,Report No. 24439-MOG Mongolia Public Expenditure and Financial Management Review Bridging the Public Expenditure Management Gap June 2002 Poverty Reduction and Economic Management Sector Unit East Asia and Pacific Region Documneni of the World Bank 4' CURRENCY EQUIVALENTS (as of end January 2002) Currency unit = Togrog (Tg) US$1 =Tg 1100 FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric System ABBREVIATION AND ACRONYMS ADB - Asian Development Bank ALOS - Average Length of Stay BOM - Bank of Mongolia CERO - Central Employment Office CMEA Council for Mutual Economic Assistance CFAA - Country Financial Accountability Assessment CSR - Civil Service Reform EBFs - Extra Budgetary Funds ECA - Europe and Central Asia EF - Employment Fund FDI - Foreign Direct Investment FSU - Former Soviet Union FPS Fiscal Policy Statement GDP - Gross Domestic Product GGHS - Good Governance for Human Security GGS - General Government Services GOM - Government of Mongolia HIF - Health Insurance Fund IMF - International Monetary Fund I-PRSP - Interim Poverty Reduction Strategy Paper LGS - Law on Government Service LSMS - Living Standard Measurement Survey MCA - Ministry Chief Accountant MDG Millennium Development Goals MNE - Ministry of Nature and Environment MOFE - Ministry of Finance and Economics MOH - Ministry of Health MOSWL - Ministry of Social Welfare and Labor MPRP - Mongolian People's Revolutionary Party MPs - Members of Parliament MTFF - Medium Term Fiscal Framework NDC - Notional Defined Contribution NGOs - Non Governmental Organizations NSO - National Statistic Office PEFMR - Public Expenditure and Financial Management Review POPS - Performance Oriented Public Sector PPP - Purchasing power parity PRGF - Poverty Reduction and Growth Facility PSMFL - Public Sector Management and Finance Law QFAs - Quasi-Fiscal Activities SAIC - State Audit and Inspection Committee SAO - Social Assistance Office SFI - State Financial Inspection SIF - Social Insurance Fund SSIGO - State Social Insurance General Office TDB - Trade and Development Bank TSA - Treasury Single Account UB - Ulaanbaatar VAT - Value Added Tax WB - World Bank SOE - State Owned Enterprises MOH - Ministry of Health OECD - Organization of European ERC - Emergency Relief Commission IMF - International Monetary Fund SPC - State Property Committee UNDP - United Nations Development Organization GTZ - Deutsche GeseLLschaft fur Technische Zusammenarbeit GmbH Regional Vice President: Jemal-ud-din Kassum Country Director: Ian C. Porter Sector Director: Homi Kharas Task Team Leader: Vera Songwe TABLE OF CONTENTS EXECUTIVE SUMMARY ....................................... The First Challenge: Strengthening Public Finances ...................... The Second Challenge: Improving Public Sector Efficiency and Effectiveness.......viii Bridging the gap......................... ............... xv 1. INTRODUCTION ................................................................................................. 1 2. ASSESSING PAST PERFORMANCE ................................................................. 5 Overview of Consolidated Government Expenditures and Revenue .............. 5 Maintaining Sustainable Fiscal Balances for Prolonged Growth: ............. 16 Poverty and Vulnerability in Mongolia ............................... 22 3. CURRENT CHALLENGES: ADDRESSING THE MAIN SOURCES OF FISCAL PRESSURE ............................................................................................. 24 Introduction..................................... .............. 24 Civil Service Pay and Employment Policies ...................... 26 Public Pension System - Issues and Options........................ 37 State Owned Enterprises and Quasi-Fiscal Activities ................ ..... 42 Corporate Governance, Accounting and Auditing Standards ........... ..... 50 Financial Sector Reform ........................................ 54 Managing Contingent Liabilities and Natural Disasters ........ ............ 55 4. STRENGHTENING BUDGET CONTROL AND EXECUTION FOR SUSTAINED FISCAL DISCIPLINE ................................................................. 59 Introduction and Summary ...................................... Strengthening Budget Execution and Control ..................... ..... 60 Modernizing Treasury Financial Management and Information Systems............... 64 Government Accounting and Reporting ........................ ...... 66 5. INTERGOVERNMENTAL FISCAL RELATIONS AND SERVICE DELIVERY .......................................................................................................... 72 Introduction and Summary ......................................... 72 Strengthening the Institutional and Legal Framework........................ 76 Clarifying Expenditure Assignments ................................ 80 Assessing Local Revenue Capabilities ....................... ........ 87 Transfer System and Incentive Structure ................................ 94 Building Capacity for Effective Decentralization.................. ..... 100 6. IMPROVING GOVERNANCE ............................................................................ 102 Introduction and Summary .................................. .... 102 Legal Parliamentary and Social Controls. ........................ ..... 102 Recent Perceptions of Public Sector Effectiveness ........................... 103 Reform Options............................................... 106 ii 7. STRENGTHENING POLICY COORDINATION PROCESSES..................... 108 The Policy Making Process ......................................... 110 Budget Fragmentation.......... ................... .......... 112 Reform Options............................ ................ 114 A Prudent Approach to Budgeting- ..................................... 117 8. IDENTIFYING PRIORITY PRO-POOR POLICY ACTIONS ........................ 122 Analysis of the Functional Composition of Public Sector Spending........................ 123 Trends in Social Sector Spending. .................................. 124 Education Sector Outcome Indicators and Resource Allocation ....... ....... 125 Reform Options......................................... ..... 131 Health Sector Strategies and Outcomes ......................... ..... 132 Resource Allocation and the Budgetary Process .................. ...... 135 Social Assistance System................................. ....... 145 Allocation of Resources in the Environment Sector................. ...... 148 Conclusion ................................................. 152 9. SEQUENCING THE REFORM AGENDA.........................................................153 Key Principles of the Reform Approach - Improving the Basics .............1 55 Application to Mongolia........ ...................... .......... 159 Institutional and Capacity Constraints to Implementation................... 161 10. SUMMARY POLICY RECOMMENDATIONS ................................................. 163 Boxes: Box 1.1: Public Sector Reform in Mongolia: Main Tenets of the PSMFL . ........... 3 Box 3.1: Civil Service Reform and Service Delivery.......... ..... ......... 27 Box 4.2: The Role of the Bank of Mongolia in Facilitating the TSA...... ....... 66 Box 4.3: Expenditure Process, Internal Financial Controls and Bank Reconciliation ..... 67 Box 4.4: Definition of Accounting Systems .............................. 67 Box 5.2: What Does Good Local Tax or Fee Look Like?............. ........93 Box 5.3: Simple Intergovernmental Finance Formulas ................. ..... 99 Box 7.2: Key Characteristics of a Functioning MTEF ................. ...... 116 Box 7.3: Prudent Budgeting Practices- Experience from the Netherlands .... ..... 118 Box 8.1: Proposed Education Sector Reforms Needed to Achieve 2015 Objectives..... 131 Box 9. 1; Risks to the Reform Program. ............................ ..... 154 Box 9.2: Key Assumptions of the Reform Program ........................ 155 Figures: Figure 2-1: Real GDP per Capita ($US) 1990-2000.............................. 5 Figure 2-2: Growth Rate in Selected Transition Economies (percent) .... .......... 5 Figure 2-3: Trends in Expenditure as a Share of Total 1993-1996 and 1997-2000 ..... 6 Figure 2-4: Exp. on Utilities and Non-Wage Costs as a Share of total 1993-2000..... 7 Figure 2-5: Mongolia Tax Composition 1993-2000 ..................... 8 Figure 2-6: Changing Tax Structure between 1993-1996 and 1997-2000 . ........... 9 iii Figure 2-7: Share of Distortionary and non-Distortionary taxes in Total Taxes........... 10 Figure 2-8: Real GDP growth 1993-2000 (percent) ............................... 11 Figure 2-9: Revenue and expenditure 1995-200 1........................... 11 Figure 2-10: Overall fiscal Deficits Exceeded 1995 Levels Through 2001 ............. 11 Figure 2-11: High Deficits Increase Domestic Borrowing and Interest Payments........... 11 Figure 3-1: Total Government Exp. as a Share of GDP 1993-1998 in Transition Economies ....................................... .......... 25 Figure 3-2: Government Wage Bill as a Share of GDP in Transition Economies ..... 28 Figure 3-3: Wages and Salaries as a Share of GDP......................... 28 Figure 3-4:Wages and Salaries as a Share of Total Expenditure 1993-2000 ............... 28 Figure 3-5: Wage bill as a Share of Current Revenue 1993-2000 ........... ..... 32 Figure 3-6: Growth in Current revenue and Wage Bill 1994-2000 .. .............. 32 Figure 3-7: Sectoral Breakdown of SOE Sales in 2000......................... 45 Figure 3-8: SOE Revenue is Concentrated in 2 Regions ....... ............... 45 Figure 3-9: Almost Half of all SOEs are Unprofitable (percent).................. 45 Figure 3-10: Both Profits and Losses are Increasing (percent): ................. 45 Figure 3-11: Half of all SOEs Pay No Tax ................ ...... ......... 47 Figure 3-12: And Just 5 Firms Pay 93 Percnet of all Taxes ............. ....... 47 Figure 3-13: Major Taxpayers Have Paid an Increasingly Larger Share ..................... 47 Figure 4-1: Local and Central Government Arrears as a Share of GDP 1999-2001........ 61 Figure 4-2: Financial Management Systems - MOFE to the Bank of Mongolia.......... 63 Figure 4-3: Status of Mongolia's Information System .................. ..... 65 Figure 5-1: Share of Local Gov. Revenue and Exp. in GDP 1996-2001 (percent).....76 Figure 5-2: Share of Local Gov Revenue and Exp. in Total, 1996-2001 (percent).......... 76 Figure 5-3: Schematic Exposition of Intergovernmental Expenditure Responsibilities... 82 Figure 5-4: Assignment of Roles and Responsibilities by Sector................ 83 Figure 5-5: Share of Local Gov. Revenue and Expenditure in GDP 1996................... 88 Figure 5-6: Share of Grants and Own Revenue in Total Local Government Revenue .... 91 Figure 6-1: Corruption Indicators for Selected East Asian Economies, 2001................ 103 Figure 6-2: Comparing the Six Indicators of Governance between 1998 and 2001....... 104 Figure 6-3: Comparing Governance Indicators Across the Region........... ........ 104 Figure 6-4: Comparing Government Effectiveness .................... ..... 104 Figure 8-1: Expenditure Categories as a Percent of GDP...................... 123 Figure 8-2: Comparative Expenditures as percent of GDP 1996-2001 ................. 123 Figure 8-3: Education Spending as a Percent of GNP, 1985-2000................................. 124 Figure 8-4: Public Health Spending as a Percent of GDP, 1990 -2000.......................... 124 Figure 8-5: Enrollment Rates Dropped Early in the Transition but are Recovering ...... 126 Figure 8-6: Enrollment Rates Increased Across all Regions ..................... 126 Figure 8-7: Basic Education (1-8) Net Enrollment by Gender and Regions in 2000..... 126 Figure 8-8: Education Sector Expenditure Composition 1996- 2000.............. 128 Figure 8-9: Share of Regions in Total Education Expenditure, 2000..... .......... 128 Figure 8-10: Av ratio of Cost per Secondary School Student to GDP per Capita by Regions, 2000 ................................................... 128 Figure 8-11: Composition of Spending on Education by Source, 1995-99.................... 129 Figure 8-12: Spending on Education by Type, 1996-99 ................ ..... 129 iv Figure 8-13: Infant mortality and Under-Five Mortality, Maternal Mortality Ratios and GDP per Capita in Low-Income Countries, 1999......... ............... 133 Figure 8-14: Public Spending on Health In real Terms (tg), 1990-2000 ...... .....1 35 Figure 8-15: Infant Mortality Rates and Budgetary Support for Health to Aimags, 2000137 Figure 8-16: Health Spending by Institutions, 1996-2000...................... 139 Figure 8-17: Spending on Other Medical Institutions as a Share of Total Health Exp. 1996-2000 .......................................... ...... 140 Figure 8-18: Share of Physicians in Total by Regions ................................ 141 Figure 8-19: Environmental Expenditure By Type (MI of Tugrog) ....... ....... 150 Figure 8-20: NRM as percent of Core Environmental Expenditures ................... 150 Figure 8-21: Expenditures for Natural resources Management ................. 150 Figure 8-22: Core Env. Exp. in the Budget (Mi. of Tg, ) Compared to Other Budget Items....................... ...................... ...... 150 TABLES: Table 2-1: Share of Tax Revenues in Total Revenues 1993-2000 ....... .......... 9 Table 2-2: Trends in Overall Fiscal Balances: 1993-2000 .......................... 13 Table 2-3: Medium Term Macroeconomic Framework ................. ..... 19 Table 2-4: Selected Economic Indicators .................................... 21 Table 2-5: Mongolia: Poverty Indicators: 1995-98 .................... 22 Table 2-6: Employment Status of Working Age Population by Educational Level,........ 23 Table 3-1: Total Number of Civil Servants, by Assignments, (000s) . ..... ........ 29 Table 3-2: Share of Public Sector in Economically Active Population .. ............ 29 Table 3-3: Mongolian Civil Service Wage Compression Ratio: 1997-2001................ 30 Table 3-4: Expenditure on Social Welfare Programs .................... 37 Table 3-5: Share of Total Cash Transfer Expenditure By Type of Payment: 1985-2001 37 Table 3-6: Social Insurance Pension System Indicators (percent)........... ..... 38 Table 3-7: Average Benefit as a Share of Average National Poverty Line, 2000........ 39 Table 3-8: Contributions - Expenditures as Percent of GDP ................... 40 Table 3-9: SOE Financial Profile According to Type of Ownership in 2000 .............. 44 Table 3-10: Before Tax Income Total for Profitable SOEs (Tg. 000s) ..... ....... 46 Table 3-12: Major Firms With Negative Gross Margins in 2000........... ..... 48 Table 3-13: Accounts Receivable Turnover for SOEs with Largest Receivables........ 49 Table 4-1: Evolution of Local and Central Government Arrears 1999-2001................ 62 Table 5-1: Demographic and Administrative Structures in Mongolia 1996-2000 .....74 Table 5-3: Aimag Expenditure as a Share of GDP and Local Government Expenditure. 81 Table 5-4: Local Government Shares in Education and Health Expenditures Share of Total ..................................................... 81 Table 5-5: Share of Local Government Revenue and Expenditures 1996-2001........... 88 Table 5-6: Shares of Taxes Allocated to Local Governments 1996-2000. (percent) ....... 90 Table 5-7: Share of Revenue per Capita in Total Expenditure per Capita in Local Governments, 1996-2001 .......................................... 92 Table 5-8: Share of Central Government Transfers in Total Local Government Revenue 1999-2001 ................................................. 95 Table 5-9: Share of Transfers in total Expenditure per Capita, 1996-2001 (percent) ...... 96 V Table 5-10: Aimag Expenditure as a Share of Local Government Exp. 1996-2000 ........ 96 Table 8-1: Ratios of Actual Average Salaries to GDP per Capita in 2000..................... 130 Table 8-2: Comparative Health Indicators for Mongolia, 1999 ........ ......... 133 Table 8-3: Differences Between Proposed and Actual Public Spending on Health at the Central and Local Levels, 1996-2000 ......................... .....136 Table 8-4: Provider Payment Mechanisms in the Health Sector ........... .....1 38 Table 8-5: Composition of Recurrent Health Expenditures (%), 1990-1997................. 140 Table 8-6: Environmental Expenditures as a Share of GDP........... ........ 150 ACKNOWLEDGEMENTS This report was prepared by a team led by Vera Songwe and comprising Malcolm Holmes (public sector management): Louise Fox (pensions and social assistance benefits): SerifSayin, and Peter Dean (budget execution and financial management); Donavan Rudisuhle (state owned enterprise reform); Bold Magvan and Xiaofeng Hua (financial sector); Christopher Thomas, Daniel Dulitzky, Mota Mayer, Gita Steiner-Khamsi-Aun. -T and Peter Russell Moock, (education sector); Rekha Menon and Son Nam Nguyen (health), Salvador Rivera, (energy); Giovanna Dore and Anthony Whitten (environment); Thilakaratna Ranaweera and Bazarvaani Darinchuluun (macro and debt projections); Tsolmon Bat-Ochir and Bold Magvan provided overall data and research assistance. Ronald Points, and Nancy Chen, oversaw integration of Country Financial Assessment and Accountability into the Public Expenditure and Financial Management Review. This document is an integrated CFAA and PER document. David Shand provided comments on the content and integration of the report. The peer reviewers for the PEFMR are Messrs./Mmes. Jeffrey Hammer, Graham Scott, Allen Schick, Debbie Wetzel, Wyaya Wikrema, Robert O'Leary. Paul Bermingham and Allen Schick provided guidance and suggestions on the structure of the report through the Quality Enhancement Review process facilitated by Vera Wilhiem. Comments from Vinaya Swaroop, lead advisor helped improve the process and content of the report. Barbara Nunberg, Amitabha Mukherjee, David Shand, and Zafar Ahmed provided useful comments. David Howarth provided inputs from the draft Country Procurement Assessment Review. Mario Reyes Vidal and John Richardson also provided comments and inputs Joost Pollack provided editorial support, and Nancy Mensah and Hedwig Abbey provided administrative and logistical support. The team would like to acknowledge substantive contributions and insights of the International Monetary Fund and the Asian Development Bank in the preparation of this report. This report was produced in a fully participatory manner with government. The team gratefully acknowledges the Government team particularly staff in the Ministry of Finance and Economy, Ministry of Education, Ministry of Health, the Cabinet Secretariat for their substantive contributions to the analytics, and workshops which helped refine the recommendations contained in this report. In particular the team would like to thank the Minister of Finance and Economy for his leadership and guidance during the process. The report was prepared under the overall guidance of Messrs./Mme. Homi Kharas (Chief Economist and Sector Director, EASPR), Ian Porter (Country Director, EACSM), and Tamar Manuelyan Atinc (Sector Manager and Lead Specialist, EASPR). EXECUTIVE SUMMARY 1. Mongolia made impressive progress in macroeconomic adjustment in the early transition years. By 1995, however, tensions arose in managing the balance between short-term macro imperatives and medium- to longer-term development challenges. Mongolia-and many other transition economies -did not pay adequate attention to the institutional underpinnings needed to combine growth with equity. This manifested itself in limited attention to the institutional arrangements necessary to put fiscal policy on a sustainable footing, support efficient and effective service delivery, and discipline the policy choices of government. 2. During the second half of the 1990s many of the gains on the fiscal side and impressive social indicators from the pre-transition period came under pressure. GDP growth averages dropped from 3.5 percent over 1993-96 to 2.4 percent between 1997 and 2001. There was a general decline in public sector fiscal and policy discipline, illustrated by a growing tendency to increase government employment and cash transfers (which together totaled over 40 percent of total expenditure by 2000), and to support ineffective state-owned enterprises (SOEs). Policy indiscipline led to a buildup of public sector and inter- enterprise arrears, and increased fragmentation of the budget and policymaking processes. Negative external conditions, including bad weather, depressed growth in the later part of the decade, further accentuating inadequacies in public expenditure management. Reasserting discipline over the budget and the civil service and improving basic financial management practices have to be of the highest priority in Mongolia. One of the keys to this is resolving the relationship between the different levels of government 3. To achieve and sustain 6 percent annual growth rates and to translate this growth into poverty reduction will require a dynamic private sector and a more disciplined and performance-oriented public sector. The public sector's contribution to growth and poverty reduction will depend on the success with which it tackles two interdependent challenges: First, it must strengthen public finances-to ensure that deficits are sustainable and the poor are protected from further burdens of fiscal austerity-by improving policy and funding discipline and basic internal financial management practices. Second, it must improve the efficiency and effectiveness of the public sector to raise the quality of services, especially for lower income groups, by designing incentive structures that lead to a more performance-oriented public sector. Progress on these two fronts is required if Mongolia is to discipline the allocation of public spending across, and particularly within, sectors to meet the Millennium Development Goals (MDG) by 2015. 4. To address growing inadequacies in public sector management, the government proposed a Public Sector Financial Management Law (PSFML) in 1997 as a major instrument for reasserting budget discipline. The law has been stalled in parliament for almost five years, however, largely because of objections to the budget decentralization it called for. Local autonomy has great political importance in Mongolia, and successive drafts of the PSFML have not yet succeeded in reaching an acceptable compromise. A new version of the law was put forward for debate in the spring of 2002. This version of the draft law is a potentially a vastly better law than the current Budget Law and related laws and its potential for improving public sector performance would be greatly increased if critical issues noted and broad principles proposed in this report are addressed. 5. This report, the Public Expenditure and Financial Management Review (PEFMR), contains a number of recommendations directed at both the substance and process of reform. It also argues that in this period of substantial change and heightened debate about the role and structure of the public sector it is useful to have a set of core principles against which reforms should be evaluated. The most important principles are: * All reforms should increase the discipline and comprehensiveness of public sector management and budgeting. ii * Proposed policy changes should be evaluated in terms of their impacts, effectiveness, efficiency and costs in the short, medium, and long terms, especially for the poor; * Every assignment of a task to a governmental unit-national or local-should clearly identify and assign the resources needed, and the authority and accountability that will accompany those resources. THE FIRST CHALLENGE: STRENGTHENING PUBLIC FINANCES Vulnerability to shocks and lagging structural reform undermined Mongolia's strong early adjustment to the transition 6. Mongolia's macroeconomic transition to a market economy was resolute and quite successful. Most transition economies needed about a decade to realize the first benefits of transition; Mongolia's real GDP growth turned positive at 2.1 percent in 1994-and peaked at 6.3 percent-in 1995. The overall deficit dropped from 14.6 percent of GDP in 1993 to 5.9 percent in 1995. Recognized as one of the "rapid reformers" in transition economies, Mongolia moved quickly in 1992-93 to liberalize prices, lower trade barriers, and undertake tight fiscal and monetary policy to ensure stabilization of the economy. These macro-stabilization policies reaped the expected benefits: growth resumed and inflation dropped from 268 percent in 1993 to 57 percent in 1995. The size of the public sector dropped from 51.8 percent of GDP in 1993 to 31.5 percent in 1995. 7. But Mongolia failed to lock in its early successes by adopting the structural and institutional reforms needed to support a truly performance-oriented public sector and create an enabling environment for private sector growth. The economic crises in the second half of the 1990s exposed these weaknesses. During the 1993-96 period of macroeconomic stability decisions on the reach of the public sector were deferred. The size and reach of the public sector continued to mirror pre-transition command economy structures, while the discipline of the command economy was eroded by continuous institutional changes. 8. Over 1996-99 falling commodity prices, particularly for copper, and successive economic shocks, including the East Asian and Russian crises, severely depressed Mongolia's fiscal position and endangered the transition. In 1996-97 export prices for Mongolia's main commodities fell sharply: gold prices fell 42.5 percent, copper by 28 percent, and cashmere by about 12.5 percent. Large revenue shortfalls resulted as corporate income tax contributions and dividend payments from major SOEs dried up. 9. To maintain public confidence and protect the vulnerable from further income erosion, the government increased public employment and cash transfer programs and bailed out failing public enterprises. Government expenditures increased from 27.3 of GDP in 1996 to 39.4 percent in 1999. With revenue rising from 24.8 percent to 34.2 percent of GDP over this same period, the deficit, which had dropped to 5.9 percent of GDP in 1995, increased to 12.2 percent in 1999. The debt-to-GDP ratio more than doubled from 41.1 percent in 1995 to 93.9 percent in 1999, with interest payments rising from 0.4 to 1.9 percent of GDP over the same period. High tax rates needed to support increasing government consumption are a disincentive to private investment and reduce labor force participation. High tax rates are a disincentive to private investment and reduce labor force participation 10. To support excessive levels of consumption in the face of falling GDP growth and external resources the government resorted to increasing tax rates. Tax revenues increased from 24.8 percent of GDP in 1996 to 38 percent of GDP in 2001, and are higher than in most other transition economies-and still increasing. While efforts to increase revenue mobilization have been successful, they are proving detrimental to private sector growth. iii 11. Mongolia has modernized its tax structure, moving from a Soviet-type scheme to a more market- oriented system increasingly based on non-distortionary taxes. In 1998 a VAT was introduced, the corporate income tax was based on a net income concept, the base of the personal income tax was bolstered, and a progressive rate schedule introduced. 12. However as expenditures increased the government raised taxes-on income, profits, social security, and payrolls-to levels that are in themselves distortionary. Personal income taxes were raised to 29 percent, and customs duties to 40 percent, in 2001. Overall corporate income tax is high compared to other East Asian countries at 40 percent, and the VAT has increased from 10 to 13 to 15 percent since 1998, well above the East Asian average of about 10 percent. High personal income tax rates deter savings and undermine investments, while high corporate tax rates stifle private sector growth. In sum, Mongolia's tax rate and structure are out of line with international norms for developing countries, and tax levels are creating a strong disincentive for private sector investment. 13. The government will have to reduce its spending before it can cut taxes. Continuous improvements in tax administration could allow for further reductions in tax rates. While Mongolia has made impressive progress in improving tax administration more is needed. 14. To consolidate gains from previous tax reforms and create room for eventual reductions in rates the government will need to focus on creating incentives that improve tax administration. It is difficult to create an efficient tax administration without an educated and trained staff equipped with sufficient resources. Adequate funds will need to be budgeted to pay good wages to tax officials and to computerize tax operations. Other tax issues that need to be addressed include continuous changes in the rules and centers of administrative oversight. 15. The external shocks of recent years and a failure to deepen structural reforms have prevented Mongolia from fully benefiting from the transition and threaten long-run stability. Government employment programs have left it with an expensive and poorly performing civil service that needs careful restructuring. Generous cash transfer programs burden the budget, largely unrecorded SOE liabilities and local government arrears could threaten its future fiscal position. A weak government- dominated financial sector that has already led to two major banking crisis's must be reformed. The country is prone to natural disasters, but has no coordinated budgeting or accounting system to handle them. 16. Central to all these problems is the absence of any comprehensive medium- to long-term strategy on the general policy directions of the government and the appropriate role and reach of government. This impedes the government's efforts to build an effective and disciplined performance-oriented public sector, improve predictability of funding flows, and manage overall balances. For hard budget constraints to bind, systems and processes must be in place to check and reward compliance or enforce controls. Balancing short-term imperatives and long-term development priorities Restructuring of the Civil Service is needed to maintain fiscal discipline and improve performance 17. Mongolia's wage bill is high compared to other transition economies. It increased from 5.9 percent of GDP in 1996 to 8.2 percent-20 percent of government expenditures- in 2000. While wage levels were low in the mid-1990s, increases totaling more than 45 percent in the last three years have brought average public sector wages to 144 percent of per capita GDP, a level comparable to other low- income countries. Further across the board increases in wages would put upward pressure on- private sector wages, reducing the competitiveness of the private sector and stalling growth. iv 18. The Mongolian civil service's main problem, however, is not wage levels, but its functional composition, the lack of a tenured professional civil service, and a general lack of incentives to perform. Frequent and poorly targeted wage increases have compressed wage ratios within the civil service from 3.1 in 1997 to 2.7 in 2001. The bonus schemes are not entirely merit based. Political patronage has eroded the civil service base, and turnover in election years exceeds 9 percent. In 2000 political pressures led to a 2.7 percent increase in public sector employment, the first since 1996. These increases, which placed substantial pressure on the budget, largely involved low-skilled workers and political appointees. 19. The government is facing considerable pressure to live up to a campaign promise to double all public sector wages by 2004 and increase pension levels and coverage. It is considering drastic and potentially damaging policies to make good on that promise, including a proposal to halve the civil service in order to be able to double the wages of remaining staff. 20. The government's civil service reform proposal also includes rapid privatization of several social sector programs. However, unless privatization is part of a well thought out reform program, it will likely amount to pushing employment and other social expenditures off budget. Real expenditure savings will accrue only if it is part of a comprehensive restructuring program. There is little real fiscal or efficiency gain to be had by redefining institutions as public-or quasi-public--enterprises that receive un- monitored implicit subsidies from the budget. Privatization of social sector services in particular or other forms of public sector service will lead to long run efficiencies only if these institutions have the authority and responsibility to determine expenditure priorities based on a realistic set of commonly agreed outcomes. In the short run however, both the likely impact on social sector outcomes and the expected savings from social sector privatization will need to be analyzed carefully. In addition any new measures need to be consistent with the principle that social sector reforms should be part of comprehensive sector strategies and not limited to short-term expenditure saving measures. 21. Hasty and random civil service retrenchment is likely to reduce the overall quality of the civil service and undermine service delivery. Instead, a pay and employment modeling exercise based on reliable data on public sector pay, pensions, and employment to assess the fiscal impact of various reform scenarios is needed. This should lead to a phased and transparent plan to right-size employment and reform compensations consistent with fiscal realities. In parallel to the modeling exercise, a change- management strategy that addresses the need for engagement, commitment and involvement of all key stakeholders should be developed very early on in order to de-politicize the process. Identification of "champions of reform" and a clear understanding of the sources of resistance to change is needed. 22. The government could make use of a wide range of short-term pay and employment measures while its longer-term civil service reform strategy is being developed. These increases should not substantially increase the wage or pension base. These include wage and hiring freezes, outsourcing non- core government functions, enforcing hiring bonus thresholds, and public sector disengagement from commercial activities. Targeted wage bonuses, and wage increases differentiated by sector or profession may also prove useful short-term tools. While some of these are temporary in nature and effect, they would give the government breathing room to discuss and build consensus for long-term structural civil service reform. Generous cash transfer programs burden the budget. 23. Social welfare programs are another source of pressure on the budget. Expenditures on cash transfer programs increased to 7.3 percent of GDP (17.7 percent of total expenditure) in 2000 from 4.4 percent in 1996. Pensions accounted for the bulk of the increase, rising from 3.8 to 5.6 percent of GDP over 1996-2000, well above the East Asia regional average. Mongolia introduced social safety net programs and institutions in the early 1990s as part of a broad range of programs developed to cope with V new social needs, including a full social insurance system of pensions, short-term income replacement benefits, and services for the unemployed. 24. The pension system encourages early retirement from the workforce, as eligibility requirements are lax. Women can receive benefits at 55 if they have 20 years of service, or 50 if they have raised more than 4 children to the age of 6. Men and women can also retire early if they work in special occupations. Life expectancy at normal retirement age is 14 years for men, 20 years for women. The number of contributors to the pension system is low and declining. In 2000 only 545,585 or 68 percent of potential contributors were reported contributing to the pension system. This compares with roughly 800,000 people known to be employed under the National Statistics Office's narrow definition of employment, which excludes most of the informal sector. 25. Short- and medium-term initiatives are required to deal with the rising cost of pensions. One basic reform recommended to reduce the cost of the system in the short run and improve equity is to immediately begin raising the retirement age for all in the labor force. In the medium term the government will have to address more structural issues that could undermine short-term efforts at improving overall fiscal balances, such as improving contributions, and improving governance structure of pension fund. The government will need to revise and restructure the pension system, develop a transition policy to put all currently working into a new system, and reform disability and survivor pensions. ... as could the liabilities of SOE's 26. Lack of budget comprehensiveness and weak accountability make it difficult to quantify the fiscal impact of SOE performance on the budget, although total liabilities of the four largest SOE's (the Erdenet copper mine, Baganuur Coal Mine, and Power Stations Number 2 and 3) in 2000 accounted for over 19 percent of GDP. The current government inherited a legacy of banking, trade, energy, mining, and transport SOEs. This portfolio merits scrutiny, as there is no systematic process for identifying and recording government liabilities arising from SOE activities. These liabilities need to be recognized, recorded and processes put in place to contain them. Government investments in many areas continue to be channeled through SOEs, most of which still benefit from implicit subsidies and capital transfers. The government continues to tolerate arrears buildups and honors SOE foreign debt obligations. In several recent instances, the government has been called on to honor guarantees made to third parties on behalf of Mongolian enterprises. 27. Although privatization is underway, the withdrawal of the state from commercial activities or the restructuring of public utilities has not been rapid enough and remains an important source of undue pressure on the budget. Failure of SOEs to honor foreign obligations could put substantial pressures on public sector expenditures. These contingent liabilities and the associated exchange risks could severely damage long-run fiscal balances and compromise policy. 28. In the short run a consolidated review of the financial position of all SOEs is recommended, followed by systematic audits of all accounts. This will enable integration of contingent liabilities into the budget. In the medium term, a roadmap for agency restructuring or privatization needs to be developed. Financial sector reforms need to be accelerated to reduce risks of banking sector crisis 29. Lackluster implementation of financial sector restructuring and modernization continue to affect the size and reach of Mongolia's public sector. Continuous banking sector crises stemming from poor management, lack of capacity, and undue public sector intervention remain a threat to macroeconomic stability. A banking sector crisis in 1996 and another in 1998, which required increased expenditure equal VI to over 3 percent of GDP, was a particularly telling example of the pressures weak financial sector management can place on the government. 30. Until, 2001 four state-owned banks held 77 percent of banking sector assets. Continued poor management, directed credit practices, and lack of compliance with international prudential norms could lead to another banking crisis. However, the sale of the Trade and Development Bank has significantly reduced government involvement in the financial sector. This is marks an important change in the banking sector infrastructure and if well managed will increase public sector confidence in the financial sector. 31. Ongoing programs to strengthen the financial sector, such as defining the role and functions of the central bank, and strengthening prudential requirements need to be accelerated. Central bank guarantees to SOEs under instruction from the government should be discouraged. Vulnerability to natural disasters calls for prudent fiscal reforms in good times 32. Mongolia's vulnerability to natural disasters poses another risk to government finances. In recent years, Mongolia has been struck by a series of calamitous events that have had a devastating effect on its citizens. The last three years have seen summer droughts and exceptionally harsh winters that have resulted in the death of millions of livestock. This has increased poverty in rural communities and forced the government to provide resources to relieve people's suffering. The 2001 snowstorm claimed over 10 percent of Mongolia's livestock and the government's estimate of monetary damages amounted to nearly 1.8 percent of GDP. 33. While the occurrence of dzuds-a combination of dry summers and severe winters-is impossible to predict, the government can take a number of measures to improve financial planning and budgeting to help address shocks. Providing realistic budget reserves for dealing with the cost of dzuds requires, in the first place, the ability to quantify those costs. Under the present system, this is next to impossible, as dzud-related expenses are charged against the normal operating budgets of at least seven budgetary entities and many SOEs. In addition, the accounting system does not take into consideration the interagency receivables that represent goods and services provided by one organization to another, as this is done in the form of barter trade. Poor accounting undermines agency balance sheets and leads to further build up of arrears. Local government arrears threaten fiscal balances 34. Accumulation of arrears, particularly by local governments poses a threat to overall fiscal balances and contributes to the general sense of indiscipline. Arrears peaked in 1999 at 2.4 percent of GDP, with local government arrears accounting for over 70 percent of the total. Local government arrears are consistently higher than those of the central government. The magnitude of local government arrears reflects in part an imbalance between mandated expenditures at local levels and the capacity of local governments to finance such expenditures. Arrears in social security, heating, and drugs are the highest. It also reflects the squeezing of deficits through the system-from the central government to local governments, and onward to companies, wage earners, and other parts of the economy. This is particularly damaging for poor pensioners, companies, as it undermines investments and proper management. 35. Policies to improve overall fiscal balances will need to address the systemic imbalances between revenues and expenditures that lead to arrears buildups and strengthen incentives for maintaining a hard budget constraint. Although the total stock of arrears is beginning to decline, there is no systematic recording of them, and local governments need to improve financial management controls. They need vii more effective information and monitoring systems and reporting mechanisms to improve their fiscal discipline and ensure that arrears are diminished in the near term. Mongolia must restructure its public finances 36. Mongolia must restructure its public finance patterns to restore aggregate fiscal balances and ensure that deficits are sustainable. Mongolia must bring its deficit levels down to below 5 percent of GDP and severely limit domestic borrowing if it is to achieve sustainable overall fiscal balances in the medium term. Managing fiscal policy to stabilize and reduce the public debt will be critical to reducing poverty, as lower debt burdens will enable the government to channel funds into poverty-reducing social programs and productivity-enhancing investments without reintroducing inflationary pressures. While the deficit fell to 5.3 percent of GDP in 2001, bringing Mongolia close to this objective, there is a need to sustain these gains by improving budgetary control and improving comprehensiveness of the budget. 37. The need for a thorough review of policy and expenditure programs is reinforced by the fact that over half of Mongolia's public expenditure in 2000 was for wages, cash transfer programs, and indirect income subsidies. Restoring fiscal balances while at the same time contributing more effectively to poverty reduction will require the government to pay early attention to reforming the civil service and cash transfer systems, and improving budget comprehensiveness to account for extra-budgetary expenditures. A prudent approach to budget formulation and improved financial management systems would strengthen fiscal balances 38. Fiscal policy in Mongolia faces a variety of uncertainties and risks led by external economic shocks and natural disasters. The overall fiduciary environment is opaque. In the face of such uncertainties it would be prudent to err on the side of caution in determining the aggregate resource constraint that must underlie budget planning. A conservative approach to determining the resource envelope will contribute to fiscal stability and a more predictable environment for the public sector. 39. The uncertainty that Mongolia faces, together with its ambitious program for dddressing poverty, requires that it take a more medium-term perspective to budget and policy making. Mongolia needs to continue developing its rolling medium-term fiscal framework (MTFF) based on macro projections suitable for fiscal stability, and to link this more effectively with the enforcement of budget control and budget execution. Countries have used a variety of frameworks-involving differing degrees of discretion-to guide medium-term budget policy. All such successful frameworks include some form of top-down assessment of resources available for public expenditure and clear strategic policy priorities, bottom-up costing of expenditure policies, and political mechanisms for reconciling these costs with available resources and priorities. Mongolia is developing processes and systems that move it in this direction. It is particularly important to ensure that these do not become paper exercises but bite on decision making at the center, sector and agency levels and in aimags and soums. In Mongolia's current phase of transition, particular attention needs to be given to making the reconciliation of demands and availabilities more policy based. 40. The absence of timely and comprehensive financial information impedes the government's efforts to build an effective financial management structure, improve predictability of funding flows and manage overall balances. For hard bi4dget constraints to bind, systems and processes must be in place to check and reward compliance or enforce controls. Better information management can be an early warning that identifies developing problems-such as arrears buildups-before they reach critical proportions. Decision makers also need reliable financial information to assess the implications of alternative policy decisions such as wage increases to improve budget discipline. viii 41. Mongolia's financial; information management, governance, and budget execution systems are plagued by problems. Budget data is unconsolidated and extra-budgetary accounts have proliferated. MOFE micromanages in excessive detail in an attempt to keep abreast of government functions that often overlap, or duplicate each other. Auditing and accounting procedures are poor and there is weak overall capacity at the center and in line ministries. These problems are compounded by a huge lack of trust between the center agencies and the line ministries, proliferation of agency bank accounts, diverse accounting practices and a shortage of information to inform the legislature and citizens on the effectiveness of service delivery. 42. The installation of a treasury, system is the most important systemic change needed. A modern treasury can give the government tools to implement its macroeconomic objectives. It can serve as center for government accounting and produce comprehensive, timely, reliable uniform financial and fiscal data. A fully functional treasury would also assist the government in enforcing the budget law and strengthening financial discipline, and serve as a basis for integrated government financial asset and liability management. 43. An effective treasury system would also improve internal accounting controls, streamline disbursement procedures, permit better management of receivables and payables, improve cash management, produce more reliable accounting data; and improve the timeliness and usability of financial reports and statements. There is a substantial degree of capacity building that needs to be undertaken to support both the central and local treasury offices in their ability to improve budget execution and cash management practices of budget entities. 44. Some other basic changes in systems and practices could substantially improve the overall functioning of budget execution and financial management practices, particularly by improving the transparency of the system and governance. These include establishing a new Accountant General position in MOFE and new line ministry positions for chief accountants to provide leadership in financial management reforms, adopting uniform chart of accounts based on international standards, and adopting systems-based audits as the dominant audit methodology across the government. Some of these measures are being implemented under assistance projects, but poor coordination impedes comprehensive reform. THE SECOND CHALLENGE: IMPROVING PUBLIC SECTOR EFFICIENCY AND EFFECTIVENESS 45. The second challenge Mongolia must meet to restore growth and improve living standards is to improve the efficiency and effectiveness of the public sector, especially for lower income groups, by designing incentive structures that lead to a more performance-oriented public sector that delivers quality services and improves outcomes. 46. Reducing poverty and increasing the assets of the poor remain Mongolia's biggest medium-term tasks. Mongolia's transition from a command to a market economy will depend on the government's success in allocating increasingly scarce resources to support its objectives of improving human capital assets and creating an environment for broad and sustained economic growth. In 2000 the government signed the MDG, which provide international benchmarks for monitoring success in key development areas. Mongolia will need to discipline its policy setting as it develops its Poverty Reduction Strategy Paper (PSRSP) and allocation processes to direct resources to areas most likely to assist in achieving the MDG if it is to achieve those goals by 2015. 47. Total spending in the social sectors increased considerably in the last half decade, rising from 13 to 19.6 percent of GDP over 1996-2000: education expenditure increased from 5.2 percent of GDP to 7.9 percent, health expenditures from 3.5 to 4.4 percent and social security expenditures from 4.4 to 7.3 percent. However, most social indicators have not recovered from substantial drops over 1990-95. In ix education, while primary school enrollment increased between 1996 and 2000 to attain pre-transition levels of 100 percent net enrollment, secondary school enrollment remained particularly low at below 70 percent gross enrollment in 2000. Dropout rates are increasing, particularly among boys. Maternal mortality increased from 120 to 197 per 1,000 live births. On average access to services remains poor particularly in the central regions. 48. Achieving the government's objective of delivering higher quality social services more efficiently requires significant changes in resource allocation mechanisms and funding systems, as the current system undermines incentives for efficient and effective service delivery. 49. Local governments are responsible for the bulk of service delivery in health and education. Local government's share of education expenditure in 2001 was 71 percent of total expenditure, while in health the local governments accounted for over 65 percent of total expenditures in the same year. Local government expenditures as a share of GDP have been rising in Mongolia over the last half decade from 10.3 percent of GDP to over 13.1 percent in 2000, while local government expenditure as a share of total consolidated general government expenditure was 27 percent of GDP. 50. As local governments deliver over 70 percent of social sector services, the effectiveness of intergovernmental arrangements greatly influences the quality of service delivery. Addressing shortcomings in intergovernmental finance is essential if Mongolia is to make progress in maintaining overall sustainable fiscal balances, improving public service delivery, and supporting equitable growth. Intergovernmental fiscal arrangements are handicapped by lack of clarity and accountability in design system. 51. Fiscal decentralization is perhaps the most difficult design issue for improving service delivery at the local level while ensuring that sustainable national macro balances are protected. The current system undermines incentives for effective and efficient allocation of resources and performance management. The need for reform is unanimously recognized within the central government as a prerequisite for improving public sector performance, outcomes, and fiscal balances. However, considerable resistance remains to improving accountability at the local government level. 52. Weaknesses in the organizational structure of intergovernmental relations are the main cause of less than satisfactory performance in social sector outcomes despite expenditure increases over the second half of the decade. The main problems that affect service delivery and intergovernmental fiscal relations include difficulty in establishing the appropriate role and size of local governments in service delivery, weak accountability structures, inappropriate matching of expenditure responsibilities with revenues, poorly designed transfer systems, and overall lack of comprehensiveness and predictability of social sector budgets at the agency level. Mongolia must establish clear lines of fiscal responsibility and accountability between its central and local governments 53. To improve the system of intergovernmental transfers the government needs to align its approach to decentralization in a way consistent with the legal and regulatory framework defined by the Constitution and other relevant laws. These laws have to be clear and not overlap with each other. Secondly, to achieve a better match between, expenditure assignments and revenue-generating capacity of local governments there is a need to strengthen accountability links between local governments and citizens. In the longer run, incentives need to be developed for local governments to raise their own revenues within the context of the budget and manage their resources in a transparent and prudent manner. In the short run, revenue sharing arrangements need to be made stable and predictable. x Addressing the unpredictable nature of transfers to local governments is essential, and gradual implementation of a formula-based transfers system will be critical to achieving the government's aim of reducing inequalities while increasing efficiency. 54. Mongolia's transition from a centralized command economy to a more democratic system in which responsibility for social service delivery is largely devolved to local governments remains incomplete. The Constitution, which establishes Mongolia as a unitary state, grants its 21 aimag (provincial) governments and Ulaanbaatar only limited autonomy. All levels of government are in principle accountable to their immediate superiors. Mongolia is, in effect, a deconcentrated state with some fiscal decentralization. However, the lines of authority and responsibility have not yet been made entirely clear legislatively, leaving room for substantial, overlap and lack of transparency and accountability. 55. The challenge faced by Mongolia and other transition economies is to find stable structures of government that are politically acceptable and at the same time help promote efficient and effective local governments. It is of critical importance that measures to reform government structures clearly take into account the appropriate scale for key services, and that roles and responsibilities are set out in a clear legal framework. Mismatches between decision-making authority and responsibility undermine service delivery 56. Numerous actors participate in the delivery of services in most functional areas. These include the central government, in some cases local governments, deconcentrated budget entities, and district units that directly provide services (such as schools, clinics, social assistance organizations). A first clear tension is that between responsibility and decision-making authority. While lower levels of government are typically responsible for delivering many services, they often lack control over key decisions that affect these tasks. For example, in education, a set of norms is used to determine expenditure allocations centrally. Decisions on the number of teachers and their wage scales, which account for over 40 percent of local government expenditures, are taken at higher levels, leaving local governments little scope for resource management. 57. Frequent budget overruns by local governments indicate a mismatch between expenditure responsibilities and sources of financing or lack of accountability for financial management. This mismatch must be addressed for local governments to deliver services efficiently and effectively. One of the principal weakness in the current division of expenditures assignments grows out of the separation between fiscal responsibility on the one hand, and the power to reduce or increase expenditure commitments on the other. The responsibilities of each level of government must be matched to its decision-making authority. The capacity of local governments to deliver services needs to be assessed prior to the devolution of responsibility. Local governments have limited revenue autonomy 58. The limited amount of revenue autonomy at the local level further complicates Mongolia's attempts at decentralization. Own revenue as a share of total expenditure per capita is below 15 percent in over a third of the aimags. In 2001 only two local governments, Orkhon and UB city, were able to meet 80 percent of expenditures from their own funds. The rest relied on state transfers for over 70 percent, on average, of expenditures. The system of tax sharing is unclear, and changes from year to year, making it difficult for local governments to predict what resources they will receive. 59. While Mongolia has taken significant legislative steps to increase the taxing power of local jurisdictions-most notably by allowing provincial land and property taxes in 1999-the gap between xi local revenues and expenditure obligations remains wide. Very high national tax rates leave little scope for raising local rates. However, improvements in overall tax administration and efforts to develop local collection capacity could increase their revenue capacity. Incentives for local governments to raise revenues could be further strengthened by increasing predictability by stabilizing tax-sharing formulas and introducing formula-based financing. Improvement of local revenues must begin with fundamental restructuring of the budgetary system to guarantee stability and promote revenue collection. Transfers are made on an infrastructure basis and appear to disregard populatibn movements, undermining equity objectives 60. Despite population decreases in many aimags, transfers to a majority of them have been increasing,'indicating weakness in Mongolia's infrastructure-based transfer system. Transfers to local governments are needed to address mismatches between their expenditure responsibilities and revenue- raising capacities. But recent growth in government employment and social services has left the central government itself seriously overextended, so that transfers to lower jurisdictions are largely ad hoc and rationed through political negotiations. Transfer levels are, thus, unpredictable, and central and local budget authorities spend much more of their time negotiating with each other than designing and administering efficient service systems. 61. Social sector outcomes in Mongolia vary considerably across regions and are generally correlated with provincial revenue-raising capacity. Illiteracy rates, for instance, are low in Ulaanbaatar, while dropout rates are high in rural areas. Transfers help address such inequalities across regions. However for transfers to be effective in reducing regional inequities they must be stable and based on clear identification of needs and objectives. 62. In Mongolia expenditure transfers are allocated based on informal negotiating processes with local officials. Evaluation of transfers is based on ministry revenue estimates and the intended expenditures of local governments. Expenditure transfers are divided between fixed and variable costs. Fixed costs transfers, for electricity, heating, transportation, water, and sewerage, are estimated on an infrastructure basis and based on experience. The level of variable costs allocated to local governments is established by norms (e.g., education costs per pupil, student/teacher ratios) that reflect standards and costing. This system of mixed transfers, undermines incentives for facility rationalization and accountability, and fails to promote equitable resource distribution across local governments. 63. Any reform of the transfer system needs to focus on improving incentives for service delivery at least cost. This should include streamlining the number of budget entities to realize ecoriomies of scale and improve incentives for efficiency, developing mechanisms to streamline funding, developing a needs- based equalization transfer system, and centralizing provision of national public goods, such as policy formulation, monitoring, and evaluation. 64. A first step toward strengthening the system of transfers involves identifying the goals the system is trying to achieve. Transfers can be used to reduce vertical imbalances and to minimize inequities between more- and less-developed regions. They can also be used to promote specific types of expenditures, or activities the central government would like to encourage. The system of transfers is closely linked with existing expenditure and revenue assignments, so it is important to be aware that redesigning the system in isolation is likely to have little impact. It is essential to consider the system as a whole, and this is particularly important for transfers. A second step is to establish in a transparent manner the pool of resources available for transfer, followed by determining a per capita formula-based system to allocate these resources, combined with a long term planning of what services will be needed and where. xii A credible top down approach to budgeting is required 65. Improvements in intergovernmental fiscal relations will invariably strengthen budget preparation processes. Mongolia continues to grapple with the legacy of a budgeting system designed purely to fulfill its role in financing the national economic plan. That the budget should discipline planning, or that the budget and its associated financial rules and regulations should be the driver of a dynamic results-oriented public sector was never a consideration prior to 1990. There has been some improvement in linking policies to available resources over the past 10 years. More is needed to place the budget at the center of a performance-oriented public sector. 66. The budget process appears to be more about control than about policy making. MOFE interacts directly with some 300 agencies and local governments on current spending. Expenditure ceilings are provided to each of these entities at the beginning of the budget process, but the ceilings often come late, and adherence to them is very patchy. Lack of prior discussion by Cabinet on budget ceilings reduces commitment to these ceilings. 67. The biggest weakness of the budget process is the apparent lack of any budget discipline in policy making. Policy development has not been paralleled by a similar exercise to identify non-priority programs or activities to be cut back or eliminated, and an increasing use of earmarking further reduces the scope for reallocation. After budget passage there remain opportunities for new proposals to be considered for funding, which means that proposals do not have to compete with each other for priority or funding. The fact that new proposals can be approved on the expectation of additional revenue compounds this problem (and appears to be condoned by the budget law): Access to supplementary during the budget year should be restricted to essential expenditures. 68. Mongolia had a relatively stable budget preparation process--on paper-until 2000. Actual practice however deviated from year to year, considerably undermining the integrity of the process. The formal processes began to change in 2001, with the introduction of the Medium Term Fiscal Framework (MTFF), and this will accelerate in the years ahead as advances are made toward the use of a full-fledged Medium-Term Expenditure Framework (MTEF). If these new processes are to support a more performance-oriented public sector that directs resources to the priorities represented by the Government Action Program and MDG, it will be crucial that they be adhered to. 69. One important recommendation of this report is that each year MOFE publicize the budget timetable before the process begins, and then follow that timetable as closely as possible. As political factors are the most likely reason for non-adherence, the commitment of the Cabinet to the timetable should be sought. Very much linked to this concept is the need for transparency in the rules for allocating resources and, particularly, reallocations during budget execution. 70. A system of forward estimates alongside the MTFF, Government Action Program, and the PRSP will give Mongolia the key tools to support a performance-oriented public sector. These are the essential ingredients -of a shift to a credible top-down approach to budgeting based on a realistic assessment of what is affordable in aggregate over the medium to longer term, and informed by policy objectives and priorities. 71. The key focus of efforts to improve budget formulation should be in strengthening this top-down approach, beginning with a prudent approach to aggregate budget formulation. Apart from the key contribution this can make to matching policy reach and resource availability, a more top-down approach is central to supporting operational performance through greater predictability of policy and funding. Another benefit of this approach is that it pushes more of the policy conflict down to the sector level where there are greater incentives to review existing policy and utilize analytical tools. In Mongolia this xiii means that considerable attention needs to be given to moving to this more sector-based approach to budget formulation. 72. Improvements in the clarity of roles and responsibilities at the central and local government levels, coupled with improvements in financial management and more a transparent budget process would provide the legislature and the citizens with adequate information to monitor and evaluate public sector effectiveness. Budgets should reflect poverty-reducing and growth strategies of the government 73. While the economy needs increased investment in all sectors to upgrade and maintain its aging infrastructure-particularly in the social sectors-the prospects for substantial additional funds in the medium term to cover all critical expenditure are weak, as public and private resources are limited Achieving the government's objective of delivering quality services more efficiently requires changing policy priorities and channeling funds to areas where they will have the highest payoff in terms of improving outcomes. These include primary and secondary education, primary health care, environmental protection, and improved access to safe water and clean air. Improving intergovernmental fiscal relations and budget processes would facilitate improved allocation of resources. Education strategies need to focus on improving secondary school enrollment and reducing dropout rates, particularly for boys 74. Public spending on education has been well protected during the transition. Education expenditures were high, at 7.9 percent of GDP, in 2000, even higher than social welfare and social assistance expenditures. The main challenge for the education sector is to improve expenditure allocations to provide more basic primary and secondary school education to the population. While enrollment rates in primary school have recovered, secondary school enrollments levels remain well below their pre- transition level. Primary school net enrollment rates dipped in the mid-1990s, then recovered substantially to 98 percent in 1999. Secondary net enrollment continues to lag with only 84 percent enrolled in lower secondary school and 35 percent in upper secondary schools. 75. Nearly 60 percent of all dropouts in 2000 were male, although in 5 of the 21 aimags more females than males dropped out, and female dropout rates also appear to be increasing in Ulaanbaatar. Dropout rates among rural students are substantially higher in urban and semi-urban areas, but the gap is decreasing. The main causes of high dropouts and low enrollment are poverty and the high opportunity costs of enrolling children in school, particularly in agricultural regions, although more than 18 percent of dropouts cite poor health and the Participatory Living Standard Assessment cited widespread frustration with the quality of educational services on offer, which in turn lowers the perceived rate of return for education. Access is an issue in many urban areas where schools are overcrowded. 