ones with two-thirds established after 2007 and also two-thirds existing in Europe. 152 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT BOX 5.4 Independent Fiscal Institutions models Independent Fiscal Institutions or agencies are public institutions with a mandate to critically assess, and in some cases provide non-partisan advice on fiscal policy and performance. Traditionally, these have fallen under two main models: fiscal councils and parliamentary budget office. The nature, objectives, composition and degree of independence of these institutions vary significantly from country to country. Fiscal Council Model: Traditionally under the statutory authority of the executive or standalone with the following main functions: A. Non-partisan analysis, review and monitoring of government’s fiscal policies, plans and performance. B. Developing and/or reviewing macroeconomic projections C. Costing of budget policy proposals (including election proposals) D. Advising policymakers on policy options Parliamentary Budget Office Model: mostly designed to support the parliament and the budget committee.May also have a costing role. 37. The right design, degree of independence and composition of the independent fiscal institution tend to depend on the needs of the country. In particular, close attention is required to the incentive structures in place and whether the fiscal rules and other institutions that take an active role in the management of resources and the budget process are actually fulfilling their legal obligations. BOX 5.5 Fiscal Institutions in OECD countries Country Institutional Model Australia Parliamentary Budget Office Austria Fiscal Council Belgium Fiscal Council Canada Parliamentary Budget Office Denmark Fiscal Council Finland Other89 France Other Ireland Fiscal Council Italy Parliamentary Budget Office Korea Parliamentary Budget Office Mexico Parliamentary Budget Office Netherlands Fiscal Council Portugal Fiscal Council Slovak Republic Fiscal Council Spain Fiscal Council Sweden Fiscal Council United Kingdom Fiscal Council United States Parliamentary Budget Office Source: OECD Database 89 https://www.imf.org/~/media/Files/Publications/WP/2018/wp1868.ashx 153 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 38. In terms of effectiveness of fiscal councils, there is evidence that their presence is associated with reduction on optimistic budget forecasts and increase in their accuracy. Furthermore, partly on the back of their influence on the budget plans, the fiscal councils tend to also enhance compliance with budget-balance and expenditure rules.90 39. As mentioned earlier, as part of the Economic Recovery Program, Mongolia created a Fiscal Council but its independence raises a concern. Though the Fiscal Council is found to be lacking the required level of independence, being under the Parliamentary Budget Standing Committee, the country plans to embark on this agenda for establishing and operationalizing this independent institution. The IMF has recommended that a prerequisite for effective fiscal councils are those that possess solid Public Financial Management (PFM) systems and good statistical governance. In this regard, the PFM reforms that the government has been implemented during the last years coupled with the ongoing support that the World Bank and other development partners are contributing to the PFM agenda and will help establish an enabling environment for the establishment of an effective fiscal council. 40. Although effectiveness of a fiscal council is not always linked to a specific institutional set-up, the most successful fiscal councils foster transparency, promote a culture of macroeconomic stability and provide non-partisan advice regarding the budget process. Due to the role that politics has often played in the budget process, a fiscal council with a greater degree of independence should be considered. While more traditional institutions have relied on reputation and established practice, a more common feature of the new generation institutions has become strict legal independence (Figure 5.2). In parallel, operational independence and the ability to operate independently should also be considered. A larger proportion of new fiscal councils also have guaranteed legal access to information and benefit from safeguards on their budget. Availability of sufficient resources to the fiscal councils would help them develop a sufficiently persuasive and thorough analysis of the budget. FIGURE 5.2 Fiscal Council: Aspects of Legal and Operational Independence Traditional New generation Legal independence Access to information Safeguards on budget Own staff attend tasks 0 20 40 60 80 100 Source: : Independent Fiscal Councils: Recent Trends and Performance, IMF. March 2018 90 https://www.imf.org/~/media/Files/Publications/WP/2018/wp1868.ashx 154 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT 41. Similarly, the main objective and tasks of the fiscal council should be clearly defined, broad and operational. Given some of the specific challenges and biases that Mongolia has faced, when designing the fiscal council, the following tasks outlined in the box below could be considered. BOX 5.6 Likely tasks of a future fiscal council in Mongolia Problem/ Challenge Tasks of Fiscal Council Unrealistic revenue and expenditure forecasts Non-partisan macro-fiscal forecasts that have led to bid deficits. Contribute to legislative debate on fiscal Limited Technical Capacity of Parliament matters, report it parliament regularly and answer to queries and questions from MPs. Fiscal rules are not followed; policies that are Monitor compliance of fiscal rules and examine agreed to are not fulfilled fiscal outcomes of government’s commitments. 