76. Any strategy for improving allocations in education must take into account the factors that affect where, how, and what kind of education services need to be delivered. These include migration, changing needs of the emerging market economy, population growth, and private sector involvement. To achieve its objective of improving access, the government will need to allocate additional resources to primary and secondary education as they become available. Government should also continue to encourage greater efficiency and cost sharing within the higher and vocational education sub sectors. Capacity at the center to effectively set, cost, implement, and monitor standards needs strengthening. xiv Health resources need rationalization 77. Mongolia's health system is struggling to deliver quality health services to the population with limited funds. Overall health expenditures were 4.2 percent of GDP in 2000, and sector funding has shrunk both in real terms and as a percentage of GDP since 1990. Despite the existence of a well- developed human and physical resource base inherited from the socialist period, the public health care delivery system provides very little value in terms of both the amount of services produced and their appropriateness and quality. Rationalizing the resource base, re-distributing the skill mix, and reallocating the funds in accordance with health needs instead of resource inputs could achieve much better results. 78. In general, the health status of the people of Mongolia is somewhat better than that of an average low-income country. However, Mongolia lags behind low and middle-income East Asia and Pacific countries in many health indicators. Mongolia's maternal mortality ratio, at about 158 per 100,000 live births, remains high and varies widely across regions. Mongolia is undergoing an epidemiological transition characterized by a decline in the share of communicable diseases and a rise in non- communicable diseases in total mortality and morbidity. 79. The medium-term challenge for the health sector is to use available health resources more efficiently to deliver quality services, and to target spending more directly to the poor. Achieving this will require significantly stronger efforts at priority setting-directing resources to areas that benefit the most needy and have the highest payoff in terms of improving health outcomes, and policy reforms-to remove obstacles to efficiency. Cash benefit and in-kind programs need to be reformed and their overall goals reconsidered 80. Social assistance benefits increased by over 25 percent over 1998-2000. These benefits are mandated by national law, but are paid from local government budgets. There are about 20 different benefits, of 3 types: short-term (e.g. maternity benefits); lump sum, such as for having twins, and benefits in kind, such as fuel for cooking and heating. All benefits are defined by these categories; there is no guaranteed minimum income or basic social assistance benefit. Many benefits have income tests, while others do not. 81. The social assistance system needs fundamental revision. The cash benefit and in-kind programs need to be reformed and the overall goals of the program reconsidered. The goal of both the cash benefits and cash programs should be to support the re-inclusion of children into families and individuals into society. This is not the case now. The system should avoid labeling people, and the need for assistance should not be pre-determined based on, for example, a medical event (e.g. multiple births). The legal ,framework should be revised to incorporate this goal. Allocations to other non-social sectors should reflect policy priorities and result in efficient spending 82. Mongolia is heavily dependent on natural resources, which contribute 13 to 20 percent of GDP. The livelihoods of over two thirds of the population are directly affected by environmental policies or events. Natural disasters such as drought and steppe fires also affect the livelihoods of Mongolians. In 2001 historically harsh weather led to the death of over 10 percent of the national herd, severely affecting the livelihoods of most Mongolians. 83. Improvements in environmental management require changes in the legislative framework and capacity building. While budget allocations for recurrent and capital expenditures on the environment are low and may require increases, resources generated from land reform measures could provide some relief. xv In the short to medium term a strategy for addressing Mongolia's most pressing environmental and natural resources management issues, which include solid waste collection and management, water pollution abatement, land management, and air and industrial pollution should be developed and costed. 84. The task of the government will be to ensure that allocations to other non-social sectors result in efficient spending, and are grounded in the development priorities of the government. The MIDGs provide benchmarks for development objectives in sectors that the government could adopt. While this PEFMR makes specific sectoral recommendations, the real message is that these recommendations should be fed into the reformed decision making, budgeting, and management systems and processes discussed above. BRIDGING THE GAP 85. Mongolia's vision for public management is laid out in the draft PSMFL. What is missing is an implementation plan for closing the gap between current practices and those necessary to support a more performance-oriented public sector. This report proposes a reform strategy that builds on the interdependence between the core systems and processes at the center of government and those at the sector, agency, provincial, and local levels. Building on experience in other reforming countries, a key aspect of the proposed approach is the provision of greater budgetary and personnel management flexibility and funding predictability to entities that meet key financial and performance management standards. This would be supported by higher quality and more transparent financial and non-financial performance information. 86. The approach is built on two broad, sometimes overlapping, clusters of basic practices. The first focuses on basic internal financial management practices, including budget and funds control, internal auditing, accounting, procurement, and asset management. The second targets basic performance management, including budget planning, output costing, and human resource management. These are supported by requirements for financial and performance reporting. Both principles emphasize the need for prudence to foster an environment that supports and demands performance and adequately sequences changes in the budget management processes. 87. There are numerous advantages to this approach. It is based on incentives, and agencies that achieve the standards receive something in return. They also have to show concrete evidence of changes in systems and practices before traditional controls are eased or new approaches adopted. The approach is staged, and places strong emphasis on building capacity in line with growing demands and incentives. It builds a momentum that puts pressure on all players to perform, while not holding back those who want to go faster. It has the potential for early wins, through self-selection by reform-minded agencies. It leverages demonstration effects. It improves the sequencing of different components of budget reform. It both facilitates external technical assistance and demands a greater consistency of approach from donors. 88. This report provides an analysis of policy options centered on stronger fiscal discipline, better pro-poor budgetary objectives, and improved budget processes and systems. It focuses on identifying the priority items that will accelerate the market transition process, already well under way in Mongolia. It identifies public policies and actions to reduce the size of government and contain debt levels, linking policies with budgets that prioritize and target government interventions, and improving systems and institutions to improve outcomes. The proposed approach would provide incentives for reform by making more flexible and predictable operating environments conditional on agencies and local governments demonstrating their capabilities in basic financial, policy, and personnel management. 89. The recommendations in the PEFMR are guided by a set of basic principles. They call for a disciplined, comprehensive public sector, and effective and balanced links between short, medium, and long-term policy changes. Roles and responsibilities should be clearly set out in a public sector that xvi matches authority (flexibility) with responsibility (accountability) at the ministry, sector, agency and sub- national levels of government. 1. INTRODUCTION Improving fiscal balances, strengthening institutions, increasing efficiency of public actions 1.1 Mongolia has realized important progress toward more efficient and sustainable public finances over the last decade, and the transition has been considerably well managed compared to other Central and Eastern European economies. However, weaknesses in the institutional and structural reform agenda threaten Mongolia's fiscal balances. Unfavorable external conditions and weak public expenditure management have eroded most of the gains of the stabilization period. In 2000 GDP growth was at an all- time low of 1.1 percent since 1993, after averaging about 3.5 percent over 1994-99. Government expenditures rose to over 40 percent of GDP in 2000. High tax rates have contributed to keep revenue up and contain the deficit, but have hampered private sector growth. These high tax rates and significant increases in public expenditure have not benefited the poor. Standards of living have dropped as GDP per capita decreased from about $429 in 1996 to $403 in 2000. Unemployment appears to have significantly increased. The depth and severity of poverty have increased, and over a third of the population lives below the poverty line. 1.2 Evidence of weak public expenditure management is reflected in high levels and composition of public spending and revenue mobilization. The size and scope of the government remain large and continue to expand. The government has remained engaged in unproductive economic activities inherited from the command economy. Unsustainable fiscal policy directions in these areas have depressed growth and hampered the pace and effectiveness of the government's poverty reduction initiatives. Public sector resource allocations are not consistently pro-poor, and have been unsatisfactory in improving outcomes at least cost. Resource allocations have concentrated on cash transfers to the population, rather than improving their productive assets. Much less attention has been paid to improving incentive structures in the institutional framework, management processes, and systems required to enhance service delivery. These shortcomings in public sector management and an overall depressed external economic environment have prevented the achievement of substantial improvements in living standards and undermined private sector growth. 1.3 To restore growth and raise the living standards of the population, Mongolia proposes to grow at 6 percent, maintain inflation at single digit figures, and progressively reduce the deficit in the medium term. This report discusses policy measures needed to ensure the successful realization of the government's program. Restoring fiscal balances needed to engender growth requires the government need to address fundamental fiscal and structural deficiencies in the budget and budget processes over the medium term. 1.4 To achieve its objectives of restoring growth, improving outcomes and reducing poverty Mongolia is confronted with the triple challenge of: * Strengthening public finance to ensure that deficits are sustainable and the poor are protected from further burdens of fiscal austerity; * Disciplining the allocation of public spending across-and particularly within-sectors to reflect the changing role of the public sector as a provider of public goods, support broad based growth, and generate employment opportunities, and, 2 Improving the efficiency and effectiveness of public sector institutions and processes to increase the chances for lower income groups, especially, to benefit from low cost quality services. 1.5 Fundamental changes in policies, systems and processes will be required to meet these objectives. The government is contemplating a set of reforms to bridge the gap between current practices and the measures required to meet its objectives of building a performance-oriented public sector capable of facing the challenges ahead. The Public Sector Management and Finance Law (PSMFL), under discussion in Parliament, defines in significant detail a comprehensive framework within which considerable advances could be made to improve performance and meet the government's development objectives. These objectives are defined in the Government Action Plan (GAP), the Good Governance for Human Security (GGHS) paper, recently reinforced in the Interim Poverty Reduction Strategy Paper (I- PRSP), and the Poverty Reduction and Growth Facility (PRGF) paper. 1.6 Implementation of a more performance-oriented public sector (POPS), whether or not this is via the PSMFL, faces many hurdles. Mongolia does have an advantage over many developing countries in that it does have a blueprint, in the form of the PSIFL, for a POPS. While the PSMFL and accompanying Implementation Plan provide some important guideposts and current efforts to reform the budget process and system are moving in the right direction, they are partial and do not grapple with the yawning gap between current practices and those practices that will be required to support a POPS. 1.7 However, the current government Implementation Plan needs to be supplemented by a range of reforms to make it consistent with current systems and practices if a more POPS is to be achieved. Consequently an approach that builds on incremental and sequential windows of opportunity to support the government's public expenditure management reform program is recommended. 1.8 This Public Expenditure and Financial Management Review contributes to this process by providing an analysis of policy options centered on stronger fiscal discipline, better pro-poor budgetary objectives, and improved budget processes and systems to address these challenges. The discussion will focus on identifying the priority items that will accelerate the market transition process, which is well under way in Mongolia. Particularly, the PER will assist the government in identifying and implementing public policies and actions to reduce the size of government and contain debt levels, link policies with budgets that prioritize and target government interventions, and improve systems and institutions to improve the quality of outcomes. It also focuses attention on the need for institutional arrangements that discipline policy choices, and on the yawning gap between current public sector practices and those that will be needed to support a more performance public sector and, ultimately, achievement of the Millennium Development Goals. Principles by which to appraise any changes in public sector 1.9 The rapid changes and emerging political and economic challenges in Mongolia often make it difficult to assess the likely impact of individual reforms. In addition to the challenges emerging from the transition to a market economy, two other factors have a significant bearing on public sector performance. The first of these is intergovernmental relations and the second is the draft Public Sector Management and Finance Law before the Parliament which proposes major reforms in public sector management and practice. There are of course many other changes being discussed, many of which are contained in other draft legislation. Against this background, the PEFMR argues that having a set of broad principles against which new (changes) reforms should be tested will help manage the transition. Such a set of principles is discussed below. 1.10 A key starting point is that a more performance-oriented public sector will be built on sound basics. This include: 3 * Foster an environment that supports and demands performance before introducing performance or outcome budgeting; * control inputs before seeking to control outputs; account for cash before accounting for accruals; * establish external controls before introducing internal control; establish internal control before introducing managerial accountability; * operate a reliable accounting system and treasury system before installing an integrated financial management system; * budget for work to be done before budgeting for results to be achieved; * enforce formal contracts in the market sector before introducing performance contracts in the public sector; * have effective financial auditing before moving to performance auditing; * adopt and implement predictable budgets before insisting that managers efficiently use the resources entrusted to them . 1.11 A disciplined public sector, another of the principles, is arguably at the core of these basics. Without a disciplined public sector - one that adheres to the rules, whether they apply to living within the budget or to employment rules - improving performance at the policy and operational levels and sustaining it is very difficult. These principles apply to the messages in all the chapters. The centrality of discipline comes through in relation to fiscal policy, civil service reform, policy formulation and service delivery. lPibadhsn fsedt reed por pg mi% h ge e .. parepm eenortayb eM 'ngO t Jddtdrfq UC&*iY14M (c) heeeu otpul. basa mment and rep or f u (8)theprearaioio sareic usiessplas b niA ,abce ai parehamenc,s tarbde anr e Av 1 I aco2b n en%:h ear I(e) the maegruon of crualbren n and iaccunteing by mougtries agproe pfatirkaproesthe borm o~ th g~rnen s woe ~ '~ 'I" *' ,, -_g - agg)ioprileationsio iptrimanagendcison o ste s ecris andiaf exestrv s;: el s tB (g) the preparationof soraec Accuinsanialbtatm ise, andarnual c and inraniae ts . 1.2 h filctertoni androches reformv weesll beiedaLuid i discplie no theime menttio ,I K Iit apre) he leappro gn td cpial chaterg gjoeu therappopiak one po te other prinipl namelyapnmt inn and agit esonsiblity to this prsipe of dcli as n thei achiev ten prepaaionoc financial,mngmn cabilitis ndaa age clExperincdoeiniatetha I.s "Geting he BsicsRigh"fAleciac,u199. I , X ,III .~h .~ .J . .. .. apprah the approac siouested,,n thfingfalhter goes furthrt lIn on of te othagerPn pstnils acieem)e dofnsi fiacial mangemeg-nt acapabities biiniageies. Eeren ce does idte ta tGtig h a ics l R ight" inftagle n i Schicsib, 1997.ertne n-cif61u .p 4 improvements in operational performance, a particular objective of the draft PSMFL, require that managers have substantial flexibility (authority) in the management of the resources at their disposal. This flexibility must be accompanied by commensurate accountability (responsibility) and the basic disciplines in financial and personnel management. 1.13 This report also argues that, in the past, GOM has perhaps given insufficient attention to balancing short, medium and longer-term considerations when it makes its decisions. The PEFMR focuses attention on how short term imperatives, particularly as they relate to stabilization policy, might be better managed to support other objectives in relation to policy and service delivery, which require a more medium to longer term approach. The commitment to poverty reduction in the PRSP reinforces the need to seek this balance. The report aims to bridge the gap between current practices and those required if Mongolia is to build a high quality public sector service. 1.14 The remaining principle that needs highlighting here is that of clarity of roles and responsibilities. Again this has application to all chapters but it is particularly pertinent to the chapter on intergovernmental relations. Perhaps the major challenge to improving operational performance is to sort out the respective roles and responsibilities of sub-national and national government. This has to be seen as a basic. 1.15 The following analysis focuses primarily on the centrality of the broad principles, with specific reference to the changes needed in the Ministries of Finance and Economy, Health, Education, and Nature and Environment to improve Mongolia's public expenditure management and build the foundations for developing a full-fledged performance-oriented budgeting system capable of delivering on the government's poverty reduction and growth objectives. 1.16 The PEFMR is organized into 9 chapters. Chapter 2 discusses past performance and provides a brief discussion of recent economic and social developments. Chapter 3, discusses future challenges, analyzes the main sources of fiscal pressure on Mongolia's budget, and proposes a sustainable macroeconomic reform path. Chapter 4, evaluates budget execution processes. Chapter 5 discusses the impact of intergovernmental fiscal relations on fiscal balances, and service delivery. Chapter 6 discusses governance. Chapter 7 analyzes budget formulation and policy coordination processes. Chapter 8 discusses public sector resource allocations, with particular emphasis on education, health, and the environment sectors. The PEFMR concludes by discussing the importance of putting into place measures to bridge the gap between existing public expenditure management practices and those needed to improve the overall performance of the Mongolian public sector, and proposes a framework within which this can be achieved. 2. ASSESSING PAST PERFORMANCE Impressive macroeconomic performance in the early years of the transition 2.1 Mongolia's macroeconomic transition to a market economy was resolute and quite successful. Four years into the transition, real GDP growth turned positive for the first time in 1994; most other transition economies needed about a decade to realize the first benefits of transition (Figure 2.2). Growth reached record highs in 1995 at over 6 percent of GDP. The overall fiscal deficit including grants dropped from 13.5 percent of GDP in 1990 to 6.7 percent in 1995. Recognized as one of the "rapid reformers" in transition economies, Mongolia moved quickly in 1992 to liberalize prices, lower trade barriers, and undertake tight fiscal and monetary policy to ensure rapid stabilization of the economy. These macro- stabilization policies reaped the expected benefits: growth resumed and inflation dropped from three- to two-digit figures. The size of the public sector in GDP dropped from 51.8 percent in 1993 to 31.5 percent in 1995. . Figure 2-1: Real GDP per Capita ($US) 1990-2000 Figure 2-2: Growth Rate in Selected Transition Economies (percent)2 US$ 500 500 - 425 429 2- 450 401 403 382 403 400 350 304 300 253 250 200 150 100 50 --3 02 -30est lit alb mon - kyr kaz - uzb - bel Source: World Bank Sima data base and staff estimates, 2001 OVERVIEW OF CONSOLIDATED GOVERNMENT EXPENDITURES AND REVENUE 2.2 The composition of expenditures and revenues in Mongolia has undergone some changes during the period since 1993 as a result of changes in structural fiscal policy and overall international economic pressures. While the composition of expenditures changes substantially average expenditures did not change substantially over the same periods. Over the period 1993-1996 total expenditures as a share of GDP averaged 40 percent while between 1997 and 2000 this average remained about the same at 39 I. 2Countries include: Estonia, Lithuania, Albania, Mongolia, Kyrgyz, Kazakhstan, Uzbekistan and Belarus. 6 percent of GDP. The internal composition of expenditures also changed reflecting changing priorities of the government and emerging social pressures resulting from the transition. 800 - 70.0- 60.0- 50.0 - U 1993-1996 C 40.0- U 1997-2000 300- 200- 100 - 0.0 - ca Category Source: Ministry of Finance and Economy, 2001 Analysis of Expenditure Trends 2.3 Current expenditures have increased dramatically over the last decade. The share of current expenditure in total expenditure increased from 57.4 percent of total expenditures in 1993 to 69.2 in 2000. Current expenditures increased on average between 1993/1996 and 1997/2000 by 21 percent. Capital expenditures and net lending decreased as a share of total expenditures accordingly from 42.6 to 30.8 percent of total expenditure over the same period. The average between 1993/1996 and 1997/2000 capital expenditures decreased by 27 percent with capital expenditures decreasing by 12 percent and net lending by 32 percent. (Figure 2.3) 2.4 The overall distribution of current expenditures has become increasingly rigid. Expenditures on goods and services have increased from 34.2 percent of total expenditure in 1993 to 50.5 percent of total expenditure in 2000, a 47 percent increase. Increases in wages and salaries account for the bulk of this increase, going from 10 to 20.8 percent of total expenditure over the period. Increases in expenditures on wages and salaries accelerated between 1997 and 2000 increasing on average by 25 percent in the latter half of the decade. The rise in wage bill expenditures was due partly to systematic increases in salaries between 1997 and 2000, partly to adjust for inflation and to increases in wage contributions. 2.5 The high share of expenditures on goods and services in the early part of the decade are a reflection of the huge infrastructure inherited from the pre-transition era. However, continuous increases in the expenditure on non-wage goods and services reflect the increasing cost of doing business in the public sector. Expenditure on goods and services as a share of total expenditure increased by 22 percent from 24 to 29 percent of total expenditures between 1993 and 2000. Expenditure on heating and other utilities account for a large share of public sector operating costs, (Figure 2.4). 7 7.0 6.0 5.0 4.0 0 1993-1996 30- ZO0I 1997-2000 2.0 I I ; I 1.0 I 00 cd0 Source: Ministry of Finance and Economy, 2001 and staff estimates 2.6 The share of expenditure on non-wage goods and services is a good indicator of the cost of public sector activity and maintenance of public facilities. Aggregate statistics indicate that expenditures on heating and electricity account for a large share of non-wage expenditures. Heating and electricity expenditures account for 26 percent of total non-wage expenditures. Slow pace of facility rationalization and recent increases in energy prices explain these high levels of expenditures. Increases in the share of expenditures on administrative operating costs reflect in part increases in non-wage subsidies to high- level officials in the form of free access to telephones, transportation, newspaper and other services. Increases in expenditure on food are mainly a result of changes in government education policy and increases in food subsidies to dormitories. 2.7 Increases in domestic borrowing in the second half of the decade and cost of banking sector restructuring lead to substantial increases in the share of total interest payments in expenditures (fig 2.3). Interest payments increased from 2.4 percent of total expenditure in 1993 to 4.4 percent. This clearly underlies the extent to which past aggregate fiscal performance has imposed a burden on current budgets. 2.8 Transfers of goods and services exhibited an uninterrupted upward increase between 1993 and 2000, going from 9.4 in 1993 to 19.8 in 2000. Expenditures on transfers increased from 11.2 to 17.1 percent of total expenditures between 1993/1996 and 1997/2000. 2.9 Subsidies to state owned enterprises decreased from 4.6 percent of total expenditures to 0.6 percent as all state owned enterprise expenditure was moved off budget between 1996 and 1997. Subsidies included in the budget are subsidies directed at the energy authority to subsidize heating expenses and to the urban transport authority for fuel expenses. 2.10 Increases in capital expenditure and net lending reflect the need to upgrade overall infrastructure of the public sector. While capital expenditure has increased substantially between 1993 and 2000, 8 average capital expenditure between 1997 and 2000 decreased compared to 1993/1996. This decreasing trend in capital expenditure can be explained by the need to reduce expenditures in the later part of the decade to restore overall macroeconomic balances, however this tendency is worrying as substantial capital investments continue to be needed to upgrade overall infrastructure networks particularly in the social sectors. 2.11 Mongolia's expenditures have been growing persistently over the last decade and are increasingly exerting pressures on overall fiscal balances. Efforts to reduce the overall fiscal deficit to a safe level must depend primarily on constraining primary spending growth. However, reforming Mongolia's expenditure composition will require fundamental institutional changes. The institutional challenges in slowing spending growth and the multiple routes that need to be followed to meet these restructuring challenges are centered on the role and reach of government. Consequently these issues can only be treated efficiently as part of a comprehensive effort to improve overall performance of the public sector. In the short to medium term however, efforts to limit growth in primary spending will need to focus on social security outlays and wages and salary expenditures and other contingent liabilities not included in the general government accounts. Analysis of Revenue trends: Tax system and Tax administration - Issues and Options 2.12 On the revenue side, sweeping reforms in the tax system and tax administration over the last decade substantially increased revenue performance, but there is room for improvement. The main problem on the revenue side, is the increasing tax burden which is already very high. Mongolia will need to reform the tax administration and continuing to adopt reforms that favor a more efficient tax regime, with fewer distortions and increased compliance. Problems with the tax incidence and the increasing fiscal measures to meet rising expenditure commitments continue to undermine performance. In the medium term and following expenditure reductions, reforms should focus on reducing tax burden to promote growth by strengthening the tax administration, broadening the tax base, reducing exemptions and improving legal framework for revenue assignments. Percent Shame of revenue sources in total 100% 80% 60% 40%- 20%. 0%_ 1993 1994 1995 1996 1997 1998 1999 2000 ETax rev and soc sec contnb.s E Non-tax revenue 0 Capital rev. and grants 0 Foreign grants Source: Ministry of Finance and Economy, 2001 9 2.13 Prudent fiscal policy in the mid 90's aided by a period of high commodity prices helped Mongolia in achieving a high level of revenue mobilization, especially on the collection of corporate income taxes and dividends. Since then balance in the tax structure has changed somewhat. 2.14 Tax revenue as a share of total revenue dropped from 76 percent to 64 percent between 1993 and 2000. Non-tax revenues increased from 2.9 to 18.3 percent over this period as dividends from state owned enterprises increased. The share of grant revenue increased somewhat between 1994 and 1998 but decreased substantially after 1998. 35 0 300 250 20a 15 1993-1996 150 o0 a1997-2oo 10.0 Source: Ministry of Finance and Economy, 2001 and staff estimates 2.15 Deep reforms in the Mongolian tax system and tax administration began as result of the falling copper prices in 1995. Faced with falling revenue and potential erosion of all the early gains from the stabilization process Mongolia in 1997 accelerated revenue enhancement reforms through changes in the tax system and tax administration processes. These reforms were very successful as revenue went from 23 percent of GDP in 1995 to 33,6 percent of GDP in 2001. As part of the reforms the VAT was introduced and increased from 10 to 15 percent replacing the sales tax in 1998. Corporate income tax was increased to 15 and 40 percent, and the brackets reduced for both corporate and personal income tax from five to two. Table 2-1: Share of Tax Revenues in Total Revenues 1993-2000 1993 1994 1995 1996 1997 1998 1999 2000 Income taxes 41.1 28.4 25.7 208 20.8 14.3 12.1 15.6 Social secunty contributions 4.1 5.6 8.4 8.4 9.5 12.6 12.2 10.7 Sales tax and VAT 10.3 10.1 8.6 9.8 12.6 15.8 18.8 18.3 Excise taxes 9.5 4.7 4.6 5.1 6.7 8.4 8.3 9.8 Taxes on specific. services 0.6 2.5 1.7 2.1 1.6 2.2 2.1 1.7 Customs duties and export taxes 9.5 6.7 5.1 6.4 2.9 0.5 2.8 5.4 Other taxes 0.9 1.8 1.9 2.6 2.6 2.7 2.8 2.8 Source: Ministry of Finance and Economy, 2001 and staff estimates 10 2.16 Currently, Mongolia's tax system consists of the Box 2.1: Main Fiscal reforms adopted in 1998-1999Main Fiscal enterprise and individual Reforms adopted in 1998-1999 income taxes, VAT, various The number of coporate income tax rates was reduced from four 15, excise25, 35, and 40 percent) to two (15 percent and 40 percent), andthe excie taes, mpot duies,and depreciation system was simplified to single rate, and the straight line export taxes, and a number of method; minor taxes. In 2000, tax * The number of individual income tax rates was reduced fromfive (2, revenue accounted for 75 5, 15, 27, and 40 percent) to three (10, 20, and 40 percent), and the percent of total general exemption threshold was replaced with a tax credit; government revenue. The sales e The base of the sales tax was broadened to include selected tax, the enterprise income tax enterprises in the food, construction, and communication sectors, and and the excise tax were the three the annual turnover threshold was eliminated; largest revenue sources, * All custom duties were eliminated, except a 15 percent duty on accounting for 48 percent of alcoholic spirits; total general government * The excise on petroleum products, spirits and wine were doubled; revenue. The enterprise income * The royalty rate on minerals was reduced at 2.2percent, from the tax and the sales tax (VAT after previous system of negotiated rates in the range of 1.5 to 12.5 percent. July1, .998)werethe wo VAT was introduced at 10 percent on July 1, 1998. July 1, .1998) were the two largest revenue. sources, * A15 percent customs duty on flour and vegetables was reintroduced accounting for 30 percent and on a temporary basis 24 percent of total tax revenue, respectively. As a result of the tax reform in 1997, excise taxes 8-1 have gained in importance in total tax revenue, reaching 10 percent in the first nine months of 1998, while import duties and export taxes declined to just one percent of total tax revenue. 2.17 Tax administration is the responsibility of the General Department of National Taxation (GDNT), which was established as a separate government agency in 1992. The GDNT is under the supervision of the Minister of Finance, and controls 4 city and 18 aimag (provincial) offices. There are 8 district tax offices under the supervision of the capital city office and about 360 sou (district) offices under the supervision of the aim*g offices and other cities. The tax administration is responsible for collecting both state and local taxes, and supervises activities of tax education and tax collection methodology. Since 1997, a major task of the tax administration has been the design and implementation of the VAT. With the technical assistance from the USAF, the GDNT prepared the action plan and drafted VAT administrative regulations. By mid- September 1998, the VAT refund system had not become fully operational as the establishment of the special refund account was delayed. The administration of the VAT has also been complicated by several ministries' ad hoc decisions to offer exemptions. 60.0- 50.0 A Dpstoetloos ad y taxes 40.0- ~30.0- 20.0- Non-distoronasry taxes 10.0. 0.0- - 1991 1982 199 1994 1995 1996 1997 1998 1999 2000 Source: Ministry of Finance and Economy, and staff estimates, 2001 11 2.18 Overall Mongolia's tax incidence is improving. However to ensure that this trend is maintained, that the poor are protected from any austerity measures and the private sector is not heavily taxed Mongolia will need to reduce expenditures by accelerating 'the structural reform agenda in order to create room for tax reductions. Failure to complement macro-fiscal reforms with institutional and structural reforms weakened transition efforts 2.19 While increases in revenue mobilization greatly contributed to restore macro-economic stabilization, Mongolia failed to lock in its early successes in macroeconomic stabilization by deepening the structural and institutional reforms needed to ensure prolonged growth. The size and reach of the public sector continued to mirror pre-transition command economy structures and the discipline of the command economy was eroded by continuous institutional changes. Figure 2-8: Real GDP growth 1993-2000 (percent) Figure 2-9: Revenue and expenditure 1995-2001 Growth drops Swelling expenditure drowns revenue mobilization efforts 45 7- 7, 40 General ' 6 5 35 4 per30- 3 ce E 2 ~nt 25- Revenue 2 20 0 15 '0 0 '0 '0 10 Q 10 1 1 1 11 C.A -~j 00 ' 1995 1996 1997 1998 1999 2000 2001 Figure 2-10: Overall fiscal : Deficits Exceeded 1995 Figure 2-11: High Deficits Increase Domestic Levels Through 2001 Borrowing and Interest Payments -16 -16 -14 -14 140 -12 12 Uo 11. -10 10 nao -8 8 o -6-- Overall balance % 6 40 -4 -0 1h. -2 -1995 1996 1997 1998 1999 2000 2001-- 2 o 0 i I I 0 -20 46 Ftuign(O Dmm) Ium 5 1993-1995 N 199&2000 Source: Ministry of Finance and Economy, 2001 12 2.20 Employment levels in the public sector remained high, with the government using public employment as an indirect subsidy in response to rising unemployment created by the breakdown of the welfare system and privatization of State Owned Enterprises. Social insurance and safety net policies in place during the transition, such as a generous pension system and subsidized housing, were maintained. Despite some privatization, the public sector continued to engage in productive activities, such as food and agricultural production, textiles, and retail trading. Quality service delivery and efficiency objectives in the education and health sectors were overlooked resulting in significant drops in outcomes. Cumbersome inter-governmental fiscal arrangements undermined performance and accountability, and the processes and systems for ensuring disciplined and accountable public actions remained weak to nonexistent. 2.21 Mongolia's failure to address non-macroeconomic concerns pertaining to the transition was in large part overlooked by the international community. It was generally believed that once macro-stability was achieved, the discipline engendered by this process would spill over into structural and institutional reforms. In most countries, this has not been the case: balancing macroeconomic aggregates has proven to be easier than defining the appropriate role of government and designing the right systems and processes to reflect this role. Many countries have yet to reach consensus on the role of the public sector; Mongolia has the advantage that this debate is alive and present. Negative external conditions -East Asian and Russian crisis-further complicated the transition 2.22 The 1997-98 East Asian and Russian crisis exposed Mongolia's underlying structural inefficiencies. Export prices of its main commodities fell sharply: gold prices fell by 42.5 percent, copper by 28 percent, and cashmere by about 12.5 percent. Between 1995 and 1998, economic growth fell from 6.3 to 3.3 percent, a 44 percent contraction (Table 2.1). Large revenue shortfalls resulted as corporate income tax contributions and dividend payments from major state owned enterprises dried up. By end- 1998 revenue collection had dropped by over 5 percentage points of GDP, while expenditures between 1995 and 1998 increased by over 33 percent. The deficit more than doubled from 6.7 percent of GDP in 1995 to 14.3 percent in 1998. 13 Table 2-2: Trends in Overall Fiscal Balances: 1993-2000 1993 1994 1995 1996 1997 1998 1999 2000 Est. (Percent change) Real GDP -3 2.1 6.3 2.4 4 3.5 3.2 1 1 Consumer prices (period average) 268.4 87.6 568 46.8 36.6 9.4 7.6 11.6 Consumer pnces (end period) 183 66.3 53.1 446 20.5 6 10 8 1 (In percent of GDP) General government revenue 30.2 31.1 25.6 248 255 27.6 272 33.6 General government expenditure and net lending 28.3 26.8 31.5 32.7 34.5 41.9 39.4 40.5 Current expenditure 25.5 18.4 237 221 231 27.2 26.7 30.1 Interest payments 1.2 06 0.4 0.7 25 1.4 19 17 Capital expenditure and net lending 25.5 19.7 16.7 139 11.4 146 12.7 10.4 Current balance 1.9 29 6.2 4 1.6 -07 -02 3.1 Overall balance -14.6 -19 8 -5.9 -79 -9.1 -14.3 -122 -6.8 Financing 59 7.9 9.1 143 12.7 6.8 Foreign (net) 73 54 11.1 8.4 11.4 6.4 Domestic (net) -3.7 1.9 -3 1 4.1 0.0 0.4 Memorandum Items Public and publicly guaranteed external debt 4/ 366 478.1 504 542 605 753 850 854 (In percent of GDP) 41.1 46 57.4 77.4 93.9 880 NPV of public and publicly guaranted external debt - - - 571 558 (In percent of GDP) - - - - 63 57.5 Debt service 41.9 66.7 64.2 56.6 41 39.3 41.3 31.5 Nominal GDP (billion togrogs) 550 647 833 817 925 1,045 Nommal GDP (miion US dollars) 1.227 1 179 1.054 972 906 970 Sources. Mongolian authorities, staff estimates 2.23 The crisis plunged Mongolia into a precarious financial situation as insufficiencies in the structural and institutional reform agenda, evidenced by public and financial sector defaults, threatened to reverse the gains of the transition. To maintain public confidence and protect the vulnerable from further income erosion, the government increased public employment and cash transfers to the population. Cash transfers and subsidies increased from 3.7 in 1995 to 6.8 percent in 1998. The government also moved to bail out SOEs and banks by sanctioning buildups in arrears. These policies increased the size of the state even further. By 2000, public spending reached 40.5 percent of GDP, higher than in most other transition economies. Renewed macroeconomic imbalances set in 1998 2.24 To accommodate the large increases in expenditure the government increased its dependence on foreign financing, and then resorted to large amounts of domestic financing as foreign financing dried up during the crisis years (Figure 2.12). Prior to 1997, over 90 percent of the deficit was financed externally; in 1999 some 40 percent was domestically financed. Increased government spending aroused fears of renewed inflation and foreign direct investment dried up, capital outflows increased, and banks all but ceased lending to the private sector. The ratio of public debt to GDP doubled over 1996-99 as government expenditures rose and the fiscal position became increasingly unsustainable. Successive severe droughts in 1999 and 2000 further depressed Mongolia's fiscal position even further leaving it highly precarious; immediate reform was necessary if gains from the early transition period were to be protected and growth restored. 2.25 Governments facing such negative external shocks and mounting domestic pressure to increase living standards are often tempted to continue weak fiscal policies, by either allowing higher deficits or disproportionately relying on excessive corporate taxes to finance undisciplined government 14 consumption. This hinders emerging growth sectors, and further worsens the well-being of the population. Mongolia must act now to restore its macro-fiscal balances to sustainable levels by taking active and decisive measures to define the functions of the public sector and reduce the size and reach of the government. Change of government in 2000 ushered in a period of political stability 2.26 In 2000 the new government began to make an effort to restore macro balances. Average inflation dropped in 2001 to 9.4 percent from 11.6 percent in 2000. However, prices of food, non-food items, and services performed differently. Food prices rose 15 percent, led by higher meat prices induced by shortages in domestics supply due to the dzud and increased Russian imports of Mongolian meat. Other factors driving up prices in 2001 were increases in electricity tariffs needed to make energy supply companies profitable, and higher user fees for water. Increases averaging 15 percent in public wages and pensions also contributed to rising prices. 2.27 The deficit decreased from 12.2 percent of GDP in 1999 to 6.8 percent of GDP in 2000 and is expected to have dropped even further in 2001 to 5.1 percent of GDP. While this adjustment is commendable, the quality of the adjustment remains weak. Principally increasing the fiscal burden on the private sector and disproportionately taxing large enterprises achieved the reduction in the deficit. Further efforts at sustainable fiscal consolidation would have to address deep-seated issues of expenditure rationalization. Government continues to rely on high taxes to restore fiscal balances 2.28 In 2001, general government revenues continued the remarkable increase begun prior to 1996. As a share of GDP, revenues increased by 2.4 percent to 36.7 percent in 2001, up from 33.6 percent in 2000. Tax revenues increased by 6.8 percent overall from 25 percent of GDP in 2000 to 26.7 percent. The rise in revenue was primarily due to an increase in business taxes and general improvements in tax collection. Indirect taxes buoyed by VAT increased in 2001 to 16.7 percent of GDP. Non-tax revenues decreased by 14.6 percent to 7 percent of GDP in 2001, down from 8.2 percent of GDP in 2000. Decreases in the revenues of the Erdenet copper and Gobi cashmere firms in 2001 due to the international slowdown were largely responsible for the drop in dividend payments. Capital revenues and grants increased to 0.7 percent of GDP in 2001, a 40 percent increase, mainly due to an increase in donor contributions to assist Mongolia with the drought and for dzud relief. Resolute and sustained structural reforms are needed to reduce expenditures 2.29 Expenditures increased in 2001 to 41.6 percent of GDP, up from 40.5 percent in 2000. Sharp increases in current expenditures and purchase of goods account for this increase. However, as a share of GDP expenditure on wages and salaries decreased to 8.1 percent from 8.2 percent in 2000. Expenditures on goods and services increased by 17.6 percent to 13.8 in 2001, up from 11.9 percent in 2000, largely reflecting increases in prices and an attempt by the government to adequately budget purchases of goods while reducing arrears. Interest payments dropped to 1.6 percent of GDP, reflecting a drop in domestically financed interest payments. Capital spending dropped substantially, to 9.8 percent of GDP from 10.4 percent. Foreign-financed capital expenditure dropped to 0.3 percent of GDP, and domestically financed capital expenditure to 3.4 percent from 4.2 percent of GDP. This drop in capital expenditure is worrying, as heavy investments are needed in the social sectors, especially health and education, to improve the overall quality of service delivery. Rationalization of expenditures continues to be difficult and ad-hoc expenditure cuts will not serve Mongolia well in the future. 15 Unstable external conditions argue for an acceleration in the domestic reform agenda 2.30 Mongolia continues to face difficult external conditions; it will have to strengthen and accelerate structural and institutional reforms if it is to build on its ongoing reform process and create an enabling environment for private sector activity. Despite poor weather and a world economic downturn, real GDP growth in 2001 held steady at 1.1 percent. The effects of poor weather and falling cashmere exports were eased by rising gold prices and increases in textile and meat and other industry exports. A 14 percent drop in oil prices accounted for the drop in imports. Gross capital formation dropped by 12.5 percent. 2.31 On the external front, the current account deficit, excluding official transfers, increased by 10.6 percent to 16.2 percent of GDP, up from 15.5 percent in 2000. The current account deficit, including official transfers, accounted for 6.9 percent of GDP in 2001, up from 5.3 percent in 2000. Mongolia's overall external performance has worsened over the last four years mainly as a result of adverse terms of trade shocks, heavy reliance of the economy on copper and gold exports, and poor weather. Gross international reserves increased to 17.3 weeks of imports ($210 million) from 15.7 weeks in 2000. Increased diversification of the economy as the government improves the regulatory framework would reduce reliance on copper as new export niches are created such as textiles, meat, and other light commodity exports. 2.32 Privatization receipts increased to $43m (0.6 percent of GDP) as the government revived the privatization agenda stalled under the previous government. Privatization receipts have been boosted by the sale of the Trade and Development Bank. Privatization of Gobi Enterprises, and the NIC petroleum consortium remain incomplete. 2.33 The stock of public and publicly guaranteed debt decreased to 84 percent of GDP ($865 million) in 2001 from 88.0 percent ($854m) in 2000. The net present value of external debt (excluding Russian debt) decreased to 56.2 percent of GDP in 2001 from 57.8 percent in 2000. A large share of the increase in debt came from concessional lending from the World Bank (WB) and Asian Development Bank (ADB). At end 2001, total public and publicly guaranteed debt amounted to US$865 million (84 percent of GDP), of which 61 percent is to multilateral creditors, 37 percent to bilateral creditors, and only 2.9 percent to commercial creditors. The largest multilateral creditor is the Asian Development Bank (36.9 percent of total nominal debt), followed by the World Bank (17.9 percent), and the International Monetary Fund (IMF) (5.4 percent). Of the official bilateral debt, most is owed to Japan (22.4 percent of total nominal debt), Germany (5.4 percent), and Russia (3.1 percent) and the annual debt service currently amounts to 4.3 percent of GDP, as the majority of debt has been contracted on highly concessional terms. Overall, excluding Russian debt, the buildup in the stock of external debt slowed and the maturity structure remains favorable. Mongolia would need to work hard to maintain this trend if future deficits are to be sustainable. 2.34 Bank lending to the private sector more than doubled from Tg 45.5 billion in 2000 to Tg 106 billion in 2001 and the share of non-performing loans in total loans decreased by over a third from Tg 23. 1billion in 2000 to Tg 9.2 billion in 2001, reflecting improved performance of the banking sector. There remained concerns, however, that this increase in bank lending may mask a weakening of the banking sector if prudential regulations and banking sector liquidity constraints are not enforced. Credit to non-banks, proxied by total performing loans, expanded by over 126 percent in 2001. 2.35 Overall, poor growth performance has kept unemployment levels high. Official labor statistics3 are not considered reliable, as there are multiple data sources with varying numbers. Unemployment was I. 3 National Statistics data 16 officially recorded at 4.6 percent at end-2000 or 4.8 percent of the labor force. These numbers may be severely understated, as the 2000 census records unemployment at over 17.5 percent. Education continues as the highest employer in the public sector, while agriculture leads private sector employment, followed by services, education, and manufacturing. MAINTAINING SUSTAINABLE FISCAL BALANCES FOR PROLONGED GROWTH:4 Low debt levels protect fiscal balances and promote growth 2.36 While reasonable debt levels have a positive effect on growth by encouraging much needed capital investments in particular, high debt levels are a hindrance to growth - reduce investments and increase interest payments. This section evaluates Mongolia's external debt sustainability by examining critical debt indicators within the context of a 10-year macroeconomic and balance of payments framework. The debt estimates and net present value (NPV) calculations are based on loan-by-loan data on public and publicly guaranteed debt as of December 31, 2001, provided by the Mongolian authorities. Mongolia's debt rose sharply in the second half of the decade 2.37 Before Mongolia launched its transition to a market economy in 1991, it maintained a very close trade and financial relationship with then the Soviet Union and former Council for Mutual Economic Assistance (CMEA) countries. This resulted in high external debt of more than 10 billion in transferable rubles to the Soviet Union, CMEA countries, and China at the end of 1990, with more than 98 percent owed to Soviet Union. The severe economic shock from the breakdown of traditional trading relationships with the Soviet Union and CMEA counties in the early 1990s, however, rapidly changed the composition and currency of Mongolia's external debt. External debt in convertible currency, which was virtually nonexistent before 1990, began to expand rapidly. The external debt stock in convertible currency, which stood at a mere 6percent of GDP in 1991, jumped to 62.4 percent of GDP in 1993. Since then, growth has been more subdued, with a cumulative increase of 22 percent during the last eight years. This is smaller than the average increase experienced by Russia, the Baltics, and the other states of the former Soviet Union.5 In nominal terms, however, the debt stock more than doubled since 1993, to pay for adjustment costs during transition, particularly heavy investments in the energy sector. During the 1990s, transferable ruble debts to CMEA countries and China were all settled, except the claim by Russia. Mongolia's debt is highly concessional 2.38 At end 2001 Mongolia's public and publicly guaranteed debt amounted to $865 million (84 percent of GDP), of which 61 percent is to multilateral creditors, 37 percent to bilateral creditors, and only 2.9 percent to commercial creditors. The largest multilateral creditor is the Asian Development Bank (36.9 percent of total nominal debt), followed by the World Bank (17.9 percent), and the IMF (5.4 percent). Of the official bilateral debt, most was owed to Japan (22.4 percent of nominal debt), Germany (5.4 percent), and Russia (3.1 percent). Annual debt service currently amounts to 4.3 percent of GDP, as the majority of debt has been contracted on highly concessional terms. Excluded from the analysis is the unresolved transferable ruble claim held by Russia relating to Mongolia's transactions with the former I. 4 Debt Sustainability analysis and Medium Term Fiscal Framework prepared jointly by the IMF and The World Bank. Analysis based on methodology applied within HIPIC framework. 5 External Debt/GDP ratio increased by 20. 1percent on average during 1995 to 2001 for these countries: source. IMF documents. Assessing Sustainability 17 Soviet Union. Given its high nominal value of $10.5 billion,6 ten times of Mongolia's current GDP, the settlement of this claim has potentially high impact on Mongolia's debt sustainability. (see Annex 2) 2.39 In terms of NPV, the total debt in 2001 amounts to $590 million, or 68 percent of nominal value. An important caveat applies. Since most of Mongolia's debt is denominated in currencies other than US dollar - SDR, Japanese yen, and euro - the US dollar value of nominal debt and NPV are sensitive to fluctuations in exchange rates. In addition, NPV is highly sensitive to changes in CIRRs. In particular, measured at 2000 CIRRs the NPV in 2001 would have been $545 million7 (Table 1). Assumptions of the Baseline Scenario 2.40 Macroeconomic projections used in this analysis are based on the steadfast implementation of sound macroeconomic policies and structural reforms, as agreed in the context of the PRGF program. Mongolian GDP is projected to grow 3 percent in 2002, 5 percent in 2003, and 6 percent thereafter. Export volume growth is also expected to grow 5 percent and 5.1 percent in 2002 and 2003, and 6 percent thereafter.8 Import volume growth is assumed to increase in line with the GDP growth rate. Given that average GDP growth since 1995 has been 3.1 percent, the base case assumption appears to be reasonable with accelerated structural reforms under the PRGF program. Given that the average export volume growth rate has been 9.3 percent since 1995, the assumed rate of 6 percent for export volume growth is conservative, reflecting the volatility of Mongolia's commodity exports. 2.41 The medium-term price projections are based on the most recent WEO assumptions for commodity prices (March 2002) and other relevant price projections. The projection shows general improvements in terms of trade up to 2007. From 2008, the terms of trade is assumed not to change. During the projection period, the togrog is assumed to remain unchanged in real effective terms. 2.42 On the financing side, a declining trend in official grant is projected to be compensated by increasing foreign direct investment (FDI) and other private capital inflows. While official grants are assumed to decrease from $95.5 million in 2001 to $73 million this gap is more than compensated by increases in FDI and an expanded inflow of worker remittance. FDI is projected to rise gradually from $43 million in 2001 to more than $70 million in 2010, and private unrequited capital inflow from $39 million to $50 million, thanks to increases in Mongolian workers abroad. 2.43 It is assumed that Mongolia will continue borrowings from multilateral and bilateral sources on concessional terms. New disbursements of project loans are assumed to increase at a much slower pace than the GDP growth-from $75 million in 2002 to $92 million in 2010. During the PRGF arrangement, additional financing is assumed to be provided by multilateral institutions to bridge the financing gap. It is assumed there will be no financing gap from 2005 on. Central government revenue excluding grants, which stood at 36.5 percent of GDP in 2001, is projected to be constant at 34.5 percent from 2002 on. (Relate it to fiscal sector)9 I. 6 This is based on the data submitted on September 17, 1997 by Russian authorities when Russia was in the process of participating the Paris club as a creditor. Total debt stock disbursed before January 1992 is $10,514.7 million or TRR 10,511.5 million. For example, $ CIRR is at 5.85 percent as of the end of 2001, as compared to 7.19 percent at the end of 2000. Copper, gold and cashmere exports volumes are assumed to grow at () percent respectively. In 2001 authorities estimated FDI at $63 million. Staff lowered its estimates because of the local practice of recording FDI on a comnutment basis and poor tracking of FDI repatriation. 18 Mongolia's debt burden is sustainable in the medium term 2.44 On the basis of the above assumptions and policies, the total debt stock as a percentage of GDP is projected to peak at 87 percent in 2003 and decline to 69 percent in 2010. Likewise, the NPV as percent of GDP is expected to decline from 57 percent in 2001 to a more comfortable 49 percent in 2010. The annual debt service burden, which stood at 4.3 percent of GDP at end 2001, is projected to continue to decline to .2.6 percent in 2010. Thus, baseline scenarios indicate that Mongolia's debt burden is sustainable in the medium to long term and Mongolia does not require debt rescheduling, barring major negative changes in baseline assumptions. Debt dynamics are most sensitive to GDP growth and export performance 2.45 Mongolia's balance of payments projection, however, is subject to considerable uncertainties, particularly with regard to highly volatile export prospects, future fiscal consolidation, the conditions of future external financing, and, most of all, the eventual settlement of the transferable ruble claim by Russia. Mongolia's export base is very narrow, largely dominated by commodities such as copper, gold, and cashmere, which are subject to sharp price and volume fluctuations. Conditions of new borrowing can also change in the medium term, forcing Mongolia to borrow on non-concessional terms. Thus, sensitivity analysis has been carried out under four different scenarios: (1) lower GDP growth rate and lower export volume growth rate; (2) deteriorating borrowing terms resulting in some non-concessional borrowing; (3) combination of (1) and (2), and, finally, (4) various terms of settlement of the Russian claim. (see annex 3 for discussion of other scenarios). 2.46 Under the combined scenario of lower GDP and export volume growth and unfavorable financing, debt dynamics worsen noticeably. The ratio of total debt stock to GDP peaks at 90.2 percent in 2004 and is expected to be 84.6 percent of GDP in 2010, 15.3 percent higher than the base case. Likewise, NPV will be 64 percent of GDP in 2010, 15 percent higher than the base case. Annual debt service burden is 3.9 percent of GDP, still bearable, but 1.3 percent higher than the base case. Overall, this scenario shows that debt dynamic is more sensitive to GDP and export volume growth than changes in financing conditions. Fiscal consolidation and accelerated structural fiscal reform required to sustain debt levels 2.47 Excluding Russian claim, Mongolia's external debt appears to be sustainable, although export-led growth is critical to continue to lower the debt service burden in coming years. At the policy level, structural reform, particularly fiscal consolidation, needs to continue to be aggressively pursued so as not to incur additional debt. Even after the PRGF program, contracting loans on non-concessional terms should be avoided as much as possible. This strategy will require expansion of FDI as a foreign capital source, which, in turn, will require structural reforms to promote private sector-led growth, including more aggressive privatization. As about half of Mongolia's loans are centered on the energy sector, the government needs to make sure that projects in these sectors are selected on a priority basis and are implemented efficiently, while adopting measures to promote proper functioning of price mechanisms in the energy sector. Last but not least, prudential macroeconomic policies, including a flexible exchange rate that adequately reflects Mongolia's external competitiveness needs to be pursued. Medium Term Fiscal Framework 2.48 Over the medium term, overall general government balances will need to fall comfortably beneath 5 percent of GDP if Mongolia is to guarantee sustained fiscal consolidation and create an enabling environment for private sector growth. Achieving fiscal consolidation will hinge importantly on 19 addressing the immediate sources of fiscal pressure identified above and more broadly reassessing overall existing public expenditure commitments in light of the new role of the government. 2.49 The high case medium term framework below assumes a modest recovery of GDP growth in 2002 of 3 percent following the low growth of about 1 percent achieved during 2000 -2001. Growth is expected to gradually increase reaching a level of about 6 percent by the year 2005 and stabilizing thereafter. The improvement in growth would be associated with an increase in investment levels especially after 2005 as current expenditures drop. 2.50 Successful fiscal consolidation for Mongolia should result in: low fiscal deficits and debt: effective demand management, by retaining sufficient flexibility to respond in an appropriate way to domestic and external imbalances, particularly weather and terms of trade shocks: and, maintaining reasonable and stable tax rates. 2.51 While the framework presented below rests on the reduction in the overall wage bill, containment of transfers and a reduction in overall taxes, the institutional challenges associated with slowing spending growth discussed in the sections above, illustrate that restoring sustained fiscal balances in Mongolia can be achieved only as part of a comprehensive effort to increase overall performance of the public sector. Table 2-3: Medium Term Macroeconomic Framework GDP Growth 1.1 1.1 30 5.0 6.0 6.0 60 6.0 6.0 6.0 6. Total revenue and grants 33.6 38.0 36.5 36.5 365 36.5 36.3 36.2 36.2 36.2 36. Current revenue 33.2 37.2 35.8 35.8 35.8 35.8 35.7 35 6 35.6 35.6 35 Tax revenue and social security contributions 25.0 290 27.8 27.9 28.0 28.1 28.1 28.1 28.1 28.1 28.1 Nontax revenue 8.2 8.1 8.1 8.0 7.8 7.7 7.6 '7.5 7.5 7.5 7.5 Capital revenue and grants 0.5 0.8 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 Total expenditure and net lending 40.5 43.3 43.2 42.6 41.9 41.4 40.9 40.3 40.0 39.6 39.4 Current expenditure 30.1 32.4 32.4 31.9 31.1 30.6 30.0 294 28.8 28.4 28.2 Wages and salaries 8.2 8.1 8.0 7.8 7.6 7.5 7.3 7.2 7.1 7.0 70 Transfers 8.1 7.9 80 7.8 74 7.3 7.1 7.0 7.0 7.0 70 Interest payments 1.7 1.5 1.7 1.3 1.3 1.1 1.1 1.0 0.9 0.9 0.8 Capital expenditure and net lending 10.4 10.9 10.7 10.7 10.8 10.8 10.9 10.9 112 11.2 11.2 Current balance 3.1 4.7 3.4 3.9 4.7 5.1 5.6 6.2 6.8 7.2 7.4 Overall balance (incl. grants) -6 8 -5.3 -6.6 -6.1 -5.5 -5.0 -4.6 -4.1 -3 7 -3 3 -3.1 Financing 6.8 5.9 6.6 6.1 5.5 5.0 4.6 4.1 3.7 3.3 3.1 Foreign Financing (net) 6.4 6.4 5.5 5.6 5.2 4.6 45 4.1 3.7 3.3 3.1 Amortization 0.9 1.7 1.5 1.0 1. 12 1.1 1.1 1.0 0.9 0.8 Pnvatization receipts 0.7 0.5 0.9 1 0 0.3 0.4 0 1 0.0 0.0 0.00 0.00 Memorandum items: __I Nominal GDP 1,045 1,131 1,23 1,361 1,514 1,685 1,876 2,088 2,324 2,587 2,879 Sources Ministry of Finance, and Fund staff estimates and projections. Conclusion 2.52 To restore growth Mongolia will need to make a sustained commitment to significant expenditure reductions by redefining the role of the public sector in the economy. The impact of such a policy on growth could be substantial, as large cuts in spending will reduce growth in the size of the public debt and positively affect interest rates, increase consumption as the fiscal burden on private sector and businesses is decreased and improve labor market conditions. Lower production costs will also increase competitiveness of the economy. 20 2.53 The benefits of undertaking these adjustments are large, as they are essential to achieving the government's primary challenge of stable growth by ensuring overall fiscal stability and -continuing poverty reduction. Without a medium-term fiscal and institutional mechanism to ensure that policy clear performance targets achieving sustained fiscal stability discipline decisions and poverty reduction will elude Mongolia in the medium run. 21 Table 2-4: Selected Economic Indicators 1993 1994 1995 1996 1997 1998 1999 2000 est (Percent change) Real GDP -3 21 6.3 2.4 4 3.5 3.2 1.1 Consumer prices (period average) 2684 87.6 56.8 46.8 36.6 94 7.6 11.6 Consumer prices (end period) 183 66.3 53.1 446 20.5 6 10 8 1 (In percent of GDP) (In percent of GDP) General government revenue and grants 1/ 33.3 30.4 33.8 27.8 29 3 27.6 27.2 33.6 General government expenditure and net lending 49 4 43 4 40 4 36.0 37.9 41.9 39.4 40 5 Overall balance -16.1 -13.0 -6.7 -8.2 -8.6 -14.3 -12.2 -6.8 Financing 1/ 16.1 13.0 67 8.2 8.6 14.3 12.2 6.8 Government debt 1/ 4/ 85.3 102.2 95.8 Broad money 32.9 25.8 198 8.8 31.7 17.5 Reserve money 36.2 26.2 135 51 8 19.4 (In millions of US dollars, unless otherwise indicated) Current account balance, including official transfers 269 -36.9 75.2 -75.5 -57.1 -68 3 (in percent of GDP) 2.2 -3.1 7.1 -7.8 -6.3 -7.0 Trade balance -3 -87 31 -120 -113 -150 (in percent of GDP) -0.3 -7.4 3.0 -12.4 -12.5 -15.5 Exports, fob 365.8 367 486 423 569 462 454 537 (Percent change) 841 -12.8 34.5 -18.8 -1.7 18.2 Imports, cif -374.5 -354.9 489 511 538 582 567 687 (Percent change) 12.8 4.5 5.4 8.2 -2.6 21.2 Gross official international reserves (end-penod) 65 92 115 98 137 125 157 191 (in weeks of next year /projected imports c.i.f) 12 9 12 11 12 14 Memorandum Items Public and publicly guaranteed external debt 2/ 374 473.7 503 8 5422 604.8 753.1 850.0 854.0 (In percent of GDP) 66.7 68 7 52.8 54.0 63 3 77.4 93.9 88 0 Debt service 2/ 419 66.7 64.2 56.6 41.0 39.3 41.3 31.5 In percent of exports of goods & services 5/ 16.4 11.0 9.0 55 73 9.3 60 Exchange rates and trade prices Togrogs per US dollar (end of period) 395 413 473 694 813 902 1072 1097 Togrogs per US dollar (period average) 449 548 790 841 1022 1077 NEER, end-penod (1995=100) 96 69 63 69 61 59 REER, end-penod (1995=100) 108 106 113 118 108 108 Terms of trade (percent change) 15 -23 3 -12 -4 7 Nominal GDP (billion togrogs) II 166 283 550 647 833 817 925 1045 Nominal GDP (million US dollars) 1/ 561 686 957 1006 955 972 906 970 Sources: Mongolian authorities and Fund staff estimates 1/ Data for 1993 used from the Table 1 and 13, Mongolia Country Economic Memorandum. Report No 16749, August 18, 1997, data between 1994-1997 from the Tables 1,16, 17, and 18, the IMF Staff Country Report No. 00/26, March 2000 , Mongolia: Statistical Annex, and data between 1998-2002 is adjusted according to Table 2, the IMF Document, December 17, 2001, Letter to the Minister of MOFE 2/ Excludes Russia's unresolved 10.6 billion transferable ruble claim. Table I1, Mongolia Country Economic Memorandum. Report No. 16749, August 18, 1997, Tables I and 43, the IMF Staff Country Report No. 00/26, March 2000 and Table 2, the IMF Document, December 17, 2001 22 POVERTY AND VULNERABILITY IN MONGOLIA Protecting the vulnerable during the transition has been difficult 2.54 Reducing poverty is the single most important development challenge facing Mongolia. During the past decade, the standard of living of the average Mongolian has worsened. In Soviet times, poverty was not widespread in Mongolia, as most FSU systems supported generous systems of social transfers. The collapse of the Soviet Union, the breakdown of monetary and trade relations among the COMECON states, and recent huge terms of trade shocks, poor weather, and the east Asian crisis have all taken a toll on the economy. Over a third of the population lives below the poverty line and inequality is rising 2.55 The 1998 Living Standards Measurement Survey (LSMS) showed over a third of the population living below the poverty line, and worsening macroeconomic conditions and two successive droughts suggests an increase in poverty in Mongolia between 1998 and 2001. Analysis of poverty trends in Mongolia is limited by the lack of available information on the poverty levels at the beginning of the transition. Only two data points exist 1995 and 1998, These do not capture the severe macro-economic shocks experienced between 1998 and 2000 either, however this date provides a general indication of the level of poverty in Mongolia. 2.56 There have been small gains in poverty reduction between 1995 and 1998. While the poverty headcount index remained largely unchanged over 1995 (36.1 percent) to 1998 (35.6 percent) the Gini coefficient rose from 31 to 35 percent. 10 This period was one of strong macroeconomic stabilization and growth, and GDP growth averaged over 3.5 percent. The poverty headcount and Gini data suggest growth benefits were distributed unequally among the population. Poverty dropped slightly in UB and the rural areas while other urban areas, aimags and soum centers, witnessed an increase in poverty. The LSMS results suggest economic growth had a beneficial impact on the rural areas while the emergence of a market economy Ulaanbaatar city benefited the urban city dwellers. Table 2-5: Mongolia: Poverty Indicators: 1995-98 Jrban 38.5 39.4 12.2 13.9 5.7 7.1 Ulaanbaatar city 35.1 34.1 10.4 13 4.5 7.4 Rural 33.1 32.6 8.9 9.8 3.6 4.4 Total 36.3 35.6 10.9 11.7 4.8 5.6 Source: Mongolia: Living Standards Measurement Survey, 1998 2.57 Household income sources and poverty are heavily correlated in Mongolia. Households dependent on public wage income, pensions, or other cash transfers are more likely to be poor than households who depend on livestock herding or crop production. The elderly, the disabled, and street children are more likely to be poor. In the rural areas livestock ownership is heavily correlated with poverty levels. Households with less than 200 head of livestock are more likely to be poor. I. 10 The two surveys are not strictly comparable. 23 2.58 Unemployment is another determinant of poverty in Mongolia. The unemployed in urban areas are 2.7 times more likely to be poor than the employed. The geographical location of the unemployed is also a determinant of the severity of poverty. Urban unemployed are 1.7 times more likely to be acutely poor than in the rural areas. 2.59 The level of human capital development is another determinant of poverty in Mongolia. Individuals with no secondary school education are 2.3 times more likely to be poor than those with vocational or higher education. Unemployment rates are over five times higher among the urban uneducated and among the rural uneducated. Overall literacy indicators have not performed well in Mongolia: over 37.7 percent of the population is illiterate. Table 2-6: Employment Status of Working Age Population by Educational Level, (1998, in percent) Undch6dr- imiY'.,ec6dar,*VocaioniJ -e,Trnaryoi. ducanded~ E'dLian'it'n. ,~~ EducaLion ,Edua~n Edbcab86, i Employed 26.7 37.1 38.0 51.6 67.5 Z Unemployed 23.3 22.9 28.3 23.0 11.3 Employed 68.3 79.9 77.0 78.3 80.1 _ Unemployed 4.9 7.3 11.5 12.6 9.6 Source: Living Standards Measurement Survey, 1998, National Statistical Office, Mongolia Increasing vulnerability and weakening kinship ties 2.60 In addition to income and economic insecurity captured by the LSMS, the 2000 Participatory Living Standards Assessment (PLSA) brought to the fore other problems of rising environmental, physical, and social insecurity and vulnerability faced by Mongolian households. Drought, harsh winters, growing over-grazing of pastureland, and poor natural resource management has weakened the capacity of households to manage shocks. Alcohol abuse and domestic violence emerged from the PLSA as major threats to the physical security of women and children and have led to weakening social cohesion and increased vulnerability. 2.61 Weakening kinship ties have increased the level of social insecurity. Inter-household transfers and kinship support were a major source of survival for many poor rural families. The increasing rise in semi-commercial forms of intra-household transfers has weakened these linkages, and the reluctance of Mongolians to enter into labor contract agreements with family members has further reduced the income of many poor rural households. Conclusion 2.62 Mongolia has not been able to reduce poverty and aggregate social indicators remain poor. No country has reduced poverty without continuing positive economic growth," and government actions to reduce poverty need to promote growth and be pro-poor. In Mongolia, this requires first ensuring macroeconomic stability, then addressing other structural impediments to growth and poverty reduction. Structural issues that require government attention to reduce poverty and sustain growth include public sector employment policies, the systems of cash transfers, public sector enterprise governance issues, access to education, health, environmental policies to safeguard rural incomes, and institutional structures, I. " Ravallion and Chen, 1996, (WDR 2000) 24 systems, and processes that govern public sector activity and ensure the provision of quality services to the population. 3. CURRENT CHALLENGES: ADDRESSING THE MAIN SOURCES OF FISCAL PRESSURE INTRODUCTION 3.1 The government's efforts to transform the public sector into a more market-oriented institution with limited reach and scope have met with mixed success. Mongolia's public sector continues to grow. While adding expenditures to reflect its new role in the market economy-health, education, and safety nets-the government remains engaged in too many commercial activities. In addition, new administrative units and expenditures categories are being created without appropriate attention to the skills or resources required to maintain these levels of expenditure. Weak expenditure controls are particularly threatening in three areas: government pay and employment policies, cash transfer programs, quasi-fiscal activities of both SOE's, and state owned commercial banks. 3.2 Consolidated government expenditure and net lending has risen steadily since 1995. Total government expenditure in 2000 increased to over 40.5 percent of GDP and is expected to remain above 41 percent in 2001. Compared with other transition economies this is high (figure 3.1). Unsustainably large spending is exerting upward pressure on debt, which has doubled over 1995-2000 from 41.1 percent to 88 percent of GDP. The continuous involvement of the public sector in fiscal-cum-quasi fiscal expenditures in the second half of the decade also weakened the stabilization effort though successful revenue mobilization efforts between 1996 and 2000 have provided the government with some respite, an immediate and sustained reduction in primary expenditures will be necessary to improve Mongolia's overall fiscal balances. Wage increases can be contemplated in the context of a comprehensive civil service reform program 3.3 One main source of fiscal pressure on the budget is the wage bill, which rose 60 percent (from 4.9 to 8.2 percent of GDP) over 1995-2000. The main problem in the Mongolian civil service today is not the average wage level, but its functional composition, and the lack of a tenured professional civil service. Short-term fiscal pressures from an increasing wage bill need to be addressed to create room for medium- term civil service reform. Reforming the cash transfer system to increase labor force participation 3.4 As in other transition economies, Mongolia introduced new institutions to protect the population from economic hardships created by the transition in the early 1990s. Social safety net programs included a full social insurance system (pensions and short-term income replacement benefits), employment assistance services, and social assistance benefits. These programs have been overly generous and poorly targeted, and overextended the government. While minimal savings will result from reforms in the cash transfer and social assistance systems in the short run, immediate action is needed to set the process of reform in motion to protect medium-term fiscal stability. 26 Divesting from commercial activities and mainstreaming quasi-fiscal activities into budget 3.5 The current government inherited a legacy of SOEs in the banking, trade, energy, restaurant, mining, and transport sectors. The portfolio of state-owned enterprises merits scrutiny, as there is no systematic process for identifying and recording government liabilities arising from SOE activities. These liabilities need to be recognized and processes put in place to address them. A banking sector crisis in 1998, which required increased expenditure equal to 3 percent of GDP, was a particularly telling example of the pressures SOEs and the banking sector can put on government. 3.6 Government spending on operations and material was over 28 percent of its total spending in 2000, and is expected to increase to 31 percent in 2001. Rising expenditures on heating and transportation are primarily responsible, but implicit government subsidies also account for an important share of the increase. 80.0- 70.0 60.0 51.8 50.0 38.6 a~40.0. 30.0- 20.0- 10.0 1 1 1 0.0 RU 0 0 . O< a 1993 0 1998 Source: Ministry of Finance and Economy, 2001 3.7 The persistently large size of government indicates that reductions in the size of the public sector at the onset of transition were the result of favorable international economic conditions rather than systematic policy decisions to reassess the role of the government during the transition to a market economy. 3.8 Medium-term economic performance will depend importantly on the choices made by the government in public administration and public finance reform. Improving efficiency and adopting policies that support the transition will require that the government clearly identify its objectives, institutional and administrative structures, and resource availabilities and requirements. Restoring fiscal discipline and budget comprehensiveness is essential for long term reform 3.9 In the short term, however there is a pressing need to ensure that fiscal balances are brought into line and made sustainable to create room for more fundamental structural reforms. The government will need to undertake measures to adjust its fiscal balances to create an environment that will allow market- oriented public expenditure management policies that will sustain a smooth transition to its longer-term 27 growth and deficit objectives. To protect medium-term fiscal balances a number of basic issues will need to be addressed, including civil service wages, pensions, and SOE policies. 3.10 The primary short-term challenge for Mongolia's fiscal policy is to reduce fiscal pressures, which could threaten long run success of the reform program, without undermining long-term public sector management reform, to ensure that future deficits are sustainable, protect the poor from further burdens of fiscal austerity, and promote pro-poor growth. The rest of this chapter discusses expenditure policies the government could consider to achieve the fiscal sustainability needed to create an enabling environment for increased private sector employment opportunities and improved public sector performance. Section A discusses civil service pay and employment, Section B discusses government cash transfer programs, and Section C discusses the impact of government contingent liabilities, particularly SOEs and other quasi-fiscal activities, on the size of government. Section D draws on recommendations from previous sections to discuss alternative sustainable fiscal policy paths the government could consider in the design of its medium-term expenditure framework. CIviL SERVICE PAY AND EMPLOYMENT POLICIES 3.11 Sustained fiscal discipline is a prerequisite to medium-to-long-term structural and institutional reform. The government will need to adopt preliminary fiscal measures to ensure the economy is on a sustainable growth path that does not rely on unduly taxing the private sector and stalling growth to finance undisciplined policy priorities. Short-term wage containment measures could provide the government with the desired headroom. 3.12 The composition, pay policies, and patterns of the Mongolian civil service do not reflect the roles and responsibilities of a market economy public sector, and are a source of substantial pressure on the budget. A wide range of potential short-term pay and employment reduction policies are available, including wage and hiring freezes, outsourcing non-core government functions, voluntary retirement schemes, enforcing hiring bonus thresholds, and public sector disengagement from productive activities. While some of these are temporary in nature and effect, these policies essentially allow the government to discuss and build consensus for long-term structural civil service reform. Agreement on a civil service reform path needs to result from discussion in government and among other key stakeholders about the appropriate role and structure of government required to deliver quality services to the population. Mongolia's PSMFL has created room for such a debate (Box 3.1) 3.13 Public spending on personnel increased dramatically in the second part of the decade and weighs heavily on the government's finances. Between 1995 and 2000, the wage bill went from 6.3 to 8.2 percent of GDP. In 2000 Mongolia's civil service employed 132, 400 (down from 154, 400 in 1995), 17 percent of the employed population in Mongolia and 5.3 percent of the total population. Overall employment increased by 2.7 percent between 1999 and 2000 for the first time since 1994. In 2000 the wage bill accounted for 8.2 percent of GDP. 3.14 While the share of government wages to GDP prior to 1997 was in line with other small open transition economies and the regional average, between 1997 and 2000 the share of government wage expenditure rose substantially, and is now higher than in most other transition economies (Figure 3.2). This rise was largely due to systematic wage policy increases needed to adjust for inflation but also in 2000 to increases in the overall size of the civil service. . While revenue mobilization constraints may have contributed to keep expenditure levels low in some transition economies, successful structural reform policies in other better performing transition/accession countries led to a reduction in revenue mobilization as public sector consumption was reduced. 28 h m e It'è.id:jtiö arth a the 'n norý fier a I rue jtbr- en Js 4' i ea ti Pri'cateMå re,u ce 1 istraeik pritiae, Forerdsnne, pocyseisule is we,ter,, a epobcy ofgvenenDs Pu .s~iec"'torgeJMplomet4 prcegntIichysfare, mpp~Øeoeh enIyetrIsconsistent 'g wa. wl' e-reurewtspothnd éåalisita' rormpefdoeanori ritend civiltsenice comenint loperation Htihey are sivet'i cvimpromise t operabilitvco these new co s letafn wh'.policy o e isrq ie. A i ~éd irmmeeéimadel st-iiy,1n[ésfijp-itbe s,wl E arefliciè t sä èffabiè lwe'In" rice ee ie put ec on a sustahrn i requLre attention to . the ahigbagebill This is fnot tobe eér"d k 'ie eredsi dtfr WaIe andf Éi ceingve ens e pr to 'aces t e board blury t "itrument s at"å t elst a t- shor termsaings rpecra nve y key· dcis nisihåt bd e o rg reå ste.:ent componi.se real cngi tsé riéenTSoris Te Sd a p i f el ourecutostraegic pnoritiToryersonnel polica indhmaeralisc governiens' em»iagenl of a hò6 d i;effordable oricobpadle loithèashito f s;,riti pe ownard povey ted si on.CUueid .pubIiý set rrern'plby,h"ent" prauis and 'polJLieý'ijrr9. ncons" et< t~hawl'btrciiex upr and4" *acitte- the pro,'arice on?e ntrd'public Tecocmpldnent th'e I}~i.Lmrhniecvlrie rom to alig bic sector 'statfig Wýith ,policy objeefillisis'requife "A' tirst step iwithis:-dLreup sid',,Jvelpah straiegy, for,'ref&Sk1o! nludelt undert fue" fioinai'review ote cV~ ver and,bi~ildlia py and iemp0"ment mckéi 5- Z~ de .tŽ .'j' ' 1..'...' *~ teplTvient' Jl"fet& évdf~rIe.T, iref64 ý9sédd'i iiiecrI .Tii is -e'! Y" fi'Tisi;the agpeci of pertlormanice to l1whi&hc' i e' c Tifs-s"herethebaartig'ifpersonnel numbetsandLskiIls and thehU rag 00 ~ o'qaiay enics"at kat cost- sholJuld take.place. In,tus, respea anyU'1 I,........*c" Srvie soulib~ainedat.Increasing tranisparenicy in hinngthengt-mýenuiVes. The SCCIshould. ensure.a policy,' :,Ithr poesion an ratgal .s s;,d vaade, for'senion, Iåaagnai~ asd dWpåyneonhradsi" and, petrmne 'pay systemrs s'hould', p ~ t 'J t ti I;. - ' i 19I * ,,:. 29 Employment and Wage Bill Trends Mongolia's wage bill as a share of GDP is higher than in most transition economies Tajikistan Kyrgyz Republic Kazakhstan Bulgaria Romania Moldova Transition econom lea Belarus Asia Mongolia Macedonia Ukraine 0 1 2 3 4 5 6 7 8 9 10 Figure 3-3: Wages and Salaries as a Share of Figure 3-4:Wages and Salaries as a Share of GDP Total Expenditure 1993-2000 9 25% - 8 20%, 7 * 6 15% -10% 3 r& G5% 0 _ _ i_ _ _I_ _ 1993 1994 1995 1996 1997 1998 1999 2000 1995 1996 1997 1998 1999 2000 Source World Bank Sima database, World Development Indicators, and Ministry of Finance and Economy, 2001 Central and local government wages rose with central government wages doubling in five years 3.15 Increases in central government wages account for the large increase in the overall wage bill between 1995 and 2000, as they have been rising faster than local government wages. Central government wages in 1995 were 1.7 percent of GDP while local government wages accounted for 3.2 percent of the total wage bill. By 2000, however, the central government share of the wage bill to GDP had almost doubled to 3.3 percent while the local government share increased by only 53 percent to 4.9. 3.16 A comprehensive 1995 Law on Government Service (LGS) regulates the Mongolian civil service. It classifies the civil service into four categories: (a) political service includes ministers, governors and parliamentarians; (b) administrative service includes central government staff; (c) special service includes defense, security court judges and prosecutors, and central bank managerial staff, and (d) support service includes governing, executive and supporting posts in education (including teachers), science, health, culture, and administrative posts in SOEs. 30 3.17 The most important factor in Mongolia's high wage bill is that public employment often has little to do with providing efficient pro-poor public goods and services, but is used as a source of employment of last resort. Many governments in the former transition economies continue to use public employment as a way of reducing unemployment, redistributing incomes, or distributing political favors. Changes in Employment Levels Varying trends across employment categories 3.18 In absolute terms, government employment decreased by over 12 percent between 1994 and 2000. However, the numbers of political and special assignment posts have increased by 18.5 and 0.5 percent, while the number of state administration and state support service posts dropped by 33 and 13 percent respectively. However, between 1999 and 2000 there was a 2.7 percent increase in the total number of civil servants. Special assignment posts were reduced by about 13 percent and state administrative positions were reduced by 6 percent. State service employment increased by 7.4 percent between 1999 and 20002. Table 3-1: Total Number of Civil Servants, by Assignments, (000s) Political assignment 2.2 2.3 2.6 2.6 2.7 2.6 State admimstration 11.6 7.3 8 8.1 8.3 7.8 Special assignment 20.6 20.6 23.9 24.3 23.7 20.7 State support service 119.7 111.3 106.5 102.5 97.1 104.3 I'otal 154.1 141.5 141 137.5 131.8 135.4 Source: Ministry of Finance and Economy, 2000 Table 3-2: Share of Public Sector in Economically Active Population 1992 1995 1999 200C Population of working age 1134.6 1186.7 1279.3 1374.4 Economically active population 860 812.7 853.4 847.6 o/w Employed 806 767.6 813.6 809 Unemployed 54 45.1 39.8 38.6 Percent of public sector* 16.5 17.1 Labor force participation rate 75.8 68.5 66.7 62.9 Employment rate 71 64.7 63.6 60 Unemployment rate 6.3 5.5 4.7 4.6 Source: National Statistics Office 2000: *Public sector includes public enterprises 3.19 Analysis by employment category shows four broad trends in the Mongolian civil service: (i) Employer of last resort: During the early years of the transition between 1994 and 1996, with growth trends in the economy, the state was able to shed labor in the less productive State support and special assignment categories of the civil service. With the economic downturn, these trends were reversed as the state moved in to act as employer of last resort and employment increased in all categories of the civil service but the support service. Support service employment increased by over 8 percent in 2000. I. 12 Data limitations and difficulty reconciling different data sets makes long-term comparisons difficult. Data on civil service used in this section are all official data received from different government agencies, including the Ministry of Finance and Economy, the General Services Council, and the Cabinet Secretariat. 31 (ii) Continued involvement in productive activities: The state continues to be involved in productive activities. A weakening privatization program has resulted in continued state involvement in productive sectors. While the public sector is not directly responsible for all SOE salaries it continues to be involved in productive activities such as hotels, restaurant and travel activities, where it picks up the wage bill. The average monthly wage at SOEs is about Tg 98,000, almost twice as much as the public sector average wages of Tg 55,000 in 2000. (iii) Political patronage: Mongolia has had five changes in government since 1996, each accompanied by increases in the number of political and administrative civil servants as managers tended to bring additional personnel with them into the civil service. The staff of the Ministry of Finance and Economy, for example, increased by over 75 percent following the change in government in mid-2000. Frequent changes in job assignments and renewals of civil service staff deplete capacity of the civil service as there is lack of continuity and loss of experience. (iv) The civil service law allows government employees to hold multiple posts, and unofficial estimates put the number of double postings at about 5,000. Thus, even though the Ministry of Finance and Economy and the Cabinet Secretariat control the number of employees, they cannot control the creation of posts in budget entities. Cumbersome Government Administrative Arrangements 3.20 Mongolia's four-layer administrative structure is another source of pressure on civil service numbers. With 21 aimags, over 300 soums and districts, and over 1,681 bags, the creation of new organizations at the center leads to an increase in the number of employees or posts created by a multiple of at least 50. Civil Service Pay and Grading Systems 3.21 The scale and rate of remuneration due to government employees is regulated by the LGS and revised by the cabinet and parliament annually rUl . 1ngdlx6" rin~O 1997 1999 2001(est) Category Scales Posts Minimum Maximum Minimum Maximum Minimum Maximum Political service 25 13 36,570 54,648 55,313 82,655 55,313 90,520 Administrative 25 13 28,290 52,758 42,789 79,525 45,289 82,025 service Special service 12 18 6,772 52,578 0,493 79,525 40,493 81,201 Support service 25(7) 14 17,500 39,700 33,000 77,000 P3,000 .77000 compression ratios 3.12 2.50 2.74 Source: Mongolia, Cabinet Secretariat, 2000 High levels of inflation in the early 1990s eroded public sector wages 3.22 Civil service employees experienced drastic drops in real wages between 1990 and 1994 as increases in public sector wages fell behind inflation. Difficulties in maintaining qualified staff in the public sector and a favorable revenue and growth environment in the private sector pushed the government to begin increasing wages in 1995. 32 3.23 Wage adjustment policies were poorly structured. Increases were either across the board or, because of the increasing number of potentially poor, skewed in favor of lower level staff. Thus, the differential between the highest paid and lowest paid staff has narrowed since 1997, and wages of the low-level staff have caught up with and sometimes exceeded those of their counterparts in the private sector. 3.24 Increases in salary grades and promotions are discretionary and are not subject to transparent and objective evaluations. In addition to overall salary adjustments, many budget entities have resorted to promotions or creating new posts to increase the base income of staff. This policy has led to a clustering of employees in mid-range civil service categories, further depressing the morale of mid-level staff and encouraging more exits to the private sector. Wage bonus are generous and not merit based 3.25 The level of allowances, rewards, and incentives in Mongolia is overly generous and not systematically merit-based. In addition to base salaries, bonus payments are regulated by law, and can amount to 40 percent of an employee's highest (in case of two jobs) base salary. Despite attempts to regulate bonuses, the actual practice is opaque, as government has set no objective criteria. These bonuses are used to supplement employee incomes and grant political favors. Accountability to the Finance Ministry is minimal. Agreement on the size of rewards is not reached during the budget process resulting in poor planning and a buildup of arrears. 3.26 Over 2.6 percent of expenditures in this category are implicit non-wage subsidies to government employees based on the LGS. Government employees are provided transportation, newspaper, telephone allowances, and travel allowances. Anecdotal evidence suggests that these extra spending perks equal over 100 percent of some political service employee salaries. 3.27 Implicit subsidies to civil servants are poorly targeted and provide indirect transfers to the private sector. The cost of these implicit transfers in the education sector alone is about 1.2 percent of GDP. These include tuition for civil servants children and free textbooks for primary and secondary schools. The share of tuition for civil servants children has increased from 0.1 percent of GDP in 2000 to an estimated 0.3 percent in the 2002 budget. Fragmentation in civil service financial and performance management 3.28 Management of the size of the wage bill for macro-fiscal objectives rests with the Ministry of Finance and Economy (MOFE). Discussions on the level of the wage bill is included in the overall budget negotiation process between individual budget entities and the MOFE. On the other hand decisions on hiring and firing staff is done at the budget entity level with implicit oversight from the State Services Council. And finally management of the civil service database is handled at the cabinet secretariat level. This complete fragmentation of civil service management has led to poor staffing decisions increases in the wage bill and contributed to depress the quality of service delivery. The draft PSMFL if adopted will resolve some of these problems Civil Service Reform Options and Objectives Building a civil service which delivers quality outputs at least cost 3.29 Mongolia cannot afford its wage bill, and civil service performance is low. Reducing the wage bill by shrinking the government is a necessity if fiscal balances are to improve. The twin challenge of retaining and attracting quality civil servants to the civil service while shedding unproductive employees 33 will be the key to any successful medium-term civil service reform effort. The PSMFL has particular implications for civil service reform that could be used to underpin the government's personnel and staffing reform agenda, while the Law on Government Service would also need to be strengthened further. Figure 3-5: Wage bill as a Share of Current Revenue Figure 3-6: Growth in Current revenue and Wage Bill 1993-2000 1994-2000 1000 30.0 - 90.0 25.0 . 80.0- Tndh Current revenue 700 2 0 600 4)2 500 S15.0- 40.0 W e bil 1O(1 - 300 20.0 5.0- 100 1994 1995 1996 1997 1998 1999 2000 1993 199 1995 1996 1997 1998 1999 200O Source: Ministry of Finance and Economy ,and staff calculations 2001 3.30 While comprehensive civil service reform could take many years to complete, in the short run the government will need to adopt measures to contain any further growth of the wage bill in order not to endanger macr6-stability or jeopardize the possibility of longer-run reform. The need to put fiscal policy on a sustainable basis requires that immediate attention be paid to the high wage bill. Actions to achieve this objective should focus on the quantitative size of the wage bill by reducing the overall numbers and posts in the civil service. Specifically, such measures in the short run could focus on public sector employees providing services of a commercial nature, such as farmers employed by schools to supply to food local schools. These across-the-board blunt instruments are primarily useful for short-term macro- stability objectives-and are unsustainable and ineffective if they are not accompanied by fundamental qualitative reforms. 3.31 It is important to note that; short-term rightsizing not followed by comprehensive reforms could be detrimental to the overall performance of the public sector. The skills needed to introduce the necessary efficiency savings to offset the effect of staff reductions may be weak or absent; the mandatory process is undertaken in isolation of a fuller review of cross-sectoral priorities and resource allocation needs; cuts will be made to central government policy-making and implementation entities, rather than to those sectors which are comparatively well staffed. Developing a well-defined strategy for civil service reform consistent with role of government 3.32 The long-run objective of civil service reform is to increase the overall performance and efficiency of the public sector, by creating a tenured, motivated, and qualified civil service. Developing a well-defined strategy for civil service reform, which has the support of key stakeholders, is a first step for sustainable civil service reform. The strategy should specify timetables and responsibilities. A successful 34 reform strategy needs to be formulated with care and should include measures to assess the role of the public sector and downsize the civil service, improve incentive structures, enhance civil service management and accountability at all levels, and improve the legal and regulatory framework for civil service reform. Rapid downsizing would endanger long run poverty reducing objectives 3.33 The government is facing considerable pressure to live up to its campaign promise of doubling wages and increasing pension coverage. However constrained by resources and the need to maintain sustainable fiscal balances the government is considering incisive reduction policies. A proposal to halve the civil service and double wages of remaining staff is currently under consideration by the Cabinet. The government's willingness to undertake deep-rooted public sector reform is encouraging. However, some caution is required as experience suggests that rapid civil service retrenchment could have potentially damaging consequences on public sector performance and poverty. These include: younger workers with high state tenure are more likely to leave reducing the overall quality of the civil service in the short run at least: public sector job leavers appear to suffer much higher levels of unemployment than the average for their area; those at risk of unemployment are particularly younger women, and older males; younger men with lower severance payments are more likely to be employed and higher severance payments appear to be used mainly for self-employment and retirement; seniority-weighted severance packages are likely to lead to the loss of younger, higher-tenure, more educated, and more able workers. 3.34 The government recently completed a civil service survey in order to better inform its policy decisions and help is being sought to analyze and undertake further analysis of the current composition of the civil service prior to the adoption of new measures. This path is encouraged as it will increase the transparency of the process and allow the government time to appropriately access and analyze the impacts of any reforms taken and their costs both in the short and medium term. Retrenchment is a means, whose end is the efficient provision of public services 3.35 There are benefits to a well-designed civil service retrenchment program: programs that use targeting on the basis of skills and age, use multiple methods of employment reduction, and use a combination of compensation packages that include enhancements of the safety net for assisting the reallocation of workers are much less likely to exhibit problems with re-hiring. While such programs with a multi-dimensional approach tend to be expensive, there are potentially large payoffs in productivity gains and they reduce adjustment costs from well-designed programs. Well-designed programs allow for the provision of income supports especially important for young women, and could provide of start-up capital for self-employment may reduce the unemployment of older males and females. The main message is: it is possible to achieve a targeted number of employment reductions, without unnecessary cost using a well-designed model, reliable data and consensus to modulate the level of severance payments; and design the structure of severance payments as to induce the desired distributions of workers who take advantage of the program. 3.36 The government's civil service reform proposal also includes rapid social sector privatization. Experience suggests while short-term expenditure reductions may be realized, in the long run these practices amount to pushing employment and other social sector expenditures off budget. It is widely recognized in Mongolia that private sector provision of social services is poor and the legal and regulatory framework undeveloped. The trade off between short term fiscal savings if any, low quality services and reduced access of privatized services to the most vulnerable segments of society would need to be examined very carefully before such policies are adopted. In addition, the expected savings from social sector privatization will need to be analyzed carefully and modeling undertaken based on reliable data discussed more broadly. 35 An assessment of the size, level and composition of the civil service is urgently needed 3.37 The first step in any successful public sector reform agenda is to undertake a review of the civil service and assign a government unit in charge of the rationalization of public sector employment. The government has already begun to undertake some of these measures with the civil service census and the creation of a joint working group under the Prime minister's secretariat and compromising of members of the Ministry of Finance and Economy, the Ministry of Social Welfare and Labor and the,State Service council. 3.38 Such a review will allow the government to complete reforms in intergovernmental fiscal relations with a better assessment of the skills mix needed to improve public sector performance as well as identify areas of overlap in government central and local structures. 3.39 If successful, the reform could achieve a number of objectives; build an a complete inventory of the civil service and enable the formulation of clear regulations for the creation of additional posts; differentiate agencies between administrative and commercial agencies facilitating the design of a rightsizing program; create opportunity for the review of roles and objectives of agencies and assist in establishing improved control and accountability mechanisms of budget managers. Preparing for long term reform will require some short term wage containment measures 3.40 Short-term wage bill containment measures, which the government could consider, include: freeze hiring, divest from commercial activities, and in the short run cap the wage bill at 8.2 percent of GDP. The government could make use of a wide range of short-term pay and employment measures while its longer-term civil service reform strategy is being developed. These increases should not substantially increase the wage or pension base. These include wage and hiring freezes, outsourcing non- core government functions, enforcing hiring bonus thresholds, and public sector disengagement from commercial activities. Targeted wage bonuses, and wage increases differentiated by sector or profession may also prove useful short-term tools. While some of these are temporary in nature and effect, they would give the government breathing room to discuss and build consensus for long-term structural civil service reform. 3.41 , Reducing the level of employment will be a key challenge for the government, but it is needed to improve fiscal sustainability. Without comprehensive civil service reform as part of an overall policy to improve public sector performance, the gains from temporary wage bill reductions will not assist the government in building a sustainable poverty reduction strategy underpinned by a professional civil service capable of delivering high quality services at least cost. A comprehensive reform process is needed to transform the size and structure of the civil service. There are instruments and mechanisms in place to support this reform. 36 In the long run building a professional civil service is required i~i f ~ox3- Dssouruir c ie men anc1s W8ii i ffresoce equirements and 'eiffecilin one l Ac n ara sal in~ct on, ."q T '.:~ ~ ~ i~9n~C~, ~ ~apubic~~ a aroois isbae ~~~~~~~~~~~N1 reieV,4 il,avdai 1:f~i~$ iro propriate,jrrso prcdes for udgment ure Souc: Barbara bas6R4adi Europe. bli ne rion Refor and -'eanAcce io in,Cenural and Ee Ban chia Pa .No. 466.2 0004. 4 n .ess A sequenced reform strategy needs to be developed very early on 3.42 Comprehensive civil service reform is a difficult challenge even for governments with strong political support. To be successful, therefore the government will need to develop a well-defined strategy for civil service reform, which has the support of key stakeholders. The strategy should specify timetables and responsibilities. 3.43 Substantial technical assistance to support the reform process would be critical to assist the government in designing a strategy. Also vital to the success of the reform is donor coordination and sustained involvement in the reform process as many components of the reform program will take time to become effective and for successful results to be apparent to all. Consequently, the institutional aspects of the reform will have to be nurtured over the long term and the reforms sequenced in a way that early successes are achieved to build more support for the overall process. Revisions to the Law on Government Service will provide a framework for reform 3.44 Discussions around the PSMFL reflect some political consensus on the need for reform and particularly the need to revise the civil service law. More agreement needs to be reached on the pace and depth of comprehensive performance improving civil service reform. The PSMFL contains some personnel management provisions. The roles and responsibilities for general managers appear to give them some scope to manage while placing sensible restraints on them. However the law is not clear on who determines level of employees. While General managers determine the "salary fund". Political involvement in civil service promotions and rewards and bonuses should be discouraged 3.45 The authorities set out under the draft PSMIFL have the potential to make the State Service Council (SSC) a key part of operating a less politically driven appointment and dismissal system for top managers. It gives the SSC a regulatory role but leaves general managers room to manage within the laws 37 and regulations. There is a question though about the wording of the authority to "control the implementation of policy on human -sources..." (45.1.1). This would be better expressed as "promulgate policies on human resources within the authorities of the SSC" as the way it is currently expressed makes it unclear how much scope they have to be involved in the operations of HR management. The usefulness of keeping a register of potential applicants can be questioned. 3.46 While the draft PSMFL appears to allow for performance- base6tay, the law also suggests the State Service Council can put forth proposals to Portfolio Ministers for civil servant rewards. There is no point in limiting political interference in appointments but allowing so much political influence in rewarding public servants? Some checks should be put on base pay and bonuses by requiring the publication of the numbers of employees in bands that account for all benefits including salary, bonuses and payments in kind. The State Services Council could issues guidelines on pay levels and General Managers could be required to report against as part of reporting on their performance agreements. 3.47 Discussions on amending the Law on Government Service to complement the PSMFL are underway and care should be taken to ensure consistency between the two laws. In addition, the opportunity at hand to undertake these reforms in connection with the PSMFL should not be lost. The sense of urgency provided by the need to restore macro-economic balances should be exploited to launch a reform process, which aims at improving civil service performance, and engage in comprehensive reforms with the help of other external stakeholders. A comprehensive review of the civil service law should address broad issues of compensation, hiring and firing schemes, bonus systems, and training and broad human resource management. Sequencing staffing arrangements and the PSMFL 3.48 The PSMFL requires that budget governors and portfolio ministers enter into contracts that state the terms and conditions under which services are to be provided. Budget governors are subsequently free to undertake all relevant staffing decisions. The balance between local government as agents of central government and as autonomous entities will have a substantial impact on the approach to contracting. Credible contracting agreements will also depend on the information available. The parties to contracts need information on contract costs, both forecast and actual, and the skills mix required to deliver outputs. This implies costing information relevant to the letting and monitoring of contracts-and a need for good contract accounting. Relevant cost and contract accounting skills do not seem to be available, implying a need for training and recruitment. These skill deficiencies are particularly apparent at lower levels of government. 3.49 Staffing policy following a civil service functional review can be better implemented in Mongolia's current situation, where costs of outputs have not been established and staff skills not been certified. There is little to suggest (outside pilot agencies) that civil servants are ready to set contracts, enforce their terms, and judge and be judged by their performance-or that politicians would be prepared to abide by such contracts. In addition, appropriation by output is a concept full of methodological difficulty. 3.50 It seems likely that appropriations will not be output-based in the near future and intermediate solutions need to be designed to fill in the gaps while capacity is being built to implement contract-based appropriations. There are associated difficulties arising from the need to cost outputs. True output-based budgeting requires both costs and outputs to be measured for each organizational unit of government. Normally there are plenty of output measures available. The problem is choosing the right one. An output is a unit representing the delivery of something of value. However, simple delivery is not the only consideration. Other considerations are to whom, according to what criteria, at what location, with what level of accuracy and timeliness. Thus, output measures are meaningless without measures of quality. 38 Mongolia has the benefit of the Millennium Development Goals from which to draw and set clear achievable and monitorable objectives. The task ahead for staffing -'rposes will be to determine skills mix required to deliver the output and cost input mix. PUBLIC PENSION SYSTEM - ISSUES AND OPTIONS. Cash transfer programs inc. eased rapidly during the transition 3.51 Cash transfer programs are another important source of pressure on Mongolia's budget. They accounted for about 17.7 percent of government expenditure in 2000. Their share in the budget increased by over 55 percent between 1997 and 2000, from 5.1 percent of GDP to 7.3 percent of GDP. Pensions are the largest subcategory, accounting for over 75 percent of transfers, or 5.6 percent of GDP in 2000. Cash transfers to households make up 93 percent of social welfare expenditures. Family benefits are the next largest expenditure, including payments to women on maternity and childcare leave. Although social insurance transfers were planned as self-financing from earmarked payroll taxes, revenues have lagged behind expenditures. Immediate action is needed to arrest continued growth of cash transfer programs. Table 3-4: Expenditure on Social Welfare Programs Share of Share of Share of Share of Share of Share of Budget (%) GDP (%) Budget (%) GDP (%) Budget (%) GDP(%) 15.1 6.3 16.4 6.5 17.7 7.3 Source: Government of Mongolia (2001) Safety net programs introduced to cope with the transition need restructuring 3.52 Mongolia introduced social safety net programs and institutions in the early 1990s as part of a broad range of programs developed to cope with new social needs, including a full social insurance system of pensions, short-term income replacement benefits, and employment services for the unemployed. The residential care system was retained and expanded. The collapse of the Soviet Union and the inflationary spiral that followed drastically reduced the value of benefits, and the government has attempted to restore them to sustainable real levels. This commendable attempt by the government continues to be undermined by the failure to accompany increases in social insurance and social assistance with much needed structural reforms of the social welfare system, particularly the pension system. Table 3-5: Share of Total Cash Transfer Expenditure By Type of Payment: 1985-2001 11 AN0, M10,9%S. 99 ,"11,9OWN M2O0 .01 ocial Insurance Pension 60.2 67.7 85.8 86.8 77.3 78.6 73.2 77.4 79.0 ocial Assit. Pension 0.0 0.0 3.5 2.5 4.0 3.4 5.0 5.0 4.6 oc. Insurance Short-Term 18.8 15.5 4.8 4.3 3.5 3.6 4.3 3.9 4.0 oc. Assistance Short-Term 19.5 14.6 5.8 40 10.3 8.9 11.6 8.3 10.1 Social Insurance Others 0.0 0.0 0.0 0.1 0.2 0.2 0.2 0.2 0.3 Social Assistance Others 1.5 2.2 0.2 2.3 4.6 5.2 5.7 5.3 2.0 Total 1000 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (SI= social insurance, SA= social assistance), Source Ministry of Social Welfare and Labor, 2001 3.53 While reforming the social assistance and transfer system should be a medium-term effort, some short-term savings are needed to reduce fiscal pressures from systemic inefficiencies. These savings 39 should focus on a few measures that will not endanger long-run safety net policies but improve overall allocative efficiency. 3.54 The government provides cash benefits-broken down into over 24 categories-depending on whether beneficiaries are insured or qualify for social assistance (see Annex 2.A for detailed description) The Public Pension System -Issues Eligibility requirements for public pension system are lax and the formula is generous 3.55 Mongolia spent 5.6 percent of GDP on pensions alone in 2000, well above the regional average - (Armenia 3.6 percent of GDP, China 2.6 percent). Eligibility requirements for the public pension system are lax. Pensions are given at a relatively young age--only 25 percent of pensioners are 60 when they start receiving benefits and the number of pensioners is nearly twice the share of the population over 60. Women can receive benefits at 55 if they have 20 years of service, or 50 if they have raised more than four children to the age of six. Men and women can also retire early if they work in special occupations. Life expectancy at normal retirement age is 14 years for men and 20 years for women. If Mongolians behave as pensioners in most FSU countries, most are economically active for at least the first five years after "retirement," despite prohibitions on working and receiving a pension. Table 3-6: Social Insurance Pension System Indicators (percent) Pension expenditure/GDP 0.5 3.2 4.5 4.3 5.6 5.9 Pension contributions/GDP 0.2 2.3 3.0 3.0 3.9 3.8 Government transfer/GDP 0.0 1.3 1.0 1.3 1.5 2.0 Pensioners/population 20-59 26.0 27.9 21.7 20.6 21.4 21.0 Pensioners/contnbutors N/A 69.5 43.7 44.8 46.6 45.9 Pensioners/population 60+ 183.7 221.4 155.7 177.2 194.1 197.2 Average pension/average covered wage N/A 35.4 59.4 60.2 58.8 61.0 Covered wage bdi/GDP N/A 13.0 17.5 16.0 20.4 21.1 Source: SSIGO, mission estimates 3.56 The pension benefit formula is fairly generous. With 20 years of service, benefits are 45% of the pensioner's monthly average wage. Each additional year of service accrues an additional 1.5% pension. Average years of service are 33, entitling most pensioners to receive a benefit equal to 65% of average wage at the time the pension is granted. As this benefit is untaxed, it represents a nearly full replacement rate. By contrast, given Mongolia's life expectancy at pension age and survival probabilities, a private firm would require about 10 -15 more years of age and service before providing a benefit at this level based on Mongolia's current contribution rate.'3 A reduced pension is offered for those with more than 10 years of service but less than 20 years of service. In 2000, only 7% of the total stock of old-age pensioners had taken this reduced pension option. I. 13 Comparing benefits paid with an actuarially fair annuity is one method of measuring the efficiency and equity of the pension system. It is, among other things, one measure of the system generosity towards current recipients. It ignores desired inter- and intra-generational redistributions. The above calculation is very rough, and is based on IMF, (1998). 40 Ad hoc indexation policy leaves pensioners vulnerable 3.57 The majority of pensions in payment are not as generous as the formula would indicate, leaving many pensioners closer to the poverty line. The average pension benefit is currently about 60% of the average covered wage14, and 121% of the national poverty line. But this average is deceiving. 55.7% of old age and disability pensioners and 72% of survivors' pensioners in 2000 received the minimum pension, equal to about 16,000 Tgs, about half the average covered wage and about 4% above the average poverty line. 68% of old age pensioners receive a benefit below 21,000 Tgs. The reasons for this are: (a) the inflation of the early 1990s wiped out most pensions in real terms so almost all who received pensions before 1995 get the minimum pension; and (b) indexation has been ad hoc since then. For example, pensions were completely recalculated in 1999. No indexation was given in 2000, and then in April 2001 pensions were increased by factors ranging from 10 percent for the lowest to about 4% at the top. Table 3-7: Average Benefit as a Share of Average National Poverty Line, 2000 I pensions 18, 629 121.5 percent I disability pension 15,779 102.9 percent urvivor's pension 15,227 99.3 percent A pensions 11,503 75.0 percent (nemployment benefit 12,652 82.5 percent SA child care benefit 7,591 49.5 percent (SI= social insurance, SA= social assistance), Source Ministry of Social Welfare and Labor, 2001 Falling contributions threaten sustainability of system 3.58 The number of contributors to the pension scheme is low, and declining. In 2000, only 545,585 people were reported contributing to the pension system. This compares with roughly 800,000 people known to be employed in the public and private under the National Statistics Office's narrow definition of employment (excluding most of the informal sector). Several factors have contributed to the low number of contributors. Until 1995, contributions were not required for social insurance, so everyone who worked was covered. Those working in privatized agriculture and animal husbandry activities and the informal sector are mostly not covered by social insurance, as they evade these taxes. Tax evasion appears widespread, and arrears have been building up as employers fail to pay taxes. 3.59 Pension expenditures have grown about 8 percent per year in real terms since 1997 even as the number of contributors as a share of the labor force has fallen, resulting in a steady decline in contributions relative to benefit expenditures. For short-term benefits, such as unemployment insurance, a decline in coverage reduces liabilities rapidly. For pensions, however, the decline in liabilities will not appear for many years-and will probably show up in increased claims for social assistance pensions. Not surprisingly, a structural deficit has emerged in the pension fund, which has been covered by budget transfers, some of which simply transfer a shifting of arrears from one level of government to another. At the end of the first semester of 2001, social insurance contribution arrears for the general government to the social insurance budget (including health insurance contribution arrears) stood at about 4.5 percent of 2000 contributions. Two thirds of this, and most of the 2001 increase, came from the local government sector. I. " The average covered wage is the average of wages on which social insurance contributions were paid. This is somewhere between 20 and 50% lower than the average economy-wide wage. As noted above, labor force statistics are poorly collected, as they primarily reflect the formal sector. Taxes are also poorly collected, however. 41 Past efforts to restructure pension system should be revisited and process accelerated 3.60 Successive governments have recognized that the pension system was in trouble, and that structural reforms in the pension system were required. In 1999, after several tries, a new pension system that depends entirely on contributions was adopted. The system only applies to those born after 1960. Individual contribution account records are being created and recorded in notional balances in worker accounts at retirement. These balances are calculated by summing individual contributions and indexed to the growth rate of contribution wages.5 At retirement, balances are divided by life expectancy in months to get total pension. Acquired rights under the old system will be transferred to the new notional accounts. This system is similar to ones adopted in Sweden, Latvia, Poland, and Italy. The government's intention is to eventually run a surplus, and convert some portion of the notional accounts into funded privately managed accounts.'6 The target date for running a surplus is 2005. However, recent projections suggest that this target is unlikely to be reached. 3.61 The pension reform was a valiant effort. Between 2020 and 2050, projected system deficits are expected to be smaller than they would have been under the old system. Until 2010 deficits are unaffected, as all pensions will be granted under the old system. But even with the reform, substantial deficits are projected (Table 3-10). The pension system remains a source of structural fiscal pressure. These numbers do not include social assistance pensions. Expenditures on these will also grow over the period given the declines in coverage observed above.17 Table 3-8: Contributions - Expenditures as Percent of GDP Old Law DB system -16 -1.3 -2.6 -4.2 -5.3 -6.6 -6.7 -7.2 New Law (combined system) -1.6 -1.3 -1.8 -2.2 -2.5 -3.2 -3.4 -3.6 Source: Staff estimates. 2001 3.62 The new pension system suffers from at least five serious flaws, which need correction 1: (i) as the reform only begins to take effect in 2010, it will not reduce expenditures substantially until about 2020; (ii) the new system fails to raise the minimum retirement age for those covered or get rid of the special early retirement provisions; (iii) the lack of a transition mechanism for pensioners retiring after 1999 but born before 1960 makes the reforms generationally unfair. With the same work history and contributions, someone born on Dec. 31, 1960 will get a pension twice as high as someone born on January 1, 1961 (box above); (iv) the new system also fails to reform disability and survivor pensions. As a result, these pensions will be much higher than old age pensions for those born in 1960 and later; and (v) the plan also fails to consider the costs of the system as a whole, as related to benefits, and set up the notional accounts accordingly. Under the current notional account system, 15 percentage points of pension contributions are credited to a contributor's account (regardless of whether it is notional or eventually funded). In other words, despite an intention to make the system self-financing (based on contributions), the system will run a perpetual structural deficit because of design flaws. I. 15 This means in effect that real value of the sum of contributions at retirement, and therefore beginning pension, is linked to the growth of wages in the covered sector. 16 See "Basic Guideline for Implementing Pension Reform up to 2021". 17 The minimum pension, provided to those with at least 15 years contribution, will take some of the pressure off of the social assistance pension. However, it adds to the social insurance deficit since it is not properly funded. 1 See Annex 1 for detailed discussion of flaws 42 3.63 In sum, the new system is a good start, and an improvement over the inherited Soviet system. However, primarily because it fails to address the transition problem and linkages with other systems, it is unfair and too expensive. Without further reform, the system is unlikely to survive once the reforms finally take effect and will run a perpetual deficit. The target of partial funding will not be reached unless the contribution rate is increased or other funding is found. Increasing the tax rate will not improve the coverage rate, and is probably not justified given Mongolia's high poverty rate." It would be better to leave some cash in workers' pockets, and reduce the costs of the pension system. Policy Options and Recommendations Reform of the pension system will be a medium term effort; howeve an early start is required 3.64 Short- and long-term initiatives are required to deal with the rising cost of pensions. One basic reform recommended to reduce the cost of the system in the short run and improve equity is to immediately begin raising the retirement age now for all in the labor force. 3.65 In the medium term, the government will have to address more structural issues that could undermine short-term efforts at improving overall fiscal balances. The government will need to: * Revise and restructure the pension system * Develop a transition policy to put all currently working into a new system * Reform disability and survivors pensions Immediately raise the retirement age for all in the labor force 3.66 The main reason Mongolia's pension costs are so high is that Mongolia has given out too many pensions too early: there are too many pensioners below 60 years of age. Mongolia has basically kept the retirement ages from the Soviet system. including the provisions for early retirement (mothers with many children, certain occupations, blind, etc). Most other transition economies Europe and Central Asia have already started raising retirement ages by four to six months a year, and abolished special entitlement groups. In some countries (Czech Republic, Lithuania) these groups retire at the normal age, but receive higher pensions to compensate for acquired rights. In others, the retirement age has been gradually raised (Poland, Moldova) and no compensation is paid at the time of pensioning. One option for Mongolia which combines these approaches would be to: * Immediately make 55 the minimum retirement age for everyone, including all special groups, and, * Begin raising the retirement age by six months a year up to 50, and three months a year from 60 to 62 3.67 There would be a short-term cash flow effect from delaying the number of pensioners, and total cost would fall. However, under the new (NDC- Notional Defined contribution) system this would be balanced by higher pensions in the medium term.0 The main fiscal savings under the new system would come from reduced expenditures on minimum pensions (as a higher retirement age implies more years of service). Higher pensions would also reduce social assistance costs and, ultimately, poverty. This reform would also ease the transition problem, as the differences between pensions under the old and new systems would start to converge. I. 1 Current government policy is not to increase the contribution rate. However, this is inconsistent with the aim of maintaining current retirement ages and moving to partial funding. 2 Under an NDC system the delay in retirement results in actuarially fair pensions, so in the medium term, in a pure NDC system, the government is indifferent between retirement ages from a purely financial point of view. 43 Develop a transition policy for pension reform 3.68 A transition system needs to be developed to contain the costs of reform and ensure the elderly and other vulnerable members of society are protected from further drops in income. Mongolia has begun the switch from the unsustainable pay-as-you-go system to a notional defined contribution (NDC) system. 3.69 Most other countries introducing notional accounts have made gradual transitions to the new system, beginning the year of the pension reform. Mongolia should follow this lead. Two basic approaches to the transition have been followed in other notional account system reforms. The first is to give everyone a notional account, with the opening balance based on acquired rights under the old system. This in effects means that the older workers have a pension mostly based on past service, while younger ones have one mostly based on contributions. The second is to have-a transition of ten years or so, so that ensions in the first year would be 10 percent new system, 90 percent old system, the second year 20 trcent new system, 80 percent old system, etc. While either approach would work for Mongolia, the -cond approach is administratively more complex. 3.70 Compared with the current rules, the new rules will result in projected benefit cuts of 40 to 50 percent for those who retire under the new system based on current norms for retirement age and years of service, and even higher cuts for those with low contribution records (Annex 1). Those entitled to early retirement (women and those in hazardous occupations such as mining--currently about 70 percent of new pensioners) would face the largest drop in benefits if they retire at the minimum retirement age under the law. Disability and survivor pensions benefits are roughly the same, as these were not reformed. Reform disability and survivor pension 3.71 Mongolia needs to complete the switch from the old system to the new system by reforming disability and survivor pensions to make them consistent with an NDC approach. This requires three changes: * revise the formula to be based on the amount in the notional (or real) account (e.g., be based on contributions); * make provisions to convert disability pensions into old age pensions at minimum old-age retirement age; and * add in a safety net (minimum) so that those who become disabled at a young age receive an adequate pension. 3.72 The fiscal effect of such a change in Mongolia could be large. About 0.6 percent of GDP a year in savings, rising to about 1 percent of GDP a year when the NDC system is fully in place after 30 years. Results on potential savings from a change in survivor pensions are pending. (based on preliminary estimates). STATE OWNED ENTERPRISES AND QUASI-FISCAL ACTIVITIES State Owned Enterprise Activity, Financial and Energy Sector, and natural disasters are all sources of fiscal pressure 3.73 A complete assessment of the reach and scope of the public sector in Mongolia is further complicated by the lack of comprehensiveness in the budget presentation and poor classification and consolidation of implicit and explicit government obligations. This section highlights the potential risks Quasi-Fiscal Activities (QFAs) and other contingent liabilities, mostly off budget, pose to macroeconomic stability and growth. Chapter 7 highlights how the fragmentation of the budget undermines the budget formulation process. 44 3.74 Activities of State Owned Enterprises and financial institutions including the Central Bank disguise the size of the government through the use of QFAs. Such activities include subsidized lending to SOEs through state commercial banks, provision of services to the government by SOEs without corresponding budget support, implicit subsidies to public enterprises and households through low pricing policies, sanctioning of arrears, non-reimbursed on-lent state loans, and capital injections. Most of these activities are not systematically reported or recorded in the budget, as they are carried out off budget. The budget contains no statements on contingent liabilities, tax expenditures or quasi-fiscal activities. Inclusion of these activities in the budget would substantially increase the size of Mongolia's public budget. Energy sector and financial sector QFAs pose the largest threat to overall fiscal stability in Mongolia. Overview of Current SOE Portfolio State-owned enterprises are a legacy of the pre-transition era 3.75 The current Mongolian government inherited a legacy of SOEs that operate throughout the economy. Many of these firms are performing functions that are generally recognized as commercial private sector activities. Some enterprises are large and employ considerable numbers of people. Others are small and have insignificant sales and payrolls. Although many of their activities are commercial in nature, there are a few that provide essential services to the population, such as producing and distributing electricity and hot water for heating. Nearly half of all SOE's are unprofitable 3.76 Many SOEs are unprofitable and require financing in the form of government subsidies (tax arrears) or directed borrowing from domestic banks to keep them afloat. Most of these indirect subsidies are not accounted for in the budget keeping the apparent size of government artificially low. Moreover, the level of this support has been insufficient to prevent a buildup of accumulated operating losses that has manifested itself in a cascading chain of debt and arrears across the SOE sector, financial institutions, the private sector, and the government. 3.77 The SOE portfolio merits some scrutiny, not only to help ensure the proper role of government in a market-driven economy, but, more importantly, to help predict and provide for unforeseen expenditures and revenues in the budget planning process. At present, there is no systematic process for identifying and recording future contingent liabilities arising from events at SOEs. 3.78 Mongolia's SOEs face many of the risks that affect private firms operating in a market economy. These include macroeconomic factors, such as rising interest rates, devaluation of the national currency, and increases in unemployment. Sector-specific risks, such as the deterioration of the world market price for a commodity exported by a firm, or increases in the foreign cost of critical inputs also affect profitability. External economic events are most likely to affect firms that have exposure to foreign trade, such as enterprises participating in mining, textiles, and meat production, as they may receive lower prices for their exports. Also, firms earning local currency but with debt denominated in foreign currency will be affected by deterioration in the value of the Togrog. 3.79 Natural disasters can also have devastating effects on firm profitability and government revenues. The recent dzuds caused the death of millions of livestock and brought much hardship to communities dependent on the herds and agriculture for their livelihood. In 2000 and 2001 for example, the government relied on SOEs to absorb some of the cost of providing fodder to vulnerable areas without due attention to the cost of such unforeseen expenditures on their balance sheets. 45 3.80 SOE performance affects the budget directly or indirectly in the following ways: firms owe less in taxes to the government; subsidies required by unprofitable firms may increase; firms may lay off personnel to make ends meet, increasing the number of unemployed; the government may be called upon to make good on privately contracted or publicly guaranteed debt. For example they may receive direct credit from State-owned banks, which will then require recapitalization. The misallocation of resources stemming from inappropriate pricing policies, production inefficiencies and over consumption threatens long run stability of the budget. General Characteristics of SOEs The number of SOEs has declined but large SOEs are yet to be restructured 3.81 Over the last six years there has been a 27 percent annual decrease in the number of SOEs from over 300 in 1996 to 118 in 2000 as firms were divested or closed down. The government's Program for Privatization of State Property in 2002 contemplates the complete or partial privatization of an additional 19 enterprises and 25 more are to undergo restructuring in preparation for privatization. However, the privatization programs are yet to reach the 18 most valued enterprises in Mongolia. Most SOEs are in the infrastructure and trade sectors. 3.82 SOE activities are concentrated in four principal areas. In 2000, these were mining (16 firms), trade (7), energy (26) and transportation (3). These 52 firms accounted for 89 percent of SOE sales in 2000. In 1999, the most recent year for which data is available, these four sectors accounted for 75 percent of SOE jobs and 81 percent of their payrolls. The remaining 66 firms operated in a wide variety of sectors, including food production and distribution, toll road administration, telecommunications, health and education, insurance, and defense. 