42. To complement the role of the fiscal council in monitoring compliance with fiscal rules and to help counter the political influence over budget spending, certain legal provisions have been found effective elsewhere. In Peru, for instance, there are statutory numerical restrictions on budget in election years. In particular, non-financial expenditure of the general government executed during the first seven months of the year cannot exceed 60% of the budgeted amount for the whole year and deficit of the combined public sector during the first semester of the fiscal year cannot exceed 50% of the budgeted deficit for the whole year.91 5.3.2. Transparency and Citizen Engagement 43. Internationally, emerging evidence indicates that countries that have undertaken budget transparency efforts as part of their overall PFM reforms are more likely to respond to people’s needs in ways that will improve their lives. Among its main benefits, budget transparency when paired with citizen engagement efforts has the potential to lead to greater fiscal credibility and performance (Alt and Lassen 2006), as well as cheaper international credit (Glennerster and Shin 2008), improved allocations of resources (Reinikka and Svensson 2004) which will translate into better provision of public services, in particular in social sectors (health, education etc.); while limiting corruption and wasteful spending that will have a positive impact in overall good governance outcomes. 44. Brazil is one of the leading countries in engaging citizens in public matters and in leveraging technology to allow them to participate in the decision-making process on how public resources are allocated and used. In 1988 Brazil pioneered the use of Participatory Budgeting in several cities of the country, a consultative mechanism that allowed citizens to participate and decide how and where to use part of the local budget. The new participatory tool allowed local governments to understand and respond to the most immediate needs of the citizens while bringing a transparency angle on the use of the local resources. The participatory budgeting experience in Brazil evolved and incorporated over time the use of internet voting (combined with off-line venues) to expand the geographical coverage and the number of participants in the process. In 2013, the state of Rio Grande do Sul run the referendum on state-level spending priorities called “State System for Citizen Participation in the Budget Process”. The participatory budgeting process allowed citizens to influence the allocation of public spending by involving them in 91 http://documents.worldbank.org/curated/en/225811468297875669/pdf/NonAsciiFileName0.pdf 155 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue the formulation of investment projects to be financed with public resources. The new participatory budgeting included also a monitoring feature that allowed Civil Society Organization (CSOs) to oversee the implementation of those processes. Turnout in the State participatory budgeting vote is around 15% of the eligible voters, reaching over one million participants on a yearly basis, making the Rio Grande do Sul one of the largest participatory budgeting process in the world in terms of number of participants. The use of ICT tools boosted the participation of youth and other segments of the population and more importantly exact a high level of institutional responsiveness (all citizens’ proposals originated by PB are incorporated in the state’s budget). Replicated in more than 1400 cities and states around the world the use of PB as one of the participatory government innovations have shown direct impact on reducing mortality rates, promoting health care spending and creating more vigilant CSOs that contribute to strengthen external public oversight (Touchton and Wampler, 2013; Gonçalves, 2014). 