3.83 The majority of SOE economic activity is based in Ulaanbaatar, where 84 of the 118 SOEs are based. They account for 68 percent of SOE revenue, and in 1999, they provided 40,821 jobs, 74.4 percent of 54,877 SOE jobs. Orhon is the second most important region; largely because of the presence of the Erdenet copper mine, the largest and most profitable SOE in Mongolia. Copper is the highest foreign currency earner in Mongolia. Erdenet with Tg. 180 billion in net sales in 2000 contributed over 87 percent of SOE revenue in the region. Erdenet also provided 86 percent of SOE jobs in Orhon in 1999. Orhon is also home to four SOEs in the energy, auto, road, and defense sectors that account for a little over 1,000 jobs. Table 3-9: SOE Financial Profile According to Type of Ownership in 2000 (Tg. 000s) ~ ~FLTS' Sd ',T I-N~ 6e I 50% or less 7 93,788,727 6,485,387 2,154,85 4,330,532 51% to < 100% 25 346,626,388 66,981,802 31,206,985 35,774,819 100 % 86 317,306,382 (15,053,998) 9,198,401, (24,252,399, Total 118 757,721,497 58,413,191 42,560,241 15,852,952 Source: Mongolia State Property Committee and Staff estimates 46 Overview of SOE Characteristics Figure 3-7: Sectoral Breakdown of SOE Sales in 2000 Figure 3-8: SOE Revenue is Concentrated in 2 Regions 5% All others 11% 27% Minng Transport 33% 19% 68% Trade Energy 7% *Ulaanbaatar (68%) UOrhon (27%) 0 All 19 others 30% Figure 3-9: Almost Half of all SOEs are Unprofitable Figure 3-10: Both Profits and Losses are Increasing (percent) (percent): BTI of Profitable and Unprofitable SOEs 100,000,000 48 80,000,000 60,000,000 40,000,000 - 70 X 20,000,000 (20,000,000) (40,000.000) *IPmave ffn ENegative NOT) - profitable firms EBTI- unprofitable firms Source: State Property Committee, 2001 and staff estimates.,2001 Wholly state owned enterprises are less profitable than joint venture enterprises 3.84 The overwhelming majority-86, or 73 percent-of the 118 SOEs in operation in 2000 was wholly state owned. The government holds more than 50 percent of the equity in another 25, giving it a controlling interest in 94 percent of the firms. Government equity holdings of over Tg. 547 billion in 100 percent owned firms represent 54.3 percent of total SOE equity and the sum of all government holdings including minority interests amounts to almost 80 percent of all SOE equity. 3.85 Enterprises that are wholly government owned have tended to under perform compared to those with majority private capital stakes. Wholly owned government firms account for only 42 percent of SOE sales. Wholly owned state enterprises lost money in the aggregate for the year 2000; their before tax income was a negative Tg 15 billion, which reduced the positive before tax income generated by the other two groups by a little over 20 percent. 47 Profitability and economic viability of SOEs 3.86 The performance of SOEs is linked to the size and reach of the public sector from revenue, liquidity, debt obligations, and employment perspectives. Negative SOE performance in any of these areas would contribute to increases in the size of government as it steps in to absorb shocks. Almost half of Mongolia's SOEs are loss making 3.87 Total losses of unprofitable firms increased between 1999 and 2000, with negative before tax income deteriorating by 73 percent from over Tg. 16 billion to about Tg. 27 billion. During this same period before tax income from profitable firms, excluding Erdenet increased slightly by 3.4 percent. When Erdenet is included, profitable firms more than doubled there before tax income, from Tg. 40.3 billion to Tg. 86.3 billion. Government relies on very few firms for revenue 3.88 Most of the revenue that accrues to the government from SOEs comes from a handful of enterprises that are overburdened, contributing on the average over 50 percent of their total before tax income in taxes and paying 93 percent of all SOE taxes. INEG itself contributed an exorbitant average of nearly 84 percent of its total before tax income in 1999 and 2000. These firms subsidize a large number of unprofitable SOEs. If the government were to privatize all the others, there would be little difference in the tax revenue stream it receives. Furthermore, the maintenance of a large portfolio of under performing firms only increases government exposure to potential liabilities while offering few prospects for increased revenue or employment, as competition and entry are stifled. Table 3-10: Before Tax Income Total for Profitable SOEs (Tg. 000s) Profitable SOE's Total BTI for all Profitable Firms 40,318,441 86,333,463 114 1 Less: Erdenet BTI 6,070,7811 50,935,407 739.0 Net BTI for Profitable Firms w/o Erdenet 34,247,664 - -3-5,398,056 3.4 All SOE's M Before Tax Income - All SOEs 24,173,60 58,413,192 141.6 Lss: Erdenet BTI 6,070,781 50,935,407 739.0 Net BTI without Erdenet 18,102,825F 7,477,785, -58.7 Source: Mongolia State Property Committee and Staff estimates 48 Figure 3-11: Half of all SOEs Pay No Tax Figure 3-12: And Just 5 Firms Pay 93 Percnet of an Taxes Breakdown of Taxpaying Firms in 2000 _____________ % of Total SOE Taxes Paid by Firms in 2000 8% U Paid no Tax - (49%) MErdenst Pant (55%) 5% U <40% - (38%) 4 MINEG(18%) 49%51aMnoinTlcm(1% 38% 0>=40% <50% - (5%) 55% lShlrAt(5%) BrekdwnofTotlOBTa Pametal 99 TaesPad2y0irs0n000 ErErdenttlaPlant% 5INgIEG(1% Mongoliann eleele(11% 55%iShjIrAItlt% 18gobbi(% D >=50% - (8%) Eol(% OAI Others (7%) Source: State Property Committee and Staff estimates, 2001 Breakdown of Total SOe Tax Payments (1999 vs 2000) Erdenet Plant -- INEG Mongolian Telecom c2000 Slujir Alt 19 Gobi All Others 5 10 15 20 25 Tg x 1,000,000,000 Source. Mongolia State Property Committee and Staf estimates 3.89 Seventeen enterprises paid no tax at all in 2000 in spite of reporting positive before tax incomes of Tg.3 billion, which at the average tax rate of 51.1 percent would have brought in Tg 1.5 billion in tax revenue to the government, or about 0.15 percent of GDP. Enterprise Arrears, subsidies and price controls affect profitability 3.90 Subsidies and price control policies affect SOE profitability across the board and increase risks of bankruptcy. In a number of cases, lack of profits is directly linked to firm pricing policies. The costs of their products or services often exceed the revenues paid by customners.2 This is particularly true in the energy sector (Table 4.15). Energy sector QFAs are related principally to poor pricing poor and the toleration of arrears including these indirect subsidies explicitly in the budget would increase the size of government by 1 percent of GDP. 21 Gross sales margins do not reflect production costs as a result of controlled and unadjusted prnces. 49 3.91 The financial statements submitted by the SOEs do not explicitly account for indirect Table 3-11: Major Profitable Firms That Paid No subsidies provided by the government. They Tax in 2000 include a line item called non-operating income, but it is not clear precisely what sort of items are B ril!ay -transport 1,503,160 included in this account. It is interesting to note NDS 491,34 that when this item is removed from the financial stShSG -Energy 425,378 statements of the 70 enterprises that were on amcale -Cashmere 125,629 profitable in 2000, 16 of them turn unprofitable. 3rdenet TETS-Ming 105,467 Source State Property Committee and Staff estimaes, 2001 3.92 Most SOEs have litie staying power inI terms of cash and other quick assets that could be used to cover short-term liabilities in the event of a sudden reversal in business conditions. For example, Monrostsevnet, a large industrial concern, has current liabilities of Tg. 8.8 billion and quick assets of only Tg. 1. 1 billion. That would only be sufficient to pay off about 13 percent of its short-term obligations in the event a sudden downturn in business interrupted its normal cash flow strearn. Erdenet, in spite of its size and profitability, only has a liquidity ratio sufficient to pay down about 41 percent of its current liabilities, assuming its receivables are all collectible. Most firms with healthy balance sheets have a liquidity ratio of over 50 percent of current liabilities. Table 3-12: Major Firms With Negative Gross Margins in 2000 (Tg.000s) etalurical plant 3,100,003 5,860,619 189.1 (2,760,616 aganuur coal mine 22,515,530 24,788,058 110.1 (2,272,528 avhan diesel station 308,99 1,511,00 489.0 (1,202,007 uvspt diesel station 398,043 1,312,11 329.6 (914,076 utul cement and chalk 2,646,725 3,555,33pi 134.3 (908,611 bi-Altai diesel station 314,62 1,201,32 381.8 (886,706 ower Station 4 36,534,28 37,411,261 102.4 (876,971 ayanhongor diesel station 308,883 1,116,06 361.3 (807,184 alanzadgad DST Energy 399,868 992,564 248.2 (592,697 ornod DS 1,539,300 2,041,337 132.6 (502,037, rdene zam 935,135 1,358,514 145.3 (423,380) STsShSG 8,374,367 8,737,797 104.3 (363,430 hivee Ovoo 2,937,545 3,170,758 107.9 (233,213 alaih DS 441,211 622,857 141.2 (181,646 hariin gol 7,626,757 7,795,908 102.2 (169,151 ower Station 3 15,597,820 15,733,595 100.9 (135,774 BBEH system 1,473,613 1,575,476 106.9 (101,863A Source: State Property Committee and Staff estimates, 2001 3.93 Mongolia's SOEs are plagued by huge problems of arrears, particularly between the large energy stations. Energy QFAs pose a particular threat to overall fiscal balances. The three firms that are of the most concern, Power Station 4, Baganuur, and EstShSg-with arrears ranging from 9 to 12 months worth of sales tied up in receivables-are those the government cannot afford to shut down. Rising arrears in these SOEs imply government intervention to cover the international obligations of these firms. These risks are not explicitly budgeted for, and could result in a substantial rise in budget expenditures. On average, all other SOEs, except Mongol Amicale and Shivee Ovoo, have about three to four months of sales in receivables. The large arrears in the system also impact tax revenues. 50 Viability of SOEs 3.94 The government would need to inject over 1.5 percent of GDP into firms with negative gross margins or increase sales prices of outputs to bring all SOE's to a zero profit margin status, according to staff estimates. However, this is an expensive and inefficient option for all firms. Other options the government could consider include closing some of the non-profitable SOEs, and liberalizing the sales prices of SOE outputs. These are difficult decisions the government will need to consider in the privatization and restructuring process. 3.95 In 2000, there were 23 SOEs with a total negative gross margin of Tg 14 billion: 1.3 percent of GDP. For firms with persistently negative gross margins, every sale leads to further revenue losses. It is important to distinguish such firms from those with simple operating losses, as the approaches taken to restore market conditions in each situation are different. While one set of firm's needs liberalization of prices, the others need to improve overall efficiency by tightening up expenditures. However, where the cost of goods is greater than the sales price, something must be done to reduce costs or usage of materials and labor, or the firm will only compound its losses the more it sells. Table 3-13: Accounts Receivable Turnover for SOEs with Largest Receivables ower Station 4 770,037 37,486,460 0.3 12.3 rdenet copper plant 23,732,587 20,191,253 2.0 1.3 nergy authority 1 19,788,809 2.9 aganuur Coal Mine 11,539,020 19,152,276 6.5 10.2 stShSG 2,678,736 16,296,330 19.4 9.3 NEG rail transportation 12,618,718 15,712,914 6.0 7.4 IC Petroleum Company 6,568,258 8,281,564 1.7 2.0 TAT Airlines 5,786,483 6,478,343 2.0 2.0 ongolian Telecommunication 4,407,257 6,142,069 2.2 2.6 ower Station 3 2,106,079 5,950,362 1.7 4.6 BTsShSg 6,429,026 5,642,792 3.6 2.9 on Amicale Cashmere 5,699,876 4,673,915 13.3 10.6 hariin gol 2,578,335 4,003,636 4.6 6.3 hivee Ovoo 2,083,944 3,732,966 11.5 15.2 Gobi Cashmere 4,193,270 3,704,701 2.1 2.0 Source: State Property Committee and Staff estimates, 2001 SOE Employment 3.96 SOEs absorbed over 7 percent of the total workforce in 2000, employing over 54,000 people. The chart below shows the distribution of SOE employees across sectors. The table below lists the major SOE employers in 1999. As can be seen, in 1999 15 of these enterprises accounted for 75 percent of SOE employment and nearly 82 percent of SOE payrolls. At Tg. 79,814 the monthly cost of an employee was at the higher range of the scale for these companies. However, this was only slightly higher than the average of Tg. 73,846 for all SOEs. 3.97 Some of these firms, such as,Power Station 4, Baganuur, UBTsShSg, MIAT and Hutul Cement and Chalk lost significant amounts money in both 1999 and 2000. In the absence of continued subsidies, it is not clear whether they can continue to operate without reducing personnel. If the subsidies required to offset the operating losses are not sent by the government, there could be a secondary liability generated in the form of legally mandated termination pay. 51 3.98 In the event of privatizations, the cost of providing termination pay to these enterprises and retraining lay off workers would also contribute to increasing the size of government. Publicly guaranteed debt of SOEs -emerging fiscal liabilities 3.99 Failure of SOEs to honor foreign obligations could put substantial pressures on public sector expenditures. In several recent instances, the government has been called on to honor guarantees made to third parties on behalf of Mongolian organizations. The most recent of these involves Buyan Holding, a cashmere garment manufacturer that signed a US$18.8 million equipment purchase contract with Marubeni Corporation of Japan. When Buyan failed to honor the terms of the agreement the Japanese seller initiated legal action to recover the amounts from the guarantor, Mongol Bank. If the court rules in Marubeni's favor, the government will face a major non-budgeted expenditure that will draw scarce resources away from essential services. Last year, there was a similar situation involving a Mongol Bank guarantee of US$1.4 million for the purchase of a passport machine by a government agency. Once called, these contingent liabilities, and the associated exchange risks could severely damage long run fiscal balances and compromise policy directions. Table 3-17: Major SOE Employers in 1999 UB railway 13,408 14,010,90 Erdenet plant 6,266 6,910,00 Mongolian Telecommunication 4,543 3,180,142 Monrostsevmet 3,024 2,856,908 IC Petroleum 2,501 1,811,845 Robi Cashmere 1,632 1,896,922 ower Station 4 1,404 1,399,459 aganuur Coal mine 1,341 1,453,708 hariin Gol 1,331 1,074,476 NEG rail transport 1,152 1,218,324 IAT Airlines 1,100 1,488,86 BTsShSg 1,045 668,714 ower Station 3 987 788,282 ongol Post 85 474,31 lutul Cement and Chalk 773 377,62 Total 41,35 39,610,47 Total SOE Salaries 54,877 48,629,47 % of all SOEs 75.4% 81.5 Source: State Property Committee and Staff estimates, 2001 CORPORATE GOVERNANCE, ACCOUNTING AND AUDITING STANDARDS 3.100 Corporate governance is an unfamiliar concept in Mongolia2. Under communism, all significant enterprises were owned and operated by the state. During the early period of market transition, the enterprise sector suffered a massive decline as links with the former Soviet Union disappeared. State assets including state enterprises are now administered by the State Property Committee (SPC), which reports to the prime minister. Considerable privatization has occurred; and more is planned (currently Gobi Cashmere, NIC). 3.101 State enterprises submit their audited financial statements to SPC by 10 February (41 days after the year ends). The chairman of the board and chief accountant signs the financial statements. SPC I. 22 Mongolia Country Financial Accountability Assessment, 2002. 52 publishes an annual compendium of state enterprise financial statements by mid-April.23 It also sets performance targets for board members, who may be dismissed for failing to achieve targets or may share in profits when targets are exceeded. Generally there is little public accountability in the state enterprise sector. Documents are available on request but are not widely disseminated. 3.102 SOEs in Mongolia are audited by the private sector. However, the findings of a survey of financial management reporting in the private sector showed that financial reporting is better in the public than in the private sector, large entities are better than small ones, and 9,000 of the entities identified did not submit financial statements. Only 25 percent of the entities surveyed met international accounting standards, and less than 10 percent used computer software for maintaining their accounts. Many of the entities surveyed did not have their own accountants but hired them temporarily for specific purposes such as producing final accounts. 3.103 Local private auditors audit most donor projects including World Bank projects. However, the audit reports had not been submitted for most projects prior to June 2001. For audit reports received subsequent to June 2001, no separate opinion was given on the use of statement of expenditures for withdrawal of funds as required in the DCA. The project audits are limited to an examination of documentation supporting fund releases and payments from special accounts as opposed to an audit of the financial statements. The audit reports as currently presented do not meet the Bank's minimum requirements. 3.104 In general, entities had weak primary records, accounting records, and financial skills, with many entities operating without accountants or with untrained accountants. This is not the case for the large SOEs, which employ internationally acceptable accounting and audit firms. The situation is worse in rural areas than it is in Ulaanbaatar, as well trained accountants refuse to accept the inconveniences of working in rural areas. 3.105 Among the reasons for poor financial controls are shortages of skills aggravated by the transition to new accounting requirements, an absence of standard reporting formats and instructions, unwillin 9ess of companies to disclose information, and the inability of enforcement agencies to enforce discipline. 3.106 The lack of good financial audits for the SOEs also constitutes serious risks to the government. Expected privatization revenues could be severely overstated based on inaccurate SOE balance sheets. 3.107 Independent audit is relatively new to Mongolia. In the early post-communist years, professional and business participators founded a Union of Auditors. The present Union of Auditors is its successor and is a representative body for professional auditing firms. The audit law was passed in 1997. It defines the entities to be audited, describes typical audit activities and requires audits to be carried out by licensed audit companies. MOFE approves the issue of audit licenses. The following principles are enunciated: * Audits are to be conducted according to international standards2 * Auditing is to be independent * Auditors are selected by those who employ them I. 2 Known as "Production, Finance and Economic Indicators of Economic Entities with State Property or State Participation". 24 Inspections and audits were seen as compliance oriented. Inspectors were aware of the problem but had not been able to resolve it. In some cases entities have incurred penalties (60,000TR (transferable Ruble) for the accountant and 310,OOOTR for the entity, rather than submit financial statements. The law on penalties is to be revised to revent such actions by enforcing submission of statements in addition to the penalties imposed. There is no specific mention of ISAs in the current version of the law. A specific mention is now being considered in proposed amendments to the law. 53 * Auditors are to treat the information acquired during audits as confidential * Clients have a duty to make available the necessary information for the audit 3.108 The law also sets out the arrangements for examining applicants who wish to become certified public accountants and for the issue of qualifying certificates. The law sets out manner in which audit companies are to be established (both MICPA and MOFE have to approve). Audit companies can be disciplined (their rights to conduct audits can even be curtailed) if they break the law. Proposed amendments to the law would give more functions to MICPA and less to MPAC. Missing from the law is the right of an auditor to address shareholders in the case of resignation and the duty of a newly appointed auditor to consult the previous auditor on the circumstances of resignation. Law does not establish indeed the principle that the auditor reports to the shareholders of the company. 3.109 Current problems experienced in private sector audit include: under-cutting on audit fees to a level where a viable audit is not possible, company directors who regard audit as a burden, audits known to be of variable quality but no general means of monitoring and improving quality, and there are often times cases of auditor negligence which result only in reprimand. 3.110 Clearly, the audit profession is at an early stage of development. ISAs have been translated into Mongolian but knowledge and application of ISAs is quite limited. Similarly, concepts of internal control are not widely understood. Internal audit is rarely found outside the largest companies. 3.111 In the future, the need for a strong private sector audit environment would be great. As Mongolia growth picks up, buoyed by a leaner public sector and an acceleration of the privatization agenda the need for private sector auditors and accountants would become even more pressing. A peer review of private sector auditors is recommended as a preliminary step to supporting government build capacity. In addition, most donor projects including bank projects are audited by the private sector. Weak capacity in these sectors has compromised project performance in many sectors. Reform is Underway ... 3.112 Energy regulation improved substantially in the last two years as the government de-monopolized the sector and strengthened the Energy Authority's regulatory role. The government increased energy tariffs in January 2001 for the first time in five years.2 Low tariffs had that weakened the Energy Authority's balance sheet and left it in substantial arrears to the coalmines. The Authority also reached agreement with other key SOEs to document and agree on a time-bound plan to pay down inter-enterprise arrears. Instead of having one energy company responsible for distribution and transmission, the sector has been liberalized and 18 companies now provide these services, while the Authority has been transformed into a regulating and implementing agency for government energy policy. 3.113 The main function of the reconstituted authority-renamed the Energy Regulatory Authority will be to set tariffs, ensure compliance with government policies and regulations, and set standards for energy production and distribution. The government is working on a comprehensive energy sector reform to be presented to donors in the spring of 2002. 3.114 To avoid the very real possibility of experiencing financial drains that are not provided for in the annual budget, it is important that the government take steps to identify and quantify the risks it faces. Then it should move systematically to eliminate them, or at least contain them within predetermined boundaries. I. 26 Energy prices were liberalised in the fall of 1996. 54 ... But more is needed 3.115 The government should take the following specific steps to reduce its exposure to unbudgeted expenditures arising from known commitments and obligations and unforeseen contingent liabilities, first by imposing restrictions on the issuance of new guarantees and other contingent liabilities: and, * Complete the execution of the plan to divest all SOEs that are performing functions that traditionally belong in the private sector. * In the interim, require management at all SOEs to disclose any possible contingent liabilities that might exist within their organizations, such as contractual obligations to staff or third parties as well as any guarantees that might have been issued. * Monitor and improve accounting practices of SOEs * Establish registry of contingent liabilities * Undertake a comprehensive study to assess the degree of risk faced by SOEs because of domestic and or international economic shocks. * Set up transparent on budget contingency fund for dzud and other predictable natural disasters. * Set up a tracking system that can be used to follow potential contingent liabilities on a regular and systematic basis. * Develop training program on accrual accounting for all SOEs * Establish method for calculating benchmark price of utilities such as energy and develop time bound action plan to implement revise pricing policy * In the energy sector develop system to improve collections and reduce arrears * Performance based management contracts would reduce production inefficiencies. (Transmission and distribution losses in the energy sector could be substantially reduced). * Discourage non-cash settlement of tax liabilities. Treatment of SOE management under the draft PSMFL remains inadequate 3.116 The PSMFL has some provisions, which could improve overall management of SOEs however; lack of clarity in the roles and responsibilities of government in SOE management could comprose long- term reform. There is no clear governance and accountability framework with well-specified roles for the government ministers involved in SOEs the boards and the general managers, in this draft law or the Law on State and Local Property, which has many weaknesses. This problem is particularly acute for SOEs that are not 100 percent owned by the government.'There are issues with non-government shareholder roles and rights for these SOEs. Basically the draft law ignores minority shareholder rights and allocates all critical shareholder powers to the government including monitoring performance, making key policy decisions and appointing the board members. 3.117 There is an important role missing for the governing bodies and that is to ensure that the SOEs operate in accordance with its Statement of Intents. That is the key accountability documents and the main instrument for controlling SOEs. Fully owned SOEs should do Statements of Intent. The way Statements of Intent are used in this law interferes with the rights of minority shareholders in partly owned SOEs and the implications of that should be considered. 3.118 The law usefully prohibits the State Property Committee and Minister of Finance from interfering in the operational management of the SOEs and rightly specifies their main functions to be agreeing on the Statements of Intent (note "agreeing" is too vague- can they refuse to accept one?) and monitoring. The monitoring and other roles however are not clearly specified between the Minister of Finance and the State Property Committee. The Minister of Finance has a role in monitoring the overall financial performance of the SOEs and the State Property Committee is responsible for monitoring their 55 performance- where do the boundaries lie? The Minister of Finance receives the Statements of Intent only when they are finalized which is not consistent with the roles awarded him. (29.5). FINANCIAL SECTOR REFORM Three banking sector crises weighed heavily on fiscal balances 3.119 Mongolia has suffered three banking sector crises 1994, 1996, and 1998. The cost of the 1994 restructuring was estimated at over 2 percent of GDP. The cost of the 1996 restructuring exercise as measured solely by issuance of government reconstruction bonds, amounted to Tg 46 billion (7.8 percent of GDP). Other costs, such those relating to retrenchment, have not been quantified. Subsequently, BOM provided exceptional liquidity support to RB, AG and ITI amounting to about Tg 5 billion. In 1998 three state-owned banks holding 21.4 percent of total bank assets once again became illiquid and insolvent, the Reconstruction Bank, ITI Bank, and the Agriculture Bank. The direct costs to the budget of the 1996- 1998 crisis persist, and are estimated at over 0.6 percent of GDP in 2001. Financial sector fragility continues to pose a threat to overall fiscal stability 3.120 An extremely low intermediation level characterizes Mongolia's financial sector. Net banking system assets accounted for 20 percent of GDP and the ratio of private credits to GDP was only 4 percent by end-2000, with no penetration by foreign banks and very high state participation. 3.121 The banking sector is highly concentrated. The five largest banks shared 87 percent of the urban market in 2000, while a single bank controlled nearly all the rural market. This reflects lack of a sound and safe multi-tiered financial service network. 3.122 There were 12 commercial banks operating in Mongolia as of end-2000; the 3 largest were state owned, and four state-owned banks shared 80 percent of the market. The BOM has continued to strengthen its prudential guidelines for banks, raising the minimum capital requirement for new banks to Tg 4 billion (approximately $4 million) on July 1, 2001. This requirement is to be effective only in 2004, and the BOM licensed four new banks in 2001, thus continuing to endorse inefficient allocation of scarce capital and human resources in an already heavily over-banked economy. 3.123 Financial sector restructuring and modernization remain key components affecting the size and reach of the public sector in Mongolia. Continuous banking sector crises stemming from poor management lack of capacity, and undue public sector intervention contributed to increased government expenditures in the past. 3.124 Last year's extraordinary expansion of credit to the private sector is yet to lead to serious problems partly because many of the loans are rolled over, and this may also reflect a welcome return of public confidence in the banking system. Nevertheless, unless the recent easing of monetary conditions is reversed, a continuing rapid expansion in money and credit could renew inflationary pressures. Reforms have accelerated in 2001-2002 3.125 Mongolia has strengthened its laws governing the seizure of collateral on non-performing loans, which has helped restore the willingness of banks to undertake new lending. Parliament passed legislation in January 2001 to enforce this legislation. 3.126 Privatization of the Trade and Development Bank in May 2002 has substantially reduced the share of the public sector in the banking system from 60 percent in 2001 to only 25 percent. The 56 Agricultural Bank has turned profitable after 12 months of restructuring under receivership of the BOM and external management, and now has a positive capital balance. It is expected to be privatized in 2002. But more needs to be done 3.127 The state continues to own a substantial share in the banking sector, exposing it to further costs to the budget. Sixteen commercial banks are licensed despite the small size of the economy and BOM's limited supervisory capacity. The government has direct participation in three banks. Currently, two of the five largest banks (Savings Bank and Agricultural Bank) are fully government owned; they hold 22 percent of total assets. The three banks with government ownership have about 25 percent of total assets. The thirteen privately owned banks account for 75 percent of bank assets, of which Trade and Development Bank accounts for 34 percent and Golomt for 15 percent. Reform Options 3.128 The government has begun taking steps to improve the overall soundness of the financial sector by strengthening the legal framework, policy, and supervisory capacity of the BOM, and divesting from the banking sector. 3.129 To prevent any further deterioration in the financial position of banks and avoid further pressures on the budget the government needs to ensure that the newly adopted banking sector regulations are appropriately enforced and timely sanctions imposed for non-compliance. In addition the BOM needs to ensure that all banking sector credit are allocated in a competitive way and no guarantees are provided to commercial entities by the central bank. In addition the BOM should continue to manage overall interest rate level, keep inflation in check, supervise and enforce IAS accounting and financial standards in the banking sector. 3.130 With donor assistance efforts are underway to improve the overall soundness of the financial sector. Most recently to assist Mongolian government to tackle with financial sector reform the World Bank and the Asian Development Bank have agreed to jointly implement the Financial Sector Adjustment Credit program and the Financial Sector Technical Assistance projects during 2001-2004. The main objectives of the FSAC include: improving the payment system network, raise minimum capital requirement, and complete Agricultural Bank restructuring program. The Financial Technical Capacity building project aims to support institutional capacity building of the Mongolian banking sector to reduce systemic risks and enhance financial intermediation through (i) improved commercial bank risk management; (ii) modernization of payment systems and instruments; and (iii) Savings Bank resolution. ADB, UNDP, JICA and GTZ are is supporting Micro-and rural finance projects, USALD is working with the government to strengthen Rural finance network and improve management of the Agricultural Bank; SIDA is assisting with on bankers training programs. . MANAGING CONTINGENT LIABLITIES AND NATURAL DISASTERS 3.131 The largest area of risk to the government in terms of contingent liabilities is in dealing with natural disasters. In recent years, Mongolia has been very unfortunate in having experienced a series of calamitous events that have had a devastating effect on its citizens. The last two years have seen summer droughts and exceptionally harsh winters, which have resulted in the wholesale deaths of millions of livestock. This has increased poverty in rural communities and forced to the government to provide resources in order to relieve rural hardship. During the 1999-2000 dzud, 2.4 million livestock were lost and the total damages are estimated to have reached 95 billion Togrog. Livestock losses for the year 2000-2001 were comparable to the previous year at some 2.1 million animals. 57 3.132 The Government of Mongolia has only limited resources available to satisfy the considerable needs of the populations affected by dzud. The expenses that are incurred in the process of bringing relief to affected populations are the result of an Implicit Contingent Liability for which no specific provision has been made in the annual budget. The liability is implicit because there is no express legal requirement for the government to provide a predetermined benefit to compensate the victims of dzud for the damages they have suffered. Rather, it is the government's perception that its role as a protector of its citizens creates a moral obligation to relieve their suffering when an unforeseeable disaster strikes. Furthermore, the people expect this from their government and many requests for assistance have been received by the various governmental organizations tasked with disaster relief. The amount of relief required is unfortunately far beyond the government's needs. 3.133 The government's true liability is unclear. However, herders want assistance from the government. Up until 1990, the livestock belonged to the government and the herders were government employees. In order to provide a realistic reserve in the budget for dealing with the cost of dzud, it is first necessary to be able to quantify the actual expenses. Under the present system, this is next to impossible, because the costs are distributed throughout the government, as dzud-related expenses are charged against the normal operating budgets of a great many budgetary entities and state-owned enterprises. Also, the present accounting system may not take into consideration the interagency receivables that represent goods and services provided by one organization to another, perhaps without even generating the proper paperwork. 3.134 One of the additional cons'equences of the contingent liabilities arising from dzud and other disasters is the creation of a debt chain that builds up and cascades through the system, causing cash flow shortages at various levels. For example, numerous government and private organizations are called upon to help on short notice. One of these might be the state petroleum company, which will provide fuel on an emergency basis, so that supplies and personnel can be transported quickly to the site of the disaster. These organizations are supposed to be reimbursed later from the budgets of the organizations whose mandate it is to provide disaster relief. When these organizations provide goods and services without being fully compensated, they experience cash flow problems and sometimes they are unable to pay their suppliers, such as the power company. This creates a chain of insolvencies that ultimately translates into increased government subsidies for unprofitable state owned enterprises. 3.135 The occurrence of dzud and other natural disasters is very difficult to predict. However, the government can take a number of steps to increase its preparedness. First, it can study the patterns of expenditure regarding how past disaster relief efforts were financed and secondly plan accordingly for such contingencies. The State Permanent Emergency Commission has a good database with statistics on emergencies from the past. Efforts to computerize this information have met with limited success. properly but at present, there have not been sufficient resources available to develop it and link it into existing government information networks. 3.136 In the aggregate, it will be necessary to create reserve funds to budget for those items that cannot be reliably estimated within the normal agency budget framework. However prior to this an estimation of the real costs is needed and the management systems for dzud relief better developed. Multiple organizations dealing with dzud relief efforts 3.137 The six principal government organizations tasked with providing goods, services and funds to the dzud affected areas are the following: State Emergency Commission, State Reserve Agency, State Board for Civil Defense, Ministry for Food and Agriculture, State Veterinary Service of Mongolia, Ministry of Health, and Ministry of Social Welfare and Labor. 58 3.138 No funds are explicitly provided for dzud relief. In most cases, each organization draws from its own budget or accesses special funds that have been established to deal with a broad variety of expenditures that could not have been foreseen during the normal annual budgeting process. The two main funds that are maintained centrally at the Ministry of Finance have a legal foundation under Articles 10 and 11 of the Budget Law of December 21, 1992 and are authorized by Parliament based on estimated needs for the forthcoming year. 3.139 The first of these funds is the Government Reserve Fund. Its primary purpose is to provide funding for relief for populations and communities affected by natural disasters. Among the items it is intended to support are activities such as the repair of damages caused by natural disasters, containment of epidemics of contagious diseases and the financial assistance of other budgetary entities who have exhausted their own operating budgets as a consequence of their participation disaster relief. 3.140 The other fund is the Central Budget Unallocated Expenditure Fund. It is used to provide for reimbursement of expenses relating to the secondary consequences of major disasters, like dzud. Like the other fund, it is used to fund a very wide variety of activities or procurements that arise as a consequence to unforeseen or unbudgeted events. 3.141 The distinction between the respective roles of these two funds is somewhat blurred and the official descriptions clearly bear this out. A review conducted last year of the actual items charged against the fund revealed that many of these were foreseeable to the point of being candidates for inclusion in the budgets of the respective agencies. The very broad interpretation of the types of items that can be financed out of these two funds makes it difficult to know whether the overall reserves will be adequate to deal with the relief needs. The current structures creates a situation whereby dzud relief must compete for scarce funds with a great many other needs, and this could result in a critical shortfall of resources at times when they are urgently needed to prevent the imminent spread of a communicable disease. 3.142 In view of the enormous impact that dzud has had on the Mongolian economy in recent years, it would be preferable to set up a dedicated reserve fund exclusively for dzud costs, focusing on feed supplements, veterinary medicines and transportation. Estimates could be made regarding the actual cost of funding widespread, effective relief and then this figure could be adjusted to reflect the probability that dzud will occur in the forthcoming year. State-owned enterprises heavily taxed in dzud years further endangering fiscal stability 3.143 As a consequence of poor planning in dzud years, numerous state-owned enterprises (SOEs) are called upon by the government to provide inputs for assisting with the overall task of dzud relief. In particular, SOEs in the areas of energy and transportation sector are most likely required to provide assistance. dzud expenditures of SOEs are not reimbursed by the government leaving SOEs financially weakened. As part of the privatization and restructuring program, such practices need to be contained. Recommendations for Controlling Contingent Liabilities Conduct a comprehensive audit of all contingent liabilities throughout the government 3.144 The first step will be to request all budget entities in the country to report any contingent liabilities that they might be aware of. By distributing the Budget Circulars regarding Contingent Liability Disclosure all budget entities in the government will have a formal means of communicating their contingent liability information to the Ministry of Finance and Economy. 59 3.145 In areas of higher risk, a pro-active approach is advisable. The Ministry of Finance and Economy could send representatives to visit organizations that are of types where contingent liabilities are likely to be present. These persons could work with the organization's accountants and management to help identify any potential problem items that may have been overlooked during the budget preparation exercise. 3.146 Particular attention should be focused on state-owned enterprises, both wholly owned and those where the state has only a partial stake. Companies that have been recently privatized should be scrutinized also, since the government could be held responsible for any undisclosed obligations that were not brought to the buyer's attention at the time of sale. 3.147 Another area that merits examination is the area of government pensions, and also, of the pension liabilities of privatized companies. There are currently 248,000 pensioners and the pension has been running at a deficit for some years. At present, it is estimated that the deficit for 2001 will reach Tg 21 billion, up from only Tg 8.6 billion in 1998. It may be necessary to perform a detailed actuarial study to determine the precise extent of the liabilities involved. 3.148 If possible, an attempt should be made to budget for those expenditure items that represent transactions that are part of the normal course of doing government business, even if they arise unexpectedly. These are things such as; social welfare support to national minorities; interest to the BOM on fiscal deficit recovery loans; streamlining regional economic development activities; service fees charged by banks; expenses related to attending visiting dignitaries; and mass media expenditures. 3.149 These items should be included in the annual budgets of the entities that procure or use these goods or services whenever possible. 4. STRENGHTENING BUDGET CONTROL AND EXECUTION FOR SUSTAINED FISCAL DISCIPLINE INTRODUCTION AND SUMMARY 4.1 Improving financial management, information and budget control systems is key to strengthening fiscal balances. Fiscal policy in Mongolia faces a variety of uncertainties; led by external economic shocks natural disasters, and high levels of contingent and explicit liabilities. In the face of such uncertainties it would be prudent to err on the side of caution in determining the aggregate resource constraint that must underlie budget planning. For this adequate information systems are required. Mongolia's fiduciary environment is weak and this impedes sound and disciplined budget management. The uncertainty that Mongolia faces, together with its ambitious program for addressing poverty, requires that it continue developing its rolling Medium-Term Fiscal Framework (MTFF) based on macro projections suitable for fiscal stability, and to link this more effectively with the enforcement of budget control and budget execution. Medium Term Expenditure Frameworks (MTEF) can provide some incentives for sound management of expenditure, enforcement of hard budget constraints, based on the availability forecasts across the medium term, the transparency, accessibility and uniformity of that information. 4.2 Information is the life blood of budgetary, resource allocation and financial management. Financial management systems provide decision makers and the public sector with a set of tools to support, fiscal stability, prioritize expenditure and effectively implement policy choices, and improve efficiency in the use of public resources.27 4.3 The absence of timely and comprehensive financial information.impedes the government's efforts to build an effective financial management structure, improve comprehensiveness of the budget and manage overall balances. For hard budget constraints to bind, systems and processes must be in place to check and reward compliance or enforce controls. Better information management can be an early warning that identifies developing problems-such as arrears buildups-before they reach critical proportions. Decision makers also need reliable financial information to assess the implications of alternative policy decisions such as wage increases to improve budget discipline. 4.4 The installation of a treasury, system is the most important systemic change needed to strengthen budget control and execution and improve fiscal management. A modern treasury can give the government tools to implement its macroeconomic objectives. It can serve as center for government accounting and produce comprehensive, timely, and reliable financial and fiscal data. A fully functional treasury would also assist the government in enforcing the budget law and strengthening financial discipline, and serve as a basis for integrated government financial asset and liability management. 4.5 Improving budget execution and reporting on budget implementation will strengthen the role of central government, the legislature and civil society in monitoring and evaluating government programs. Timely and adequate information on the financial position of budget entities will allow MOFE to enforce I. 27 Public Expenditure Management Handbook, World Bankl999 61 fiscal restraint and trigger disciplinary rules where necessary. Public availability of budget information will also enhance the accountability of local budget governors. Improving financial management systems is central node of the PSMFL and essential if the government is to achieve its twin objectives of growth and poverty reduction. 4.6 The first step in the implementation, execution and monitoring phase of budgets is the creation of transparent, seamless, timely and accurate flows of information from, Parliament to the Ministry of Finance, through budget entity financial management offices to the treasury and finally to the Central Bank through the TSA. Figure 4.2 below outlines a proposed process for Mongolia. 4.7 Improving information systems is not enough to improve overall performance of the public sector. Implementing and improving information systems should however be seen in the wider context of institutional building and policy reform changes. While improving the information base will improve knowledge of decision makers on key aspects of budget implementation, policy choices will be needed to impact effect real change. VA 6foMoiigo1W's1Viasparationsks ;~accuniin& Aiitingt~aar~y~'chai of accot InswiW an effeccon- *~~~ ~~ 4~J ; _ u4 g ~ ial erb ,, * Existence;o ff udge fli ext g ua rantsanduasi-fisca acuivities which are nd tracIed i~ thsystem.Budget frpmentation imiaris an uko level of fiscal isk th in bidgeCdeficittargethg and in ensuring transparency in the use S of,public funds '1".. , p * S ttlemenctoftfansactioris incash, which could resul inurecorde cash leakgs. STRENGTHENING BUDGET EXECUTION AND CONTROL 4.8 Adequate financial management, govenance and budget execution systems in Mongolia are hampered by a number of problems which include, poor capacity, duplication of functions, excessive centralization of budget control functions at the MOFE, poor auditing and accounting procedures and weak overall capacity both the center and at the line ministries. These problems are further compounded by a huge lack of trust between the center agencies and the line ministries and lack of adequate information to inform the legislature or citizens on the effectiveness efficiency of service delivery. The present system suffers from the lack of proper delegation of and accountability for handling budget management. The presence of arrears (Figure 4.1), albeit dropping, budget overruns and 62 misappropriations2 attests to weaknesses in the present institutional arrangement and lack of adequate information management systems to manage deviant conduct by budget entities. 1.40 1.20 1.00 - 080 - loca 0.60 - central 0.40 0.20 1999 2000 2001 Source: Ministry of Finance and Economy, 200 and Staff estimates 4.9 Improving financial management systems is central node of the PSIEFL and is essential if the government is to achieve its twin objectives; growth and poverty reduction otherwise fiscal. The PSMFL relies on the availability of effective and credible rules and most importantly information systems that permit the decentralization of financial management activities, such as budget control and budget execution to the line Ministries or budget governors financial management offices while the Ministry of Finance and Economy and other oversight institutions such as the office of the Auditor General, the Parliament and the public regulate, attest, enforce compliance, train and monitor implementation of policy objectives endorsed by government, given available resources. I. 28 Report of the Auditor General of Mongolia, 2001 63 Table 4-1: Evolution of Local and Central Government Arrears 1999-2001 Of which Total Social Health Clean Comm Arrears Salary insurance insurance Electncity Heating and unicati Food Medicine Other waste expences water Year Level of Government Share of GDP 1999 Central Government 0.54 0.03 0.16 003 0.14 0.06 002 002 001 0.01 008 1999 Local Government 1.26 0.06 0.38 0.06 0.32 0.13 0.05 0.04 001 0.02 020 2000 Central Government 0.57 000 0.09 000 001 003 001 000 0.00 000 0.42 2000 Local Government 106 003 0.13 000 003 030 004 0.01 005 0 13 034 2001 Central Government 023 0.00 0.04 0.00 000 0.02 0.00 000 002 0.01 0.13 2001 Local Government 033 0.01 0.06 0.01 001 0.07 001 001 0.01 0.02 0 13 Share of Total 1999 Central Government 100 5.0 300 4 6 25 3 10.4 3 6 3.2 1 1 1.3 156 1999 Local Government 100 5.0 30.0 4.6 25.3 10.4 3.6 3 2 1.1 1 3 15.6 2000 Central Government 100 0.9 15.5 00 2 2 5.4 1.2 0 8 0.0 0.0 740 2000 Local Government 100 3.2 119 0.2 2 6 28 0 3.5 10 4.4 12.6 32.6 2001 Central Government 100 1 3 17.3 0 1 1 5 10.7 0.5 1 9 7 7 2.2 56.8 2001 Local Government 100 1.6 16.8 1.8 3.7 21.0 3 3 2 0 3.6 5 6 40.5 Source: Ministry of Finance and Economy, 200 and Staff estimates 4.10 The government has begun to adopt policies to bridge the divide between financial management practices and budget discipline needed to improve overall performance of the public sector and ensure fiscal sustainability. But more remains to be done to strengthen the overall basic financial management environment. This chapter examines the current state of financial management practices in Mongolia, assesses the availability of the basic information systems needed to facilitate adequate fiduciary controls and provides recommendations29 for improving current practices which would pave the way for medium term implementation of the PSMFL. 4.11 Government employment programs have left it with an expensive and poor performing civil service that needs careful restructuring. Generous cash transfer programs burden the budget, largely unrecorded SOE liabilities and local government arrears could threaten its future fiscal position. A weak government-dominated financial sector that has already led to one banking crisis must be reformed. High tax rates are a disincentive to private investment and reduce labor force participation. The country is prone to natural disasters, but has no coordinated budgeting or accounting system to handle them. 4.12 Central to all these problems is the absence of timely and comprehensive financial information, which impedes the government's efforts to build an effective and disciplined performance-oriented public sector, improve predictability of funding flows, and manage overall balances. For hard budget constraints to bind, systems and processes must be in place to check and reward compliance or enforce controls. I. 29 Recommendations and discussions build on a rich body of literature such as IMF reports, the CFAA, ADB Technical Assistance notes and KPMG Consultant reports. 64 FigC2einncaIWaW R I MoFE n So M.i o F winc anIcnmyKM eors 02 A PB I 0 P - 1 a m Buge ET-it 4.13 The main deficiencies in the financial management systems and practices which create incentives for poor performance include: lack of a single treasury account, proliferation of budget entity bank accounts, proliferation of budget entities, weak and untimely financial reporting, use of diverse charts of accounts and informal uncoordinated consolidation practices, absence of attestation audits for financial statements, lack of official public disclosure policies, poor enforcement of auditor general's recommendations, weak administrative courts and inadequate parliamentary oversight. 4.14 Basic changes in systems and practices could substantially improve the overall functioning of the budget execution and financial management practices particularly by improving the transparency of the system include3o establishing a new position of accountant general in MOFE and new line ministry positions for ministry chief accountants to provide leadership in financial management reforms, abolishing primary budget governors and setting up portfolio ministries as central budget governors, and adopting systems-based audits as the dominant audit methodology across the government. Under different assistance projects some of these measures are already under implementation but a fundamental lack of the understanding of the comprehensive reform process persist as most of the activities are performed in a piece-meal fashion. 4.15 Nowhere is donor coordination more important than in the design and formulation of assistance to the government for improving financial management and budget execution. The reforms in these areas are difficult and are at the heart of effective fiscal policy and service delivery objectives. Poorly coordinated efforts or attempts at quick short run fixes should be avoided, as they would certainly be more costly and ineffective in the long run. Most importantly however is the need to build first and foremost understanding and capacity at the center to implement these reforms. 4.16 The PSMFL requires these basic capacities, and goes further. First and foremost it requires a competent Treasury, and then providing effective leadership to government accountants and financial I n Lt Soure A Msnunary of minac reondtions ofPthi reports prviedi0Ane. 65 managers in improving entity-based financial management systems. It also requires accrual base financial reports and output-based cost accounting across government. The PSMFL has extensive requirements for MOFE. Given its past performance in establishing the treasury, the MOFE will not be in a position to meet these basic requirements at a speed that will allow successful implementation of PSMFL. That poses a serious risk to successful PSMFL implementation in line ministries and budgetary entities. 4.17 Detailed recommendations for establishing a treasury have been provided by various IMF missions and by the IMF's resident treasury advisor. The World Bank and the ADB have provided detailed recommendations for establishing basic financial accountability and reporting arrangements. Both group of recommendations are the correct first steps in improving financial management reforms. MODERNIZING TREASURY FINANCIAL MANAGEMENT AND INFORMATION SYSTEMS 4.18 The anticipated change of greatest sirificance is the installation of an integrated financial management information, or treasury system. A modem treasury can give the government tools to implement its macroeconomic objectives and monitor progress on the poverty front. It can serves as center a for government accounting and produce comprehensive, timely, and reliable financial and fiscal data for policy discussions and civil society consultation. 4.19 If successful, a treasury system will improve the effectiveness and efficiency of service delivery, by providing better predictability on the availability of funds for key social programs. It will also improve internal accounting controls, streamline disbursement procedures, permit better management of receivables and payables, improve cash management, produce more reliable accounting data, and improve the timeliness and usability of financial reports and statements. 4.20 However, Mongolia is still far from the moment when these advantages can be enjoyed. Some of the early building blocks are either not in place, or not firmly enough in place. The Department of Treasury at MOFE lacks staff, expertise and an authoritative voice. There is no single treasury bank account in which the government's cash resources are concentrated. Procurement of a suitable treasury system has encountered significant delays, and government accounting staff are not yet a disciplined team, as government has not yet introduced a uniform chart of accounts. Certain accounting controls (e.g., accounting for commitments and the issue of MOFE warrants to control the timing of expenditure) do not yet exist. 4.21 Over the last five years, the IMF provided significant amount of guidance to MOFE's efforts to establish the treasury. Recently, IMF's recommendations became structural benchmarks in the PRGF program and IMF has provided a new treasury advisor to assist the MOFE in implementing these recommendations.32 There are still serious delays in implementation, however. The MOFE's top management has yet to fully own and champion the treasury concept33; therefore the treasury lacks strategic leadership, management, and staff necessary if the treasury is to become a capable organization4. This problem has persisted during many ministers and governments, it cannot be solved by foreign advisors writing more reports or by conditionality. I. 31 IMF FAD report 32 Annex 3 presents the strategy for treasury development recommended by IMF's resident treasury adviser. 3 The Minister of Finance and Economy and the head of the Treasury recently attended an international seminar on Treasury Management and there appears to be renewed enthusiasm for the reform. 34 At the time of writing this report, the Treasury had 10 staff, while finance inspectors had 250 staff. 66 4.22 Without a competent treasury, MOFE cannot improve financial management at the budget entity level, nor could meet any of the other requirements of an efficient modem performance oriented public sector, or the PSPML. An ineffective treasury also poses also great risks to financial accountability. 4.23 The essence of the problem is that the treasury concept is perceived as conflicting with the political will for devolution and decentralization. It is seen as "re-centralization," as the treasury and the treasury single account concepts are about providing central discipline to government financial management. Apparent resistance to centralized accounting comes not only from MOFE, but also fromother ministries and agencies. The current system gives agencies a choice on banking services, freedom to allocate own revenues to any spending item and even some de facto flexibility over budgetary resources across line items. FIgure 4-3: Status of Mongolia's Inormation System Undef*dOv0ep Don. Not ha. ~r~er Rrikt ~ ~*F~1 fyaffvt Human Resources Mgt. Information Syatem 1*rt Rcruitment and Selection *P cPteonnel M anagem ant *Portfolib oPosition Management anQ~tIt M44 -id. ~qI~n~I Salary Adminiatration *Tranng and evermdi ~tl~~spI.ioP~r~l~*m~lt-Compensation andSB.enfits *Paeyroll Project Management -Staff Retirement -Project Manegement -Management Reporting *Management Reporting Travel Integrated Database *T rlp Planning/Authorization investments Management Travel Advances *Portfolio Management iTrnveg Expenses HCuurrency Trading *Mgt. Reporting RPesone Management *Management Reporting Borrowings Revenue System (Inputs Revenue Date) :Bond Purchases/Redemptions -Corporate & Personal Income Tax *u rrency/interest Rate -Sales & VAT Tax :Accruals/Amort*zationa ACustom s Financial Analysis *Expen*e Processing aExcise -Rinn alysti 'Management Reporting *Road transport.. etc :Fin*Staft Rerecrement Mnaaegemennt Reporting Treasury Reference Model, World Bank 2000 4.24 Lack of enthusiasm for the Treasury Single Account (TSA) may also stem from the authorities confusion and frustration with conflicting advice from donor agencies on the appropriate implementation strategies." Agencies do not trust the idea of a treasury single account, because they are not certain of the quality of payment services that would be provided by treasury offices or banking services by BOM, which is not a transaction-oriented service bank. Rules on the handling of own revenues are also not clear, and agencies fear treasury and MOFE may exert line item control on them as they do on budgetary resources. The treasury could alleviate agency fears by allowing them to continue with their own accounting and payment function, contracting a single commercial bank where agencies could keep their zero balance accounts, and establishing clear rules for special revenues. Legislation is needed to determine which banks qualify to serve as payment centers for the treasury. 1. 35 The Cabinet on May 24, 2002 adopted a resolution selecting the TSA concept based on centralized payment processing through treasury consisting of central office at the MOFE, local offices at aimiags, UB city and its districts, and treasury representatives in soums mostly based on existing accounting cadre. 67 o 1t t4{ un~~R f ~ r~' ~'"1 hfsq onld,' uPlare,i ffr: exiend . oce7gutintara s roia a d r i g?errs commini a Wi a -' te nearetk locatffies' S e a g nissë tleg xp thtreans ir$ th rieir asryf cnfqpam Hä. 3 h h Cen B s su h- fi -. Åiný' - as A" L~ baccru.o Se tgorn t He qr h ce and-Economyexp tea henn e nagn manae enì'dë iti syst rl e dhrdTó u enti es r~ - o rh ngivi pay e altnisartongth build upofearrears inoic ancsk pa.y an claing st fac'*lit *ae i i ieorcp s rgangzatonatrture foriltigssårys'rquised I paymetsPQh 'c a Iné paymntps.semsenthe'er r.awBand.l ty .al 6fit,o,tantesr 4.25c e a at'ddhecenteraitonal ctier o mest hoprymfbildin herdqua.lsone npistliesndabudet cnerens, i seinrs workdt,shsérang qf!et EsalsigaTAi nepnier cenaton goborae bymuthe Ith W andc6t~ ADB. & ~hd d 4.26 he rsponsbiliy fo manann eaccurant ing sstemand1 eportng fina uncaeulth ise un e r e t h e l a tl i t of f s S p e n d g e ,tp n d ii t u c n ib Oe are st t ré a s u r y h a r e p a y e ns kTras¾eerncftäS 94r4 Bank Tåchna1Öero5,5åijnhnu ffie f'r*jPj-ý' 1 1 ben establishemod. popEd bcoueris littla lioyetodevo utio n and anenthed stem.fi gure42) Under thios ysnt mn try budget t acetfilindappro,vepaymer. hesepaymets slips aie t n senr nkthe auntsw rq as a pas threci fo cn anue urs MO E hgermenitasn id atetund i,n the treasFy ne this ifda'te technology a!dnidoan esorce andkU fo 0 milln) t o forhensta BankiunderahTS ansystem and2w 006l the iio) lan raework fitsy o ply thesrolsof retai (50*5*6000=1.5mil- li o bake o hegvemetHoevrth1MiscyofFiai adEqn mämst'ensiiieith,at thé n&cessary fipn#ial rnmanagernentfand Eontrol:svsäéms are. 9dhered t,Ölgibudi iernit ie p r: to: making payrnentå.:('" jýThe4q s eo anälecrom tr kpyneland clearnng'systém4faidility-failiates'.fficn pept of, receipts u~nAeSres thttet tsr a acsra and ä*nlyjiä?räianon on balancesaåhd transadfoi i' fs accounis pnrio to s u bnnng-bä h is rnenrs to the céntral Bank&'."S'rjX x, *tZ .** In Monoha tesel1inkåges"are yet4 to, be understoödlandl th sjystis ,tre notinRaebhatherauyadathe OCW.rI'Bn ko-r6?h 'ptojecrsý7 uneä~ 6 iiit 111iaI nformardin systemrs ati tih rasr a new inter-bank;', p9ynenIsystems at'the Cetra-Bank. ,Lj*''-tJI, 4=6mJllion), and around fifty local professional staiff (50*5*6000-1 .Smiillion). 68 bank accounts used by budget entities and has issued instructions that no entity may have more than one, These are prelininary moves associated with the establishment of a treasury single account, which is expected to improve financial control and cash management. 1ýteidetedeafflforaMÉ 2 ahispitohithl dthy l u« $u'ahjuge uiå,,btf théérlovennnlhWöa governmentihastheir own bankk acount., he allcated6amount is uomatically'uransferred a.ihe beginning fo die. monthiffon Tirisiv téthe-budget units'brikacounts.iThis.syteni'commits a great deal oast',t6 the,spending. unt,'brk accöun äbforeihéicash isFer.,ne' c'&ieiainvyiolates t basic aontrl princpals that are srblisein ae resur- Singie A&dun. As nnys! aredu e toubepaid for' salaries and wäges or fi-- gods and 2 eic,a ent 13rders ie Ñeedandie.acumntr thé b'Lidiët',u!nit,aimäg -år the soui$proves' the. payrnentrder for,paymei 6y e bank..Tge,mon,e fio'the,payment óder is)geneiallydrw'iniriah to pay the. .. - . . -1. -1i .- -. ..,Pq 1 . i.withii th' .. ,tem to ' , iha .: th . eI od ill obligationt' Thére appears,tobe no controliwithMri thetem tof ssure ihat total -purchasés for the,pend wil nol exceed :tikllocatio.lT'husrhthe opp'rtunity exists'larrears1to buil'd up. There' is alsoAot aformal writuen! .piocedure:for'con&olf- dgt'spending.: . ' L- .,. 4.28 The principal reporting device is a multi-sheet form budget entities complete each month and each quarter and send to MOFE. The completed financial reports are received by Treasury and aggregated to provide government-wide reports. The financial report does not require accountants to reconcile changes in a budget entity's cash and bank balances between the beginning and end of a reporting period with the surplus or deficit shown by fiscal reports for the same period. Nor does it require regular bank reconciliation. The result is that financial statements are inaccurate. liccounting r i an expenies when cash s,either received asrevenues or when rec gnze ittevn e d_d5u cash isaida'sependitres All durchaeï.f gdš'or'ser-vices. regard e ortheialue. would be exensed to the budget There ould rnot be'6aiablan eèt s eall;.,agset p,rchases are expensed-t- th5 budge.1Liabiliues or; :Rcivableswou!d not,exisjsince these ir: not cash ransactions,.nd do not,get recognied.until the transaction. resuftsin ipreäeprvor payient,of'cash.Finåncia,stateients:would oinistlof å Cash Flow taterént, which wbuld. réflåcthechngé icash position during~a,reporphä, period:The" a,iemeni of!Reveues: and Expenses iould '' ' 1, 4.': ', .'.: .& ;,- q 1. essentially 6den eé fr&n the Budgetiatements. . .: rl .I. Modified casi pccoubing isrnotconsidrd in thisj instance since ii s notécnsidered much of an improver ent over cashbasis accouning . p . .. - - . .. . Xecrual accountng recogmzes a transacuon henit occurs as opposed to «'ben the cash changes' hands FOT Uarnplea Trade"PayableexKLs 'then an i rem is,purhsdedand the'r&eiingentiy hasa legal obiga'tion.to pay tor, the'goods or services. Likey,ise,, revenuesrekecognized%Z,heri.egallvaarned. The enity wold be required to éstaibhsh a capitalizauon policv hreb' all goods purchasedover a certain- value or within other criteria would not' b e'xpened b't &öuid b 1 :Modified 'acc-ual aIcoäintingis a.level,.bey~dffshbasis in that it contains certain eléfients of accrual accounting. 'ut doe not>.conain the copnleterecogriiuglpf on-financiál.insä"ti ns it d s not recogrizethe purase of gdas a: se nor es gni-zeé i- subseueéni ýiindni .ihiig alue of the s dprdciaon Most. conservativeapproacheso died aérua'l'accoutithä do'not recognize accruals fär sulh reyenues as taxes or fees 'at,are oing to the government We rec rimiß tht the GOM'adopt modified äccröal aecountngto include all td paå bls. slariesandÄ ages social se crity'andiriteresreceivabIe or paÉableondebt 'i ' ' - . ya..~~ ~ ~ ~ ., aI. , ý1 69 4.29 The consolidated government-wide fiscal report prepared by MOFE showing receipts and payments for central government is presented to the State Great Hural as a statement of annual budgetary execution--the government's annual financial statement -after the end of each year. 4.30 Donor implementing agencies are required to maintain separate project accounts as per the requirements of the Financial Covenants of DCAs. In Mongolia, however, systems to maintain such separate accounts have not been developed. In most cases, information on project expenditures are extracted from the government ledgers to prepare project accounts. Although the Accountancy Law stipulates that financial statements be prepared in conformity with International Accounting Standards (IAS), compliance with IAS remain weak, partly due to the lack of capacity and enforcement. An acceptable set of financial statements should include a project balance sheet, a statement of sources and uses of project funds, and a Special Account statement. Financial statements that include all project expenditures financed by the Bank as well as the government are often not prepared. Lack of acceptable accounting systems and poor compliance with accounting standards have posed serious risks to the Bank portfolio. Requirements of PSMFL The law requires budget entities to produce accrual-based financial statements in accord with international public sector accounting standards. MOFE is required to consolidate agency statements to produce government-wide financial statements, and SAIC is to audit these financial statements37. The law also requires budget entities to cost their outputs. Accounting Practices of Budget Entities 4.31 Mongolia has a good legal framework for accounting, however actual practice varies widely. Mechanisms for enforcement are lacking. The Accounting Law of 1993 has recently been replaced by a new accounting Law based on international Accounting Standards. The law is administered by MOFE, which regulates the maintenance of accounting records and the preparation of balance sheets by business entities and organizations in both the private and the public sectors. The Law makes the Minister of Finance responsible for providing professional and methodological direction to Mongolian accounting arrangements. 4.32 The Accounting Law also sets out financial reporting requirements. These require that the Minister of Finance shall approve standard accounting forms and methods for business entities and organizations. Budget entities produce accrual-type accounting information (receivables, payables, inventory etc.) and data on fixed assets. But the registers from which reports are taken are not well designed. Consequently very small and much larger assets, current and fixed assets are all reported together and without adequate analysis by asset class or location. The existence of such accounting data provides a possible basis for an accrual accounting system for budget entities38 4.33 Significant inaccuracies exist in cash-based reports, and financial reports bear only superficial resemblance to conventional financial statements in developed countries. Reporting formats will have to be redesigned. Major difficulties should be anticipated in asset accounting and revaluation. The transition of accounting in budget entities to international standard accrual system should be made gradually in the notes attached stating the deviations in existing treatments from the existing standards. The transition to I. 3 To be discussed later under section 3. 38 The Public Administration Reform Project (ADB) has taken five budget entities and helped them to make the necessary accounting changes to produce accrual accounting statements 70 the full accrual accounting in budget should be driven by the needs in more reliable and timely information for users carefully weighting the costs and benefits. 4.34 The Fiscal Technical Assistance Project located in MOFE has drafted a new chart of accounts for government designed to operate on either the cash or modified accruals basis. The chart has been piloted and is nearing completion. It contains brief accounting standards (for both cash and accruals bases). Adoption of the new classification scheme and accounting manual would provide a basis for budget entities to produce uniform accounts and for MOFE to produce fiscal reports in compliance with the new Government Finance Statistics Handbook. This will help Treasury produce improved government-wide statements. Role of Chief Accountants 4.35 At present, government accountants lack adequate instruction, supervision and leadership. A regulation was passed in June 2001 on the appointment, dismissal, and qualifications for chief accountants. A budget governor makes an official request to MOFE when a chief accountant is to be appointed or dismissed. Thus budget governors appoint chief accountants with MOFE having the power of approval or non-approval. This is the first step in creating a more unified and effective accounting force. 4.36 Generally each budgetary entity has a chief accountant and there is no hierarchical relation between the chief accountant of a ministry and chief accountants of subordinate bodies. Each chief accountant is responsible directly to MOFE and particularly to the treasury, which now lacks the authority to provide leadership, instruction, and supervision to chief accountants. Local governments practice assigning one accountant to several budget entities. A typical soum would employ one or two accountants to cover the accounting function for both the soum budget and accounting for financial operations of soum budget entities. Reform options 4.37 A new position-Accountant General-needs to be created under MOFE, preferably as part of treasury, to provide strong centralized leadership to chief accountants and take responsibility for directing government financial management, modernizing the government's financial management systems, and strengthening financial reporting. The functions of Accountant General should include issuing government-wide accounting policies, standards, establishing financial management policies, regulations, and monitoring the establishment and operation of financial management systems. The Accountant General should monitor resources required to effectively operate, maintain, and enhance financial management systems and make recommendations to ministry heads on the administrative structure of ministry financial management activities. He should approve the appointment of ministry chief accountants and oversee their performance, monitor the financial execution of the budget, and ensure that the government has a highly qualified cadre of financial management professionals. 4.38 In addition to the Accountant General, it might be useful to create a hierarchical relation between the chief accountant of a portfolio ministry and chief accountants of subordinate budgetary entities. Creating such a middle layer of authority between the treasury and a budgetary entity will ease treasury's responsibility and help the line ministries manage their portfolios more effectively. In such a framework the treasury would issue system-wide regulations and ministry chief accountants oversee the 71 implementation of these regulations in entities in their portfolios. A ministry chief accountant would also need to develop portfolio-specific financial management regulations. 4.39 The responsibilities of ministry chief accountants would include developing and maintaining integrated accounting and financial management systems at entities under their authority, directing, managing, and providing policy guidance and oversight of all ministry financial management personnel, activities, and operations, approving and managing financial management systems design and enhancement projects, developing budgets for financial management operations and improvements; overseeing recruitment, selection, and training of personnel to carry out ministry financial management functions, implementing ministry asset management systems, and monitoring the financial execution of ministry budgets in relation to actual expenditures. 4.40 The Accountant General position needs to be established immediately in MOFE. Ministry chief accountants could be formed as ministries implement other requirements of PSMFL. The Office of Accountant General could establish a group of 10 to 15 chief accountants and train them for chief accountant positions. Role of Government Auditors and Audits 4.41 Insufficiency, poor capacity and duplication of both the internal and external audit functions has systematically resulted in the poor use of funds at the budget entity level. Both the internal and external auditors limit their job to enforcing compliance with administrative rules and rarely look into issues of efficiency or effectiveness. In addition the absence of clear and transparent legal and administrative institutions to enforce recommendations of the auditors has further contributed to weaken the system and increase inefficiencies, as evidenced by rising arrears and high budget overruns. Strengthening the functions of the auditors is important if governance issues are to avert. 4.42 The typical form of audit is inspection aimed at discovering instances of non-compliance - detailed examinations to discover whether regulations have been followed. Such inspections act as a deterrent to misappropriation, leakage, and corruption. The introduction of risk assessment and systems- based audit would increase the effectiveness of compliance checks and would bridge the current gap between inspecting and auditing. 4.43 There are three layers of financial inspection in central government. At the lowest layer central budget governors have one or two financial inspectors, at the middle, the MOFE has the State Financial Inspection (SFI) body with around 250 staff, at the top is the independent State Audit and Inspection Committee (SAIC) with 300 staff, reporting to the parliament. 4.44 The mandate of SF1 is also quite wide. It covers about 4000 budget entities and state enterprises that are also under the SAIC's jurisdiction. SAIC is responsible for checking, budget revenue collection and expenditure, use of state property, state owned banks including the Bank of Mongolia, state enterprises, foreign financed loans and grants, use of government special funds, and the ethical behaviour of about 200 key officials (excluding members of parliament, the President, prosecutors, and judges). 4.45 The two bodies do very similar work. But they operate separately and without coordination. They use similar inspection methodologies", impose penalties, and give instructions for reimbursement of I. The framework for sharing responsibilities between the Accountant General and ministry chief accountants is provided in Annex 4. 4 SF1 inspects at a more detailed level than SAIC (primary documents, prices, suppliers etc). 72 money that has been wrongly applied. They do not share audit reports and findings, their annual work programs are not aligned and they rarely follow up each other work. Although their experience in inspection allows them to have a good understanding of the state of internal controls in audited entities, neither SAIC nor SFI address systems-related weaknesses in their reports. 4.46 The lack of attestation audit (giving an opinion on the reliability of financial statements) implies a possible lack of integrity of financial statements. In addition, performance audit (the type of audit that establishes whether value for money has been achieved) does not yet exist. SAIC is taking steps to establish the legal authority for such audits. 4.47 Although there are more than 500 financial inspectors, budget entities do not have internal auditors. There is little local knowledge of internal control (as opposed to the current concept of abiding by rules set outside the entity), hence very little appreciation and need for internal auditors. They would need them under the new regime to assure managers that the entity's internal controls are working. Recommendations 4.48 SAIC and SFI should adopt systems-based auditing starting with the assessment of internal control systems of the bodies which they inspect. Systems-based audit should replace compliance inspection as the dominant audit methodology across the government. This would requires extensive training for around 500 auditors. 4.49 SAIC should proceed with its plans to carry out attestation audits. It should first develop the expertise and should as soon as is practicable thereafter provide an opinion on the annual financial statements of government presented to Parliament annually. 4.50 SAIC should proceed with its plans to carry out performance audits by developing the necessary expertise and applying it to carefully selected audits. SFI's aimag operations should base their methodology on internal auditing principles. SFI central operations should concentrate on two areas, developing internal audit methodology for use at the aimag level and coordinating its work more closely with SAIC. 4.51 The types of work carried out by SAIC and SFI are so similar that far greater co-ordination of work programs would be advantageous. The case for merger might be explored. 4.52 In line with budget entities playing a more independent role under the proposed PSMFL internal audits should be established in entities where they can be expected to yield immediate benefits (e.g. large state enterprises and hospitals). Requirements of PSMFL 4.53 The PSMFL requires the SAIC to audit the financial statements of budgetary entities and the whole of government, which means focusing on attestation audits. The PSMFL has no direct recommendation for SF. Although the PSMFL has no direct recommendation about internal audits, the portfolio minister concept implies a need to develop internal audit units at ministry and at entity level. 73 5. INTERGOVERNMENTAL FISCAL RELATIONS AND SERVICE DELIVERY INTRODUCTION AND SUMMARY 5.1 Developing a system and incentive framework to meet the twin challenges of achieving sustainable fiscal balances and quality service provision is at the center of the debate on the role, size, and function of Mongolia's public sector. The outcome of this debate will have a substantial impact on the pace and direction of reform-and on the speed with which the government achieves its objectives of increasing growth and reducing poverty. Mongolia has been grappling with decentralization for some time. Balancing the imperatives of fiscal sustainability with the potential for radically improving the operational performance of government by following the principles of the PSMFL requires explicit management of the inevitable tension between them. A centralized budget process is more likely to yield fiscal restraint-but decentralized service provision can deliver better quality services to the population when local government accountability rules are clearly established and capacity upgraded. 5.2 In the mid-1990s, as the central government tried to maintain macroeconomic stability in the context of dramatically contracting output, lower levels of government were asked to take on increasing responsibilities for providing services, often with limited ability to raise financial resources and with limited capacity and decision-making authority. Local government's share in general government expenditure rose to an all-time high of 35 percent. Financial relations between the levels of government worsened, however, as local governments began to build up arrears. By 2000, local government's share of expenditure dropped to about 28 percent as the central government increasingly assumed more expenditure responsibilities. 5.3 While much has been done to reform the public sector in the last 10 years, it is not yet an effective market economy-enforcing institution. The centerpiece of the government's reform agenda is the PSMFL, and its main thrust calls for a shift from input-based budgeting to output- or performance-based budgeting. Service delivery functions under the PSMFL will be devolved to the local providers and overseen by budget governors. This will require reforms in the organization of government, the mechanisms for strategic priority setting, the budget formulation and execution process, financial management systems, and the civil service. Broadly, PSMFL relies on a system of strict financial management controls, and a lean core public sector. But PSMFL involves significant risks, and the government will need shorter-term strategies to bridge the gap between current practices and the new system. It must strengthen administrative controls and define a clear and transparent rules-based system to guide public sector activities, performance, and size. 5.4 Local governments in Mongolia spend the bulk of their resources on-and are the principal providers of- primary and secondary education and health care. These services are central to improving the living standards of the poor and increasing the assets of the population. Intergovernmental rules undermine incentives for improved service delivery 5.5 The anticipated benefits of decentralization in Mongolia are being undermined by a lack of clarity over roles and responsibilities in the legal and institutional framework. Incentive systems for service delivery are poor due to conflicts between expenditure and revenue assignments and inefficient 74 and inequitable transfer mechanisms. This chapter will address each of these issues and their links to the design of an institutional framework that combines incentives for macro-stability with improved service delivery, particularly to the poor. 5.6 Before the transition Mongolia had a strong central government with extensive powers, while local governments had few independent expenditure or revenue powers. Under the 1992 Budget Law Mongolia began restructuring its public sector, and it has since implemented substantial changes in its fiscal structures. Intergovernmental reforms have awarded more budget authority, including ownership of local assets, to local governments. However, the center continues to wield considerable authority over the local governments. Three major weaknesses undermine incentives for strong intergovernmental fiscal relations 5.7 Three major problems have hindered Mongolia's efforts to build a decentralized institutional framework that embraces more market economy characteristics and devolves service delivery to local governments with clear roles and responsibilities. First, there are large asymmetries between expenditure responsibility and decision-making authority. Second, poor accountability systems govern the multiple local government administrative units, webs of budget agents, and administrative nodes responsible for service delivery. Third, unpredictable funding and sharing arrangements from the center undercut already weak local government revenue-raising capacity. 5.8 This cumbersome and non-transparent framework has given local governments few incentives to rationalize expenditures, increase service delivery efficiency, or support regional equity. The multiplicity of lower-tier government levels and agencies has also prevented the nascent financial management institutions from disciplining budget execution effectively. 5.9 Several measures could improve public sector performance and fiscal discipline. On the legislative side, there is a need to reform government structures while taking into account the changing role of government, Mongolia's changing demographic structure, and appropriate scales for service delivery. This should include streamlining the number of budget entities, clarifying expenditure assignments, and consolidating service delivery agencies. The legislature should also strengthen institutions and systems that support accountability, such as local parliaments, accounting, and treasury offices. Matching expenditure assignments with authority would improve local government accountability. Improvements in the revenue sharing system to provide local governments with more revenue-generating authority and more stable tax-sharing systems would increase fiscal discipline at lower levels. The government should also develop mechanisms for streamlining funding (single-pipe funding), use equalization formulas for the distribution of transfers, and consolidate the provision of national public goods, such as policy formulation, monitoring, and evaluation at the center. Local government structure and economies of scale 5.10 Under central planning the administrative arms of the national government carried out functions and delivered services at the aimag and, often, soum levels. Following Soviet withdrawal, many transition countries have devolved authority by creating new layers of local self-government. These have often, however, been based on political considerations, rather than effective service delivery or economies of scale. Increasing pressures to improve fiscal balances and lack of capacity at lower levels of government have led to movements to reconsolidate local governments in many FSU countries. Poland, Albania, and Hungary, for example, have adopted measures to reduce the number of sub-national governments. 5.11 Widely varying regional populations and demographic changes during the transition have created further problems for Mongolia (Table 5.1). Aimag populations vary widely-from 12,300 in Gobi- 75 Sumber to 113,000 in Uburkhangai-and migration into urban centers, UB city and Orkhon in particular has led to significant population shifts in recent years. Table 5-1: Demographic and Administrative Structures in Mongolia 1996-2000 Number of Average population per soum Change 4imtags Population ('000s) Soumns (.000s) .1996t2000 1996 1997 1998 1999 2000 1996 1997 1998 1999 2000. Arkhangai 104 104 104 971 97.5 19 5.5 5.5 5.5 5.1 5.1 -6 3ayan-Ulgii 92 94 96 91.1 94.6 14 6.5 6.7 6.9 65 6.E 3 Bayankhongor 91 91 92 84.8 85.3 20 4.5 4.5 4.6 4.2 4.3 -6 ulgan 64 65 66 61.8 62.6 18 3.6 3.6 3.7 3.4 3.5 -2 bi-Altai 75 75 75 63.7 63. 20 3.7 3.8 3.7 3.2 3.2 -15 iobi-Sumber 13 13 13 12.2 12.3 1 12.8 13.1 13 1 12.2 12.3 -4 khan-uul 91 93 94 83.3 84.8 4 22.7 23.1 23.6 20.8 21.2 -7 mogobi 49 49 50 50.6 51.1 14 3.5 3.5 3.6 3 6 3.7 5 Dornd 86 84 84 75.4 74 14 6.1 6 6 5.4 5.3 -13 Dundgobi 53 54 55 51.5 51.3 15 3.6 3.6 3.6 3.4 3.4 -4 Zavkhan 107 106 105 90 87.1 25 4.3 4.3 4.2 36 3.5 -18 3rkhon 67 71 73 71.5 76 1 67.3 70.8 72.5 71.5 76 13 Jburkhangat 114 116 117 111 113 18 6.3 6.4 6.5 6.2 6.3 -1 Umuoi46 46 46 46.9 46.9 14 3.3 3.3 3.3 3.4 3.4 3 ukhbaatar 60 60 60 56.2 55.9 13 4.6 4.6 4.6 4.3 4.3 -6 elenge 104 106 107 100 101 17 6.1 6.2 6.3 5.9 5.9 -3 mv 112 114 114 99.3 98 27 4.2 4.2 4.2 3.7 3.6 -13 vs 103 102 101 90 86.8 20 5.1 5.1 5 4.5 4.3 -16 obd 92 93 93 86.8 878 17 5.4 5.5 5.5 5.1 5.2 -4 uvsgul 121 123 124 119 120 23 5.3 5.3 5.4 5.2 5.2 -1 Khentii 76 77 78 70.9 71.4 21 3.6 3.7 3.7 3.4 34 -6 4imag Total 1,717 1,734 1,744 1,614 1,621 335 5.1 5.2 5.2 4.8 4.8 -6 Ulaanbaatar City 629 646 669 760 787 9 69.9 71.7 74.3 84.5 87.4 25 I'otal 2,347 2,380 2413 2,74 2 344 6.8 6.9 7 6.9 7 3 Source - National Stanstical Office ofMongolia, 2000 5.12 The government has begun discussions on consolidating local government administrative structures in the context of the overall public sector reform agenda and drafted a regional development concept paper. Developing a consolidation strategy is still, however, hampered by a desire to legislate poles of economic activity. 5.13 The challenge for Mongolia as discussions on improving service delivery and strengthening financial discipline evolve will be to find stable government structures that are politically acceptable and help promote efficient and effective local governments. It will be critically important that reform measures clearly take into account the appropriate scale for efficient services delivery and that principles such as matching the roles and responsibilities of different actors are set out in a clear legal framework and matched by adequate resources. 5.14 Some countries have addressed the need to balance economies of scale and changing demographics by aggregating smaller localities. Where aggregation is not politically possible, sub- national links across levels have been forged to pool service delivery. Mongolia has used inter-soum hospitals to pool health sector resources, and this structure could be expanded to many other areas, including tax agencies, police, and education. Political will is needed. 76 Recent trends in Sub-national Revenue and Expenditure Measuring the effective degree of decentralization in Mongolia 5.15 Decentralization transfers authority and responsibility for public functions to subordinate governments, and covers political, administrative, fiscal, and market decentralization-all of which can overlap. There are also different degrees of decentralization: under deconcentration lower levels administer central government decisions; delegation transfers some responsibility for decision making to lower levels, and devolution transfers full responsibility for decision making, financing, and management. 5.16 The most common measures of decentralization are local government expenditures as a share of GDP, and local expenditures as a share of general government expenditure. Local government expenditures rose from 10.3 percent of GDP in 1996 to over 13.1 percent in 2000 and are estimated at 14.1 percent in 2001 (Figure 5.2). This is well above the average of about 10 percent shown by other transition economies. Local government expenditures as a share of consolidated general government expenditures fell over the same period from all time highs of 31 percent to about 26 percent, moving closer to other transition economy levels. 5.17 While Mongolia has devolved substantial administrative authority to aimags and soums, decisions on budget spending and allocation of appropriated budgets are made at the aimag level fiscal decentralization has been much more limited. Over eighty percent of the aimags depend on the center for fiscal resources. Local governments have very little or no revenue raising capacity. 5.18 But local government revenues have failed to keep pace with expenditures. Local government revenue as a share of GDP increased between 1996 and 2000 from 5.9 percent of GDP to 6.4 percent. However local government revenue as a share of total revenue has dropped substantially over the same period from 23.5 percent in 1996 to 17.5 percent in 2000. 5.19 While both revenue and expenditures of the local governments as a share of GDP have increased over the period, expenditures have increased faster than revenue. There appears to be a shift in expenditure responsibilities from local governments to the center in 1999, but a renewed increase in responsibilities being shifted to local levels in 2001. Revenue, as a share of either GDP, has not matched increases in expenditures indicating that resources have not been transferred to local governments to match expenditure assignments. 5.20 But Mongolia has not devolved authority in line with fiscal responsibility. While the data show that local expenditures in Mongolia are considerable at over 13 percent of GDP, local governments administer much of this expenditure in a de-concentrated way. To fully assess the degree of decentralization, therefore, it is necessary to consider fiscal decentralization comprehensively by assessing the existing structure of government and institutions that support accountability and responsibility arrangements at all levels of government. The next sections will examine these issues. 77 Figure 5-1: Share of Local Gov. Revenue and Exp. Figure 5-2: Share of Local Gov Revenue and Exp. in GDP 1996-2001 (percent) in Total, 1996-2001 (percent) 16 14 35 12 30 1 25 8 20 6 15 4 10 I 0 1996 1997 1998 1999 2000 2001 1996 1997 1998 1999 2000 2001 , Local Government Revenue GDP U Total local gov. exp. % total expenditure F Local G o expenditure mn GDP li Total Local Gov. revenue % total revenues Source: Minstry of Fmance and Economy 2001. 5.21 An overview of intergovernmental relations in Mongolia suggests there is a centralized approach to intergovernmental fiscal relations, local governments are assigned a set of expenditure responsibilities, and their share of revenue to GDP is relatively small. The system of intergovernmental relations is consistent in practice with the Constitution, which sets the central government as the principal agent, and the local governments as agents of the center rather than independent entities. Any reforms to the set of rules and regulations governing these relations in the future will need to be consistent with these principles. 5.22 Mongolia recognizes these issues, and the govemnment and legislature are seeking ways to bring local revenues and responsibilities into better balance. This process will have to be guided by a few key principles, including matching expanding responsibility with sufficient authority and revenue generating capacity, increasing the predictability of local government funding, and more clearly defining roles and responsibilities. Any new system will have to be based on adequate-and more transparent-d- mkances..ffic .pbecossile withoutcreases in iequality if the.central govemnmen-keeps rayté responsibtyforfii,iinng while.local governments take a.er respnsiiitfor,spedin disions mnputst and implenmetadaonom social:sjervices may .nhereni be, moretdifficttro decentrahize tlian.'others.1 Economic, theory suggests that' redstbuonma,bsbecr otbhihrlvesf"genenbcue labór'mobilir wi ¶mk -attemps(bÿ9oe jurrådictions-to;changeftlie distribution of iicome'self-defe atng as the,poor gavtate to'are'of-highrliátilion, while>jhh rich custer in areas of.lòEredistrilion.Silieven if iFëcnraI'oNrnments takeprinary-respanrsibLiÅyÿ far fiiianriéngf atsfety nels, an -aaÍs th iera H'c d e biffor'ra~ers e',l0cal' governmens'may i hase n informatiol advantae in s ieeenin apptpropohatton.. i,I 'L nt Desen'trizonf ol7theae scor is aha complicated because of the need for effeciive .eferral cr ielsfrrheathpous which praide,baic snervces' to highehogiyo sp,itae Unless hese int limags re ponsidired areftiljy decentralizationcan resiiülnIiideotera:ionofJ6me aspecis of the eices- provie as may hvehappened n th.P hpmes BovPa,and.ami'a. ;Theab d leveli of the sub-nationagoernment tohich.responsibilities auredecentraized: Econistsf ften.. aåéuUnited>States e i'n . :;acw¶.a. LsD'6. L -']' e;h ahliPý, att soia s"ycsr etfntIy'hienali rn xponen,ce sugges.that efficiency gamnsin the pro,vision.soi-ercsrqun manlze "winfi entra, governmient'devolves resppnsibi1btes to.the -commumtyror ficiry.:lëvel bu rarelywhendthey! ar e provmcisorrégonslrn every instaice,t 's imprtant that-revenues ocal services.foIlow .4~~ K respons bii nesfor theindelivery , . a: .e Ihs Th x se of c n4 ity ~btliztionard oversight"i ncreases inalaarive efficiency j .the,ere asocial- sevcsan only take,p ace.if-rgoreiaccurate local intormation can reach decision-makers, arnd if there arer pmiechamsmswhereby!"heée decision malrs areheld' accou tor'their performance In Coloinia iaccountabilty.io4 constitentsapushed Iocaltmayors to concentrate more on tranung and 'hing'effécave civil srnInNdruias1étizil #mmini aversidh ar dt of jo lasi helpedj'motr.ite civil srvantÉb Ir' Ét e n fincome is impotant. If nicme is distrid-ited mfrùneve with Jursdicuons..tan >across théemrdecentrahzaonocould' b egualmng if local authornes;have the capacity-to Ir e'cmé lf oradsEte'eqtuty abjectveséf:ibé&enter On the other band if there are iag l * uuadifferences tnjcapacrty1 ar usdco n,share the same equuty obevssme1unaonal~ .'oerets 'éay effectivelntarget the poor..GC,eital¯goverimen.ts may o there fors have, tager povertyi funds themselves arcreatstronger mceves for sub national governmnents to dons. ' :- o','i'!'Q ' i uc P1 ' l ine Pe As'me-um't' 'o e World ' 200i s'fiiefuf ovq1pn-Iv iéssnænro oductio aM' is' i Bak 00 i 'i ~ ~'rI~iPt f : . ,ip: i ~ itnct triuioivad lo.ý I-an ' 80 5.29 Aimags and soums may take independent decisions in matters of socio-economic life in their jurisdictions, and higher-level authorities are prohibited from deciding matters under the jurisdiction of local bodies. Local governors must approve decisions, however. Local governments also enjoy some budget autonomy. They are allowed to spend the resources available to them from transfers and own revenue, however they do not have any decision making powers on the level of taxes or tax-raising powers. These bodies are also free to legislate on local matters within the framework of the Constitution when national legislation is lacking or unspecific. Governors of sub-national units have a dual allegiance: they represent the state, implement state policies at the local level, and are responsible for the "proper observance of national laws and fulfillment of the decisions of government and the respective superior body in his/her territory."41 But they also have to observe and implement the decisions of their constituencies, as represented by the hurals. A governor can veto decisions by the local parliament, but the hural can override this veto by a simple majority, in which case the governor must obey or offer to resign. There is a need for clarity in the roles and responsibilities between different levels of government 5.30 The key issue for the system of intergovernmental finance is accountability: to whom are the different levels of government responsible? By its very nature, the current structure of intergovernmental finance incorporates two competing directions of accountability. The first is "vertical accountability": lower levels of executive government bodies are held accountable to higher levels. The second is "horizontal accountability" in which legislatures are held accountable to the citizens and voters who have elected them. 5.31 Duality of accountability directions undermines effectiveness of the system. To improve the effectiveness of intergovernmental finance, a determination must be made as to which type of accountability takes priority. Emphasis on vertical accountability implies a system of intergovernmental finance that places greater emphasis on deconcentrated responsibilities. Emphasis on horizontal accountability means devolving responsibility and decision making to lower levels of government. International experience indicates that the most appropriate structure is highly conditioned on the capacity of local governments to carry out responsibilities. 5.32 There have been significant attempts to increase clarity in assigning roles and responsibilities in the public sector in Mongolia. In 1996 amendments to the Law on Government clarified the roles of the ministries, executive agencies, regulatory agencies and implementing agencies in the overall budget framework, but some overlaps and lack of clarity persist. MOFE's role, for example, is not well specified, which has resulted in it sometimes acting as both implementing and regulatory agency. 5.33 The allocation of responsibilities between different levels of government and intergovernmental fiscal relations needs more clarification. While responsibility for fiscal management is defined by the Budget Law, which is broadly based on international standards, ambiguities and overlaps in responsibility assignments and lack of clarity regarding funding and expenditure responsibilities often leads to weak overall fiscal management and inefficiencies. Budget officials, for example, can increase expenditures (with the exception of salaries, administration, and defense) if additional revenue becomes available in the budget implementation process without any consultation with the center of the impact of such increases on medium-term expenditure forecasts or any discussions within the budget framework on the best use of additional resources. I. 41 (Article 61 (1) of the Constitution) 81 Any new public reform legislation should define accountability structures and directions of intergovernmental finance while ensuring consistency between laws 5.34 Clarifying accountability directions is one of the most difficult intergovernmental finance design issues. It requires reaching consensus on the type of accountability that is to be paramount and the degree of decentralization that accompanies it. This requires a highly participatory dialogue among all the interested parties and stakeholders. While the five years spent debating the PSMFL have provided such a forum, there remain many competing perspectives, political pressures to maintain existing centers of authority are strong, and compromise rare. 5.35 Any reforms embodied in the draft PSMFL under consideration or other laws being discussed will need to address this issue centrally by clearly assigning accountability and reporting responsibilities to all levels of government. The PSMFL provisions will have an impact on hosts of existing laws, and particular effort should be made to consolidate these laws to avoid creating more duplication and further undermining the very principles of clarity in roles and responsibilities the law seeks to promote. CLARIFYING EXPENDITURE ASSIGNMENTS 5.36 Two critical components of any good system of intergovernmental finance are the clear assignment of expenditure responsibilities to different levels of government and the effectiveness with which these responsibilities are carried out. The challenge for Mongolia is making the roles of each level clear with respect to regulation, financing, and provision of service, and providing the incentives and checks and balances necessary to support such assignments. Sub-national expenditure trends by aimag and sectors 5.37 Local government expenditures at 11.3 percent of GDP in 2000 are above the regional average of 10 percent, but have been falling. A closer look at the breakdown of expenditures by aimags, however, indicates that over a third of local government expenditures is accounted for by UB city. Most other aimags account for less than 1.1 percent of local government expenditures (Table 5.1). A comparison between Tables 5.1 and 5.3 indicate that population in some aimags has decreased by over 18 percent- while their expenditure shares have been stable, or decreased only slightly. In Gobi-Altai, Dornod, and Uvs, for instance, population changes have been significant, but expenditure shares have risen. Functional classification of local government expenditures 5.38 The central government is exclusively responsible for defense, foreign policy, social security, and justice. All other expenditure functions are shared among the different tiers of government under a fiscally decentralized/devolved arrangement. The assignment of expenditure responsibilities is generally not by function, but by institution. For instance, the state government directly controls tertiary institutions in the health and education sector (universities and specialist hospitals), but delegates its responsibilities for primary and secondary education and health to the lower tiers of government (see table on expenditure share by government levels in health and education). 82 Table 5-3: Aimag Expenditure as a Share of GDP and Local Government Expenditure Aimag expenditure as a share of total Aimag expenditure as a share of local Aimags and UB City government spending government ex enditure 1996 1997 1998 1999 2000 1996 1997 1998 1999 2000 Arkhangai 1.2 1 0.9 0.9 1.1 3.2 3.2 2.9 3.1 3.5 Bayan-Ulgn 1.2 0.8 0.9 0.9 1 3.2 26 2.9 3.1 3.1 Bayankhongor 1.2 1 1.1 0.9 1.1 3.2 3.2 3.6 3.1 3.5 Bulgan 1.2 0.9 0.9 0.9 0.9 3.2 29 2.9 3.1 2.8 Gobi-Altai 1.1 0.9 09 0.9 0.9 2.9 2.9 2.9 3.1 2.8 Dornogobi 0.9 0.8 0.9 0.8 0.9 2.4 2.6 2.9 2.7 2.8 Dornod 1.1 0.9 0.9 0.9 1 2.9 2.9 2.9 3.1 3.1 Dundgobi 07 0.6 0.6 0.6 07 1.8 1.9 2 2 2.2 Zavkhan 1.4 1.1 0.9 11 1.1 37 3.5 2.9 38 3.5 Uburkhangai 1.4 1.3 1.1 1.2 1.1 3.7 4.2 36 4.1 3.5 Umnugobi 0.9 0.7 0.6 0.7 0.8 2.4 2.3 2 2.4 2.5 Sukhbaatar 1.1 0.8 0.7 0.8 0.9 2.9 2.6 2.3 2.7 2.8 Selenge 2.2 1.8 1.7 1.6 1.8 5.8 5.8 5.5 5.5 5.7 Tuv 1.9 1.7 1.5 1.2 1.5 5 55 4.9 4.1 4.7 Uvs 1.1 0.9 0.9 1 1.1 29 2.9 2.9 3.4 3.5 Khobd 12 09 0.9 0.9 1.1 3.2 2.9 2.9 3.1 3.5 Khuvsgul 1.7 1.3 1.1 1.1 1.3 4.5 4.2 3.6 38 4.1 Khentii 1.3 1 09 1 1.1 3.4 3.2 2.9 3.4 3.5 Darkhan-uul 1.5 1.4 1.3 1.5 1.4 4 4.5 4.2 5.1 4.4 Ulaanbaatar 12.4 10 10.7 9.3 9.7 32.7 322 34.9 31.7 30.5 Orkhon 1.1 1 0.9 0.9 1.2 29 3.2 2.9 3.1 3.8 Gobi-Sumber 0.2 0.3 0.3 0.3 0.3 0.5 1 1 1 0.9 Total 37.9 31.1 30.7 29.3 31.8 ,100 100 100 100 100 Source: Ministry of Finance and Economy and staff estimates 5.39 The two most decentralized sectors in Mongolia are education-in which local governments spent 70 percent of expenditures in 2000-and health-with 63.3 percent of expenditures undertaken at the local level (Table 5.4). Local government education expenditure fell from 78.7 percent in 1996, mainly due to a shift in the provision of tertiary education services from local governments to the center. Other national functions, such as development and design of school curriculums, have also shifted to the center. Table 5-4: Local Government Shares in Education and Health Expenditures Share of Total 1996 1997 1998 1999 2000 EDUCATION 78.7 76.1 70.5 68.7 70.2 Education of kindergarten (0-3 year) 100.0 100.0 100.0 98.5 100.0 Primary and secondary school (8-16 year) 100.0 100.0 100.0 994 99.5 Vocational training and production center 9.8 9.5 7.0 7.6 8.4 Vocational secondary education 8.4 9.0 23.9 35.0 11 8 Tertiary education affairs 4.1 4.4 2.0 2.7 04 Other 28.2 27.9 20.9 10.1 12.6 HEALTH 87.8 86.4 86.3 644 63.3 General hospital services 92.9 91.5 91.3 60.6 61.0 Inter Soum'and Soums'hospitals 100.0 100.0 100.0 1000 100.0 Health resorts (spa) 57.8 67.3 54.5 51.6 37.5 Hygiene and Epidemiology 68.8 70.3 77.2 89.3 94.3 Other 12.1 2.8 Source: Mongolian Authorities and staff estimates 5.40 Local government health expenditures dropped sharply over 1996-2000 compared to education. Expenditure for tertiary health services (general hospitals) has shifted to the central government, while curative services, such as epidemiology services and hygiene, have been moved to local governments. Preventive services are increasingly assigned to the central government as well. 83 5.41 The share of expenditure assignments between levels of governments in other sectors has remained relatively stable over time with slight decreases in the expenditure assignments for general public services from 27 percent to 31 percent over 1996-2000. Aligning authority and responsibility 5.42 Numerous actors within and across government levels participate in the delivery of services.. These include the central government, in some cases regional administrations, local governments, and units that directly provide services (such as schools, clinics, social assistance organizations). A first clear tension that arises with respect to expenditure assignments is that between responsibility and decision- making authority. While lower levels of government are typically responsible for delivering many services, in many cases they do not have control over key decisions that affect service delivery. For example, in education, decisions on the number of teachers and the wages they will be paid are taken at higher levels, leaving little room for maneuver in terms of resource management. At the same time, central ministries and agencies often lack the funding to fulfill their oversight mandates. These imbalances undermine incentives for effective and efficient service delivery. Figure 5-3: Schematic Exposition of Intergovernmental Expenditure Responsibilities Central Government coordination[ MoFE I c Line Ministries provides funding control service delivery I0 Budget entities financed from central budget 8 tAin !. Governmients o !g coordination . Budget, GOv nd, i controls service deliver provides funding .... .. . . o Budget entities flianced from ainag budget 5.43 Budget entities are broadly divided into three main types: * General budget governors (the Ministry of Finance and Economy at the center, and aimag finance departments at.regional level). These distribute budgets to budget entities. * Central budget governors. These receive funds from general budget governors and distribute them to their subordinate units (budget entities). * Primary budget entities. These are budget entities that receive their funds directly from a general budget governor. 84 5.44 The shared responsibility across levels is reflected in the budget allocations. Line ministries typically control only that part of the budget that serves to fund their supervisory function. The costs of implementing sector policies are borne directly by MOFE, which allocates funds to the budget entities under the direct control of the central government and to the aimags. MOFE has to deal with roughly 300 budget entities in a direct budget relationship. Figure 5-4: Assignment of Roles and Responsibilities by Sector Roles and Responsibilities by government level Functions Regulatory framework Funding Service provision Total expenditure General governmental services General public services Central Central/Local Central I Local Defense affairs and services Central Central Central Pubhc order and safety affairs Central Central/Local Central ILocal Education Kindergarten (0-3 year) Central Central/Local Local Primary and secondary school (8-16 year) Central Central/Local Local Vocational training and production center Central Central Central / Local Vocational secondary education Central Central/Local Central I Local Tertiary education affairs Central CentralDLocal Central / Local Other Central Central/ ocal Central / Local Health Hospitals General hospital services Central Central/Local Central / Local Inter Soum' and Soums' hospital Central Central/Local Local Health resorts (spas ) Central Central/Local Central / Local Hygiene and Epidemiology Central Central/Local Central / Local Other Central Central/Local Central/ Local Ssocial security and social welfare Social secunty affairs Central Central Local Social assistance fund Central Central Local Other Central Central/Local Central / Local Housing and community amenity Central Recreation and culture Recreation, sports, culture, and arts Central Central/Local Local Radio and TV, information Central/ Local Central/Local Central ! Local Fuel and energy Energy enterpnses and other Central Central Central / Local Agriculture, forestry Veternary affairs Central/Local Central/Local Central/ Local Forestation expenses Central Central Central / Local State stock office Central Central Central / Local Mining and mineral resources, manufacturing, construction Geological institutions Central/ local Central Central Scientific technological institutions and other Central/Local Central Central Transportation and communication Road foundation Central Central Central ? Local Urban transport Central Central Central Other Central Central/ Local Central I Local Source Government authorities and slaf conposition 5.45 The principal weakness in the current division of expenditures assignments grows out of the separation between fiscal responsibility and the power to reduce or increase expenditure commitments. Local government's have limited budget autonomy, but no control over the levels of resource transfers received from the center. However, they have the right to shift resources across expenditure categories in the budget once they receive their budget allocations from the center. 5.46 Aimags generally spend 30 percent of their budgets on education. While the aimags are responsible for providing quality education based on norms set by the central government, there is no 85 systematic mechanism to ensure they have adequate resources to deliver quality services. Schools devote over 60 percent of expenditures to wages, heating, electricity and water but the prices of all these costs are fixed from the center and account for over 50 percent of total expenditure in education. 5.47 Under current legislation and practice, local governments have the authority to determine the distribution of fiscal resources across expenditure categories. They are also required by law to balance their budgets. At the same time, however, they often must abide by sectoral norms and have little power to determine costs in different sectors. This imbalance is most striking in social protection, education, and health. These imbalances result in unfunded mandates and generate arrears. Spreading policy implementation and service delivery responsibility across agencies dilutes accountability 5.48 A second critical weakness is confusion about which level of government is meant to provide what services. Current legislation is not clear as to which level of government is meant to fund, regulate, and execute each service. Each level of government has some authority to act in every sector; this is particularly true in education, health, social welfare and labor, and, to some extent, environment. The Constitution and the Law on Government Administrative Territories are ambiguous on the areas of overlap. Quality supervision monitors for education and health, for example, are funded by local governments, but designing, setting and overseeing quality standards is assigned to the central ministries. 5.49 This lack of precision makes separating "delegated expenditures" from "own expenditures" virtually impossible. This discretion in delegation makes a system of intergovernmental finance unwieldy, unpredictable, and virtually impossible to implement. 5.50 Due to lack of capacity at the local government level, the same local staff are often asked to carry out both delegated and local functions and it is likely to be very difficult to separate out basis which tasks are delegated and which are local. Improving the clarity between delegated and own functions is necessary to improve the functioning of the system. School directors, for instance, are often asked to act as soum government accountants (local task), and school inspectors (central task) to teach courses The funding of these services is assigned by different levels of government and should be carried out by different levels of soum administrative staff. Options for reform Devolution of expenditure assignments should be matched by equivalent transfers of responsibilities and accountability for decision making 5.51 International experience suggests there are a number of principles underlying successful expenditure assignment (Box 6). In theory, it is possible for different levels of government to exercise different functions in a sector. For example, one level of government may be responsible for setting policy standards and oversight, a lower one may be responsible for providing or administering such activities, and yet another for producing and distributing the service. The key lies on matching responsibilities to decision-making authority. 5.52 The present decentralization of ownership and financing in education to the soum level probably does not encourage economies of scale, as soums are too small to realize these benefits. It generates high unit costs, as soums are free to open new facilities to receive increased funding from the central government; there are few incentives for soums to cooperate and share services, which would reduce local costs. 86 5.53 One option would be to concentrate the financing and production of education in the aimags- in effect creating local education districts at that level. Aimags would own all facilities (and be responsible for all maintenance), employ all staff, and pay all bills. This would help realize economies of scale in production and make decisions on facility rationalization easier. Consolidation of infrastructure and sharing of facilities (such as libraries or teacher training) would be encouraged, as the savings could be plowed back into the sector. 5.54 Consolidation would enhance policy coordination at the center and improve the performance of MOFE and other central policy coordination ministries. MOFE could then concentrate its efforts on improving budget process and performing supervision functions. 5.55 Consolidation of tertiary institutions under the budget authority of the sector ministries could also increase cost efficiency. Currently in Mongolia there are 178 institutions of higher learning, of which only 36 are certified. This is an average of one higher-level institution for every 2,800 persons between the ages of 17 and 29. There are an equally high number of vocational training institutions. These institutions receive unmonitored budget support. Re-centralization could free up resources and increase quality as these facilities are rationalized and the center oversees quality. 87 Macete obetie o fisa decentStrizathJ?ion ildsie'efintiéall sorgofmesots vi a'responsive arid k aco natPe:v erm ent,psalequnableiprov iiono servcesto..citizen in/'>, diferet.jrd ct4ionsda4id ës7t do1ir oicönébbitwåyeandéoh1ch4eädj tSverhaentlsou1néorpon thefÉiohprgvio of Še~~~z i e nt t ' dit rov Pojotivernmupt~sis PH eqi äfrCEsh t onitveres t d p taccpyers objeclsy aossie,Tis;adcistrlbesto aqchevd.,bikeng- hocaiion of rcserie at spoesdt 9fboifbl go ernen: Ya like iftäbreštinpnffr enändGovision of ser"vit Éi'ifse.g., jutrtsdac o pitnd piron ir'c goic Ievie stbii a n r esye obije unicipae y,s io ate o nthe u ipaes r,idinal iciehey pr th ,ision o f pbie services reqis thatcdi oven lsam ufv tenets an pieetocost f .åxpr,yvsi as clae s pssibc e icha s isoca t -in s PExéridötursuneta-9ygyernment fo qiyo come ebiiualizationioreasons,ei suchasasocial'wthtos elsFr oe arf _l e eho,n .usare ,dom nlof cntrago rnmenut for ger fa bief c s - I '.. 'ii t·j",.. ' j theat loclt or~h reg i naonal t e toryanno ssignam publicmsef;tics natre,iberuse t aea ou a Ir h Éi d o r s end't s'r ticiei un er risio r sbifce, e.g. nertat n S f p'' loa *rn nt hich eashav homasnpiïttäcl'praå v tae Expend uregior D sticale znd no gby e 0 * 400 900 1400 1900 2400 2900 3400 GDP per capita (PPP, International dollars) Source: World Development Indicators 2001. Survival Rates for Girls and Urban Mongolian Children are Higher. 8.40 In Mongolia, female infant and under-five mortality rates are 29 and 26 percent lower respectively than those of males. Rural infants and children fewer than five are 1.5 times more likely to die than their urban counterparts. The leading causes of child mortality are: acute respiratory infections (55 percent), other infectious and parasitic diseases (19.5 percent), birth trauma (17.5 percent) and gastrointestinal disorders (8.4 percent). Maternal Mortality Ratio Varies Widely Across Regions and is Higher in Rural Areas Despite High Access to Professional Birth Attendants. 8.41 Mongolia's maternal mortality ratio, at 150 per 100,000 live births, remains high and varies widely across regions. There is a significant rural-urban disparity with maternal mortality in the countryside 61 percent higher than that of the capital city for the 1996-1999 periods. The major causes of maternal deaths are complications during delivery, post-partum complications, and unsafe abortions. Access to antenatal and delivery care is high. Between 1994 and 1998, 90 percent of pregnant women received antenatal care from health professionals. In the same period, 94 percent of all births occurred in a hospital or a clinic and were attended by health personnel. Most abortions (95 percent) were performed by doctors. High maternal mortality despite good coverage of maternal health services suggests problems with quality of care as well as with management and referral of obstetric emergencies. While Communicable Diseases Persist, Non-communicable Diseases are an Increasing Problem. 8.42 Communicable diseases continue to be a public health problem despite a decline in their share within total mortality. The incidence of TB remains high - estimated at 205 per 100,000 populations. Sexually transmitted diseases (STDs) tripled in incidence between 1987 and 1997. Zoonotic diseases such as brucellosis, anthrax, and bubonic plague are re-emerging. Non-communicable diseases are on the rise. In 1999, cardiovascular diseases and cancer were the largest causes of death and respectively accounted for 32 and 20 percent of total mortality. 136 RESOURCE ALLOCATION AND THE BUDGETARY PROCESS Public spending on health has decreased in both real terms and as a percent of GDP. 8.43 Prior to 1990, government expenditures on health was around 5 percent of GDP or 8 percent of total government expenditures. After the collapse of the Soviet Union and the introduction of the market economy into Mongolia, there has been a considerable reduction in public expenditure on health both in real terms and as a percentage of GDP (Figure 8.14). For example, at 18,173 togrog (or approximately 17 US dollars) per capita, public spending on health in 2000 was around 4.2 percent of GDP compared to 5.4 percent in 1990. In constant dollars, public spending on health in 2000 was only 73 percent of the 1990 level. However, as a share of total Government expenditures, spending on health has increased from 9.3 percent in 1990 to 10.6 percent in 2000. 500- 450- 400- .4 350 - 300 - 250 - 200 , 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Year Source: National Statistical Office of Mongolia The Budgetary Process Structure of the Health Sector 8.44 Mongolia's health sector is organized as a five-step referral system based on the administrative structure of the country. Specialist and general hospitals are in the capital and aimag centers; soum and inter-soum hospitals at the sub provincial level; family health clinics and family care practices and mobile clinics at the city, aimag, soum and bag levels, and national public health service facilities with offices at all levels of government. Budgets are based on historical data and input-based norms and standards 8.45 Each national and local level heath facility develops its own budget based on a line item budget using established input-based budget norms and within the ceilings provided by the MOFE. The budget proposals for all health facilities at an administrative level (soum, aimag etc.) are consolidated by the local government finance office and forwarded to the next higher level of public administration for 137 approval. Once it reaches the national level MOFE office, these budgets are summarized along with the budget for the MOH and other central programs. This forms the formal budget proposal for the health sector. There is therefore little incentive to rationalize or close facilities. Also, at each level there is a tendency to inflate the budget to compensate for possible reduction in funding at the higher level. 8.46 The input-based budgeting approach also leads to large discrepancies between proposed and actual budgets, with actual spending often exceeding the proposed levels. Such differences (expressed in percent terms) are also much greater at the central than at the local government levels (Table 8.4). Table 8-3: Differences Between Proposed and Actual Public Spending on Health at the Central and Local Levels, 1996-2000 1996 1997 1998 1999 2000 Central government Budget 2,3402 3,354.6 5,636.9 13,217.9 14,680.8 Actual 2,730.7 3,900.8 4,493.3 12,706.0 16,853.1 Change (%) -16.7% -16.3% 20.3% 3.9% -14.8% Local government Budget 19,056.2 23,915.2 27,8368 22,547.1 26,871.2 Actual 19,659.3 24,678.7 28,421.3 22,952.3 29,098.5 Change (%) -3.2% -3.2% -2.1% -1.8% -8.3% Total Budget 21,396.4 27,269.8 33,473.7 35,765.0 41,552.0 Actual 22,390.0 28,579.5 32,914.6 35,658.3 45,951.6 Change (%) -4.6% -4.8% 1.7% 0.3% -10.6% Source: MOH data Local budgetary officials have no incentive to prepare budgets reflecting national priorities. 8.47 The budgetary process (assigned to MOFE through local government finance office) in Mongolia is separated from the technical and administrative process (assigned to the central MOH). As is seen, the MOH therefore has only a very limited control over the budgetary process and thus cannot enforce that national health priorities are reflected in the budget. There is no correlation between budgetary allocations and health needs. 8.48 Local budget allocations vary much from aimag to aimag and from year to year. With the input- based budgeting approach, there is little correlation between aimag needs as represented by their health status or poverty measures and total health expenditures, government subsidy per capita and health insurance financing per capita (Figure 8.15). 138 60 50 40 > i 30 - 20 - 10 0- 0 5 10 15 Government budget allocation for health by aigmag (per capita, thousands) Source: MOH data Sources of health funding 8.49 Prior to 1990, there were two main sources of health financing in Mongolia - aid from the former Soviet Union and central government's budget. Aid from former Soviet Union amounted to about 40 to 50 percent of the total health budget, while the rest of the health expenditures were financed out of allocations from the central government's budget to the MOH. The MOH was responsible for the allocation of resources within the sector based on set norms and existing infrastructure. 8.50 With the introduction of user fees in, 1992 and the Health Insurance Fund in 1994, the sources of health financing have been diversified, although government contribution (raised through general taxation) still remains the largest source of public finances in health. Contributions to the Health Insurance Fuhd by employers and individuals as well as user co-payments and fees have become increasingly significant sources of income for the health sector. The HIF also derives a large part of its funding from premiums allocated for vulnerable groups by the central government. In 2000, the HIF contributed to 37.8 percent of total health expenditures. The share of external assistance for health as a percent of total external assistance is small. Between 1990 and 1998, only 2.3 percent of total external aid was allocated to the health sector. In 2000, donor funding for health was around US$ 79.28 million or 18 percent of the national health budget. Health services are funded by multiple sources. 8.51 Health facilities receives funding from multiple sources (Table 8.5): (i) line-item budget for fixed costs from the central and/or local government (depending on the level of care) and based on a number of factors which include historical budgets, inputs-based norms and standards (number of health personnel, number of hospital beds etc.); (ii) budget for variable costs based on a prospective payment method; and, (iii) user co-payments or user fees. 139 The health insurance system is not financially sustainable 8.52 "Vulnerable groups", who receive free health care under the HIF scheme and constitute the majority of the enrollees, are not defined based on income or other measures of poverty. As a result, there is substantial subsidization for people who are not necessarily poor. On the other hand, the unemployed and homeless are not covered by health insurance. Nomadic populations have difficulties in registering with local aimag insurance offices. Collecting premiums from compulsory and voluntary employees is problematic, as premiums are paid on an annual basis and many people cannot afford it at one time. The shortfall in health insurance funding is complicated by late government payments. All inpatient care (including spas) provided by the public sector is covered under health insurance - there are no positive or negative lists. The setting of premiums, co-payment, and reimbursement rates are unrealistic as information on the real costs of different health interventions are not available. As the result, the health insurance scheme is financially unsustainable. Table 8-4: Provider Payment Mechanisms in the Health Sector Sources of funding and cost item funded Type of Central budget SSIGO Local budget Out-of-pocket service provider Feldsher posts (outpatients) Variable cost FGP Percentage of drug costs on Capitation (45% from (outpatients) essential drug list local budget, 55% from an ADB loan) (contract between FGP and local health department) Soum hospitals (out patients) Percentage of drug costs on Fixed cost and part of essential drug list variable costs Soum hospitals (inpatients) Variable cost based on a Fixed cost and part of Co-payment (10%) standard tariff variable costs or user fees Aimag and district hospitals Fixed cost and part Percentage of drug costs on Fixed costs and part of (out patients) of vanable cost essential drug list variable cost Aimag and district hospitals Fixed cost and part Variable cost based on a Fixed costs and part of Co-payment (10%) (inpatients) of variable cost standard tariff variable cost or user fees Centrally funded hospitals and Fixed cost and part Percentage of drug costs on institutes (outpatients) of variable cost essential drug list Centrally funded hospitals and Fixed cost and part Variable cost based on Co-payment (10%) institutes (inpatients) of variable cost standard tariff or user fees. Private health facilities (outpatients) Private health facilities Variable cost Co-payment (10%) (inpatients) (flat rate per patient day - -or user fees usually half the rate paid to Private______________ doctors______ _ public facility at same level) Fee for_s7,7c7e771 Private doctors _________ ____________ ___________Fee for service Note: * Fixed costs are for utilities (heating, water and sewage, electricity). Variable costs include salaries, drugs, food, communication costs, stationary and other items Investment cost covers equipment. * As per regulation, fixed costs and funds for increase in the salary of civil servants are financed from the central and local budgets while vanable costs are supposed to be covered by health facility's revenue (comprised of HIP reimbursement and own revenue generating from rent and service fee). In practice, state budget also covers part of the variable costs through financing the budget deficit. * HIP mainly finances health inpatent care, while outpatient care is financed by public budget (mostly from local budget) and some from own resources of the facilities. * Certain in-patient services (emergencies, infectious disease treatment, STDs and childbirths) are free of charge. * Non-vulnerable groups are subject to a 10% co-payment of variable costs for their in-patients services. At tertiary hospitals, certain services have to be paid according to a fee schedule. I. 6 Six vulnerable groups are (i) children under 16, (ii) pensioners, (iii) mothers with children under 2, (iv) people with levels I and 2 disabilities, (v) military personnel, (vi) people receiving income support from the state 140 a Transfer from central budget to HIF are for vulnerable groups as well as HIF deficits Uses of health sector funding Consolidated information on uses of health Sector funding is lacking. 8.53 While there are multiple sources of funding in the health sector, Mongolia has not yet established an adequate National Health Accounts System to generate information on sources of funds and expenditures by type of service. Therefore it is difficult for example, to determine how much resources are being allocated to primary health care despite this being a priority area for the sector. Health spending is skewed towards hospital-based services at the central level 8.54 Before 1990s, Mongolia's health system gave priority to hospital-based services. As a legacy of that period, Mongolia has a comparatively extensive hospital infrastructure despite efforts to reduce this infrastructure for better efficiency.". There are over 22 general and 15 specialized hospitals in Mongolia, which concentrate in aigmag centers and Ulanbaatar. Average length of stay for in-patients remains long at 11.3 days. 8.55 As a result, most health spending goes to hospital-based services, which is a pattern typical of countries in transition. In 2000, general and specialized hospitals received 69 percent of the health budget. Lower-level health facilities on received 18 percent while 6 percent went to policy and public health services, 3 percent to investment and 2 percent to spas and other health facilities. Even at 69 percent, expenditures on specialized and general hospitals have declined from a historic high of 89 percent in 1996. This is due to privatization of some general hospitals in the mid-1990s and consolidation of other facilities. With 3 general and 12 specialized hospitals (out of a total of 37 tertiary care facilities in the country) based in Ulanbataar, health budget allocation is skewed towards toward the central and specialized hospitals and away from aimag and soum hospitals and clinics. 100 80% 70% 60% 50% 40% 30% 20% 10% 0% 1996 2000 1998 1999 2000 NGeneral hospital lInter Soum and Soums 0 Other Source: Ministry of Finance and Economy and Staff estimates I. 61 In the last decade, the number of hospitals was only slightly reduced (5%) - from 432 in 1990 to 412 in 2000. However, there have been a significant reduction in the number of hospital beds per 1,000 population during this period. In 2000, there were 17,974 hospital beds in Mongolia, or 7 5 per 1,000 population (a 40% reduction from that in 1990). Nevertheless, even after this reduction, Mongoha still has the same ratio as Finland. 141 Fixed costs account for a large share in public expenditures on health 8.56 Fixed costs are for utilities such as heating, electricity and water. They take up a large share in expenditures in Mongolia's health sector. For example, in 1997, fixed costs accounted for 53 percent of recurrent health expenditures or 30 percent of total public spending on health. For comparison, public spending on salaries for health personnel and drugs was respectively more than 2 and 3 times lower than spending on utilities in the same year. Table 8-5: Composition of Recurrent Health Expenditures (percent), 1990-1997 1990 1991 1992 1993 1994 1995 1996 1997 Salaries 41.4 457 32.5 21.1 280 308 25.0 24.5 Drugs 14.8 12.6 134 11.3 20 167 138 15.6 Food 114 9.5 10.2 8 6.7 7.0 6.1 6.5 Utilities 32.4 322 43.8 59.6 45.2 455 55 1 53.4 Total 100 100 100 100 100 100 100 100 Source. MOFE Public spending on non-essential services still persist 8.57 Interventions in the health sector include pure public goods (e.g., health information, disease surveillance, communicable diseases control, etc.), pure private goods (e.g., treatment of non- communicable diseases) and a host of interventions in between. As a rule, public goods should have a better claim to support from the public budget. In Mongolia, it is difficult to ascertain what share of public expenditures is being spent on public goods. However as Figure 5 shows, more public budget is still being allocated to institutions such as spas, which are pure private goods, and of little medical value compared to medical research. 8.0- 70- 6.0- 50- 40- 30- 2.0- 1.0. 00- 1996 2000 1998 1999 2000 a Heah - resom (spa) * Hygiene and Bpienuailogy o3 Medical reseich msutos C National programns * Ministry of heahh end Social welfare 142 Low public spending on capital investments 8.58 Capital investments in the health sector are the responsibility of the central government. Though the capital stock is in need of rehabilitation, investment expenditures remain low. Investment in health amounted to 0.2 percent of GDP in 2000, or 3.0 percent of total health sector expenditures. While investment picked up slightly in 1997 to 3.3 percent, it has since dropped. In 1999 investment in health was 40 percent lower than public spending on spas and health resorts. High doctor-to-population ratios, poor skill mix and over-concentration of doctors in urban areas 8.59 Prior to 1990, the country had a relatively high facility-to-population and personnel-to- population ratios (for instance, the doctor-to-population ratio in 1990 was 29 per 10,000 - only second to that of Italy). With the introduction of health reform, steps have been taken to streamline the health workforce. Despite such efforts, there is still relative overstaffing of physicians in the health system. The current ratio of 25.4 physicians per 10,000 population (a reduction of 12% from the 1990 level) is still high compared to other countries of same level of GNP per capita. Sixty percent of doctors are in Ulaanbaatar, where only 25 percent of the population resides. 