45. Brazil also pioneered the use of budget transparency portals when launched the first one in 2004. Since 2010 it became one the main channels through which citizens can monitor the financial implementation of government programs (how the public money is being used) in order to increase the transparency of public financial management in the country. The “portal da transparencia” (http://www. transparencia.gov.br): which is easily accessible to the public (that is, without requiring a username or password) contains technical budget and other financial on five broad categories 1) direct spending by federal government agencies through contracts and tender processes; 2) all financial transfers to states, municipalities and the federal district; 3) financial transfers to social program benefactors; 4) administrative spending, including staff salaries, staff travel expenses and per diems and office expenditures; and 5) information on all government official credit card spending. Important data sets, such as those on income and spending, are updated on a daily basis and data related to money transfers between different levels of government - is updated weekly. 46. In India, improved fiscal transparency prompted the publication of detailed information on the status of budget implementation. This made possible for researchers and think tanks to analyze that data and to provide the evidence needed to initiate a dialogue with the Government of India and to exact action from the latter to keep its promise to target resources to benefit the citizen, thereby strengthening the accountability for public service delivery.98 47. In Mexico, organizations such as Fundar, an independent Mexico City-based think-tank, analyzes fiscal data and produces detailed reports on specific policy areas and programs such as agriculture subsidies and other sectoral issues of public interest. The think tank has been also a major driver of increased fiscal openness and transparency in the country. Moreover, the work of Fundar has not been limited to use and disseminate budget information, but also to provide legislators with a steady supply of information and analysis on budgetary issues that inform national budget debates in Congress helping strengthen external accountability in the country. 5.3.3. Stabilization vs. Sovereign Wealth Funds 48. Traditionally, resource rich countries to set up and operate a stabilization fund. Cyclical resource related revenue is often accumulated in the stabilization fund. Money is transferred back to the treasury account when there is a shortage of revenue collection. The stabilization fund is a special fund in the budget and the Ministry of Finance manages the fund. Resource rich countries also create sovereign wealth funds to distribute wealth across generations. An example of this is the saving funds. 92 http://www.internationalbudget.org/wp-content/uploads/LP-case-study-NCDHR-one-page-summary.pdf 156 INSTITUTIONAL FRAMEWORK FOR IMPROVED FISCAL MANAGEMENT BOX 5.7 Stabilization Funds around the world  Russia: The stabilization fund uses non-resource primary balance is used to maintain the sta- bility of the budget. In case of revenue shortage, the Ministry of Finance determines how much resources it will use from the stabilization fund.  Mexico: The Oil Revenue Stabilization Fund targets structural primary balance and maintains a stabilization fund. The Ministry of Finance is responsible in determining how much out of this fund is used.  Chile: The law mandates the government to transfer proceeds for more than 0.2% of GDP to the Economic and Social Stabilization Fund. An amount equivalent to 0.2% of GDP is transferred to the pension fund (even when there is a deficit), the remainder is transferred to the stabilization fund. This fund is used to finance future budget deficits.  Norway: The country maintains the target of the structural deficit to no more than 4% of GDP as this is the expected rate of return of the country’s wealth fund. 49. As mentioned earlier, Mongolia has enacted legislation to create both a Fiscal Stability Fund as well as a savings fund called the Future Heritage Fund (FHF). Given the current context, however, given the situation where the country continued to run large fiscal deficits, international best practices suggest that the dangers of borrowing at a high interest rate and attempting to save simultaneously. Thus, currently this SWF arrangement may not seem suitable Mongolia: that is, saving while borrowing at higher interest rate. Accordingly, the new FHF is not expected to become functional until after 2020. 50. Sovereign Wealth Funds vary significantly on macro fiscal objectives, asset allocation and structure. They are designed based on country-specific needs and they can evolve as needs evolve over time. BOX 5.8 Types of SWFs  Stabilization Funds: Used to insulate the economy from commodity price volatility. Highly liquid portfolio assets. Chile (Economic Stabilization fund), and Russia (Oil Stabilization Fund) have set up these kinds of SWFs.  Saving Funds: Meant to share wealth across generations. Examples include the Abu Dhabi Investment Authority and Russia’s National Wealth Fund.  Pension Reserve Funds: Used to cover pension related contingent liabilities.  Development Funds: Used to devote money to development projects, usually infrastructure.  Reserve Investment Funds: Reduce negative carrying costs of holding reserves. 