8.60 There is also a mismatch between physicians' available skills and the needs/priorities of the health care system. Medical training was oriented towards early specialization and as a result, there was both a surplus of specialists and a shortage of general practitioners working in primary health care. The Government has tried to address this problem by reducing the intake of medical students, providing education in/orientation to family medicine, and offering in-service training for feldshers. In 1993, a new category of doctors called family physicians was introduced into the health care system. 60 - 57 54 50 - 40 - 30 - 20 - 16 14 14 14 11 10 10 -6 5 Western region Hangay region Central region Eastern region Ulaanbaatar region 01996 02000 Source. Ministry of Finance and Economy and staff estimates Reform Options 8.61 Mongolia's health system is struggling to deliver quality health services to the population with limited funds. Overall health expenditures are 4.2 percent of GDP in 2000 and sector funding has shrunk in real terms since 1990. Despite the existence of a well-developed human and physical resource base inherited from the Socialist period, the public health care delivery system provides very little value in terms of both the amount of services produced and their appropriateness and quality. Much better results 143 could be achieved by rationalizing the resource base, re-distributing the skill mix and reallocating the funds in accordance with the health needs instead of the resource inputs. 8.62 The prospects for additional funds in the health sector in the medium term are poor, as available public and private resources are extremely limited. The medium term challenge for the health sector is to use available health resources more efficiently to deliver quality services, and to target spending more directly to the poor. Achieving this will involve significantly stronger efforts at (a) priority setting - directing resources to areas which benefit the most needy and have the highest payoff in terms of improving health outcomes, and (b) policy reforms - to remove obstacles to efficiency. 8.63 The fact that the government is struggling to maintain an oversized human and physical infrastructure for health while many health needs are not met makes a strong case for a departure from the input-based system. Such health resources need to be critically appraised in the perspective that health spending should aim to meet the population's health needs rather than to cater to inputs based on the provider's perceived needs and preferences. This may imply a major restructuring and rationalizing of existing physical and human resources and the introduction of new budget principles based on outputs which take into consideration factors such as morbidity mix, poverty level, access to health care, etc, in different populations 8.64 There is still an incomplete understanding of the nature of health, its determinants and health needs in Mongolia. An exercise methodologically assessing the country's needs for health services and costing out such needs would form the basis for output-based planning and budgeting. On the basis of priority health needs and their costs, targets and timelines would be set for resource allocation to strike a balance between expenditures on curative and preventive services, tertiary and primary care, urban and rural areas. Resources should be directed more strongly to national health needs to make sure they are adequately funded. National Health Accounts (NHA) should be developed to track sources, financing agents and uses of health resources. A functioning NHA system would help tracking whether health priorities are being adequately funded. 8.65 Reform of the inter-governmental fiscal relations should be a key part of the overall strategies. Incentives should be provided for sub-national government to improve cost-effectiveness of their own expenditures, while ensuring access for the poorest. It is desirable to design a mechanism for flow of funds in line with the health system hierarchy and the patient referral scheme. 8.66 The roles, responsibilities and authorities of the MOH, other health-related agencies need to be clarified. MOHSW should have a bigger role in the decision-making process for health budgeting. Mechanisms for local governments to comply with national health goals and strategies should be developed. National health development plan should include sub-national targets and goals in order to assist local government in planning and budgeting for health. The financial management system, currently in the pilot phase, should be expanded to provide information on the real cost of health services. 8.67 Further refinement of the health insurance design is needed to (i) reduce government contribution, (ii) improve premium collection from insurers (iii) improve equity. Mechanisms should be identified to attract more voluntary enrollees as well as to reduce self-adverse selection among them. The definition of vulnerable groups should be revised on the basis of income or other measures of poverty. To better target the poor, the levels of premiums could be based on income. Cost data should be systematically collected so that the actual costs of providing individual interventions can be determined. On this basis, the benefit package, premiums and co-payment rates can be reset. To provide incentives for efficiency, it is important to replace the method of reimbursing hospitals based on the number of bed with an appropriate payment system with realistic reimbursement rates. For example, lump sum payment based 144 on a combination of different variables such as previous utilization, local morbidity mix, past technical quality and financial audit. 8.68 To improve the efficiency of health services, the government must to develop a national health services plan based on the heath needs assessment. This service plan will form the basis for: rationalization of the physical infrastructure; assist in the identification of services which should be provided by the public sector; and provide alternative modes for service delivery and options for contracting out. National facility and equipment master plans as well as maintenance programs should also be formulated. On this basis, programs for rehabilitation of selected health facilities and targeted closure underutilized ones could be carried out. 8.69 A national human resource master plan, which identifies the required distribution pattern and skill mix, should be developed. Such a plan would help the government set realistic targets in the training of general practitioners and other allied health professionals. Incentives should be provided to deploy doctors and other health professionals to work in remote rural areas with a shortage of health care providers. New modes of payment can help address the lack of motivation among health care workers. The pilot of capitation payment for Family Physician Groups is the first right step toward this goal. Licensing and continuing education should be strengthened. Using the PSMFL to Set Priorities, Improve Incentives, and Service Delivery in the Health Sector. 8.70 Changing financing flows is another important component of the health reform agenda. The implementation of the proposed PSMFL provides an important opportunity for the Government to confront these issues and develop solutions in a systematic and comprehensive manner. The new approach requires the Ministry and the Government to resolve the current fragmentation in financing and the gap between policy and budget. It also requires the Ministry to set affordable strategic priorities through the preparation of a Sector Strategic Business Plan, with monitorable indicators and costs attached to programs. 8.71 As discussed above, the PMSFL is an ambitious reform, which will take several years to implement. As one of the champion agencies for reform, the MoH will need to have a transition strategy. A sequenced approach to implementation will be needed. A suggestion on how to frame and sequence the reform is outlined below. It is important to emphasize that; these suggestions are being tabled for the purpose of assisting the MoH come up with its own plan. - 8.72 The first step in implementing the PSMFL in the health sector is to identify the products (outputs) of the sector, the priorities among products and sub products, and the costs of these products. These could include: (a) Services to People: These are services provided in hospitals and clinics for an individual which cannot be consumed by more than one person (e.g. an operation, a consultation, etc.). This is the bulk of products provided by the health service today. In most countries, these services are generally provided with some public subsidy to most of the population through insurance schemes or direct provision. Priority setting rules to divide money among services and recipients is critical. (b) Services to Providers: These are intermediate inputs to (a), and are services, which are provided to providers by the state because of economies of scale. They include, for example, pathology and blood bank services, but they also include in-service staff training. Ideally, these should be provided on a cost-recovery basis, (the cost would be included in the price of (a). These are ideal for outsourcing or privatization in the medium term. 145 (c) Services to Society - public goods: These are not unique to an individual or health unit. They are high priority for public funding. They include health information, research, sanitary inspection centers, facilities accreditation, etc. (d) Ministry Services and Policy: These include administration, training, policy advice and monitoring, financial management and audit. 8.73 The Minister will be required to 'purchase' these products from either the public or the private sector by contracting directly for each. The critical policy decision for the Ministry is to develop a new institutional framework for managing these funds and purchasing efficiently. 8.74 The Ministry strategic plan will be the starting point for setting priorities among the four output groups, and within output groups among recipients and outputs. For example, if the Government priority is basic services to the poor at soum-level, the Ministry needs to purchase more of these products and less of other products. 8.75 Products (c) and (d) are the simplest to purchase, since Ministry staff or central budget agencies deliver them. The products in-group (a) cannot be purchased using current financing and accountability arrangements. In particular, reform of the fragmented financing scheme is required to produce a single financing stream for these services before contracting can begin. Developing this reform is not a simple proposition. The Ministry should begin now to develop a strategy for achieving this reform and managing risks. 8.76 The first step is to consolidate all current financing for health care services into one budget. The next step is to consider how to distribute financing in a purchaser-provider mode. Broadly speaking, three options can be considered. Financing options proposed under the PSMFL 8.77 Funds can be disbursed from the HIF to aimags according to a simple formula reflecting demographics and any special conditions. This would in effect require each aimag to start their own health insurance fund in its department of health. It would be the most decentralized option. The HIF would perform monitoring and benchmarking functions (financial and service quality) to help improve accountability. It could also provide guidance and oversight on contracting between the aimag department of health and the facilities. Aimags would then in effect become regional health maintenance organizations, responsible for managing the provider network and delivering the most cost-effective basket of services to their clientele. 8.78 Funds can remain in the national HIF. The HIF would conduct contracts directly with each provider (according to an agreed pricing structure). This would be the most centralized option. Aimag health departments would be limited to quality assurance and monitoring functions for their territory, as well as insuring adequate asset management as owner of the facilities. This would provide the cleanest purchaser/provider split. 8.79 Funds can be disbursed to aimags according to pre-agreed prices and ceilings on resource use by facilities. This would be similar to the system in place today, but it would include the full cost of all service provision, including fixed costs (and eventually capital charges). Responsibility for managing the provider network would still need be placed somewhere (the aimag is the most obvious place). The HIF would have to monitor compliance with contracts. 8.80 It is important to state at the outset that that there is no correct option. Mongolia's small population argues for a more centralized approach to realize economies of scale and insure equity. 146 However, the vast territory as well as political and social traditions may argue for more decentralization. The key point at this stage is that a major reform to the financing system needs to be prepared which creates this single pipe. The Ministry should make the preparation of this reform a high priority. Strategic documents should be prepared, and consultation with stakeholders should begin. Ministry efforts to implement the Government Plan of action proposal to create an independent health insurance agency needs to be broadened to include resolution of this issue. Any trade-off between performance and decentralization must be explicitly addressed. 8.81 In all cases, these difficult issues will take time, investment, and training to resolve. In particular, pricing outputs is a complex problem. SOCIAL ASSISTANCE SYSTEM 8.82 The social assistance system could become another source of fiscal pressure if expenditures continue to rise. Social assistance benefits increased by over 25 percent over 1998-2000. These benefits are mandated by national law, but are paid from local government budgets. There are about 20 different benefits, of 3 types: short-term (e.g. maternity benefits); lump sum, such as for having twins, and benefits in kind, such as fuel for cooking and heating. All benefits are defined by these categories; there is no guaranteed minimum income or basic social assistance benefit. Many benefits have income tests, but others do not. For example, social assistance maternity benefits (equal to 1 minimum wage), are supposed to be provided to all mothers who have carried a child 196 days or more, who are not eligible for social insurance. Likewise twin's benefit is a lump sum provided to all parents of twins within 3 months of birth. However, reimbursement of 100 percent of the cost of prosthesis is only provided to "very poor families", while the benefit for taking care of a disabled child is provided to a family member who is unemployed. 8.83 Average social assistance benefits are smaller than social insurance benefits, and related to the poverty line. It is important to note that not all legally required benefits are paid on time to eligible recipients. Localities may not allocate enough money in their budgets to meet their legal requirements. This happened, for example, in 2000, when the number of beneficiaries fell sharply because of arrears and budget shortfalls at the local level. It is not clear what criteria are used in the benefit rationing process. In field visits, a number of different criteria were offered, indicating a rather non-standard approach to implementing the law. 8.84 The introduction of contribution requirements for social insurance benefits coincided with a substantial increase in the number of social assistance benefits paid, especially social assistance pensions. As a result, social assistance benefits have risen as a share of total expenditure since 1995. In the case of old-age pensions and maternity benefits, the eligibility requirements are basically the same for both systems; except for social assistance pensions the individual should not be eligible for social insurance benefits. For old-age pensions, there is an additional requirement that the recipient be unable to support themselves, although there is no disability or working capacity test for, for example, a 45 year-old woman who has had four children. For disability, the threshold is the loss of 50 percent or more of work capacity - a low threshold. It appears that the response to the introduction of contribution requirements for social insurance benefit eligibility was to shift people over to the social assistance rolls without consideration as to whether these categories of people are the neediest in society. Given the growth of the uncovered sector, this trend can be expected to increase; especially as the pension reform takes hold at the end of the decade. 8.85 The government for a range of vulnerable individuals provides residential social care. A social protection system with excessive use of institutional care is one of the legacies of the Soviet approach. It has been well established that the approach of removing a child or an adult from the family and the 147 community is more expensive per client served than more inclusive approaches which are designed to support individuals within their own families and mainstreaming them as much as possible 8.86 It is not surprising that as poverty and social dislocation increased in Mongolia, the number of people in care increased as well as families in crisis were not offered alternative support. The policy framework reinforces this institutional care approach. The increase in social care is particularly noticeable in UB, in response to the emergence of street children. The origins of the street children phenomenon have not been studied definitively (UNICEF, 2000). Small sample studies report material deprivation is the largest factor (not enough food to eat at home, no dwelling), followed by abuse and or family conflict. However, there are now more than 1000 children deprived ,of parental care that are in some form of residential care, and a significant number remain on the street part or full time. The quality of care in NGO and state programs varies substantially (UNICEF, 2000). National standards for residential care are not in place, nor are there a capability for monitoring them if they were. Policy Coordination and Management 8.87 The objective of the MOSWL is "providing appropriate, decisive, professional support and advice for the ministry and the action program of the government" toward the goal of "creating sufficient living conditions for citizens where they can work effectively and strengthening social welfare service for them"62. The ministry of social welfare and labor is responsible for setting policy and supervising policy implementation in the areas of employment, labor protection, social insurance, social assistance, and has overarching responsibility for poverty reduction. 8.88 The functions of the MOSWL are spread out among a number of independent subordinate agencies. The most important are: (a) The State Social Insurance General Office (SSIGO), (b) The Labor Inspection Agency, (c) The Central Employment R Office (CERO), Social Assistance Office (SAO), (d) The National Committee for Children, and (e) National Center for Rehabilitation and Training. 8.89 The roles and responsibilities in MOSWL are divided between the central and local government agencies. The central level (the ministry and subordinate agencies) is directly responsible for the provision of all social insurance benefits, residential social care, and employment services. The provincial level on the other hand is responsible for the provision of client needs assessments, social assistance benefits, day programs, and some social care facilities (primarily UB city's). 8.90 No standardized system exists for administering social protection expenditures across beneficiaries or aimags. The ministry administers the budget for its staff and other administrative expenditures. 8.91 Most of the supervision for effective implementation of policy by subordinate agencies is carried out by MOFE. The ministry exercises limited overseeing functions over the budget of the subordinate agencies. MOFE determines the provincial budgets, and the respective budget processes determine the expenditure at that level. During the process of determining provincial budgets, MOSWL issues spending guidelines (norms), which can help to shape the negotiation over the total provincial budget. However, these norms are not binding. Aid projects are mostly outside the budget process. 8.92 The multiplicity of agencies providing social security or assistance benefits is augmented by the existence of three quasi-independent funds: the Social Insurance Fund (SIF), the Employment Fund (EF), and the Social Assistance Fund (SAF). While legislative entitlements and ministry decisions, endorsed by I. 62 MOSWL internal document. 148 the cabinet, play the largest role in shaping the budgets of the SIF and EF the independent Councils also have a policy and overseeing role. For example, the Social Insurance Council may direct the SSIGO to prepare regulations on key implementation issues in the pension reform, or request draft legislation. The Council is also supposed to monitor the budget implementation. 8.93 Fragmentation and multiplicity of funding agents have weakened the otherwise well designed functional structure of the MOSWL intended to support an integrated social welfare policy. In practice, therefore MOSWL does not seem to be yielding the expected results of poverty reduction. It should be noted however that it is a newly formed ministry, still finding its way. Due to the fragmentation caused by multiplicity, policy formulation and evaluation are organized primarily according to expenditure type (SI, SA, employment programs, and others), with inadequate attention to the effects of the whole system. 8.94 Performance of the overall civil service is compromised. The fragmentation of functions and budget agencies in MOSWL, like in education, health and especially MOFE has created substantial capacity problems. A fine layer of skilled staff is spread very thinly among a vast number of agencies. No benefits from scale or skill concentration are realized. Options for Reform 8.95 The system needs a fundamental revision. Not only do the cash benefit and in-kind programs need to be reformed, but the overall goals of the program need to be considered. Three issues need to be tackled: 8.96 The goal of both the cash benefits and cash programs should be to support individuals' inclusion in families and society. The system should avoid labeling people, and the need for assistance should not be pre-determined based on, for example, a medical event (e.g. multiple births). The legal framework should be revised to incorporate this goal. 8.97 The system should be simplified. Consistent with above, the policy should encourage case workers not look for a category to put a vulnerable individual in, but rather look for needs and solution ("one right of assistance"). Benefit entitlements should not be created which mirror the social insurance scheme, since these reduce incentives to contribute and may not reflect true vulnerabilities. 8.98 In keeping with the directions of the proposed PSMFL, financing for social assistance (programs, institutions, and benefits) should move away from an input basis. Mongolia needs to develop a program to reform social care financing so that the state is the purchaser, but not necessarily the provider. The provider could be NGOs, communities, or families. In general, day programs and inclusive care in eth family is much more affordable than residential care, and Mongolia needs to develop a reform program to move in this direction. This also implies improvements in the gate keeping function (the process by which people are assigned to care and removed from care). 8.99 Finally, the social assistance and social care system should be the system of last resort in the case of poverty. Prevention of poverty, especially among the vulnerable is much cheaper than paying benefits. One of the side effects of Mongolia's fiscal crisis and response appears to be a growing exclusion of portions of the population form other social services because the do not have a formal sector job or residency papers, or they have certain characteristics (age, ethnicity, mild disability, etc.). This is a common problem throughout the FSU, as countries struggle to find new ways to ration scarce resources and provide incentives for efficient allocation. It is an issue which requires Government-wide attention, however. The MoSWL could take the lead in helping work through these issues. 149 8.100 Employment promotion programs. The new Employment Promotion Law sets an ambitious agenda. The first task is to collect information on who are the unemployed and how they related to labor demand, for the purpose of setting priorities in program implementation. A full implementation needs to be developed, with monitorable indicators. The agency also needs to: strictly define what training is; * use demand driven mechanisms to chose training providers; * implement cost sharing mechanisms as much as possible, especially for the employed and self-employed; * avoid wage subsidies or strictly limit their use since they tend to encourage substitution of one worker for another rather than net job creation; * avoid government provided and administered micro-finance as experience under the NPAP was problematic; * set strict rules for public works, with matching funds required. 8.101 Expected outcomes should be specified and monitored at all levels. This is especially important for programs, but is also relevant for cash benefits. Analysis of other functional public expenditures 8.102 In the future the task of the government will be to ensure that allocations to other non-social sector priority sectors result in efficient spending, and are grounded in the development priorities of the government. The MDGs provide some benchmarks for sector objectives in some sectors such as environment, transport and energy. The section below will provide a detailed assessment of allocation patterns and direction in environment sector and brief reviews of expenditure trends in other government sectors including defense. ALLOCATION OF RESOURCES IN THE ENVIRONMENT SECTOR 8.103 Public expenditures on the environment and natural resources management in Mongolia are about average compared to other Asian countries as a percentage of GDP, public expenditures, and in per capita terms. Between 1996 and 2000, environmental expenditures followed an erratic trend, and fell almost 50 percent during the 1998-99 economic recession. 8.104 Mongolia is heavily dependent on natural resources, which contribute 13 to 20 percent of GDP. The livelihoods of over two thirds of the population are directly affected by environmental policies or events, including land management, air pollution, and solid waste management. Natural disasters such as the Dzud, drought, and steppe fires also affect the livelihoods of Mongolians. In 2001 historically harsh weather led to the death of over 10 percent of the Mongolian herd stock, severely affecting the livelihood of most Mongolians. 8.105 Current budget allocations for recurrent and capital expenditures on the environment are low and should be increased-as savings are generated from other expenditure cuts-so that MNE and the government can properly address Mongolia's most pressing environmental and natural resources management issues. These include solid waste collection and management, water pollution abatement, and air and industrial pollution. 8.106 Core environmental expenditures refer to discrete budget allocations whose sole, or at least primary, purpose is either to provide environmental public goods or to address adverse environmental impacts. Examples include monitoring, analyzing, and disseminating information on environmental quality and pollution sources; setting and enforcing environmental standards; and institutional strengthening and capacity building. Core environmental expenditures by natural resource management 150 agencies include budgets for national parks, forest protection; budgets for integrated conservation- development projects and social forestry projects; and expenditures on reforestation. Examples of other projects and programs under this definition include government-sponsored industrial pollution control programs and sewerage and sanitation projects. 8.107 Over 1996-2000 the budget allocation for core environmental expenditures increased by about 33 percent, from Tg 1,3 million to Tg 5,3 million. The upward trend was reversed only in 1999, when the budget allocation suffered a 48 percent cut during a period when both government spending and the economy contracted. However, the decline of environmental expenditure was more rapid and stretched across a longer period. Core environmental expenditures declined twice: from 0.2 percent of GDP in 1996 to 0.17 percent in 1997, and from 0.42 in 1998 to 0.27 percent in 1999. 8.108 Over 1996-2000 58.2 percent of the budget allocation to MNE went to Environmental Management Quality, 39.2 percent to Natural Resources Management, and 1.7 percent to Research & Development. Expenditure allocations among these three categories of expenditures varied significantly from year to year (Figure 8.19). This reflects government efforts to strike a balance between expenditures for brown and green environmental problems. 151 Table 8-6: Environmental Expenditures as a Share of GDP Mongolia GDP (TG millions) 646,559 832,636 817,393 873,679 1,045,000 Core Environmental Expenditure (TG millions) 1,315.00 1,398.60 3,459.80 2,348.50 5,309 Core Environmental Expenditure as percent of GDP 0.2 0.17 0.42 0.27 0.5 Per capita Core Environmental Expenditure (US$) 0.53 0.56 1.13 0.63 1.48 EMQ as a percent of Core Environmental Expenditure 0.18 0.15 0.25 0.08 0.22 NRM as a percent of Core Environmental Expenditure 0.02 0.02 0.17 0.19 0.28 Source- Ministry of Finance and Economy, and Ministry of Nature and Environment i Eine#txlE ditureUShiftiig woreena A lteExenis uto Figure 8-19: Environmental Expenditure By Type Figure 8-20: NRM as percent of Core Environmental (MI of Tugrog) Expenditures 90 so 8 70 7 A 60 H6 100 f 40 Reoet5ai90 II0 4 r 20P o--sc 40 0 e n 19 1997 1998 199 2000 SuEnv Quality o NRMeR&DOExp for Forest 1996 1997 1998 1999 2000 Figure 8-21: Expenditures for Natural resources Figure 8-22: Core Env. Exp. in the Development Budget Management (Ml. of Tugrog) Comparison to Trends in Other Budget Items. 190 OReorestalon 90.0 rmie bconriti 80.0 rr 700 DRehab after A60.0 60( 60.0 C0lGeologicai 50 0 - 40 ~PmspeMtn 0 40( 3.!r . 00-- --L 30 61-and Inventory 311k 20(I 200C-*A* 10( R Land 0OoW"w Monlronn 0 199 199 199 199 200 ig ~ ~ i~~o Source: Ministry of Finance and Economy, and Ministry of Nature and Environment 8.109 Mongolia's industrial pollution levels are higher than those recommended by the Millennium Development Goals, and respiratory diseases are its leading cause of death. However, between 1996 and 1998, the budget allocation for industrial pollution accounted for about 30 percent of costs, and the remaining 70 percent was covered by contributions from the private sector. Since 1998 there has been no 152 government allocation for industrial pollution abatement, and the private sector and bilateral donors have provided the needed funding. 8.110 Despite being identified as one of the main environmental issues for Mongolia, solid waste collection and disposal received no allocations from central or local budgets over 1996-2000. In 2000, due to the seriousness of the solid waste situation, the international donor community made available Tg 1.1 million in foreign loans and grants to assist the government. More effort and resources will need to be dedicated to address the problem in a sustainable manner in the medium term. 8.111 Natural Resources Management includes seven subcategories: (i) Reforestation; (ii) Biodiversity Conservation; (iii) Rehabilitation of Ecosystems after mining; (iv) Geological Prospecting; (v) Plant and Vegetation; (vi) Land Inventory Management; and (vii) Land Condition and Quality Monitoring. The budget allocation for all the seven categories has increased steadily, going from Tg 172.2 million in 1996 to Tg. 2.9 millions in 2000. 8.112 While Reforestation and Biodiversity Conservation received increased allocations throughout the period, Geological Prospecting, Plant and Vegetation; and Land Inventory Management only started receiving allocations in 1998, Land Condition and Quality Monitoring from 1999, and Rehabilitation of Ecosystems after Mining in 2000. These shifts reflect the government's attempt to balance its responses to brown and green environmental problems. 8.113 Data for environmental expenditures by line agencies other than MNE is very limited. It is common for developing countries to initially give most environmental responsibility to core agencies, with "mainstreaming" into line agencies coming at a later stage. 8.114 Core Environmental Expenditures in the Development Budget. Information on core environmental expenditures in the development budget is very limited, as the Ministry of Finance keeps data on development programs only, not on individual projects within programs. The only relevant data covers forestry investments, which rose 20 percent increase from Tg. 109.2 million in 1996 to Tg. 539.7million in 2000. 8.115 Comparison To Natural Resources Revenues. Natural resources contributed substantially to Mongolian government revenues, with oil and mineral.resources being particularly important. Over 1996- 2000 revenues from natural resources contributed an average of 2 percent to GDP, while the average core environmental expenditures share of GDP was around 0.3 percent-about 17 percent of resource-based revenues. While some reinvestment is obviously needed to ensure the sustainability of natural resources and the protect benefit they provide, earmarking resource-based revenues for environmental protection would be inappropriate. 8.116 These findings indicates that the current budget allocations for environmental expenditures need to be increased to ensure sustainability of current expenditures trends, support the renewed effort of the MNE to undertake R&D activities in the medium term and improve environmental management at the aimag level in the medium to long term. Policy coordination and management processes. 8.117 The MNE was reorganised in 1998 with an administrative structure at central, regional, and local levels suited to address the above interventions. The minister of nature and environment as member of the cabinet is acting as a driving force for formulating policy and implementing necessary measures. Decisions on environmental issues are to be taken at the lowest appropriate level. Thus it is important to have knowledge and capacity at regional and local levels. Environmental policy formulation and 153 implementation of measures should be based on a participatory approach, involving users, planners, and policymakers at all levels. 8.118 Management structures especially at the local agency levels are weak and coordination with the national authorities on overall policy directions is largely ad hoc. An accountable management structure with the necessary human and funding resources to perform the coordination, management, and performance monitoring of the environment and natural resources management sectors is missing. 8.119 Agency staff at both the local and central government level lack training and the salary structure does not attract qualified staff with the required professional capacity to meet Mongolia's need for efficient environmental and natural resources management and enforcement of laws. Budget preparation capacity at all levels is weak. MNE suffers from inadequate human and financial capacity to fully perform all responsibilities assigned to it. The loss of professional staff to the private sector has exacerbated this problem. 8.120 Relations between the central coordinating agencies on environmental policy are weak, as other agents do not perceive the ministry of nature and environment as being the principal focal point in the coordination and design of environmental policy based on government directives. 8.121 Since 1995, satisfactory progress has been made on most of the issues mentioned above, taking into consideration the short time period since MNE was established. Nevertheless, Mongolia has just started on its journey toward more efficient management of its nature and environment, and many more efforts are needed to strengthen the institutional capacity and performance. CONCLUSION 8.122 Mongolia's medium-term resources are insufficient to address all these issues simultaneously. In the short to medium term, therefore, the government needs to develop a pro-poor policy mix that enables actions in each of the sectors to improve overall outcomes for the population. An effective mix of health and environmental policies for example or health and transportation policies could improve living standards and raise enrolment rates. 8.123 Social indicators in Mongolia have dropped from the significantly high levels observed in the early 1990s. Despite substantial improvements in the mid-1990s they remain worse than transition economy averages. Expenditure is high, and funding to the social sectors, which shrank over 1990-96, has since recovered. Funding in other economic and general government sectors also increased during this period. While the economy needs increased investments in all sectors to upgrade and maintain its aging infrastructure-particularly in the social sectors, the prospects for substantial additional funds in the medium term to cover all critical expenditure are weak, as public and private resources are limited Achieving the government's objective of delivering quality services more efficiently requires changes in policy priorities and funding processes to enable existing resources to be channeled to areas where they have the highest payoff in terms of improving outcomes. These include primary and secondary education, primary health care, environmental protection, and improved access to safe water and clean air. 154 9. SEQUENCING THE REFORM AGENDA 9.1 Mongolia faces many hurdles in putting a more effective and performance-oriented public sector into place The government's detailed Implementation Plan provides important guideposts for this process, and its reform efforts are beginning to move the budget process in a direction more consistent with the principles of a performance-oriented public sector. Past Reform Efforts have Lacked Overall Government Commitment 9.2 The reform effort remains tentative and un-sequenced, however, and has stalled on three fronts. There is still no real commitment across the government on the public sector reform agenda, partly because of frequent personnel turnover in key positions. Differing reform philosophies and approaches among donors have also impeded progress. These have manifested themselves in tensions between achieving short-term fiscal objectives and medium-term structural reforms, and a lack of realism about the pace and capacity of the government to implement donor-proposed reforms. Reform initiatives have also been compartmentalized along the main lines of public administration,63 so that activities that should have been linked were cordoned off into separate specializations. Consensus Around the PSMFL and the MDG's is a Formidable Asset 9.3 The PSMFL provides Mongolia with a vision of where it wants to take public sector management, and its commitment to the MDGs provides a rationale for the process. What is missing is the how: a realistic sequencing plan based on the current realities of public sector management in Mongolia, and a coherent approach for getting from here to there. The Implementation Plan needs to be supplemented by a range of reforms to systems and practices to build a more performance-oriented public sector. I. 63 One set of reforms has sought to rationalize budget decision-making, another has targeted civil service practices, still others have sought to improve public procurement, cash management, accounting systems or other parts of the machinery of government. Thus activities that should have been linked were cordoned off into separate secializations Opportunity, Strategy, and Tactics in Reforming Public Management; Allen Schick, PUMA/SGF (99)4. 155 However, the PSMFL is not without risks foer mai9.n hsRktt thetifemen m tiorögäa el tes toýtensions; witlfch nng en z tion 1 as da _eIla e p -t äi&I 6cz o - n- ' fluiigy lä pp en AWe'cid the L, 'urcesa oes e6~rvice tethe peopl è r s ecall cnKt ared ,éathg xiiti 1o,~nd~hi ~sgiiafl rpp eis C51 -,ý 44 rtensiv b t hu,dey tlëteveisafionnsu tan TTrnn at n nsi tW 2oplemnt t#élthé',,0 EA pposawibi t p e e co to es m-qj ý,0 -Wver"'11dl t ssiflenntl-, rgiér a s ithr ty'olld i b ~e~r t e eè i n er. l ed åtttKin a ic tpaymi s a u nc Wsc illpartualy b o u aoli "Ilel sfpitiitå r e ' - ~ ,,, ~m r 119 iäiro'iiii th g vr e , yieäan thé sfmäö ManagiengeW, rrl o a apocth ti s eformncei1iobealaundt 9.4 Thschptrarueora app?,roaht eomta engags talthe plyers intepblcsco (and m-an nyfrma tot yfcsn n htte cieathheitinerstsaspubic official (Boxa8) Te arocehoas e n of w ii d ca stie n the-ä pproahisarnesing and Macining tefortrsof and fhemplaers an aroah th atd retwarseen rfrenaces-hc nld the lack of a realistic budget that is implemented as originally legislated, and a management culture that promotes fidelity to formal rules-and those required to support the successful operation of the PSMFL. 9.5 This approach is based on experience in a range of reform programs around the world. It draws particularly on the approach in some countries of providing agencies that adhere to key financial, performance management rules, obligations greater budgetary flexibility, and increased predictability on the amount and timing of funding. In Mongolia the approach needs to be broadened to integrate policy making, planning, management, and budgeting at the outset. It will need to embrace intergovernmental relations and civil service reform as well as the budget process. This integration needs to happen at the 156 center of government, a"d at the level of each sector and agency. One of the strengths of the approach is that it will focus attention on the interaction between these processes at the agency level and at the center. i.4d kuyliz umpt& -Rheaintere ia lare,14~ff?4 , R1 .aa.a,c nn.4Ir il g1 'Fl 4g ktdMlnt f,U1l'N4ICenfri elk,s ' 915 , ., Z%f t. 5 interes.ts t b. '.7 ~J', * C ;~ .2>.PK4i .~~~~~ ~~~~~ Sburdait5ezae roiitcunrsoreso n a 'i I~ ~ . 4 IDE m dcranzati pfcc: a.f * i~~t' RhR~ 4%~' WEMftijR44ediIng,btro%' susimabifisaI.pIThvfV.. '* I a t *'~~''tt' tV 4t11>1. 44Sdnd uma reourc maagepen polciet ''~ 2 '24L RAI' ^02 T P'4 A. Oil .,- kll 'h. Jt ,j~0," Tuhrfyo'odCjb I-i ;A sys$tem raCeAadJappbdpuihsgodpromne*-7 * ii'I ff!;F: mea f '~r.L .i f~,:';*; .y ~ ~ ', , isesrvce 'o,il6at'&r& A yd Ortce- oiy ang*m( re u4' es Ti Tv . Y I ehnaqs,re9, ' ' ' * ~ ''' " pubic ect r.Thekey ele et ofLivtheofapp roh a ar e , es igne toree R dicpln in thepuli st ori policy~ ~ ~ ~~~~I maig plnig bugein (icudn fiacaiaaeet,adpesnefaaeet e ee en suporin thi dicpln is machn auhrt and1 aconaiiy Suprtn ths in' turn .i-IIs higher~~"o"j quait and moetasaetfnnillnomtoln h eelpeto o-iaca perormnc information 9.7 Change will be leveraged by all players F-.;recognizn thgastobreledfmacopaiv approach to ea strngheing db&isln,atching(ndepadng utoit ndaconablty.n 157 improving the information base. Rebuilding discipline will go hand-in-hand with expanding and matching authority and accountability. Creating a Disciplined Top-Down Process as a Basis for Improved Accountability 9.8 The chapter on budget formulation contains a set of recommendations designed to strengthen the links between policymaking, planning, budgeting, and management. These recommendations aim to strengthen the disciplining of policy making, planning, and budgeting at the center of government. The emphasis is on hardening the budget constraint and disciplining a more clearly articulated set of policy priorities. From the perspective of individual ministers, line agencies, and lower levels of government a shift to a disciplined top-down process will create the conditions for greater predictability in policy and funding, greater responsibility for allocating program resources, and greater flexibility over the management of resources. From the perspective of the center this, in turn, provides a base for holding individual ministers and agencies accountable for improving operational performance-the delivery of the policy priorities of government efficiently and effectively. The challenge is for all the players to recognize this as a win/win situation. 9.9 While reform at the center can help create the conditions for better performance, there are a number of other practices highlighted in other chapters that need substantive reform if a fully disciplined base on which PSMFL can be launched is to be put in place. Making Progress on Basic Management Practices 9.10 There are two broad, sometimes overlapping, clusters of practices on which the principles of the approach are built. The first focus on basic internal financial management practices, including budget and funds control, internal auditing, procurement, and asset management. The second targets basic performance management, including budget planning and output costing. These are supported by requirements for financial and performance reporting. Increases Agency Authority and Responsibility 9.11 What distinguishes the approach is that improvements in these components are linked to the provision from the center of greater flexibility, predictability, and access to revenue generated. Although flexibility and predictability are two of the central elements underpinning agency performance-and are crucial to matching and expanding authority and accountability-Mongolia should be working on all available points of leverage. 9.12 A particular strength of the principles of flexibility and predictability is that they apply across the full range of public sector practices. Giving public sector managers flexibility over, for example, management of financial and human resources, also delivers authority by imposing a discipline on all other players not to interfere in the exercise of that flexibility. In Mongolia, this is particularly pertinent to personnel management. Flexibility in the policy area is a difficult issue but the proposed shift to a more sectoral, top-down approach to budgeting enables clearer lines of authority and responsibility. Cabinet would share authority and responsibility for overall policy priorities - and choices and individual ministers for program choices and allocations. 9.13 Predictability's significance ranges from policy, funding and staff movements, to predictability in the budget timetable, in how proposals for funding arising outside the budget context are dealt with, and in adherence to service-wide rules for personnel management. MOFE would be the agent of the center negotiating budget and financial management agreements. The General Services Administration would ensure that personnel management and staffing are consistent with the Cabinet's service objectives. 158 Another strength of this approach is that it demands that the center work together to ensure a coherent approach to reform. 9.14 Three other features distinguish this approach from many other reform strategies: (i) agencies, rather than particular reform measures, are the building blocks of reform; (ii) pockets of reform energy hidden in the bureaucracy need to be mobilized, and (iii) MOFE must champions the approach. The Agency Rather than the Measure should be the Focus of Reforms Agency Reforms should be Comprehensive 9.15 This principle is based on the lesson that reforms that concentrate on individual components (e.g., performance budgeting, devolving authority across the board, even medium-term expenditure frameworks) rather than on managerial systems and behavior often prove disappointing. Therefore, rather than focusing on one reform module at a time, the approach focuses on agencies. Several reform modules are combined to form a package for each agency formalized in a resource agreement or through the budget process. In Mongolia applying this principle will mean: * developing consensus on the government's objectives * selecting an agency or line ministry responsible for delivering a service to the public, identifying the main impediments to its effective and efficient functioning, and * developing a set of administrative procedures, rules for financial management, incentives for civil service performance, and accounting systems which in combination would improve performance. 9.16 The version of the PSMFL under consideration in parliament proposes the more conventional approach. But the argument here is that this will not work without tackling the current pathologies of the system. Using the approach outlined here can do this most effectively. 9.17 The state of current practice in the public sector in Mongolia poses a particular challenge to this approach, because the required strengthening of discipline in the public sector will be seen by many as a recentralization of authority and responsibility. It is in fact a strength of the proposed approach that it simultaneously tackles the discipline side and the enlarged authority and greater predictability so important to agency performance. Perhaps most importantly, it attempts to turn the current reality into an incentive for agencies to reform themselves. This is what the approach offers Mongolia: making the availability of a more flexible and predictable operating environment for program agencies conditional on each program agency raising the quality of its financial and management practices. Pockets of reform energy hidden in the bureaucracy need to be mobilized 9.18 The principle here is based on the notion that successful reform does not just happen by itself: it depends on leaders who exploit openings and give impetus and direction to change.65 There are, in any country, pockets of reform energy and this approach facilitates bringing this energy into the open. The current approach to pilots in Mongolia draws on this principle but does- not appear to take full advantage of the opportunities offered by the more comprehensive approach proposed here. 9.19 An essential requirement for reform is active support for management reform from the senior managers of program agencies. Agency managers who demonstrate leadership and willingness to adopt the reform are rewarded with less interference from the center. The offer of a de-controlled package, formalized by a memorandum of understanding and a resource agreement, promises the more active I. 65 Opportunity, Strategy, and Tactics in Reforming Public Management; Allen Schick, PUMA/SGF (99)4. 159 senior managers relief from what they usually perceive to be oppressive detailed line item control and interference in personnel management decisions, along with a more stable planning environment associated with a more medium-term perspective on policy and funding. Ministry of Finance and Economy must Champion the Reform 9.20 Reform on the scale required has significant implications for the role and behavior of the center. MOFE's role in controlling expenditures, allocating resources, and supporting efficiency will need to change. MOFE's role today is perceived to be command and undue controls. It specifies and determines levels of expenditure, regulates allocations of funds, and leads the budget allocation process. This approach cannot coexist with the devolution of authority to central line agencies, envisaged by government. If managers are not given authority and predictability the basic environment to support a performing public sector will not be in place. With Substantive Support from Sector Experts 9.21 Therefore, while the direction of reform proposed by the PSFML will impact particularly on program agencies (line ministries and budget agencies), the creation of the practices that will enable them to deliver better performance will depend crucially on the performance of MOFE. It is unrealistic to expect that budget office staff, used to maintaining accounting and control frameworks and procedures, can immediately shift to being facilitators of a performance-oriented public sector. To kick-start these processes MOFE will need to recruit sector expertise to assist it and reforming agencies in identifying gaps and developing strategies to fill them. While MOFE is receiving various forms of support aimed at improving its role, they are not well coordinated and linked to the reform effort. In addition, the fiscal policy department central to the success of the reform process lacks support. 9.22 Sector experts to assist with gap identification in reform agencies other than MOFE will also need to be identified. Sector experts will rarely be experts in all reform areas, so it is important that they be supported by specific expertise in each reform area. A key feature of the sector work is that it not only provides each agency with a feel for its gap-filling tasks, but also provides the budget office with an understanding of the changes it should seek before easing detailed controls and increasing predictability to an individual agency. This is particularly important, as the budget agency must not been seen as raising the bar after the event. 9.23 Considerable effort needs to go into managing the change process in the budget office, and such efforts will need to be fully integrated with the work of the proposed PSMFL Implementation Task Force. MOFE will need to draw together all the lessons from developing the approach to build the performance and financial management regulatory framework to support the PSMFL. MOFE assesses performance and regulates processes 9.24 While budget office staff may not be able to lead on specifics reforms, they must participate from the outset. Rather than continuing to be involved in the management of detailed budget lines in budget preparation and allotment, their role will change to the interpretation of agency performance against output specification, review costs underpinning agency budget requests and make inputs into policy development. Rather than being experts on the costs of activities they will need to become experts on the value of activities, particularly in relation to their impact on the policy objectives of government. 160 The Advantages of the Approach It is based on clear incentives and incremental awards for performance 9.25 There are numerous advantages to this approach. It is based on incentives and agencies that achieve the standards receive something in return; but they also have to show concrete evidence of changes in systems and practices before traditional controls are eased or new approaches adopted. It is staged," and places strong emphasis on building capacity in line with growing demands and incentives. It builds a momentum that puts pressure on all players to perform, while not holding back those who want to go faster. It has the potential for early wins, through self-selection by reform-minded agencies. It leverages demonstration effects. It improves the sequencing of different components of budget reform. It both facilitates external technical assistance and demands a greater consistency of approach from donors. It allows for sequencing and consensus building 9.26 The approach creates an opportunity for Mongolia to reach a consensus on the priority of reforms, and their sequencing. It also aligns the interests of the central coordinating ministers and their agencies, line ministers, and agencies and lower levels of government. These interests range from maintaining aggregate fiscal discipline, through ensuring that budget processes contribute to resources being allocated to priorities, to a regulatory framework that supports and demands the efficient and effective provision of services to the people of Mongolia. It places strong emphasis on capacity building and upfront agency commitment 9.27 The approach also builds on the experience of those agencies willing to-and capable of- making more rapid progress in moving to a more performance-based approach. This has implications for the current approach to piloting the implementation of reforms and argues for implementation over a number of years. Perhaps the most telling argument to come out of the piloting experience to date is that currently only the responsibilities and obligations of general managers are spelt out, not their authority or rights. This approach forces more explicit attention to matching accountability and authority. Another major benefit from the perspective of agencies is that the pilots would, under this approach, reap the benefits of changed practices as and when they achieved standards, rather than piloting the PSMFL while continuing to operate day-to-day under the current flawed system. APPLICATION TO MONGOLIA 9.28 The key point about this approach is that it provides a way of achieving consensus around the reform agenda. It also has the potential to breathe new life into the pilots as it takes them from essentially learning exercises to ones driving the establishment of the new system. The elements identified above as components of a well functioning budget and financial management system provide the core of many of the needed changes in Mongolia. This is particularly true for basic financial management. There is a real urgency to putting in place such core components of a sound financial management system as funds control, procurement practices, and internal audit. I. 66 One very important lesson from the Thai experience is that rather than being a once-and-for-all exercise of meeting standards and receiving maximum flexibility, flexibility can be gradually increased as progress is made in moving toward the standards. 161 Improving financial management processes to discipline the budget 9.29 Significant attention has been given to reducing arrears, but this has largely involved MOFE assuming and funding liabilities. Much less attention has been paid to the underlying causes of the problem. The components relating to funds control and budget planning must incorporate elements dealing with arrears. This will involve a stronger focus on commitment controls and agency discipline during budget execution. 9.30 Complementing this focus, the budget-planning component must explicitly incorporate elements to ensure the realism of funding for individual budget units within each agency. Ultimately this is determined through political processes, but MOFE must become an advocate for more realistic budgets. Any agreement by MOFE to fund an agency's arrears must be accompanied by explicit improvements in the agency's capacity to avoid arrears in future. 9.31 As has been emphasized in earlier chapters, MOFE also has to move away from its current practice of negotiating budgets with some 300 different entities, and shift to negotiating budgets at the sector level. The entity emphasis of the PSMFL also demands close attention to the relationships between agencies and parent ministry planning, budgeting, and management systems and processes. The centrality of a disciplined priority setting process argues strongly for an early focus on building integrated strategic planning, budgeting, and management at the sector level. The approach outlined here, particularly for funding predictability over the medium term, could be particularly effective. Output costing and specification and other day-to-day activities central to budget execution, such as funds control, offer opportunities to link standards and greater within-year flexibility and predictability at the agency level. The simple rule here should be to determine where the authority lies-primarily at the sector level for budget planning and sector regulation, and primarily at the agency level for funds control during budget execution-and target the quality of components, memoranda of understanding, and resource agreements at that level. 9.32 Components such as procurement, internal audit, and asset management should be targeted at the sector level, at least initially, with the approach being seen as a key catalyst for shifting to a sector- wide approach to budgeting and strategic management. Improving intergovernmental arrangements 9.33 A third crucial issue in Mongolia concerns intergovernmental relations, with all their current ambiguities. It is quite apparent that the greatest concern lawmakers have over the PSMFL is the re- centralization it implies. This is an issue that requires the earliest and highest level attention. The reality is that current levels of performance at the local government level are so poor that many will see re- centralization as the only route to achieving performance levels implied by the PSMFL, even in the medium to longer term. 9.34 All of these elements are, or should be, concerns to local government. If they are to continue to receive funds directly from MOFE to produce PSMFL outputs, they will need to improve the full range of their practices, and negotiations around greater flexibility and predictability should provide a powerful incentive to lift their capability. If, at the other extreme, local governments were to become the agents of central government ministries in delivering services, they would still have to be party to such an exercise. A checklist used to assess the readiness of local districts in Ghana to manage donor funds directly is attached as an illustration of an approach one country has taken (Annex 2) Again, the strength of this approach lies in its emphasis on incentives. 162 Build a professional civil service 9.35 Another feature of Mongolian practice is the weak state of personnel management. Reforming personnel practices needs to be intimately linked with the budget flexibility quid pro quo. Before that can begin, however, the politicization of the civil service has to be somewhat abated. 9.36 The interaction between the proposed approach and the sequencing of implementation of the PSMFL at the agency/sector level needs some thought. The original Implementation Plan and the recent Issues Note both indicate that the health and education sectors should be the last to adopt the PSMFL. This position is understandable given their complexity, particularly in terms of intergovernmental relations. However, it is important to draw on the "reform energy," at least in the health sector, to drive a more performance-oriented approach at the sector level. Again, the proposed approach is attractive because it can be employed to keep up the momentum for change, so that when the time is right to formally adopt PSMFL, these key sectors are already operating more -strategically and with a focus on results. INSTITUTIONAL AND CAPACITY CONSTRAINTS TO IMPLEMENTATION. Improvements in processes lead to improvements in capacity 9.37 Lack of capacity is often cited as a constraint on the performance of the public sector (as has been noted elsewhere in the report). While there is some truth in this, particularly at lower levels of government, it pales into insignificance against the adverse impact of poor practices. Improve practices and the capacity will, to a significant extent, be there to respond. The problem here is very much related to another issue, that of the accusation that those arguing for public sector reforms that focus on improving performance are seeking to impose first best OECD practices on developing countries. Again this is no doubt sometimes true (and has been argued in the case of Mongolia). However, for the most part, such reform advocates are arguing that performance is a basic, whatever the current level of performance67. Incremental management changes could have substantial long term impacts 9.38 If a country is assessed as performing at the level of 2 out of 10, raising that to 2.5 represents a 25 percent increase. While this still falls well short of the performance expected of well functioning budget system, it would represent significant progress. It is arguable that the major problem for the PSMFL may be that its successful implementation is likely to require a country having its budgetary practices operating at least 6 out of 10-and Mongolia is some distance away from a 6. 9.39 Capacity has to be thought of in relation to current and plausible short- to medium-term future levels of performance. In virtually all countries capacity, narrowly defined as skills, knowledge and ability, is significantly greater than the quality of performance in the public sector. Thus the challenge is to reform practices so that they can catch up to and then extend existing capacity. The proposed approach has much to commend it in this context, particularly because it is built around incentives. It can contribute to a focus on the key practices requiring reform and associated capacity needs. Even for local government, where capacity is more genuinely a constraint, experience with the use of this win-win approach in similar situations is that it creates a demand for capacity building from those requiring the capacity. This is the basis of ownership and sustainability. 67 Performance here is in terms of the three dimensions of budgetary outcomes: aggregate fiscal discipline, resources allocated to prionties, and efficient and effective resource use. 163 The PSMFL and MDG as solutions to kick start Mongolia's reform Impasse 9.40 Mongolia is now in a Catch 22 situation on capacity. Agencies enmeshed in a web of external controls have little incentive or opportunity to develop internal management skills--and the absence of such skills is used by central agencies (and donors) as a reason for not relaxing detailed controls. 9.41 The PSMFL involves a different approach to control and regulation, one that emphasizes performance. To move from the current system, which is characterized by arbitrary and ad hoc control, to a performance-based control system will require close attention to the full range of practices in the public sector. 9.42 Closing the gap will be a challenge, but the approach suggested here provides the government with a framework for deepening and accelerating a process that is already under way. Early success in this approach has the potential of increasing support for the reform effort. This will need to happen if the poverty reduction and growth objectives laid out in the Poverty Reduction Strategy and measured by improvements in the Millennium Development Goals (MDG) indicators are to be realized. 164 Civil Service Pay and Employment Policies Srtg/cin * Arbitrary recruitment and promotion practices * AsesrlofcvlsrieBgncvlseiefutoaleiwadpyad * Poor mecentive structure. Low remuneration for skilled workers compared to epomn td odvlpcmrhniesrtg private sector counterparts while low skilled workers fare better than in private * Seuppossnaplcynlssurtocoditeoiyavcendtfigdesos acrosssector * Increasing size, poor composition, of civil service, Sre Acos * The wage bill as a share of GDP is high and rising re s ro mtiosre Bedin need aeri t * Coineve scteoposunertion bisdtwr o slle workers cmae oepomn td odvlpcmrhniesrtg * Increasing size, poor composition, of civil service, * Contan weil es g o mne civil service reor is esge * The wage bill as a share of GDP is high and rising implemented * Civil service composition biased toward low skill workers * Freeze hiring or adopt one-three replacement policy until strategy is developed * Divest from all commercial activity * Contract out low skilled services * Improve civil service and payroll databases * Weak incentive systems impede creation of qualified tenured civil service * Reduce ceilings on wage bonuses to 25-30 percent and increase compliance 0 Monetize all benefits * Make rewards performance-based * Proliferation of government administrative units reduce accountability mechanism * Consolidate central government adnistrative units at the local level and monitoring of employment * Law on Government Services is not enforced fully and does not determine clearly * Review and revise Law on Governent Service to amend remuneration structure, the level of employees employment policies and practices " Training programs are not well developed planned and sequenced 0 Based on the results of the functional review support the training and skills upgrade of S core civil service staff Publc Pension System Strategy/Actions A Eligibility requirements for public pension system are lax and formula is costly Develop timehe to revise and restructure the pension system * Reduce size of cash transfer programs by reforming pension system * Raise effective retirement age, reform disability and survivors transfers * Poor design of pesion system leaves many pensioners vulnerable u Develop a transition policy to put all currently working into a new system * Abrupt and unfair transition from the unsustainable pay-as-you-go system to a * Put entire working population in the new system within one year, with transition notional defined contribution (NDC) system provisions to account for acquired rights SLow retirement age for current and fuiture retirees creates large fiscal burden, 0 Raise retirement age (to reduce number of expected minimum pensions) low pensions Raise retirement age by 6 months per year for women and men up to 62 F Raise retirement age by 6 months per year for everyone " Falling contributions threaten sustainability of the system 0 Remove all entitlements to early retirement * Failure to make disability and survivors pensions consistent with old age I Reform disability and survivors pensions pensions 165 -MY -: C~EONDMNA 1OfS4NWES-IONS1 State Owned Enterprises and Quasi-Fiscal Activities * Many SOEs are unprofitable and require government subsidies (tax arrears) and * Improve comprehensiveness and transparency of the directed borrowing from domestic banks to keep them afloat * Divest from productive activities, improve SOE governance and introduce a time- * Quasi-fiscal activities and other contingent liabilities are not systematically bound plan to reduce level of inter-enterprise arrears accounted for 0 Set up a tracking system to follow potential contingent liabilities on a regular and systematic basis. 0 Prepare and disseminate a policy statement regarding which obligations the government will commit itself to pay for, and which others it will categorically refuse to pay for * Impose restrictions on the issuance of new guarantees and other contingent liabilities Corporate governance, accounting and auditing standards Strategy/Actions * Lack of financial management skills and capacity results in weak corporate accounting and reporting * Improve accounting and auditing standards. Introduce mandatoiy annual auditing of SOES * Accounting systems of SOEs lack transparency and do not explicitly account for 0 Adopt and implement AIS standards and guidelines indirect subsidies provided by government 0 Develop training program on accrual accounting for all SOEs 0 Poor private sector audits 0 Discourage non-cash settlement payments * Carry out peer review of pivate sector auditors * Poor management of large and inefficient energy sector SOEs Establish method for calculating benchmark price of utilities and develop time-bound action plan to revise pricing policy. including use of line tariffs, if needed, to protect low-income consumers. In the energy sector develop time-bound action plan to improve collections, reduce losses and arrears 0 As a first step toward private sector participation, introduce performance-based management contracts * Car out per review of private sector auditors FInancal Sector Reform Strategy/Actions * Financial sector fragility continues to pose a threat to overall fiscal stability * Ensure that banking sector regulations newly adopted in accordance with international best practices are appropriately enforced * Impose timely sanctions for non-compliance * Reforain from directed credits * Strengthen governance, internal control systems, and internal and external audit in __________________________________________________ financial institutions Contingent Liabilities and Managing Natural Disasters p Stratrti/Actions a All implicit liabilities of the government have not been identified t Restrict issuance of new guarantees and other contingent liabilities by GOM and BOM * Costs of managing natural disasters frequently passed on to SOEs and not 0 Nothing here about natural disasters integrated into the budget Set up a dedicated reserve futnd exclusively for dzud costs * No risk management tools to deal with disasters 0 Develop risk management and insurance system 166 lad D S?A!ND' 'La 1JVN.%-' Budget execution and control Strategy/Actions * Budget execution and control: * Install a Treasury and Financial Management system * Treasury financial management information systems implementation- Adopt the Treasury Single Account " Government Accounting and Reporting: * Close all budget entity bank accounts * Train treasury staff and entity accountants * Improve predictability of the flow of bunds * Establish a uniform and integrated Chart of Accounts for accounting, budgeting and reporting of public sector financial activities and adopt IFAC accounting standards for public entities * Strengthen the SAIC in carrying out attestation audits. * Enhance Parliamentary oversight of all audit reports. * Establish treasury and controllership functions to manage various financial activities pertaining to government accounting, cash management and financial reporting. * Lack of a functiomng government accounting information system and appropriate * Develop internal audit methodology for budgetary institutions and aimags accounting and auditing standards * SAIC and SFI should adopt systems-based auditing starting with the assessment of * Lack of a single treasury account and a uniform chart of accounts with an effect on internal control systems of the bodies wluch they inspect financial control and cash management * Estabhish a fully functional treasury system based on Treasury Smngle Account * Consohidate chart of accounts, enhance training of accountants and auditors Governiment auditors and audits Strategy/Actions * Absence of financial statement audits of govemnment finances affects quality and gmovement prsenite ton prhamdeopnosohnulfnacattmnso credibility of financial reporting gvrmn rsne oprimn Structure of government: Strategy/Actions * Local government accountability reporting structures should be consistent with Constitution and other legislation * Multiple government levels leads to fragmentation and duphcation of * Initiate pooling of service provision responsibilities, and undermine economes of scale * Streamline government structure, number of budget entities; strengthen role of line * Accountability for service delivery is compromised by fragmentation of ministries responsibilities * Train local governments in basic financial management systems and processes * Coordination between line ministry and local governments is * Re-centralize provision of all public goods including monitoring, evaluation and inspection functions * Introduce top-down budgetary systems Legal framework Strategy/Actions Laws are inconsistent and overlapping * Any new public reform legislation should clearly define accountability structures and * Lack of clarity in assignments of roles and responsibilities directions of intergovernmental finance while ensuring consistency between laws * Lack of accountability due to multiple enforcement agencies 167 rog- -tt1 7- -W 4 -~~ rT -FIX'RE W-__ Matching roles and responsibilities _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ * There are mismatches between local expenditure authority and revenue capacity Deoliti o nd a ssigmts shouldob mat e n r * Existence of arrears which indicates inconsistency between spending aspirations Ironsibilitie and cntbli r disiomaking and resource availability dntrueftive ancil rules airtion systes that perit Revenue assignment * Responsibility for tax collection is determined randomly by the central government annually * Incentive structure does not encourage effective revenue transfers by de- Develop incentive system to strengthen tax administration at local government levels concentrated tax administrators to central government Transfers StratgActions * Intergovernmental transfers are unstable and unpredictable * Redesign formula for allocating transfers from central government to local governments * Intergovernmental transfers do not fully address resource disparities across local * Develop needs-based per capita transfer equalization formula consistent with tax sharing governments formula or assignments * Funding flows based on infrastructure undermine incentives for facility * Single-pipe funding for service delivery in social sectors and monitor financing rationalization arrangements to identify changing resource needs and implementation capacity of local * Transfers to local governments based on a fixed and variable costs basis, not governments needs-based Agree on purpose and need of transfer, before transfers are made Capacity for effective decentralization StrategyActions " Small local governments lack technical expertise to effectively manage Create framework for decentralization combined with support to local governments to decentralization give them incentives to strengthen their abilities MOFE overst i r l a e c c t a Need header Strategy/Actions S Poor enforcement of auditor general's recommendations and inadequate * Develop improved, timely, and easily accessible government information system parliamentary oversight Improve transparency and accountability in budget preparation, execution, and D There is a growing perception among the Mongolian population that corruption oversight. is increasing Build a strong, effective, and well trained judiciary system with adequate powers to D Procurement processes are not transparent enforce laws. Develop a consistent and transparent code of ethics for the publc sector Strengthen parliamentary oversight and independent audit and investigative bodies C Improve procurement administration by separating regulatory function from purchasing function * Improve licensing practices 168 SJMRYOF REO MNATi 0 ANDCCT[6NS_i -~ey 'r- ~- - z 7. tte ii4o hngPHe oordhiatioin ocesses Policy making process Strategy/Actions * Policy volatility is damaging ability to consistently deliver quality services at 6 Publish budget timetable least cost * Prepare fiscal policy statement derived from Medium-Term Fiscal Framework * Policy making is not adequately disciplined by budget realities 6 Develop Medium-Term Budget Framework consistent with poverty reduction growth * Weak commitment to budget policies leads to unrealistic resource allocations objectives * A tendency to add new policies and programs without balancing cutbacks 6 Strengthen top-down-budgeting processes * An almost exclusive focus on the short term in the annual budget process 6 Allow adequate time for local government budget discussions *Ceiling setting process is not credible Budet fra entation Strategy/Actions " Fragmentation of decision making Decision making is often ad hoc, with Rdc h ubro ugtette assocated igh ransation ostsin buget mking* Reuceishe uro budget timetable * Centralize policy-making role to line ministries * Lack of comprehensiveness of the budget Strengthen coordination in MOFE Develop and implement training strategy for line ministries in areas of program * Off-budget funds, extra-budgetary funds, guarantees, and quasi-fiscal activities are development and costing not tracked in the system. Budget fragmentation imparts an unknown level of * Put forward estimates at the center of decision making fiscal risk in budget deficit targeting and in ensuring transparency in the use of * Move toward expenditure ceilings for current and capital expenditures public funds * Lack of transparency and apparent arbitrariness in actual allocations Increase transparency in budget allocation process * Lack of transparency in bottom-up budget consolidation process undermines Publicize budget timetable early in budget process budget process where it is most important Policies and programs subject to limited review or evaluation of their performance Increase involvement of monitoring and evaluation divisions in budget preparation and or continuing relevance execution process Introduce internal control systems and auditing Education sector: Strategy/Actions " Need to increase enrollment rates, reduce disparities between rural and urban Efficient use of resources should underlie all future strategy education outcomes, and close the gender and income gaps to provide the economy * Increase enrollment rates at the secondary level with well trained labor that will improve productivity and fuel growth * Reduce dropout rate of male students. * Enrollment rates have improved for primary school but remain low at the secondary level " male enrollment rates are systematically lower across regions and grades " Mongolia spends more on education than most low income countries Health sector Strategy/Actions * Maternal mortality ratio varies e across regions and is ugher in rural areas * Shift to result-based budgeting system (to rationalize allocations and ensure adequate e Whle commumcable diseases persist, non-communicable diseases are an funding of national health priorities) increasing problem * Reform the HIF (e.g. identify mechanisms to attract more voluntary enrollees, revise the * Local budetary officials have no incentive to vrevare budgets reflectine national definition of "vulnerable groups" based on income or other measures of poverty, revise 169 priorities the benefit packages, reform the payment system to providers) * Input-based budgeting system with many inherent weaknesses (e.g. 0 Develop National Health Accounts (to track sources financing agents and uses of health mismatch between health priorities and resource allocation, lack of resources) priontization, excessive focus on tertiary care, distinct urban bias) 0 Rationalize the infrastructures and human resources based on national health services * Health Insurance Fund not financially sustainable and human resources master plans * Lack of a good system to keep track of national health expenditures a Introduce new budget principles based on outputs that take into consideration factors tInefficient physical infrastructure and human resources in the health sector such as morbidity mix, poverty level, access to health care. * National Health Accounts (NHA) should be developed to track sources, financing agents and uses of health resources * Design a mechanism for flow of funds in line with the health system hierarchy and the patient referral scheme * MOHSW should have a bigger role in the decision-making process for health budgeting. Mechanisms for local governments to comply with national health goals and strategies should be developed Social protection system Strategy/Actions * Unclear criteria for benefit rationing process. * System should be simplified. Benefit entitlements should not be created that mirror the social insurance scheme, since these reduce incentives to contribute and may not reflect true vulnerabilities * Financing for social assistance should move away from an input basis * Develop a program to reform social care financing so that the state is the purchaser, but not necessarily the provider Environment Strategy/Actions * Mongolia's industrial pollution levels are higher than those recommended by the * Strengthen capacity to plan within a short-to-medium term budget framework; Millennium Development Goals, and respiratory diseases are its leading cause of * Strengthen capacity to identify key targets for the use of both routine and development death. budget allocations; * No strategic planning undertaken in MoE. * Improve policy coordination between National and regional offices; * Poor policy coordination between National and regional offices * Publicly disclose relevant information about MoE overall management process, both at local and national levels, of MoF allocated funds Energy Sector Strategy/Actions * Energy tariffs are regressive, energy access by low-income consumers is deficient * Establish energy tariff policies based on economic cost of supply and willingness to pay and declining, particularly in the rural and peri-urban areas. principles. * Develop a sustainable action plan to reverse decline of energy access in areas outside Ulaanbaatar. Improving the basics rst Strategy/Actions * Past reform efforts have lacked overall government commitment * Develop change management process and strategic plan * Identify champions of reform * Develop agency sequenced implementation plans * Establish agency benchmark elements of success and devolve authority against these benchmarks 170 *Reassert discipline in the public sector by matching authority and responsibility *Create a disciplined top-down process as a basis for improved accountabriity * Make porss on basic management practices Institutional and capacity constraints Strategy/Actions * Agencies enmeshed ma web of external controls have little incentive or * Agency reforms should be comprehensive opportunity to develop internal management skills-and the absence of such skills * Pockets of reform energy hidden in the bureaucracy need to be mobilized is used by central agencies (and donors) as a reason for not relaxing detailed * MOFE must champion the reform controls * MOFE assess performance and regulate processes * The current system, which is characterized by arbitrary and ad hoc control * Improve financial management processes to discipline the budget * Improve intergovernmental arrangements * Build a professional civil service ANNEX 1: DEBT SUSTAINABILITY ANALYSIS 1 ANNEx 2: CASH TRANSFERS AND SOCIAL PROTECTION 9 ANNEX 3: EDUCATION 33 ANNEx 4: PUBLIC EXPENDITURES REVIEW: NATURE AND ENVIRONMENT 45 ANNEx 5: SUMMARY OF RECOMMENDATIONS FOR PUBLIC SECTOR FINANCIAL MANAGEMENT STRENGTHENING. 55 ANNEx 6: RESPONSIBILITIES OF ACCOUNTANT GENERAL AND MINISTRY CHIEF ACCOUNTANTS 66  ANNEX 1: DEBT SUSTAINABILITY ANALYSIS Introduction 1. This debt sustainability analysis (DSA) is prepared jointly by the IMF and World Bank staff and evaluates Mongolia's external debt sustainability by examining critical debt indicators within the context of a 10-year macroeconomic and balance of payments framework. The debt estimates and net present value (NPV) calculations are based on loan-by-loan data on public and publicly guaranteed debt as of December 31, 2001, provided by the Mongolian authorities. In addition, a debt reconciliation process was completed with multilateral creditors, which account for more than 60 percent of debts. 2. The analysis is based on the methodology applied within IPIC framework. According to the methodology, the NPV estimates are calculated using the average currency-specific Commercial Interest Reference Rate (CIRR) for six-month period ending December 31, 2001. 3. Before Mongolia launched its transition to a market economy in 1991, it maintained a very close trade and financial relationship with then the Soviet Union and former Council for Mutual Economic Assistance (CMEA) countries. This resulted in high external debt of more than 10 billion in transferable rubles to the Soviet Union, CMEA countries, and China at the end of 1990, with more than 98 percent owed to Soviet Union. The severe economic shock from the breakdown of traditional trading relationships with the Soviet Union and CMEA counties in the early 1990s, however, rapidly changed the composition and currency of Mongolia's exiernal debt. External debt in convertible currency, which was virtually nonexistent before 1990, began to expand rapidly. The external debt stock in convertible currency, which stood at a mere 6percent of GDP in 1991, jumped to 62.4 percent of GDP in 1993. Since then, growth has been more subdued, with a cumulative increase of 22 percent during the last eight years. This is smaller than the average increase experienced by Russia, the Baltics, and the other states of the former Soviet Union. I In nominal terms, however, the debt stock more than doubled since 1993, to pay for adjustment costs during transition, particularly heavy investments in the energy sector. During the 1990s, transferable ruble debts to CMEA countries and China were all settled, except the claim by Russia. 4. Excluded from the analysis is the unresolved transferable ruble claim held by Russia relating to Mongolia's transactions with the former Soviet Union. Given its high nominal value of $10.5 billion,2 ten times of Mongolia's current GDP, the settlement of this claim has potentially high impact on Mongolia's debt sustainability. As discussion is still going on regarding this issue, hypothetical cases are analyzed later to gauge the potential effect of various scenarios, rather than guide how the claim should be settled. 1 External Debt/GDP ratio increased by 20. 1percent on average during 1995 to 2001 for these countries: source. IMF documents. Assessing Sustainability 2 This is based on the data submitted on September 17, 1997 by Russian authorities when Russia was in the process of participating the Paris club as a creditor. Total debt stock disbursed before January 1992 is $10,514.7 million or TRR 10,511.5 million. 2 The Structure of External Debt and Relations with Creditors 5. At end 2001 Mongolia's public and publicly guaranteed debt amounted to $865 million (84 percent of GDP), of which 61 percent is to multilateral creditors, 37 percent to bilateral creditors, and only 2.9 percent to commercial creditors. The largest multilateral creditor is the Asian Development Bank (36.9 percent of total nominal debt), followed by the World Bank (17.9 percent), and the IMF (5.4 percent). Of the official bilateral debt, most was owed to Japan (22.4 percent of nominal debt), Germany (5.4 percent), and Russia (3.1 percent). Annual debt service currently amounts to 4.3 percent of GDP, as the majority of debt has been contracted on highly concessional terms. 6. In terms of NPV, the total debt amounts to $590 million, or 68 percent of nominal value. An important caveat applies. Since most of Mongolia's debt is denominated in currencies other than US dollar - SDR, Japanese yen, and euro - the US dollar value of nominal debt and NPV are sensitive to fluctuations in exchange rates. In addition, NPV is highly sensitive to changes in CIRRs. In particular, measured at 2000 CIRRs the NPV in 2001 would have been $545 million3 (Table 1). 7. Under the PRGF program, the Mongolian government and BOM agreed not to incur any external arrears during the program period. In 2001, however, an arrears of $4.8 million occurred to Russia, as the parties failed to agree on a method of payment. In the subsequent negotiation in March 2002, both parties agreed to settle this arrears, and a new obligation of $5.5 million coming due in 2002, with a combination of 60 percent cash and 40 percent in kind by the end of the year. Recently, a dispute occurred regarding commercial debt of $18.8 million extended by Marubeni Corporation of Japan to a local textile firm called Buyan Company. As the authorities dispute Marubeni's claim that Mongolian government guaranteed this debt, a court in London is reviewing this matter. This disputed claim is not included in the current debt analysis. Assumptions of the baseline scenario 8. Macroeconomic projections used in this analysis are based on the steadfast implementation of sound macroeconomic policies and structural reforms, as agreed in the context of the PRGF program. Mongolian GDP is projected to grow 3 percent in 2002, 5 percent in 2003, and 6 percent thereafter. Export volume growth is also expected to grow 5 percent and 5.1 percent in 2002 and 2003, and 6 percent thereafter.4 Import volume growth is assumed to increase in line with the GDP growth rate. Given that average GDP growth since 1995 has been 3.1 percent, the base case assumption appears to be reasonable with accelerated structural reforms under the PRGF program. Given that the average export volume growth rate has been 9.3 percent since 1995, the assumed rate of 6 percent for export volume growth is conservative, reflecting the volatility of Mongolia's commodity exports. 9. The medium-term price projections are based on the most recent WEO assumptions for commodity prices (March 2002) and other relevant price projections. The projection shows general improvements in terms of trade up to 2007. From 2008, the terms of trade is assumed not to change. During the projection period, the togrog is assumed to remain unchanged in real effective terms. 10. On the financing side, a declining trend in official grant is projected to be compensated by increasing foreign direct investment (FDI) and other private capital inflows. While official grants are assumed to decrease from $95.5 million in 2001 to $73 million in 2010, this gap is more than compensated by increases in FDI and an expanded inflow of worker remittance. FDI is projected to rise 3 For example, $ CIRR is at 5.85 percent as of the end of 2001, as compared to 7.19 percent at the end of 2000. 4 Copper, gold and cashmere exports volumes are assumed to grow at () percent respectively. 3 gradually from $43 million in 2001 to more than $70 million in 2010, and private unrequited capital inflow from $39 million to $50 million, thanks to increases in Mongolian workers abroad. 11. It is assumed that Mongolia will continue borrowings from multilateral and bilateral sources on concessional terms. New disbursements of project loans are assumed to increase at a much slower pace than the GDP growth-from $75 million in 2002 to $92 million in 2010. During the PRGF arrangement, additional financing is assumed to be provided by multilateral institutions to bridge the financing gap. It is assumed there will be no financing gap from 2005 on. Central government revenue excluding grants, which stood at 36.5 percent of GDP in 2001, is projected to be constant at 34.5 percent from 2002 on. (Relate it to fiscal sector5 12. On the basis of the assumptions and policies, the total debt stock as a percentage of GDP is projected to peak at 87 percent in 2003 and decline to 69 percent in 2010. Likewise, the NPV as percent of GDP is expected to decline from 57 percent in 2001 to a more comfortable 49 percent in 2010. The annual debt service burden, which stood at 4.3 percent of GDP at end 2001, is projected to continue to decline to 2.6 percent in 2010. Thus, baseline scenarios indicate that Mongolia's debt burden is sustainable and Mongolia does not require debt rescheduling, barring major negative changes in baseline assumptions. Sensitivity Analysis 13. Mongolia's balance of payments projection, however, is subject to considerable uncertainties, particularly with regard to highly volatile export prospects, future fiscal consolidation, the conditions of future external financing, and, most of all, the eventual settlement of the transferable ruble claim by Russia. Mongolia's export base is very narrow, largely dominated by commodities such as copper, gold, and cashmere, which are subject to sharp price and volume fluctuations. Conditions of new borrowing can also change in the medium term, forcing Mongolia to borrow on non-concessional terms. Thus, sensitivity analysis has been carried out under four different scenarios: (1) lower GDP growth rate and lower export volume growth rate; (2) deteriorating borrowing terms resulting in some non-concessional borrowing; (3) combination of (1) and (2), and, finally, (4) various terms of settlement of the Russian claim. 14. Lower GDP and Export growth: This scenario shows that, with GDP and export volume growing at 3 percent per year, as compared to 6 percent in the baseline scenario, the debt stock/GDP peaks in 2004 and remains at 86 percent of GDP in 2010, rather than declining to 67 percent in the base case scenario. The same is true for NPV, which will increase to 61 percent of GDP in 2010, compared to the current 57 percent ratio. Annual debt service burden decreases very slowly, to 3.2 percent of GDP in 2010, as compared to 2.6 percent of GDP in the base case. 15. Less Favorable Financing: This scenario assumes that additional financing during 2002 and 2004, and $20 million each year thereafter are contracted on non-concessional terms. The detailed terms are assumed to be 2-year grace period, 8-year maturity and interest rate higher than CIRR rate by 2 percent (a grant element of --- percent), instead of concessional terms assumed under the baseline scenario (a grant element of --- percent). While NPV and annual debt service burdens get higher at 51.4 percent of GDP (2.4 percent higher) and 3.1 percent of GDP (0.5percent higher) in 2010, the ratio of total debt stock to GDP in 2010 turns out to be 67.8 percent, lower than 69.3 percent in the base case. 16. Lower Growth and Less Favorable Financing: Under the combined scenario of lower GDP and export volume growth and unfavorable financing, debt dynamics worsen noticeably. The ratio of total s In 2001 authorities estimated FDI at $63 million. Staff lowered its estimates because of the local practice of recording FDI on a commitment basis and poor tracking of FDI repatriation. 4 debt stock to GDP peaks at 90.2 percent in 2004 and is expected to be 84.6 percent of GDP in 2010, 15.3 percent higher than the base case. Likewise, NPV will be 64 percent of GDP in 2010, 15 percent higher than the base case. Annual debt service burden is 3.9 percent of GDP, still bearable, but 1.3 percent higher than the base case. Overall, this scenario shows that debt dynamic is more sensitive to GDP and export volume growth than changes in financing conditions. Sensitivity analysis with Russian claim settlement 17. The single largest uncertainty in the Mongolian DSA, however, remains the terms of settlement of Russia's transferable ruble claims. For purely analytical purpose, we applied hypothetical upfront discount rates of 80, 85, and 90 percent to the original claim, based on precedents in other countries.6 The remaining claim is assumed to have terms of 20-year maturity, 5-year grace period and an interest rate of 3 percent in dollars, lower than comparable CIRR of 5.85 percent. In addition to the heavy upfront discounts of 80 to 90 percent, Naples debt reduction terms were applied to all bilateral and commercial debts, including the discounted Russian claim. As an open economy with an exports/GDP ratio around 60 percent and fiscal revenue well over the 15 percent threshold, the criteria of 250 percent for NPV/government revenue ratio was used to test HIPC eligibility. Given the enormous size of the Russian claim, the result shows that, even after debt reduction with Naples terms and upfront discounts, Mongolia will become eligible for the HIPC Initiative with 80 or 85 percent upfront discounts. For Russian claim alone, the final discounts amount to 93.3, 95, and 96.7 percent. Sensitivity analysis was also carried out on the terms of repayment for the remainder of the Russian claim. When the interest rate was raised to 5 percent, still lower than CIRR, the NPV of the Russian claim increases by 14 percent over the base case, and Mongolia becomes eligible for HIPC, even with 90 percent upfront discount. This result highlights the critical importance of settling the Russian claim in a manner consistent with Mongolia's debt sustainability, taking into account the implications for debt to other creditors. 18. Mongolia and Russia have been discussing the Russian claim at the government level. Mongolia disputes Russia's claim on the grounds that bilateral trade before 1991 was based on artificial prices that favored Russia, and that a large portion of debt was imposed on Mongolia for Soviet military presence. In addition, Mongolia claims that it has large counterclaims, in part related to environmental damage. So far no agreement on the total size of claim or method of settlement has been reached. Conclusion and policy implications 19. Excluding Russian claim, Mongolia's external debt appears to be sustainable, although export-led growth is critical to continue to lower the debt service burden in coming years. At the policy level, structural reform, particularly fiscal consolidation, needs to continue to be aggressively pursued so as not to incur additional debt. Even after the PRGF program, contracting loans on non-concessional terms should be avoided as much as possible. This strategy will require expansion of FDI as a foreign capital source, which, in turn, will require structural reforms to promote private sector-led growth, including more aggressive privatization. As about half of Mongolia's loans are centered on the energy sector, the government needs to make sure that projects in these sectors are selected on a priority basis and are implemented efficiently, while adopting measures to promote proper functioning of price mechanisms in the energy sector. Last but not least, prudential macroeconomic policies, including a flexible exchange rate that adequately reflects Mongolia's external competitiveness needs to be pursued. 6 Vietnam settled its transferable ruble debt with Russia in 2000 by obtaining 85 percent reduction of the principal and agreeing to pay back the remainder for 23 years at an interest rate of 5 percent (EBS/01/43). Russia also applied 90 percent upfront discount to its debt to Nic's debt to Russia Nicaragua (EBS/ /). TABLE 1: BALANCE OF PAYMENTS 1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 19981/ 1999 2000 Actual Est. Est Est. Est. Trade balance -579 -140 -626 -8.7 -7.9 -3.3 -874 30.2 -120 -113 -140 Exports f.o.b. 444.8 346.5 355.8 365.8 367 485.6 4234 5685 462.3 454.3 535.8 Copper 1602 155.1 188.7 2576 174.5 199.1 124.8 119.2 160.3 Imports, c.1.f. 2/ -1024 -487 -418 -375 -375 -489 -511 -538 -582 -567 -676 services, net -72.6 -58 -25.4 -31.1 -33.8 -23.4 -6 -7.3 -11.5 -8.4 -10.6 Receipts 53.1 26.5 35.2 26.8 43 42.6 55.7 80 77.9 99.6 102.2 Payments -153 -32.3 -60.6 -57.9 -76.8 -66 -617 -87.3 -894 -108 -114 Income, net -22.7 -13.3 -14.3 0.4 -10.1 -144 Official interest payments 3/ -9.4 -9.3 -7 6 -92 -7.9 -9.5 private unrequited transfers -2.7 -0.1 -3 -2.8 6.2 4.2 2.5 7.4 8 Current account balance, excluding official transfers -651 -146 -90.7 -39.9 -44.7 -52.2 -101 12.8 -129 -124 -151 Public unrequited transfers, net 7.4 41.6 385 71 766 79.1 63.6 61.5 53.2 67.1 98.6 Official grants 7.4 41.6 38.7 667 61.9 70.9 562 54.8 45.8 60.2 86.2 Other official transfers, net -08 43 14.7 82 7.4 6.7 7.4 6.9 12.4 Financial and capital account 583.5 106.9 30.7 -9.6 -32.6 19 41.2 52.5 116.4 58.6 68 5 Direct investment 2 7.7 7 9.8 15.9 20 18.9 30.4 40 Loans - - - 93.7 109.1 90.6 Medium- and log-term, net 516.7 130.3 863 325 11.5 50.9 56 109.8 937 97.2 80.6 Disbursements 537.4 1393 1121 54 63.9 962 93.4 1423 1225 1737 151 Public sector - - - 122.5 1018 71 Private sector - - - - 719 80 Amortization -20.7 -9 -25.8 -21 5 -52.4 -45.3 -37.3 -32.5 -28.8 -766 -66.4 Public sector - - - -28.8 -317 -154 Private sector - - - - -44.9 -55 Short-term, net 66.8 -23.4 -57 6 -49.8 -51 - - - 0 11.9 10 Currency and deposits, net - - - 40 -22.7 -12.5 Commercial banks, net -8 1 -9.3 -18.1 4 -229 -135 Other - - - - 02 1 Trade credits, net - - - 6.6 -12.9 -97 Other, net - - - -42.9 -45.3 -40 Errors and omissions 7.3 -94.6 8.9 5.2 -5.8 -546 -304 -69.6 -69.6 15.4 33.9 Overall balance -53.3 -92.1 -12.6 26.7 -6.6 -8.7 -26.1 57.2 -28.8 17 34.1 Financing 53.3 92.1 12.6 -26.7 66 8.7 26.1 -57 2 28.8 -17 -20.8 Increase in net official reserves (-) 53.3 28.8 6.1 -22.6 -10.1 5.1 20.1 -64.3 15 -30.5 -35 3 Use of IMF credit -9.5 -1.9 6.8 -1.3 4.2 1.6 Increase in gross official reserves (-) 14.6 22 -71.1 16.3 -34.7 -36.9 Arrears accumulation (+)/ payments (-) (net) 63.3 -11.5 -ll -39.8 -2.4 5.9 -1.6 13.8 2 -3.1 Accumulation of new arrears (-) - - - 13.8 158 0 Repayment of outstanding arrear (-) - - 0 -13.8 -31 Exceptional financing/ rescheduhng 18 6.9 56.5 59 0 8.7 0 138 17.' Financing gap 0 0 0 0 0 0 Memorandum items: Curent account balance, excluding official transfers (in percent of GDP)2/ -5.5 -10 1.3 -13.2 -13 7 -172 6 TABLE 1: BALANCE OF PAYMENTS 1990-2000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1/ 1999 2000 Curent account balance, including official transfers (in percent of GDP) 2/ - - - -7.8 -6.3 -7 Imports, f.o.b. - - - 495 482.1 606.3 Exports. f.o.b. (annual growth) 323 -12.8 34.3 -18.8 -1.7 18.2 Imports, c.i.f. (annual growth) 2/ 32 45 5.4 82 -2.6 212 Export volume growth 6.9 11.3 294 -0.7 5.6 7 Import volume growth 2/ 20.9 1.8 14.5 16.1 03 17.7 Export price index change - - - -183 -6.9 105 import price index change - - - -6.8 -2.9 2.9 Trade balance (in percent of GDP) - - - -12.4 -12.5 -155 Net official reserves (end-period) 493 205 4.6 267 37 2 67 9 54 896 74.9 105.4 140.6 Gross official reserves (end-period) 4/ 177.4 126.3 41 1 65 4 92.4 1149 97.8 137.2 1247 156.8 190.9 (in weeks of next year/projected imports, c.i.f.) 2/ - - - 11.4 11.9 14.2 Outstanding arrears (end-period) 63.3 51.8 40 8 1 1.4 7.3 09 13.8 15.8 0 Debt service (in percent of exports of goods and services) 3/ 121 118 63 7.3 9.3 6 Public and publicly guaranteed external debt stock 3/ 503.8 542.4 604.8 753 850 854 NPV of public and publicly guaranteed external debt 3/ - - - - 571 558 Public And publicly guaranteed external debt stock (In percent of GDP) 3/ 528 54 637 774 93.9 88 NPV of public and publicly guaranteed external debt( In percent of GDP) 3/ - - - - 63 57.5 GDP (in millions of US dollars) - - - 972 906 970 Sources: Mongolian authorities; and Fund staff estimates and projections 1/ In 1999,BOP coverage was expanded due to a new private sector survey. Thus, 1998 data is not directly comparable with subsequent years 21 Includes the import of an aircraft valued at $28 million in 1998 3/ Includes payments on CMEA debts that have been converted via fomal intergovernmental agreements 4/ Includes monetary gold, and valuation effects. 5/ Assumes implementation of staff pRsals. TABLE 2: MONGOLIA: DEBT BURDEN RATIos (in Millions of USD) 2001 2002 2003 2004 2005 2006 2007 2008 200920102011Av Av Bas Case .NPV of debt-to-exports ratio 102 106 113 112 106 102 98 96 94 91 89108 94 3. NPV of debt-to-revenue ratio 226 229 230 226 217 209 199 191 183 174 166222183 .Total debt service-to-Exports Ratio 6 6 5 5 6 6 6 6 6 6 6 6 i Total debt service-to-Revenue Ratio 11 12 11 12 14 13 13 12 12 12 12 12 13 Note: External sustainability ratios include government and government-guaranteed Private non-guaranteed debt (PNG) not included 7 TABLE 3: SELECTED EcONOMIC INDICATORS Nominal GDP (1999):$906milhon 6 6.7153285 8.379562 11.29927 13.708029 12.846715 Population (1999):2.38 nullion 41.1 41.1 41.1 411 41.1 41.1 40 40 40 Quota:SDR 51.1 million 6 6 6 6 6 6 1995 1996 1997 1998 1999 2000 2002 2003 2004 Real GDP 6.3 2.4 4 3.5 3.2 1 1 4 5 6 Consumer prices (period average) 56.8 46.8 36.6 9.4 7.6 11.6 6 5 5 Consumer prices (end period) 53.1 44.6 20.5 6 10 8.1 6 5 5 General government revenue 25.6 24.8 25.5 27.6 27.2 33.6 32.7 32.6 326 General government expenditure 31.5 32.7 34.5 41.9 39.4 405 39.8 39.2 38.6 Current balance 6.2 4 1.6 -0.7 -0.2 31 3.1 3.3 4 Overall balance -6.7 -8.2 -86 -14.3 -122 -6.8 -7.1 -6.6 -6 Debt to GDP normalized 6.0 67 8.4 11.3 13.7 12.8 Net domestic credit to government -2.1 2.1 -3.4 41 1.3 -0.7 -2.6 -13 -0.9 Government debt 1/ - - - 85.3 102.2 958 97.6 97 945 NPV of total public debt - - - - 71.7 66 62.4 60.8 58.3 Net foreign assets 87 69 63.8 -31.6 96.9 33.6 168 - - Net domestic assets 124 -1.4 -6.3 56.5 -23.6 -17.9 -7.6 - - Domestic credit -21.6 133.4 -5.1 60.1 -11.1 -8 -4.4 - - Credit to enterprises 18.1 -20.1 13.2 18.5 -39.7 295 33.9 - - Broad money 32.9 25.8 19.8 8.8 31.7 17.5 11 1 - - Reserve money - 36.2 26.2 13.5 51.8 194 10.8 - - Broad money velocity (GDP/BM) 2/ 5.4 5 5.5 5 4.3 4.1 3.9 - - Annual interest rate on central bank bills (percent) 3/ 62.4 42 35 23.3 11.4 8.5 - - - Current account balance, excluding official transfers -52 -101 14 -129 -124 -167 -159 -155 -159 (in percent of GDP) -4.3 -8.5 1.3 -13.2 -13.7 -17.2 -14.7 -13.4 -12.7 Current account balance, including official transfers 26.9 -36.9 75.2 -75.5 -57.1 -68.3 -76.5 -70 1 -74.2 (in percent of GDP) 2.2 -3.1 7 1 -7.8 -6.3 -7 -7.1 -6.1 -5.9 Trade balance -3 -87 31 -120 -113 -150 -132 -124 -126 (in percent of GDP) -0.3 -7.4 3 -12.4 -12.5 -15.5 -12.3 -10.7 -10 Exports, fob 486 423 569 462 454 537 601 647 693 (Percent change) 84.1 -12.8 34.5 -18.8 -1.7 18.2 9.4 7.7 7 Imports, cif 489 511 538 582 567 687 733 771 819 (Percent change) 12.8 4.5 5.4 8.2 -2.6 21.2 49 5.2 6 2 Financial and capital account balance 19 41 53 116 59 68 81 81 89 (in percent of GDP) 1.5 3.5 5 12 65 7 1 7 5 7 7 Gross official international reserves (end- period) 115 98 137 125 157 191 231 247 266 (in weeks of next year /projected imports c.i.f.) 11.7 9.4 12.3 11.4 12 14.2 15.4 15.5 15.7 Public and publicly guaranteed external debt 4/ 504 542 605 753 850 854 1009 1089 1171 (In percent of GDP) 41.1 46 57.4 77.4 93.9 88 93.6 94.2 931 NPV of public and publicly guaranted external debt - - - - 571 558 639 679 724 (In percent of GDP) - - - - 63 57.5 59.2 58.7 57 6 Debt service 64.2 56.6 41 39.3 41.3 315 37.4 40.9 36.5 8 TABLE 3: SELECTED ECONOMIC INDICATORS In percent of exports of goods & services 121 11.8 63 73 9.4 6 53 5.4 45 Exchange rates and trade pnces Togrogs per US dollar (end of period) 474 694 813 902 1,072 1,097 1,198 1,226 1,255 Togrogs per US dollar (period average) 449 548 790 841 1,022 1,077 1,179 1,212 1,24 NEER, end-period (1995=100) 95.9 69 63 69 60.8 59.3 - - - REER, end-period (1995=100) 1082 1059 1132 1176 107.5 108.1 - - - Export prices (US dollar, percent change) 26 -218 -34 -183 -6.9 105 47 3.1 2 Copper price (US dollar, percent change) 402 -36.9 19.3 -38.1 -58 18.4 13.3 6 34 Import pnces (US dollar, percent change) 9.6 1.7 -6.5 -6.8 -2 9 2.9 0.3 0.5 1 6 Private saving-investment balance (in percent of GDP) 81 47 16.2 6.5 59 -02 0.1 05 01 Terms of trade (percent change) 15 -23.1 3.3 -12.3 -42 7.4 44 2.6 0.4 Nominal GDP (billion togrogs) 550 647 833 817 925 1,045 1270 1,401 1,559 Nonunal GDP (nullion US dollars) 1 227 1 179 1 054 972 906 970 1,078 1,156 1,257 Sources. Mongolian authorities 1/Includes IMF, guarantees and arrears 2/Seasonally adjusted figures for broad money velocity from 1997onwards 3/Annuahzed yield on end-period auction of 14-day bills. 4/Excludes Russia's unresolved 10 6 billion transferable ruble claim 9 ANNEX 2: CASH TRANSFERS AND SOCIAL PROTECTION7 Introduction. 1. This report is part of the Mongolia Public Expenditure Review project. It is based on successive missions to Mongolia in 2001, and reports prepared by other donors. The paper (a) reviews the system outcomes, in terms of labor market performance and poverty, (b) outlines the main government programs and policies in the sector and assignments of roles and responsibilities, benchmarking these where possible against other countries, especially transition economies, and (c) offer recommendations for reform which would improve the efficiency of expenditure relative to outcomes as well as increase the targeting. 2. The transition with an extensive social protection system based on Soviet practice. The main elements were: * a guaranteed job and living quarters; * a universal, generous pension system, and * residential institutions for children in difficult circumstances, the infirm, the elderly and persons with disabilities. 3. The latter two were not designed encourage inclusion. They were designed to keep vulnerable individuals out of the labor force and, in the case of the institutions, out of society. The system prevented abject poverty when a guaranteed job was the safety net for most of the population (those of working age and their dependents). This was been challenged by the transition, because of: (a) the decline in the formal sector wage employment and the consequent emergence of the informal sector, and (b) the economic decline and fiscal collapse, which both increased poverty and reduced the means of the state to alleviate poverty. Similar to other transition economies, Mongolia in the early 1990s introduced new social protection institutions. Social safety net programs were introduced as part of a broad range of programs developed to cope with new social needs, including a full social insurance system (pensions and short term income replacement benefits), employment services to assist the unemployed, and a set of social assistance benefits. The residential care system was retained and expanded. This chapter reviews the performance of these institutions, focusing primarily on the cash benefit system, which makes up over 90% of government spending on social protection. As the benchmark of a social protection system is its success in preventing poverty and vulnerability through reducing and mitigating poverty-creating social risks, we begin with a brief review of the current poverty indicators. The social protection problem in Mongolia 4. Poverty and vulnerability indicators. According to the 1998 survey, about 36% of the population of Mongolia is poor. The poverty rate (headcount) is higher in urban areas (39%) than rural (33%), especially in urban areas outside of UB (45%). Correlates of poverty include unemployment and a high ratio of dependents to earners - large families and single parent households. Unemployment was measured at 15% in the household survey, although it is only 6 percent if only those registered at the This paper was prepared by Louise Fox. Comments welcome to LFox@worldbank.org 10 employment offices are counted. However, in small cities and towns, unemployment among the poor is reported in the household survey to be over 50%. The lack of jobs in the outer cities (resulting primarily from factory closings) has triggered a steady migration to JB. These migrants live in squatter (ger) areas on the edge of the city, with less access to urban and social services. Most of these workers find employment in Mongolia's growing informal sector. 5. Employment and the labor force. It is difficult to get a clear picture of the labor market, the structure of employment and the number of unemployed in Mongolia. Formal statistics show the working age population growing by from 1.1 million people in 1992 to 1.3 million people in 2000, but the number of economically active and the number of employees remained roughly constant at about 850 and 810 thousand respectively over the same period (Government of Mongolia, 2001). This indicates a withdrawal from the labor force of some 200,000 people (in other words, an LFPR moving from 77% of the working age population to 65%), and/or a large increase in unemployment. With respect to unemployment, official figures show that the number of registered unemployed has gone down from 54 thousand to 39 thousand. However, most of those who consider themselves unemployed do not register at unemployment offices. The 1998 household survey recorded self-defined unemployment levels of 11.7% in rural areas and 30% in urban areas, accounting for some of the missing working age population. Note that the definition of employment and unemployment is this survey is unclear (FIDE, 1999), so it is hard to compare the ratios. 6. Several reports have drawn attention to the growth of Mongolia's urban informal sector as a source of employment and income. Anderson (1998) reports that because of the collection techniques employed, official labor force data do not report roughly 75-85% of the informal sector. In UB, he estimates that about 35% of the total working age population works in the informal sector (some would also have formal jobs, and some would be above working age, so these numbers are not strictly comparable with the numbers above). Bikales, Khurelbaatar, and Schelzig (1999) estimate that over 60,000 people nationally in urban areas are employed full or part time in this sector, including a significant number of pensioners. They estimate that this sector produces 13-30% of GDP. Most of those working in the informal sector have responded that they are "unemployed" in this survey, indicating that not only the authorities but the individuals concerned consider themselves "unemployed", despite earning a significant income. 73% of households in UB who report that the head of the household is "unemployed" also report informal sector income. In the 1998 household survey, urban households reported that 16% of household income came from a non-farm business, and the percent of income from this source increased with total household income pr capita - that is to say that it was positively correlated with household income per capita. This is not surprising, since Bikales, Kurelbaatar and Schelzig estimate that the average annual value added per worker per month is over 150,000 tgs (excluding part time taxis) - nearly 5 times the average wage. 7. What can be concluded from this data patchwork? First, there probably has been some decline in labor force participation in the working age population, but probably not as much as would be implied by the official statistics if it were assumed that all those not employed in the formal sector or registered as unemployed have withdrawn from the labor force.8 Data on the unemployment rate, according to a conventional behavioral definition of not having worked but having looked for work, is not available. Proper labor force survey data is not collected to measure this trend. At the same time, significant net job creation has taken place in urban areas in the informal sector (non-farm business), suggesting an active labor market. This informal sector income has been important in keeping urban households out of poverty, and has probably been an important source of economic growth, especially in the service sector. 8 All FSU countries have seen a decline in labor force participation, owing to (a) early retirement packages encouraging withdrawal from the labor force, (b) discouraged worker effects, and (c) declines in female LFPR in women of childbearing age. 11 TABLE 1: HOUSEHOLD INCOME BY SOURCE, 1998 SHARE OF INCOME BY SOURCE Urban Rural average poorest average poorest Money income: (salary or government 54.6 71.1 14.8 25.8 transfers) of which: pension 6.4 15.3 4.1 5.8 of which: social benefits 0.9 3.8 0.6 3.4 Other income: 45.4 28.6 85.2 74.2 of which: from non-farm business 15.9 4.9 2.3 0.4 of which: consumption of own products 4.5 4.5 51.7 41.3 Source: Government of Mongolia, 2001 8. Structure of household income and the role of cash transfers. In general poor households can be characterized as having a lack of entrepreneurial income - either business income or home consumption. Cash transfers from the government are an important share of household income for urban households, making up nearly 20% of total income. In both urban and rural areas, pensions are a more important source of income than other social benefits. Although these transfers make up a larger share of household income in poor households, it is impossible to conclude from this data whether these transfers are, on average, progressive (i.e. whether more funds go to poor households than to richer ones). Government programs 9. Expenditures on social welfare programs were 7.3% of GDP in 2000. 93% of this expenditure is for cash transfers to households, representing 17% of total government expenditure. Cash transfers have been rising steadily as a share of government expenditure. Pensions are the largest, accounting for over 3/4 of total cash transfers. Family benefits are the next largest expenditure; including payments to women on maternity and child care leave. Although social insurance transfers were planned to be self-financing (i.e. financed from earmarked payroll taxes) revenues have not kept up with expenditures. As a result, social welfare expenditures are a constant source of fiscal pressure. TABLE 2: EXPENDITURE ON SOCIAL WELFARE PROGRAMS 1998 1999 2000 Share of Share of Share of Share of Share of Share of Budget GDP Budget GDP Budget GDP 15.1 6.3 16.4 6.5 17.2 7.3 Source: Government of Mongolia (2001) TABLE 3: SHARE OF TOTAL CASH TRANSFER EXPENDITURE BY TYPE OF PAYMENT, 1985-2001 Type of benefit 1985 1990 1993 1995 1997 1998 1999 2000 2001 est. SI Pension 60.2 67.7 85.8 86.8 77.3 78.6 73.2 77.4 79. SA Pension 0.0 0.0 3.5 2.5 4.0 3.4 5.0 5.0 4. SI Short-Term 18.8 15.5 4.8 4.3 3.5 3.6 4.3 3.9 4.C A Short-Term 19.5 14.6 5.8 4.0 10.3 8.9 11.6 8.3 10.1 I Others 0.0 0.0 0.0 0.1 0.2 0.2 0.2 0.2 0.3 A Others 1.5 2.2 0.2 2.3 4.6 5.2 5.7 5.3 2.C otal 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.C (SI= social insurance, SA= social assistance) 12 10. According to Mongolian laws, the Government provides two types of cash benefits: (a) social insurance benefits (provided to those who are insured) and (b) social assistance benefits (provided to those who qualify). The main benefits are listed below, (see Annex 1 for detailed description). Social insurance: * pension: old-age, permanent disability and survivor's * short term benefits: sickness, temporary disability, and maternity benefits, funeral grant, work injury and illness benefits, unemployment benefit Social Assistance: * social assistance pension: old age, permanent disability, and survivors * one time and short term benefits: maternity, child care, infant grant, twins grant, guardian benefit 11. Services in kind are also provided by government social welfare providers, including residential care for children and adults, job brokerage services, rehabilitation and prosthesis services, and other services for vulnerable groups according to community decision. A range of social care services is provided by NGOs, primarily in UB. Social insurance benefits are supposed to be financed by an earmarked tax (contribution), while social assistance programs are financed primarily by local budgets (according to national legislation). A mix of state and local budgets finances social care programs, and employment services by a mix of central and local funding, and since 2001 also by the Employment Fund (services to non-insured workers). TABLE 4: SOCIAL INSURANCE CONTRIBUTION RATES Program Formal sector employees Self-employed Employer Employee Total Pensions 13.5 5.5 19.0 9.5 Short term benefits 1.0 1.0 2.0 1 Work injury benefits 1.0-2.0-0.3 - 1.0-2.0-0.3 1 Unemployment benefits 0.5 0.5 1.0 - Health care insurance 3.0 3.0 6.0 800 Tg per month. Total 19.0-21.0 10.0 29.0-31.0 11.5 + 800 Tg 12. The public pension system. The public pension system provides monthly benefits to eligible recipients. By international standards, eligibility requirements are lax. Pensions are given at a relatively young age (see annex 1). Only 25 percent of pensioners are 60 at the time of receiving benefits. Women can receive benefits at 55 if they have 20 years of service, or 50 if they have raised more than 4 children to the age of 6. Men and women can also retire early if they work in special occupations. As a result, the number of pensioners is nearly twice the share of the population over 60 (see table 6 below). Life expectancy at normal retirement age is 14 years for men, 20 years for women. If Mongolians behave as pensioners in most FSU countries, most are economically active for at least the first 5 years after so-called "retirement", despite prohibitions on working and receiving a pension, often in the urban informal sector or in agricultural activities. 13. The pension benefit formula is fairly generous. With 20 years of service, benefits are 45% of the pensioner's monthly average wage. Each additional year of service accrues an additional 1.5% pension. Average years of service are 33, entitling most pensioners to receive a benefit equal to 65% of average wage at the time the pension is granted. As this benefit is untaxed, it represents a nearly full replacement rate. By contrast, given Mongolia's life expectancy at pension age and survival probabilities, a private firm would require about 10 -15 more years of age and service before providing a benefit at this level 13 based on Mongolia's current contribution rate.9 A reduced pension is offered for those with more than 10 years of service but less than 20 years of service. In 2000, only 7% of all current old-age pensioners had taken this reduced pension option. 14. The majority of pensions in payment are not as generous as the formula would indicate, leaving many pensioners closer to the poverty line. The average pension benefit is currently about 60% of the average covered wage , and 121% of the national poverty line. But this average is deceiving. 55.7% of old age and disability pensioners and 72% of survivors' pensioners in 2000 received the minimum pension, equal to about 16,000 Tgs, about half the average covered wage and about 4% above the average poverty line. 68% of old age pensioners receive a benefit below 21,000 Tgs. The reasons for this are: (a) the inflation of the early 1990s wiped out most pensions in real terms so almost all who received pensions before 1995 get the minimum pension; and (b) indexation has been ad hoc since then. For example, pensions were completely recalculated in 1999. No indexation was given in 2000, and then in April 2001 pensions were increased by factors ranging from 10 percent for the lowest to about 4% at the top. TABLE 5: AVERAGE BENEFIT AS A SHARE OF AVERAGE NATIONAL POVERTY LINE, 2000 Share of poverty line Type of benefit Average benefit per capita SI pensions 18629 121.5% SI disability pension 15779 102.9% Survivor's pension 15227 99.3% SA pensions 11503 75.0% Unemployment benefit 12652 82.5% SA child care benefit 7591 49.5% 15. One of the main characteristics of the social insurance system in the transition is the declining coverage rate. In 2000, only 545,585 people were reported contributing to the pension system, thus gaining years of service toward a pension. The compares with roughly 800,000 people known to be employed under NSO's very narrow definition of employment (excluding most of the informal sector). Several factors have contributed to the low number of contributors. First, until 1995, contributions were not required for social insurance, so everyone who worked was covered. As a result, the marginal return to contributions for those who already have a long service record is low. Second, agriculture and animal husbandry activities have been privatized, and these sectors are-mostly not covered by social insurance, as they avoid and evade these taxes. Third, as noted above, an informal sector has grown up which also avoids and evades taxes. Fourth, arrears have been building up as employers fail to pay their taxes (although this problem reduces revenue more than coverage since employers may pay for some portion of the year). Finally, tax evasion appears to be high even among those classified as employees in the formal sector (or more correctly, their employers). Note that some of those in the labor force who did not contribute in 2000 will still receive full pensions. All those who worked at least 20 years before 1995 (e.g. most of those born before 1958), will receive the minimum pension regardless of whether they have contributed since 1995, and if they worked at least 10 years they will receive a partial pension on retirement. Comparing benefits paid with an actuarially fair annuity is one method of measuring the efficiency and equity of the pension system. It is, among other things, one measure of the system generosity towards current recipients. It ignores desired inter- and intra-generational redistributions. The above calculation is very rough, and is based on IMF, (1998). 1o The average covered wage is the average of wages on which social insurance contributions were paid. This is somewhere between 20 and 50% lower than the average economy-wide wage. As noted above, labor force statistics are poorly collected, as they primarily reflect the formal sector. Taxes are also poorly collected, however. 14 TABLE 6: SOCIAL INSURANCE PENSION SYSTEM INDICATORS enioors 1990 1995 1998 19 2000 2001 (proj.) PENSION ExPENDffuREI/GDP 0.5 3.2 4.5 4.3 5.6 5.9 Pension contributions/GDP 0.2 2.3 3.0 3.0 3.9 3.8 Government transfer/GDP 0.0 1.3 1.0 1.3 1.5 2.0 Pensioners/population 20-59 26.0 27.9 21.7 20.6 21.4 21.0 Pensioners/contributors N/A 69.5 43.7 44.8 46.6 45.9 Pensioners/population 60+ 183.7 221.4 155.7 177.2 194.1 197.2 Average pension/average covered wage N/A 35.4 59.4 60.2 58.8 61.0 Covered wage b1l/GDP N/A 13.0 17.5 16.0 20.4 21.1 Source: SSIGO, mission estimates 16. Pension expenditures have been growing at about 8% per year in real terms since 1997 even as the number of contributors as a share of the labor force has been falling, resulting in a steady decline in contributions relative to benefit expenditures. For short term insurance benefits such as unemployment benefits, a decline in coverage reduces liabilities rapidly. In the case of pensions, as noted above, the decline in liabilities will not appear for many years - until the pensions are actually paid." Not surprisingly, a structural deficit has emerged in the pension fund, which has been covered by budget transfers. Parts of these transfers are simply a shifting of arrears from one level of government to the other. At the end of the first semester of 2001, the stock of social insurance contribution arrears for the general government to the social insurance budget (including health insurance contribution arrears) stood at about 4.5% of total contributions in 2000. 2/3 of the total stock (and most of the 2001 increase) came from the local government sector. 17. Pension reform. Successive governments have recognized that the pension system was in trouble, and that structural reforms in the pension system were required. In 1999, after several tries, a new pension system was adopted. The reformed system only applies to those born after 1960, however. Under the new system, pensions depend on entirely on contributions. For these contributors, individual contribution account records are being created, and pensions under the new system will depend on the notional contribution balances in workers accounts at retirement. Notional balances are calculated by summing up individual contributions, which are indexed to the growth rate of contribution wages.' At retirement balances are divided by life expectancy in months to get total pension - a notional defined contribution system. Acquired rights under the old system will be transferred to the new notional accounts. This system is similar to ones, which have been adopted in Sweden, Latvia, Poland, and Italy. The Government's intention is to eventually run a surplus, and convert some portion of the notional accounts into funded, privately managed accounts13. The target date for running a surplus is 2005. However, recent projections suggest that this target is unlikely to be reached even by the end of the decade. 18. The main reason that the reform will not reduce project expenditure growth for at least 20 years is that the reform only affects persons born after 1960, so most pensions granted for the next 10 years will be under the current (old) rules. Once the new system takes hold, the growth of expenditures can be " Some of these savings will prove elusive as lower coverage will probably show up in increased claims for social assistance pensions. 12 This means in effect that real value of the sum of contributions at retirement, and therefore the beginning pension, is linked to the growth of wages in the covered sector. 1 See Government of Mongolia (1999). 15 expected to reduce, owing to the benefit reductions in the new formula. Compared with the current rules, the new rules will result in projected benefit cuts for of 40-50% for those who retire under the new system, but according to current norm for retirement age and years of service , and even higher cuts for those with low contribution (service) records (see Annex1.) Those entitled to early retirement (women and those in hazardous occupations - currently about 70% of new pensioners) will have the largest drop in benefits if they retire at the minimum retirement age provided under the law. A minimum pension is provided, but this is also lower than the current pension. Disability and survivors' pensions benefits are roughly the same, as these were not reformed. TABLE 7: COMPARISON OF REPLACEMENT RATES RECEIVED BY PERSONS BORN IN 1959 AND 1960 (life expectancy at retirement age is assumed to be 15 years) Years Of Service 20 30 40 Replacement Rate For Cohort Born In 1959 45% 60% 75% Replacement Rate For Cohort Born In 1960 20% 30% 40% Source: Wiese, 2001 The pension reform was a valiant effort. Between 2020 and 2050, projected system deficits are smaller than they would have been under the old system. But even with the reform, substantial deficits are projected (see table 8). The pension system remains a source of structural fiscal pressure. Note that these numbers do not include social assistance pensions. Expenditures on these will also grow over the period given the declines in coverage observed above.14 TABLE 8: PENSION SYSTEM PROJECTIONS, 2000-2050 Contributions - Expenditures as % GDP (percent) 2000 2010 2020 2030 2040 2050 2060 2070 Old Law DB system -1.6 -1.3 -2.6 -4.2 -5.3 -6.6 -6.7 -7.2 New Law (combined system) -1.6 -1.3 -1.8 -2.2 -2.5 -3.2 -3.4 -3.6 Source: Wiese, 2001 19. Reform issues. The new pension system suffers from at least five serious flaws, which need correction. First, since the reform only begins to take effect in 2010, it will not reduce expenditures substantially until about 2020 (until a substantial number of pensioners born before 1960 begin to die). The pension system will continue to be very expensive for Mongolia over the medium term. The second flaw is the failure to raise the minimum retirement age for those covered under the new system (and to get rid of the special early retirement provisions). This is a problem for two reasons. First, if people continue to retire at the minimum retirement age but get new system benefits, (and, based on the experience of other countries, it is reasonable to expect that most will), their pensions will be too low to prevent poverty". While many pensioners will continue to work to supplement their pension (either formally or informally), they will eventually stop working and then be at great risk of poverty. Putting the most elderly at risk of poverty defeats the purpose of a public pension system. The second reason that the failure to raise retirement ages is a problem is that it contributes to the large gap between the benefits 1 The minimum pension, provided to those with at least 15 years contribution, will take some of the pressure off of the social assistance pension. However, this benefit increases to the social insurance expenditures and therefore the s wstem deficit. It will be even lower than shown above, because their life expectancy will be higher. 16 under the old system and the new system. If pensioners were not able to take a pension before 62, the difference in expected average benefits would fall. 20. The third flaw is related to the first and second flaw and is the lack of a transition mechanism- affecting pensioners retiring after 1999 but born before 1960. As a result, the reform is generationally unfair in a very dramatic way. With the same work history and contributions, someone born on Dec. 31, 1960 will get a pension twice as high as someone born on January 1, 1961 (see box above). No open, political system will be able to withstand such a sudden, dramatic, and arbitrary change. If this problem is not addressed before 2010, the reform will most likely be overturned. Going back to the old system would be very expensive at this time, and a new pension crisis will emerge. In principle, transition measures could be put in place to delay the effects of the reform, thus smoothing out the difference." However, making this adjustment would prolong the current structural deficit, delaying the deficit reducing effects of the reform even further. 21. A fourth flaw is the failure to reform disability and survivor's pensions. As a result, these pensions will be much higher than old-age pensions for those who are born in 1960 and later. Aside from being expensive, this discontinuity will cause tremendous pressure on the disability system, as many will try to be certified as disabled in order to collect a higher pension. Given the high correlation of disability with age, fraud will be difficult to detect and control. 22. A fifth flaw is the overall failure to consider the costs of the system as a whole, as related to benefits, and set up the notional accounts accordingly. Under the current notional account system, 15 percentage points of pension contributions are credited to the contributors account (regardless of whether it is notional or eventually funded). This basically means that three costs have to be covered by the remaining 4 percentage points (or the state budget): (a) administrative costs; (b) the costs of disability and survivor's pensions, and (c) the costs of the minimum guaranteed pension. Since administrative costs are about 2 percentage points of the covered wage bill (and without administrative improvements will go even higher), this leaves only 2 percentage points to cover the rest. Given the failure to reform disability and survivor's pensions, this allocation is woefully inadequate. In other words, despite an intention to make the system self-financing (based on contributions), the system will run a perpetual structural deficit because of design flaws. 23. In sum, the new system is a good start, and an improvement over the Soviet system. However, owing primarily to the failure to address the transition problem and linkages with other systems, it is unfair and too expensive. Without further reform, the system (a) is unlikely to survive once the reforms finally take effect; and (b) will run a perpetual deficit. The target of partial funding will not be reached unless the contribution rate is increased or other funding is found (unlikely also). Increasing the tax rate will not improve the coverage rate, and is probably not justified given Mongolia's high poverty rate." It is better to leave some cash in worker's pockets, and reduce the costs of the pension system. 24. Other social insurance benefits. Short-term insurance benefits have been declining slightly as a share of total expenditure since 1993. This probably reflects the declining coverage. The number of beneficiaries fell by about 20% after 1995, when contribution requirements were introduced. Benefits formulas are fairly standard, with higher income replacement ratios for longer service. The length of the unemployment benefit is 76 days, the sickness benefit 78 days (about 3.5 months) which is fairly short by international standards. After sickness benefits expire, insured can qualify for a disability benefit, however. With over 15 years of service (contribution), sickness and unemployment benefits are 75% of 16 This proposal has been outlined in Wiese (2001). 17 Current Government policy is not to increase the contribution rate. However, this is inconsistent with the policy to (a) maintain current retirement ages and (b) move to partial funding. 17 the wages. Since these benefits are not taxable, incentives for abuse are strong and close monitoring is needed. 25. Social assistance benefits. National law mandates social assistance benefits. They are funded by a special transfer from the central budget to local budgets, based on estimated eligibility levels in each locality." There are about 20 different benefits, of 4 types: (a) long term benefits (pensions); (b) short term benefits (e.g. maternity benefits) (c) lump sum benefits (e.g. twins benefits), and (d) benefits in kind (such as fuel for cooking and heating). All benefits are essentially categorical; there is no guaranteed minimum income or basic social assistance benefit. Many benefits have some sort of income test, but others do not. For example, social assistance maternity benefits (equal to 1 minimum wage), are supposed to be provided to all mothers who have carried a child 196 days or more, who are not eligible for social insurance. Likewise twins benefit is a lump sum provided to all parents of twins within 3 months of birth. However, reimbursement of 100% of the cost of a prosthesis is only provided to "very poor families", while the benefit for taking care of a disabled child is provided to a family member who is unemployed (no income test is specified - so this could be someone working in the informal sector). 