157 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue 51. Over the next few years, the main tasks to by undertaken by the Mongolian government are to establish a SWF entity according to the law, including staffing, training, setting up investment mandates, and other subsequent measures to complement the new FHF law. The government of Mongolia could benefit from exposure to international experiences from other resource rich countries like Chile that have managed both savings and stabilization funds. Technical assistance from the donor community to set up operating rules, internal procedures, training of personnel and others related to the establishment of and the operationalization of the SWF should will also benefit the government significantly and will enable the government to be ready to begin when the economic context is more favorable to make the FHF functional. 5.3.4. Concluding Remarks 52. Both the establishment of a well-functioning and independent fiscal council and of a stabilization fund will help promote fiscal discipline. As previously discussed, there are many potential institutional set-ups for these two entities, but in both cases the government should insure insulation from politics. Prior incarnations of SWFs that failed and attempts to circumvent fiscal rules in the past by policymakers make it imperative to create technical institutions that will be free of electoral or political calculations. 53. It is also important to note that both the SWF and the fiscal council will be complementary to the ongoing PFM reform efforts. These shall be implemented in tandem with other reforms including in tax policy and administration, civil service, PIM, sectoral spending and pension, transparency and accountability including citizen engagement, and other areas that have been discussed at great length in other chapters of this report. 158 REFERENCES REFERENCES Baer K., & Borgne E. L., 2007. “Tax Amnesties”. Working Paper, IMF. Belinga, V.; M. Melou, and J-P. Nganou, 2016. “Does Oil Revenue Crowd Out Other Tax Revenues? Policy Lessons for Uganda”. Policy Research Working Paper 8048. World Bank. Bird, R., 2010. Organizing to Tax. Ernst & Young, 2013. GAAR rising Mapping tax enforcement´s evolution. IMF, 2008. Debt Sustainability Analysis. IMF, 2008. Selected Issues and Statistical Appendix. IMF, 2011. Revenue Mobilization in Developing Countries. IMF, 2016. Public Investment Management Assessment (PIMA) for Mongolia. IMF, 2017. Debt Sustainability Analysis. IMF, 2017. Mongolia Article IV. International Budget Partnership. “India: Budget Tracking to Give Dalits a Fair Share of Development”. URL: http://www.internationalbudget.org/wp-content/uploads/LP-case-study-NCDHR-one-page-summary.pdf Lienert, I. & Fainboim, I., 2007. “Reforming Budget Systems Laws”, PFM Technical Note No. 2, IMF. Lustig, N., 2016. A distributional analysis of selected tax and transfers policies. Lustig, N., 2016. “Commitment to Equity Handbook. A guide to estimating the impact of fiscal policy on Inequality and Poverty”. Tulane University. Montsame Mongolian News Agency, 2017. URL: http://en.montsame.mn/economics/8273. OECD, 2016. Anti-Corruption Action Plan Fourth Round of Monitoring Mongolia. Progress Update. URL: https://www.oecd.org/corruption/acn/OECD-Mongolia-Progress-Update-Oct-2016-ENG.pdf Parliament of Mongolia, 2016. Mongolia’s Sustainable Development Vision 2030. Rodriguez, S. & J. Yang, 2018. “Mongolia: Distributional impact of taxes and transfers”. Background paper of the PER. Sunley, E.; J. Gottschalk, and A. Watson, 2010. “Mongolia: The Fiscal Regime for Mining – A Way Forward”. Policy Paper. IMF. The Asia Foundation, 2017. Survey on Perceptions and Knowledge of Corruption. URL: https://asiafoundation. org/2017/10/18/political-party-financing-mongolia-road-grand-corruption/ UN, 2015. The World Population Projections Revision. UNDP, 2018. “Towards a Professional and Citizen-centered Civil Service in Mongolia” Project Document. Verena, F. & and Sue H., 2017. “Mongolia: non-technical drivers of fiscal consolidation and potential implications for WBG engagement.” (not published) 159 GROWING WITHOUT UNDUE BORROWING Enhancing Efficiency of Spending and Revenue World Bank, 2009. Mongolia: Towards a High Performing Civil Service. World Bank, 2010. Risk-Based Tax Audits: Approaches and Country Experiences. WHO, 2011. Drug regulatory system assessment of Mongolia. World Bank, 2012. “Public Expenditure Review for Peru: Spending for Results”. URL: http://documents. worldbank.org/curated/en/225811468297875669/pdf/NonAsciiFileName0.pdf World Bank, 2015. Report on Rationalizing the Existing Public Investment Portfolio. World Bank, 2015. Public Financial Management Performance Report (PEFA). World Bank, 2016. Development and Introduction of a Methodology for Public Investment Portfolio Cleaning. World Bank, 2017. Africa’s Pulse. World Economic Forum, 2016-2017. Global Competitiveness Index. Zeldin, W., 2010. Mongolia: Constitutional Amendment Procedures to be Discussed by Parliament. URL: http://www.loc.gov/law/foreign-news/article/mongolia-constitutional-amendment-procedures-to-be- discussed-by-parliament/. 160 5th Floor, MCS Plaza Building, 4 Seoul Street, 14250 Ulaanbaatar, Mongolia Tel: +(976) 70078200 Web: www.worldbank.org/mn Twitter: WorldBankMgl