26. Average social assistance benefits are smaller than social insurance benefits, and related to the poverty line. It is important to note that not all legally required benefits are paid on time to eligible recipients. This is because localities have some discretion on eligibility. If this discretion is used unwisely, if the MOF-approved amount underestimates the eligibility level, or if central transfers are late, arrears develop. This -happened, for example, in 2000, when the number of beneficiaries fell sharply because of arrears and budget shortfalls. It is not clear what criteria are used in the benefit rationing process. In field visits, a number of different criteria were offered, indicating a rather non-standard approach to implementing the law. 27. The introduction of contribution requirements for social insurance benefits coincided with a substantial increase in the number of social assistance benefits paid, especially social assistance pensions. As a result, social assistance benefits have risen as a share of total expenditure since 1995. In the case of old-age pensions and maternity benefits, the eligibility requirements are basically the same for both systems, except for social assistance pensions the individual should not be eligible for social insurance benefits. For old-age pensions, there is an additional requirement that the recipient be unable to support themselves, although there is no disability or working capacity test for, for example, a 45 year-old woman who has had four children. For disability, the threshold is the loss of 50% or more of work capacity - a. low threshold. It appears that the response to the introduction of contribution requirements for social insurance benefit eligibility was to shift people over to the social assistance rolls without consideration as to whether these categories of people are the neediest in society. Given the growth of the uncovered sector, this trend can be expected to increase; especially as the pension reform takes hold at the end of the decade. If unreformed, the social assistance system could become another source of structural fiscal pressure. 28. Social care. The government for a range of vulnerable individuals provides residential social care. Most care is financed by the central government. There are over 30 public care institutions in Mongolia. The number of persons in care has more than doubled since 1993. This is a trend observed in other countries of the former Soviet Union (Fox and Goetestam, 2001). The main reason for the increase in placements is the increase in poverty and family break-up associated with the transition. Few alternatives to residential care are available in Mongolia, and these alternatives are operated by NGOs, mostly in UB. Local authorities recommend placement in care. Family members may request placement, or placement may be enforced. Standards of care are poor. State institutions tend to be under-funded and 1 Local -governments are free to top central transfers for SA with their own funds , but they are not allowed to spend the earmarked transfer on other expenditures. Both MoF and the Social Assistance Fund do the monitoring of this. 18 the staffed poorly trained for the job. Few care homes have treatment plans or exit strategies. Most adults die in care. 29. A social protection system with excessive use of institutional care is one of the legacies of the Soviet Union. In the Soviet system, families were encouraged to use institutional care for the disabled, the aging, children who had lost a parent or were born to young single mothers, for example. Institutional care was the dominant model, instead of trying to keep the family member in the community and participating in school, work, or leisure, alongside others. It has been well established that the approach of removing a child or an adult from the family and the community is more expensive per client served than more inclusive approaches which are designed to support individuals within their own families and mainstreaming them as much as possible (Tobis, 2000). However, making this approach effective requires institutions and skills that were not developed in the Soviet period, such as social work training, family support programs, and foster care and guardianship development programs. It is not surprising that as poverty and social dislocation increased in Mongolia, the number of people in care increased as well as families in crisis were not offered alternative support. 30. The policy framework reinforces this institutional care approach. The Social Welfare Law defines eligibility for residential care. This law only specifies the right of the public authorities to place an individual in care. It does not specify any obligation to search for alternatives, the goals of the care, or the rights of the individual in care. Among those "entitled" to custody service are: (a) orphans (defined as those with one or fewer natural parents present, including those whose parents have divorced and therefore one parent is not present); (b) triplets (or higher multiple births) up to the age of 4; (c) disabled children. Most of these children could be helped to remain at home if adequate support were provided. Providing such support is cheaper than providing care, as well as better. Particular developmental damage is done to young children placed in care. 31. The increase in social care is particularly noticeable in UB, in response to the emergence of street children. The origins of the street children phenomenon have not been studied definitively (UNICEF, 2000). Small sample studies report material deprivation is the largest factor (not enough food to eat at home, no dwelling), followed by abuse and or family conflict. However, there are now more than 1000 children deprived of parental care that are in some form of residential care, and a significant number remain on the street part or full time. The quality of care in NGO and state programs varies substantially (UNICEF, 2000). A national standard for residential care are not in place, nor is there a capability for monitoring them if they were. 32. Programs to assist the unemployed. To facilitate job searching and labor exchange, Mongolia established the Employment Service, with a central office and branch offices in each district of UB and aimag center. Until the new Employment Promotion Act was passed (April 2001), these offices primarily served the insured unemployed. Under the new act, they serve all unemployed and the "potentially unemployed". They register the unemployed, distribute unemployment benefits to those eligible, collect information about vacancies from employers and offer it to those searching for a job, offer workshops and support services, finance training and retraining for the unemployed and "potentially employed", offer employers wage subsidies, develop micro-finance programs, and arrange public works. These programs are financed by contributions to the Unemployment Benefit Fund (see table above), by budget revenues, and by an earmarked tax on foreigners working in Mongolia. The activities and programs to assist those not eligible for benefits are new ones, just mandated under the new Employment Promotion Law. 33. Enactment of this Law brought a large increase in responsibilities to the Employment Service. It is too early to judge the effectiveness of the service in delivering the programs - for most program, implementing regulations are still under preparation. However, the policy framework is rather generous in its promises of subsidies and benefits. It therefore leaves open broad scope for wasting money. The 19 broad scope also suggests that some rationing among potential clients will occur. Criteria are necessary to ensure transparency. It is critical that evaluation criteria and a monitoring system be established as soon as possible to ensure the programs are actually adding value, rather than replacing private funds. Roles and responsibilities 34. The Ministry of Social Welfare and Labor is responsible for setting policy and supervising implementation in the areas of employment, labor protection, social insurance, social assistance, and has overarching responsibility for poverty reduction. The Ministry has number of subordinate agencies and centers. The most important are: * The State Social Insurance General Office (SSIGO), responsible for collecting the earmarked payroll tax for social insurance, disbursing benefits and managing 5 social insurance funds (a) pensions (old age and disability), (b)sickness and other short-term benefits, (c) work injury, (d) unemployment, and (e) health. SSIGO is governed by the National Social Insurance Council. * The Labor Inspection Agency, responsible for insuring the implementation of the Labor Code, including occupational safety and health. * The Central Employment Office (CERO), responsible for implementing employment promotion measures, including registering the unemployed and offering employment services. * Social Assistance Office (SAO), responsible for monitoring the disbursement of state social assistance benefits, operation of the social assistance "fund" (see below), training local social welfare staff, and the operation of 13 long term care homes * The National Committee for Children, secretariat to the National Counsel for Children and responsible for several children's centers and ensuring child protection * National Center for Rehabilitation and Training, which provides prosthetic services and vocational training. The Minister appoints the heads of all these centers, in consultation with the PM. 35. The central level (the Ministry and subordinate agencies) is directly responsible for the provision of: * all social insurance benefits * residential social care * employment services 36. The subnational level is responsible for the provision of: * client needs assessment and social assistance benefits * day programs and some social care facilities (primarily UB cities) 37. Budgeting and expenditure management. Social protection expenditures are administered various ways. The Ministry administers the budget for their staff and Ministry administrative expenditures. The Ministry also exercises limited oversight functions over the budget of the subordinate agencies, although the MoF provides most of the oversight. MoF determines the subnational budgets, and the respective subnational budget processes determine the expenditures at that level. During the process of determining subnational budgets, MoLSW issues spending guidelines (norms), which can help to shape the negotiation over the total subnational budget. However, these norms are not binding. Aid projects are mostly outside the budget process. 20 38. At the national level, there are three quasi-independent funds: (a) the Social Insurance Fund, a full-fledged extra budgetary fund, with earmarked taxes and the right to carry-over surpluses, governed by legislation, the Ministry, and the independent Social Insurance Council, a tri-partite body; (b) the Employment Fund, also a full-fledged extra-budgetary fund, governed by the Ministry and the Employment Council, another tri-partite body, and (c) the Social Assistance Fund, a fund in the accounting sense only, as cash expenditures in social assistance come from local budgets, and expenditures on national social care institutions are provided directly to these units on an input basis. This "Fund" is mostly an accounting agency, pulling together the information on local decisions. 39. While (i) legislative entitlements and (ii) Ministry decisions, endorsed by the Cabinet, play the largest role in shaping the budgets of (a) and (b) above, the independent Councils also have a policy and oversight role. For example, the Social Insurance Council may direct the SSIGO to prepare regulations on key implementation issues in the pension reform, or request draft legislation. The Council is also supposed to monitor the budget implementation. In practice, this has resulted in a somewhat defuse and ineffective accountability for social insurance policy and its fund oversight." It has also sometimes created policy deadlock, if the Council does not take a decision on a key issue. 40. Organizational structure. The MoSWL was formed in 2000, when it split off from the Ministry of Health. The object of the MoSWL is "providing appropriate decisive professional support and advice for the Ministry and the action program of the Government" towards the goal of "creating sufficient living condition for citizens where they can work effectively and strengthening social welfare service for them" (MoSWL internal document). This includes monitoring developments in the social protection and labor sectors, policy formulation, review, monitoring, and evaluation, and public administration leadership and capacity support in the labor and social protection areas. The Ministry is organized into four departments: * public administration and management * policy coordination and implementation * information and monitoring * strategic planning department * external relations (including foreign aid projects) 41. In principle, this broad functional structure should support an integrated social welfare policy. In practice, it does not seem to be yielding these results, although it should be noted that it is a newly formed Ministry, still finding its way. Policy formulation and evaluation is organized primarily according to expenditure type (SI, SA, employment programs, etc.), with inadequate attention to the effects of the whole system. Expenditure decisions are input driven, mostly according to entitlements. Changes in entitlements tend to be rather ad hoc, leading to politicization (for example, pension indexation, which is not governed by rules or an MTEF). There is no evidence that, for example, that the implications for social assistance policy of the decision to move to a contribution based system was ever considered. As the Ministry was not involved in the 1999 reform (this was formulated and supported primarily through SSIGO), capacity in this area in the Ministry was not developed or supported. 42. As with other Ministries, there is only the weakest of links between policy, performance, and resources. In some cases this is impossible. For example, in social assistance, the Ministry has no control or whether local government provides the legally required benefits from their budgets. Nor do they have 1 For example, the 1999 pension reform was mostly formulated in SSIGO, not in the Ministry. Likewise, at some point the SINC decided to permit SSIGO to accept contributions in kind, without consultation with MoF or the tax authority, even though the Government eventually bears the costs of these decisions through budget transfers to assure that the pensions are paid. 21 tools to monitor whether the social assistance system is helping the poorest (household surveys are infrequent). The new Employment Fund was formulated based on only the barest of information on the operation of the labor market, the behavior of firms, or the value of the interventions. 43. The Ministry is aware of these organizational weaknesses. A high priority has been placed on addressing them, and support has been received from the ADB in the form of a large policy and institutional development loan. One of the first activities will be a preparation of a medium term business plan, which will address some of these key institutional and policy issues. 44. The implementation of the PSMFL will be a challenge for the Ministry. One of the first questions will be - what does this Ministry purchase? Which social welfare services are purchased by the subnational governments, and under which rules? This will require establishing clear accountabilities with respect to the subordinate agencies and the local governments. Most likely, the role of the Ministry in monitoring, audit, and evaluation will need to be strengthened. 45. Of the subordinate agencies, the most important is SSIGO. Implementation of the pension system has been a substantial challenge for SSIGO. As the new system bases benefits on lifetime contributions, SSIGO had to build and operate whole new automated record management system, to replace the manual record system of workbooks, which in the Soviet times was actually maintained by employers. With donor support, this has been substantially completed, although there remain problems in some soum offices, where electricity is unreliable or unavailable and therefore the offices are not computerized. However, as the new pension system is not yet in force, the new information system has not yet been really tested. 46. In addition to paying benefits, SSIGO has the responsibility for collecting contributions. With the exception of some problems related to the policy of collecting contributions in kind (never a good idea anyway), the contribution collection function seems to work. The number of contributors has been dropping, but there is little information on the reasons for this. However, having two agencies collecting taxes on the same base necessarily involves duplication of functions and inefficiencies. These issues should be addressed in the medium term. Government strategy 47. The Government strategy, as expressed in the 1999 Program of Action (POA), and in the subsequent "Interim Poverty Reduction Strategy Paper and the Social Security Development Program agreed with the ADB, is threefold: (a) to gradually increase pensions (the POA proposes a doubling on nominal terms over a four year period), to be financed through increased coverage, improved collections, and real revenue gains through wage increases in the covered sector; (b) to improve the quality of and coverage of social welfare services and benefits through revising the benefit structure, training social workers to better support families, and overhauling the system of social care; and (c) developing and implementing pro-poor active labor market policies such as training programsi micro-finance schemes and public works. 2 Public Sector Management and Finance Law, a bill in Parliament which would fundamentally restructure the budgeting and expenditure management functions. Based on the ideas behind public sector reforms in New Zealand, the law envisages that each minister would use the budget to "purchase" services according to a service agreement. 22 ISSUES AND RECOMMENDATIONS 48. Overall spending. Mongolia is spending over 7% of GDP on public social welfare programs. At a macro level, this is an adequate amount of public funds, given Mongolia's income level, demographics, and competing needs. Most low-income countries with Mongolia's demographics spend a much lower amount. On pensions alone, Mongolia spent 5.6% of GDP in 2000. To give two comparisons of low- income transition economies with comparable demographics, Armenia spends 3.6% of GDP, China 2.5%. The issue in this sector is not funding adequacy, it is how the funding is allocated. 49. The social insurance system absorbs the bulk of the spending, and most is on pensions. Spending on short-term benefits is about 4% of total spending. Unemployment benefits are well designed, with good incentives for return to work. Sick leave benefits could be subject to abuse, and should be monitored. Reform in this sector should focus on pensions. 50. What are the prospects for the promised pension increases? The Government's pre-election promise of significantly higher pension was a worthy ideal. Minimum pensions are low. However, the high number of pensioners makes this ideal costly. Depending on the rate of inflation and the time frame, this policy objective could cost 3-4% of GDP. 51. The plan for financing this increase also looks weak. The Government is correct to worry about the declining revenue in SSIGO, caused by (a) declining coverage, and (b) low collections on existing coverage. However, coverage declines are a feature of all transition economies, and are not easily reversed. Farmers and herders have avoided payroll and income taxes for centuries, in all market economies. There is probably some scope for increasing collections with better tax administration. Consideration should be given to shifting the collection and administrative function over to the Revenue service to achieve this goal. However, it is unlikely that such a move would yield the amount of revenues needed. 52. If wages increase in the private sector (through, for example, improved productivity and growth), the Government should be able to take in more revenue, but in this case, pension costs would go up as well, as new pensions are linked to previous years' wages. Increasing wages in the public sector is not a viable option to pay for increased pensions, as the funds to pay for such an increase are also not available. 53. Finally, increasing all pensions without other reforms to cut future expenditures raises equity issues. Pension increases have to financed out of public funds, either with increased taxes or reduced expenditures elsewhere. Working families with children in urban areas (the main tax payers, especially of payroll taxes) are also poor. Raising taxes on this group (or lowering the level of public services delivered to them through expenditure cuts) could worsen their poverty. 54. If reforms were undertaken today to reduce the growth of expenditures, pension increases would become possible in the future. Three basic reforms are recommended to reduce the cost of the system and improve equity between the old and new systems: * begin raising retirement age now for all in the labor force; * develop a transition policy to put all those currently working and contributing into the new system; and * reform disability and survivor's pensions. 55. Note that under the ADB financed TA, an analysis of the fiscal effects over 50 years of these types of reform has been prepared, although not all the reforms suggested here were analyzed. This type of analysis is critical for pension system reform. In developing a reform package, the Government should use such a model to review all options. 23 56. Retirement age. The main reason Mongolia's pension costs are so high relative to demographics is that Mongolia has given out too many pensions too early. As a result, there are too many pensioners relative to the number of people aged 60+. As noted above, Mongolia's retirement age is very low, relative to life expectancy at retirement. Mongolia has basically kept the retirement ages from the previous system, including the provisions for early retirement (mothers with many children, certain occupations, blind, etc.). Most other transition economies Europe and Central Asia have already started rising retirement ages by 4-6 months per year (see table 9 below). Many have also abolished the early retirement groups. In some cases, these groups have to retire at the normal age, but at the time of retirement they receive higher pensions to compensate for acquired rights (Czech Republic, Lithuania). In other cases the retirement age is gradually raised (Poland, Moldova) and no compensation is paid at the time of pensioning. One option for Mongolia, which combines this approach, would be to: 57. Immediately make 55 the minimum retirement age for everyone, including all special groups, and begin raising the retirement age by 6 months per year up to 60 and 3 months per year from 60 to 62. 58. No estimates are available of the fiscal savings from such a reform, since this option has not been evaluated in the previous simulations.21 Based on the experience of other transition economies, reducing the inflow of new pensioners could save about 15-20% of total pension costs after 4-5 years of reform. 59. In the long term, under the new (NDC) system, savings from longer working years would be balanced by higher pensions as long as everyone contributed.2 But even in this case, there would be fiscal savings from reduced expenditures on minimum pensions (since a higher retirement age implies more years of service). Higher pensions would also reduce social assistance costs and ultimately poverty. This reform would also ease the transition problem, since the difference between pensions under the old and new systems would start to converge. TABLE 9: RETIREMENT AGE REFORMS IN THE FSU Country Polcy Action Kyrgyz Republic Raised retirement age to 58 for women, 63 for men Kazakhstan Raised retirement age to 58 for women, 63 for men Lithuania Raised retirement age to 60 for women, 62.5 for men Latvia Raised retirement age to 62 for both Moldova Raised retirement age to 60 for women, 65 for men Source: mission data 21 Analyzing the savings requires estimating the number of pensions deferred for each age cohort, and also the average size of the new pensions granted (as they will be higher for those who contribute longer). This requires a simulation model. 22Under an NDC system the delay in retirement results in actuarially fair pension, so in the medium term, in a pure NDC system, from a purely financial point of view, the Government is indifferent between retirement ages. However, from other points of view, the Government is not indifferent as increased age at retirement results in higher pensions, reducing poverty among the aging. It also saves the Government cash in the medium term. 24 TABLE 10: SUMMARY OF ISSUES AND RECOMMENDED PENSION SYSTEM REFORM ACTIONS Issue Recommended Reforms (1) Low retirement age for current e Raise retirement age by 6 months per year for women and men up to and future retirees creates large 62. fiscal burden, low pensions 0 Remove all entitlements to early retirement (2) Abrupt and unfair transition to 0 Raise retirement age by 6 months per year for everyone new system Put entire working population in the new system within one year, with transition provisions to account for acquired rights (3) Failure to make disability, e Revise formula in new system to be a notional accounts one survivors consistent with old age pensions (4) Allocation from payroll tax too 0 Reform disability, survivors low for disability, survivors, 9 Raise retirement age (to reduce number of expected minimum minimum pension, and pensions) aduimstration Allocate less to notional accounts (based on projections) (5) Cost too high, collections too 0 Consolidate tax administration authorty in one agency to improve low, no funding possible collections; keep intergovernmental arrears in check * Raise effective retirement age, reform disability and survivors to reduce future costs. * Consider funding inimum pension from budget revenues. * If funding is still not possible, revisit contribution rate. 60. Develop a transition policy. Most other countries introducing notional accounts have made a gradual transition to the new system, beginning the year of the pension reform. Mongolia should follow this lead. Two basic approaches to the transition have been followed in other notional account system reforms. The first is to give everyone a notional account, with the opening balance based on acquired rights under the old system. This in effects means that the older workers have a pension mostly based on past service, while the younger ones have one mostly based on contributions. the second is to have (say) ten-year transition period. The pension in the first year would be 10% from the new system, 90% from the old system, the second year 20% from the new system, 80% old system, etc. Either approach would work for Mongolia. However, the second approach may be administratively more complex. 61. Reform disability and survivors' pensions. Mongolia needs to complete the switch from the old system to the new system by reforming the disability and survivors' pension to make them consistent with an NDC approach. Technically, this requires three changes: a) revise the formula to be based on the amount in the notional (or real) account (e.g. be based on contributions); b) make provisions to convert the disability pension into an old age pension at minimum old-age retirement age; and c) add in a safety net (minimum) so that those who become disabled at a young age receive an adequate pension. 62. The fiscal effect of such a change in Mongolia could be large even though they would only be realized after 30 years. Weise (2001) estimates an initial effect of about 0.6% of GDP per year in savings rising to about 1% of GDP per year once the NDC system is fully in place. He did not simulate a change in survivors' pensions. 63. Reform of social assistance. The second major area for reform is the social assistance system. Here, the nsystem needs a fundamental revision. Not only do the cash benefit and in-kind programs need to be reformed, but the overall goals of the program need to be considered. Three issues need to be tackled: 25 a) The goal of both the cash benefits and the programs should be to support individuals' inclusion in families and society. The system should avoid labeling people, and the need for assistance should not be pre-determined based on, for example, a medical event (e.g. multiple births). The legal framework should be revised to incorporate this goal. b) The system should be simplified. Consistent with above, the policy should encourage case workers not look for a category to put a vulnerable individual in, but rather look for needs and solution ("one right of assistance"). Benefit entitlements should not be created which mirror the social insurance scheme, since these reduce incentives to contribute and may not reflect true vulnerabilities. c) Expected outcomes should be specified and monitored at all levels. This is especially important for programs, but is also relevant for cash benefits. 64. In addition, the financing of social assistance system should be reviewed. It is in the national interest to have an effective national social safety net, which is applied in a standard manner throughout the country. This is especially important during a reform period, to insure that reforms of national importance (such as energy tariff reforms) can be implemented without harming the poorest. This implies a national standard, and a means to insure fair and transparent implementation nationwide. 65. . Local implementation of social assistance is also beneficial as it permits a community-based approach to helping the vulnerable and destitute. The question is which roles should be taken by which level of government to insure an affordable, effective safety net? Several countries have successfully combined a national standard with local financing and implementation, ensuring equal treatment and effective poverty alleviation. These are primarily the Nordic countries. In these countries, the national government play a strong role in monitoring and adjudication to ensure that the minimum national standard is followed. However, the experience of transition economies with decentralized financing is not so good. Milanovic, et al (1998) found that the higher the level of decentralization in financing, the more likely were the social assistance benefits to be (a) poorly targeted and (b) under-funded. In part, this is because the national government does not have the capacity to perform the functions that it does in the richer, more developed and homogeneous Nordic countries. 66. National financing with local implementation also has dangers, as Mongolia has seen. The main risk is that the certification process will be too lax, since the local government does not feel directly the opportunity cost of the money (this is known as the principle that "he who decides, pays"). One approach to solve this is a matching funds approach, where each Tg. of local spending on social assistance is matched by a Tg. of state budget funds (according to specified rules). This approach tends to reduce spending by poor areas, however, as they do not have the matching funds. In any case, clear national standards have to be set (in accordance with point (b) above) and monitored. Monitoring combined with effective and efficient adjudication procedures should be a function of the MoSWL (or a state agency delegated the task). 67. In keeping with the directions of the proposed PSMFL, financing for social assistance (programs, institutions, and benefits) should move away from an input basis. Mongolia needs to develop a program to reform social care financing so that the state is the purchaser, but not necessarily the provider. The provider could be NGOs, communities, or families. In general, day programs and inclusive care in eth family is much more affordable than residential care, and Mongolia needs to develop a reform program to move in this direction. This also implies improvements in the gate keeping function (the process by which people are assigned to care and removed from care). 68. Finally, the social assistance and social care system should be the system of last resort in the case of poverty. Prevention of poverty, especially among the vulnerable is much cheaper than paying benefits. One of the side effects of Mongolia's fiscal crisis and response appears to be a growing exclusion of 26 portions of the population form other social services because the do not have a formal sector job or residency papers, or they have certain characteristics (age, ethnicity, mild disability, etc.). This is a common problem throughout the FSU, as countries struggle to find new ways to ration scarce resources and provide incentives for efficient allocation. It is an issue, which requires Government-wide attention, however. The MoSWL could take the lead in helping work through these issues. 69. Employment promotion programs. The new Employment promotion Law sets an ambitious agenda for CERO. The first task is to collect information on who are the unemployed and how they related to labor demand, for the purpose of setting priorities in program implementation. A full implementation needs to be developed, with monitorable indicators. The agency also needs to: * strictly define what training is; * use demand driven mechanisms to chose training providers; * implement cost sharing mechanisms as much as possible, especially for the employed and self- employed; * avoid wage subsidies or strictly limit their use since they tend to encourage substitution of one worker for another rather than net job creation; * avoid government provided and administered micro-finance as experience under the NPAP was problematic; * set strict rules for public works, with matching funds required. 27 References Anderson, James H., The Size, Origins, and Character of Mongolia's Informal Sector During the Transition. World Bank Policy Research Working Paper #1916, World Bank: Washington D.C., May 1998. Bikales, Bill, Chimed Khurelbaatar, and Karin Schelzig, "The Mongolian Informal Sector", prepared by Development Alternatives Inc., April 2000, (processed). Fox, Louise and Ragnar Goestam, "Redirecting Resources to Community Based Services: A Concept Paper" August, 2001, (processed). Government of Mongolia. "Basic Guideline for Implementing Pension Reform up to 2021". Informal translation, 1999. Government of Mongolia, "Initial Poverty Reduction Strategy Paper", September, 2001, processed. Available on: http://poverty.worldbank.orglprsp/index.php?view=sub&id=3483 International Monetary Fund, (Fiscal Affairs Department) "Mongolia: Policy Options for Strengthening Social Protection", December, 1997. "Review of the 1998 Mongolian Livings Standards Measurement Survey", Report prepared by FIDE International for the National Statistical Office and the United Nations Development Program, May 1999 (processed). Tobis, David, Moving From Residential Institutions to Community-Based Services in Central and Eastern Europe and the Former Soviet Union. Washington D.C.: World Bank, 2000. UNICEF, Mongolian Adolescents Needs Assessment Survey Report, Ulaanbaator, UNICEF, 2000. Weise, Patrick. "The Next Step in Mongolia's Pension Reform: A Plan for Transitioning From Pay As You Go to Partial Funding" July, 2001, (processed). World Bank, Averting the Old Age Crisis, Oxford: Oxford University Press, 1994. 28 FIGURE 1: SUMMARY OF SOCIAL INSURANCE BENEFITS CURRENT & NEW PENSION LEGISLATIVE PROVISIONS CATEGORY CURRENT LAW NEW LAW Retirement pension Law On Pensions & Benefits Provided By The Social Insurance Fund Law On Individual Pension Insurance 1/1/1995, Amended in 1995, 1996,1997 and 1999 Contribution Account Effective 7/11/1999 Formula RP = W*45% TAB P = ----------- RP = Retirement pension F W = Monthly average wage for any five consecutive years out of 20 P = (Annual)pension TAB = Total account balance F = Average life-expectancy factor at retirement TAB = pension capital of the insured, composed of imputed and registered social insurance contributions plus accumulated notional interest. The interest is calculated annually based on the average growth in the last three year average wages Length of service 20 for men and women 15 years minimum Credit for additional Pension increased by 1 5% of the monthly average wage for each year No credit years worked over 20, but by 1% of the pension amount per each year of total invalidity Incomplete service Pension is reduced proportionally to the total insurance period Not allowed (10 yrs minimum) Minimum pension Not less than 75% nmnimum monthly wage, and 50% same wage in Not less than 20% national average wage. case of reduced pension. Minimum pension increased by 0.5% of this each year of contribution over 15 Qualifying service Employment, educational period of contracted employees and civil Same as under the Current Law servants, conscnpted army service, maternity leave, paid child care period, temporary inability to work, period of prosecution and impnsonment at no fault, period of staying out of work due to an illegal dismissal, civil defense and other mobilization Normal pension age 60 for both men and women Women are allowed to retire at 55, if Same as under the Current Law. they wish so, provided that length of service is fully met. Indexation of In relation to changes in the living costs In direct relation to the inflation rate The pension benefits adjustment index and the procedure for its application shall be determined by the National Statistical Office based on the suggestion of the Social Insurance National Council. Taxation of benefits Tax exempt Tax exempt Early retirement Granted to women of 50 accumulated not less than 20 years service Not allowed and raised 4 and more children to age 6, if they wish so, underground workers, those in defined hazardous, hot and arduous conditions Invalidity Pension Law On Pensions & Benefits Provided By The Social Insurance Fund Law On Individual Pension Insurance Enforced on 1/1/1995, Contribution Account. Effective 7/1/1999 Amended in 1995, 1996,1997 and 1999 Formula Total invalidity = same formula as retirement Total invalidity = W*60% Partial Invalidity = (retirement formula)* percentage of loss of Partial invalidity = (W*.6) * (percentage of loss of capacity for work capacity for work) W = monthly average wage in the last three years, Length of pension From the date of commencing invalidity and ending with the day of Same as under the Current Law rehabilitation, or with the following month in which the beneficiary dies Minimum pension Same as the minimum retirement pension Same as the minimum retirement pension provided from Individual Account Quahfication Not less than 50 per cent loss of capacity for work permanently, or for Same as under the Current Law a long duration due non-occupational disease or injury 29 CATEGORY CURRENT LAW NEW LAW Classification of Total invalidity (70% and more loss of working capacity) and partial Same as under the Current Law invahdity invalidity (from 50 to70%) Certification Medical Labor Accredited Commission Same as under the Current Law Length of service Not less than 20 years insured service, or three years out of five, Same as under the Current law immediately preceding the date of commencement of invalidity Incomplete service Pension is reduced proportionally to the period of pension insurance, Not allowed (minimum 3 years) but its nummum is equal to minimum reduced retirement pension Taxation Tax exempt Tax exempt Survivor's Pension Law On Pensions & Benefits Provided By The Social Insurance Fund. Law On Individual Pension Insurance Enforced on 1/1/1995, Contribution Account Amended in 1995, 1996,1997 and 1999 Effective 7/1/1999 Formula . Same formula as retirement 40% monthly average wage in the last three years Survivors receive pension in proportion to certain percentages, for for one dependent, increased by 10% per each three or more -100%, two - 75% and one - 50%. member over two and more. But pension should not exceed 60% monthly average wage Credit for period of Pension increased by I per cent pension amount for each year of total No credit invalidity pension invalidity, if the deceased was on the receipt of invalidity pension Minimum pension 50% minimum wage for one dependent, 75% for two dependents and Same as the minimum retirement pension 100% for three dependents, Reduced pension should not be less than provided from Individual Account 50% same wage. Length of service Not less than 20 years insured service, or three years out of five, Same as under the Current Law immediately preceding the date of breadwinner's death due to non- occupational disease or accident. Incomplete service Pension is reduced proportionally to the period of pension insurance Not allowed (immum 5 years, continuous in the last year) Eligible survivors Born or adopted children under 16, if student children to age 19, Same as under the Current Law grandchildren, brothers and sisters under 16 without caregivers; grandchildren, brothers and sisters disabled or got disabled prior attaining age 16; parents over retirement age or disabled parents, spouse or grandparents, brothers and sisters without caregivers; any of parents or spouse not working and caring for children under 8, or grandchildren and younger brothers and sisters, also family dependents of the deceased who was on receipt of retirement or invalidity pension and who totally lost capacity for work in months preceding the death; step-parents, step-children not receiving alimony from their own parents, famly dependents of the disappeared; Length of pension From the date of death up to the day on which surviving children attain Same as under the Current Law age 16, if student children to age 19, incapacitated persons for period of loss of working capacities, survivors who have attained the pension age up to the end of the following month in which the insured breadwinner's dies Law on Pensions & Benefits Provided By The Fund Of Social Law On Individual Pension Insurance Disability Pension Insurance Against Employment Injury & Occupational Diseases. Contribution Account Enforced on 1/1/1995, Amended in January 2000 Effective 7/1/1999 Formula DP = W* Percentage of capacity loss(I) Note. This law is not applicable to injury related benefits. DP = W* Percentage of capacity loss(ll) DP = Disability pension W = Insured person's monthly average wage PCL(I) = 10% ( up to 10 per cent) PCL(l) = Percentages over 10 per cent Minimum pension Not less than 75 per cent minimum wage in case of working capacity 30 CATEGORY CURRENT LAW NEW LAW loss over 30% and more Length of pension From the date of commencing disability and ending with the day of rehabilitation, or with the following month in which the beneficiary dies Qualifying service While discharging employment duties at work place or other places; before the commencement of the general working hours or after the general finishing time in the course of arranging the work place & equipment; travel to or from an insured person's place of work Qualification Regardless of length of insured service Certification Loss of capacities detenined by Medical Labor Accredited Commission, whether cases are relevant to employment injury determined by Employer's Investigating & Registering Commissions Dependent's Pension Law on Benefits Provided By The Fund Of Social Insurance Against Law On Individual Pension Insurance Employment Injury & Occupational Diseases. Enforced on 1/1/1995, Contribution Account Amended in January 2000 Effective 7/1/1999 Formula DP = (W* 100%) for three dependents N/A DP = (W*75%) for two dependents DP = (W*50%) for one dependent DP = Dependent's pension W = Monthly average wage of the deceased Eligibility Famly dependents determined as survivors under the Law On Pensions & Benefits Provided By The Social Insurance Fund Length of pension Same as survivor's pension under the Law On Pensions & Benefits Provided By The Social Insurance Fund Social Assistance Law On Pensions & Benefits Provided By The Social Assistance Social Welfare Law, enforced on 1/1/1 999 Pension Fund, 7/1/1996 Pension Rate Minimum living level for a given period Same as under the Old Law Eligibility Men of 60 and more, women of 55 whose income is less than the Persons ineligible for social insurance pensions nmtumum subsistence level, persons suffering from total invalidity, such as. men of 60, women of 55 unable to make born mcapacitated or become incapacitated prior to age to 16, family living on his/or her own, without children and dependents on death of the breadwinner, blind, dumb, deaf persons relatives to maintain them, or if legal caregivers and dwarves; totally disabled in prosecution of civilian duties such as are elderly or disabled who are unable to support saving human lives, or in the fight with natural calamities, family them, and the ultra poor; disabled persons who dependents of those who died in prosecution of duties described have lost 70 per cent working capacity; born above, single mothers of 45 (father of 50) whose income is less than incapacitated over age 16, or become the minimum subsistence level and who have 4 and more children incapacitated pfor to age to 16 due to loss of more than 50 per cent working capacity, family dependents on death of the breadwinner; blind, dumb, deaf persons and dwarves; disabled in prosecution of civilian duties such as saving human lives, working in extremely dangerous infectious foci or in the fight with natural calamities; and their famly dependents in case of death; extremely poor single mothers of age 45 and more, fathers of age 50 Indexation of In relation to the movements in the minimum hpving level Same as under the Old Law benefits Iand more,Ifatherso 31 FIGURE 2: OTHER SOCIAL INSURANCE AND WELFARE BENEFITS LEGISLATIVE PROVISIONS Sickness Benefit Law On Pensions & Benefits Provided By The Social Insurance Fund. 1/1/1995, Amended in 1995, 1996,1997 and 1999 Rate of benefit 45 % - up to 5 years of insured service 55% - from 5 up to and including 14 years 75% - 15 years and more Be calculated from wages of the last three month full-time employment Eligibility Persons, who have paid contributions to the Benefit Fund for three months or more prior to the loss of capacity for work caused by a non-occupational disease or accident are eligible for sickness benefit Length of benefit 78 days, not exceeding 156 days in a calendar year, but in case malignant tumor etc 156 days on one occasion Payment of benefit Benefit for the first 6 days is paid by employer, starting from the 7 day it is paid from the Benefit Fund Time Limit for claims 4 weeks from the day on which the insured person became sick Law On Pensions & Benefits Provided By The Social Insurance Fund. 1/1/1995, Amended in Maternity Benefit 1995, 1996,1997 and 1999 Rate of benefit 70% of the last 12 month average insurable wages - Eligibility Mothers, including those who have given premature birth to a baby before the expiry of 196 days of pregnancy or had an abortion on satisfaction of not less than 12 month contributions of which 6 continuously prior to the maternity leave are eligible for matermity benefit Length of benefit 4 months for mothers insured on a compulsory basis, 3 months for mothers insured on a voluntary basis Time limits for claims 4 weeks from the date on which the mother received the delivery certificate Law On Pensions & Benefits Provided By The Social Insurance Fund. 1/1/1995, Amended in Funeral Grant 1995, 1996,1997 and 1999 Rate of grant 6 times minimum wage of a given period E rgibility Provided that the insured person has paid contFbutions for not less an 36 months in case of non- occupational disease or accident; If it was a death related to employment injury and occupational diseases, there is no contribution requirement If a beneficiary of social mnsurance dies, heJshe is eligible for a funeral grant. Time limit for claims 4 weeks from the day of a death occurrence Temporary Disability Benefit Law on Pensions & Benefits Provided By The Fund Of Social Insurance Against Employment Injury & Occupational Diseases. Enforced on 1/1/1995, Amended mn January 2000 Rate of benefit 100 per cent of the average wage received by the insured person prior to the disability occurrence Eligibility Insured persons, who have lost capacity for work due to temporary disablement caused by an employment injury or occupational disease are eligible for benefit, regardless of the length of insured service Certification Commission appointed by the relevant organization From the day of loss of capacity until the incapacity is rehabilitated or eligibility for a disability Payment of benefit pension is obtained Iength of benefit Should not exceed 180 days in the consecutive 12 months Law on Pensions & Benefits Provided By The Fund Of Social Insurance Against Employment Injury Funeral Grant & Occupational Diseases. Enforced on 1/1/1995, Amended in January 2000 Rate of grnt 6 times minimum wage of a given period Eligibility Regardless of the length of insured service Rehabilitation Payments The Same Law Eligibility Rehabilitation payment is provided to the insured person who has lost 30 per cent and more of capacity for work due to an employment injury or occupational disease for purposes of prosthetic appliances & orthopedics, and therapy in sanatonums Payment Expenses involved in home made appliances be reimbursed in 100 per cent for the first time, and 50 per cent for the second time, the cost of a foreign made appliance needed for the person be provided, in case of existence of an inevitable requirement for it. If an insured person needs treatment in sanatorium, the expenses can also be met by the Fund once a year. Unemployment Benefit Law On Unemployment Benefit Provided By The Fund Of Social Insurance, 9 June 1994 32 Rate of benefit 45% - up to 5 years of insured service 50% - from 5 to 10 60% -from 10 to 15 70% - 15 and more Be calculated from the last three month wages, taken account of the length of insured service Eligibility Not less than 24 months of which 9 continuously before becoming unemployed Minimum benefit Not less than 75 per cent minimum wage Length of benefit 76 days commencing on the day following the expiry of payments provided by reason of disnussal, in case of absence of such a payment from the following day of registration Maternity Benefit Social Welfare Law, Enforced On 1/1/1999 Rate of benefit Monthly benefit is equal to the rate of minimum wage Eligibility Mothers, including those giving premature birth before the expiry of 196 days of pregnancy or having an abortion, who are not eligible for the social insurance matermity benefit Length of benefit 4 months in total, including prior to and post delivery Child Care Benefit The Same Law Rate of benefit Be determined by the Government, taken account of the minimum wage, Additional benefit can be provided by enterprises concerned Eligibility Mothers with low income or non-working mothers caring their child up to 2 years (in case of twins 3 years), also students of colleges and universities If a mother, while receiving a childcare benefit, gives a second birth, her previous benefit be stopped and new benefit be provided. Multi-child benei The same Law Types of benefit 1) Lump sum benefit for brnging up 4 children 2) Annual benefit for bringing up 5 and more Rate of benefit 1) Lump sum benefit is equal to 5 times minimum wage 2) Each child receives 2000 tg up to 16 years old every year Infant benefit The same Law Eligibility Very poor families are eligible for Rate of benefit Be equal to the mnimum wage Time liut for claims Within three months following dehvery 1 TWINS BENEFIT The same Law Eligibility Mothers bnnging up twins 2) Mothers bringing up triplet or more Rate of benefit 20000 tg as lump sum 20000 per each child - Time limit for claims After three months following delivery Benefit for orphans The same Law Eligibility Parents adopted full orphans are eligible for benefit up to age 16 Rate of benefit 2 times minimum wage quarterly Remark All social insurance and welfare benefits are tax-free. 33 ANNEX 3: EDUCATION FUNDING MECHANISM OF FOR EDUCATION SECTOR OF MONGOLIA Local budget. 1. Starting point of education budgeting process is local education entities: primary and secondary schools (except special schools for handicapped), kindergarten and vocational & technical schools. They prepare their budget proposal for next financial year (which in Mongolia starts January I" and ends December 31) based on previous year level, changes in staff, changes in prices and tariffs for heating, electricity, fuel and other base prices, enrolment estimations and current year budget surplus or deficit. The budget proposal also includes request for capital expenditure (new equipment, major repair and maintenance with preliminary cost estimates). District and Soum budget officers review budget proposal and make changes (usually decrease, rather than increase) as they find necessary. At this stage a strong justification is important to save at least some budget items. 2. This process takes place some time in May- June of year preceding the financial year, for which budget is prepared. One important issue here is estimation of enrolment. Since academic year and financial year are not same, many schools use average number of students for a financial year (Total enrolment as of September 1" multified by 12, minus number of coming graduates multified by 6, plus number of coming new enrolment multified by 6 and result divided by 12). 3. Per student standard variable cost norms, which are different for different aimags and soums and type of schools used to determine total amount of fund for teachers salary, social insurance fees, instructional materials and other cost items, dependent on number of students. Since district and soum budget officers work closely with public schools and kindergarten, and have better knowledge of situation, even salary fund and related social insurance fees, calculated based on norms can be adjusted to reflect actual or more realistic number of full-time and par-time teachers. 4. District and Soum budget proposals are submitted to Ulaanbaatar city and Aimag Budget Departments, where further revisions, due to limit of expected subsidies and revenue and local priorities, take place. All local budget proposals with sectoral and cost item specifications come to Ministry of Finance and Economics (MoFE) some time in July-August of each year. MFE education budget officer works together with officers from MOSTEC to estimate total education budget (local and central). 5. Funding for Vocational and Technical Education Schools is a little bit complicated. Over the past 5 years often changes were occurring in funding procedures. Before transition, all the Technical education and vocational training (TEVT) institutions were funded from central budget. Recently, some of TEVT institutions have been upgraded to college level to offer degree courses. Some still remain in diploma and certificate granting training. There is also a group of TEVT schools, that enroll students after 8th grade of general education schools, therefore, have general education component in their curriculum. 6. Like public universities and colleges, former TEVT colleges are charging tuition fees with a little subsidy from central budget. Most of the TEVT institutions funded locally, particularly, in part of general education. For a diploma and certificate level training tuition fees may apply. However, Ministry of Labor and Social Welfare, and its local labor regulatory agencies may provide fund for employment training. As 34 a public institution TEVT schools are provided with utility cost subsidy either from central or local budget. Central budget. 7. Public higher education institutions and government agencies under the MOSTEC, like Institute of Education Development, Education Inspectorate Board and State Training Fund submits their budget proposal to MoFE directly. But before that each of public universities and colleges sets level of tuition fee for next academic in consultation with MOSTEC. MOSTEC sometimes limits tuition fee increase, for example, not more that annual inflation rate. 8. MoFE officer reviews each of budget proposal and determines the subsidy, which will cover *heating, electricity and water costs, taking into consideration last year actual performance and changes in prices and tariffs. Another important factor that is also controlled by MOSTEC, beside the tuition level, is number of new enrolment. Any public or private institution gets special permission from MOSTEC to offer new degree program, or extent existing program over the next 4 years. This permission contains agreed in advance maximum number of enrolment each year. If a college enrolls more student than this limit, additional income earned may be drawn from that college and given to central budget. This can also happen for private institutions. However, public higher education institutions have greater flexibility in revenue generation and spending. 9. State training fund has been established to provide loans and grant assistance to students, studying in universities and colleges. It provides: * Tuition loan for up to 50-60 percent of new enrolment for public institutions. It is a need-based loan equal to average tuition amount for public higher education institutions. Loan must be paid back with 1% above inflation rate within 10 years, starting third year after graduation. Release from loan may take place, if a student, after completion of school works at least 4-5 years for a public agency or government. Students apply for this loan through their school, filling standard forms, and money goes to school directly. Students of private institutions may have access to this loan only if their institutions are accredited. And students also must pay whatever is a difference in tuition fee and loan provided. * Tuition grant for orphans, handicapped and children from a very poor family. To be eligible for this grant students must provide proof of eligibility. This category of support is being expanded, as new government (2000) announced that it would provide free higher education to herdsmen children and third child of a family of which 2 already studying at universities and colleges. * Tuition waiver and stipend for high achievers. If a student shows very high performance during 2 consecutive semesters, he or she might be entitled to receive a very small monthly stipend. In come cases, tuition may be waived. * One child of a public servant can be provided tuition fee from this fund. But amount equal to tuition fee must be planned in budget of public organization, where parents (father or mother) work, so that State training fund can be reimbursed later. * Tuition, travel and living expenses loan for oversees postgraduate study. The Parliament of Mongolia established this fund, equal to about US$2 million a few years ago. In order to get loan from this fund a student must have a Government agency as his or her future employer, show high results in qualifying exams, be enrolled in post graduate (master or doctoral degree) program at an university in a developed country within specified by MOSTEC areas of study. Maximum amount of loan per year, including tuition fee, travel expenses, and living expenses, is US$16,000. 35 10. In recent years, increasing amount of fund was allocated to the State Training Fund, thus, increasing overall expenditure for higher education. The students, who have got tuition loan 5-7 years started to pay back. Current level of revenue from loan re-payment is approximately 20% of total budget for State Training Fund. Science and technology fund is also a course offinancing of higher education. 11. Its purpose is to fund the research work within specified each year thematic areas. Bulk of fund goes to research institutes under Academy of Sciences, although, universities are eligible to submit research proposals for funding. 12. A significant portion of centralized measures (cost item of central budget) is textbook printing and production for general education. MOSTEC is responsible for its implementation. MOSTEC works directly with authors of a textbook. And when a manuscript for a textbook is approved a bidding for its printing occurs among private publishers. Distribution of textbooks to the schools charged to local education authorities. Current intention of MOSTEC is to encourage private printing companies become publishers, which can handle whole process of textbook production and distribution. A cost-sharing mechanism, possibly, would be introduced in the future, as free provision of textbook creates significant financial burden. 13. Capital expenditure budget has been a very small portion of central budget, due to scarcity of available funds. It provides fund for major repair, mostly in case of emergency, like collapse of heating system and fall of building roof. To build a new facility or expansion of existing school, a very high level (Parliament approval) commitment is necessary. Universities and colleges are responsible for their capital expenditure requirements. Lack of fund for capital expenditure is somehow compensated by program and project loans and grants from bilateral and multilateral donors. Education Budget approval 14. MoFE is responsible for preparation and submission of both central and local budgets. State Budget proposal is discussed at Government Cabinet meeting and once again revised according to Cabinet decision, before submission to the Parliament. Parliament discussion shapes final configuration of budget and Parliament passes State Budget Law. It specifies how much subsidy is going to go to a particular Aimag and how much is central budget by cost item and each organization, funded from central budget. 15. Since the subsidy from central budget is determined, local governments also approve their budget by considering it at People's Representatives Khural, highest self-governing body at local level. 16. Approved budgets for central government agencies are delivered to them. For a public university or college, its Management Board, comprising representatives from government, business, faculty, graduates and students, has authority to approve the budget. Budget implementation. 17. School budget implementation is tightly controlled by higher level of authority and subject to constrains of revenue collection and transfer of subsidy from central budget. Salary payments are often delayed because; there is no available fund. Dues for utility bills often accumulated for several months, and paid directly by higher authority, when funds became available. Purchase of necessary goods and materials also often delayed and, as a rule, requires several visits to higher authority. There are cases, 36 when school suffers from lack of funding, as authorities some times spend money for other purposes. Overall, budget discipline is improving at the local level. 18. Public higher education institutions have more autonomy in terms of budget implementation. Since most part of revenue comes from tuition fee, which is usually paid in August- September, there is no problem with cash receipts. Only concern is to make sure that the students pay tuition fee, and State Training Fund transfers loans to the school. Some notes on sources of education financing. 19. Research grants amount, in average, up to 5% of a university budget. Colleges and private higher education institutions usually do not get funding from this source, as they do not have capacity to conduct research projects. 20. Many colleges, particularly, in areas of business, law and languages, offer short-term training courses for graduates and business entities. Income from such activities may account also up to 5%. 37 Figure 1 Sources of education financing Sources of funding and cost items funded Type of Central bu get Local budget Private fund'g Education Science Direct State Training Centralized Direct subsidy Parents Revenue Other Institution & subsidy fand measures and generation private Techno- students funding logy fund Tuition loan for up Public higher Research Utihty to 50% of fresh Capital City bus ticket Tuition fee; Subsidiary Short term education grants costs students; expenditure discount (50%); enterprises training, institution (electricity, Tuition grant for Dormitory heating orphans, Tuition for fee; Interest Consulting and water), handicapped and public servants earning, very poor and children. Travel; Projects 50% herdsmen's travel children; Living discount Tuition waiver and expenses for stipend for high students achievers, from Tuition for turd western 5 cluld, attending provinces university at the same time, Tuition, travel and living expenses loan for oversees postgraduate study Research 50% Tuition loan up to City bus ticket Tuition fee, Subsidiary Short term Private grants travel 50% of fresh discount (50%) Dormtory enterprises training, higher discount students of fee, Interest Consulting, education for accredited Tuition for Travel, earning, Projects institution students institutions public servants Living Founders from children expenses subsidy western 5 provinces 50% Tuition for Capital City bus ticket Tuition fee, Subsidiary Short term Public travel employment expenditure discount (50%) Dormitory enterprises training, Vocational discount training Fixed cost fee, and for Funding for Travel, Technical students general Living School from education expenses westem 5 component of provinces curriculum Public Textbook Fixed cost Travel and Subsidiary General production Per student living enterprises Education and printing standard expenses (hvestock, School Capital variable cost for crop and expenditure Capital boarding vegetable under expenditure school cultivation) project Dormitory cost students funding Additional charges for minor repair Private Per student Tuition fee, General standard Food Education variable cost Schools Public Fixed cost, Fee equal Kindergarten Variable cost to 50% of excluding 50% food of food cost, expenses Toys and material Capital expenditure Private All kindergarten _ expenses 38 FIGURE 2: DONOR SUPPORT FOR EDUCATION Project Amount Funding Agency Grants School construction (16-17 schools) $25 rullion JICA Rural School Development $0 88 million DANIDA Learning for Life (non-formal education) $1 7 million DANIDA, through UNESCO Education reform projects (training, school grants, $6 million Mongolia Foundation for Open Society (Soros textbooks, resource centers, pre-schools. school based Foundation) curriculum reform) National Pre-school Strengthening $1.48 million Save the Children and UNDP Technical and vocational education $1.7 nullion GTZ University management, planning, and quality 2 ullion Euro EU/TACIS assurance, professional development for public servants Higher education scholarships $1 nullion Mongolia Foundation for Open Society (Soros Foundation) Non-formal education for dropouts $1 32 million UNICEF Distance education (equipment) $0 43 million KOICA Lomm/Credit Second Education Development Program $18 8 million ADB Technical and Vocational Education (under $15 5 million ADB preparation) National Poverty Alleviation Program (education $1 3 million World Bank component) FIGURE 3: SCHOOL-AGE POPULATION PROJECTIONS tctions for Pre-snOecified re GroUOfrom 2000 to 2025: Nrooolia Estmatd oudaon Sav 0h Chidre and UNDP, IIAesc~~' ~'..,4 -,2660 2786 [2918 303 3185 3304 Prth "cooen'a'e~~ ir~.7JKv~57 55 155 58s 61 60 Primr si:66bl age, 6Ll i, 7 '379 332 1~331 1337 357 366 Secn' ho6a' ~ 92 91 1333 130 334 352 Moglected NFuodtaori Oenialpe Siiousaeit(lS AIIA s,,,'~ 4. .~ , *~ ~ 1329 1430 157 1651 1763 1864 Pri~cciFe ~ae,' " ~ j'~2827 27 28 130 129 Pri. P aj chb61ag&,6-1I;' . " 'f' 18 7 162 163 165 175 179 Sdci6fid*ich6bIi~217~j. 194 193 162 162 164 173 Pr e&ted Po pWiftoifaoFfoToutan All,'AA6 & 'i-'l~ '2~~ ". O 100 ]G 100 100 100 100 Primkry'schooIlentrYaJc,6 2 1~~4'~ . 2 19 19 119I 8 m ch6l~~6!i.G~,,' ?2:~ .i '6'~ A ,142 119 113 It 112 11 1 Secondar sbhoolage 244 "'"" 4 14 11.4 108 105 107 Pojected Po liWitionis % of Populafidnn2000" ","r AnALe. 100 104.7 1097 1148 1197 1242 Primrna6 schoolentr age, 6 ,' '7 I, 100 965 965 101 8 107 1053 Primar' scliool'age;.6-11. ' , -'- ~ . 100 876 873 889 942 966 S&odary,schoo',age, 12-17 '.4 'k 100 99.7 849 84.2 85 2 898 Pro'jected Growth Riate (% p.a) 00_005iA K 200- 101 ~~ 5 20120 ~ 2020_ 25,'*. t' 'h e.,~ 'L ''~i ~ ~09 0.9 09 08 07 Prty school intriage &6~ 1f A7i;''>~! ~ ~-07 0 11 1 _____-0 3 Pri&aryschool a'',:6-1 L~' _, ~~"' ~ ~ ~ - 0 1I4120 Secondai'. schbolig.e,l 1247 ",' 4"''f':7g 1 -0.1 1-3.2 1-02 .0.2 ______ 39 Po mlaton rojctions for Pre-spmecif ied Age Groui2 from 2000i to 2025: Mongolia ojected Populåtion (Male +iFemale, Thoåsal_d_s)____ Age8 -S 1 49 9 263 222 221 3222 235 245 Agen12'-15::'~ 30272 2 27 221 224 1238 A 6 7120 139 1116 1109 1140 5114 Projected Populatidn (Fema e Th o usand_)____ __ Age 8 - 11 '& 130 108 109 109 115 1920 Age 12 -15 135 1124 1005 109 110 117 Age 16 -.17 59 169 157 53 54 56 Prjected P6pulatioi a % a20 of Totl Age'8-.1,' ÁÁ 9 99 8 76 7.3 17.4 174 Age 215: , -' 10.2 9 174 72 17 072 Ag16-^17-_ 4 5 5 9 -336 135 3.5 Projected Popidatlon as %u o of Popu2at0i02n(2000in thousad Ag~M11084 4 84 84.4 89 4 93.2 Age 12,300 Ag 215 0092 6 79.8 81.2 82 4 87 5 A'gó 16'-- 17 100) 115 8 96 7 90.8 91 7 95 Prjected Growth Rate (% p.a) - 200-0 205sch,ool 2,5-0j',, 02-5' Age18-5 -34 -01 01 0118 Ag l -l5-1.5 -30 4 0.3 1 2 lAgé 16 - 17 2 9 -_ 3 6 1-1 2 102 10.7 Projected School Age Population, 2000-2025 (in thousands) 300- 250-sho 50 -+- Primary 200- school 150- age, 8-11 100-41- Lower 100- secondary 50- school a,ge, 12-15 0 - Utpper 2000 2005 2010 2015 2020 2025 secondary school Source- World Bank Popidation Projections TABLE 1: PROPORTION OF STUDENTS IN SCHOOL AND DROPPED OUT, BY REGION AND INCOME GROUP Incone Group Status All Ulaanbaatar Aimag centers Soums Rural areas Very poor In-school 862 848 826 907 90.5 Dropped out 138 152 174 93 95 Poor In-school 91 3 100 936 91 3 81 4 Dropped out 87 00 64 87 186 Non-poor In-school 89 5 990 94 2 929 73 5 Dropped out 105 10 58 7 1 265 Total In-school 89 5 95 6 90 9 92 1 77 6 Dropped out 109 44 91 79 224 Sotr